- developers - lists.mariadb.org

[Maria-developers] Updated (by Timour): Subquery optimization: Efficient NOT IN execution with NULLs (68)
by worklog-noreply＠askmonty.org 04 Dec '09

04 Dec '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Subquery optimization: Efficient NOT IN execution with NULLs CREATION DATE..: Fri, 27 Nov 2009, 13:22 SUPERVISOR.....: Monty IMPLEMENTOR....: Timour COPIES TO......: CATEGORY.......: Server-Sprint TASK ID........: 68 (http://askmonty.org/worklog/?tid=68) VERSION........: Server-9.x STATUS.........: Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Timour - Fri, 04 Dec 2009, 11:27)=-=- Version updated. --- /tmp/wklog.68.old.5257 2009-12-04 11:27:11.000000000 +0200 +++ /tmp/wklog.68.new.5257 2009-12-04 11:27:11.000000000 +0200 @@ -1 +1 @@ -Benchmarks-3.0 +Server-9.x -=-=(Timour - Fri, 04 Dec 2009, 11:27)=-=- Category updated. --- /tmp/wklog.68.old.5242 2009-12-04 11:27:02.000000000 +0200 +++ /tmp/wklog.68.new.5242 2009-12-04 11:27:02.000000000 +0200 @@ -1 +1 @@ -Server-RawIdeaBin +Server-Sprint -=-=(Timour - Fri, 04 Dec 2009, 11:27)=-=- Status updated. --- /tmp/wklog.68.old.5242 2009-12-04 11:27:02.000000000 +0200 +++ /tmp/wklog.68.new.5242 2009-12-04 11:27:02.000000000 +0200 @@ -1 +1 @@ -Un-Assigned +Assigned -=-=(Timour - Fri, 04 Dec 2009, 11:26)=-=- High-Level Specification modified. --- /tmp/wklog.68.old.5182 2009-12-04 11:26:25.000000000 +0200 +++ /tmp/wklog.68.new.5182 2009-12-04 11:26:25.000000000 +0200 @@ -50,23 +50,39 @@ The array can be created on demand. Other consideration that may be taken into account: + 1. If columns a_j1,...,a_jm do not contain null values in the temporary -table at all, create for them only one index array (and of course do not -create any bitmaps for them). -2. Consider the ratio d(a_i)=N'/ V(a_i), where N' is the number of rows -where a_i is not null and V(a_i) is the number of distinct values for -a_i excluding nulls. - If d(a_i) is close to 1 then do not create any index array: check +table at all and v_j1,...,v_jm cannot be null, create for these columns +only one index array (and of course do not create any bitmaps for them). + +2. Consider the ratio d(a_i)=N'(a_i)/V(a_i), where N'(a_i) is the number +of rows, where a_i is not null and V(a_i) is the number of distinct +values for a_i excluding nulls. +If d(a_i) is close to N'(a_i) then do not create any index array: check whether there is a match running through the records that have been -filtered in. Anyway if d(a_i) is close to 1 then a intersection with -rowid{a_i=v_i} would not reduce the number of remaining rowids +filtered in. Anyway if d(a_i) is close to N'(a_i) then the intersection + with rowid{a_i=v_i} will not reduce the number of remaining rowids significantly. - If additionally N-N' is small do not create a bitmap for this column -either. -3. If for a column a_i d(a_i) is not close to 1, but N-N' is small a -sorted array of rowids from the set rowid{a_i is null} can be used -instead of a bitmap. +In other words is V(a_i) exceeds some threshold there is no sense to +create an index for a_i. +If additionally N-N'(a_i) is small do not create a bitmap for this +column either. + +3. If for a column a_i d(a_i) is not close to N'(a_i), but N-N'(a_i) is +small a sorted array of rowids from the set rowid{a_i is null} can be +used instead of a bitmap. + 4. We always have a match if R0= INTERSECT rowid{a_i is null} is not empty. Here i runs through all indexes from [1..n] such that v_i is not null. For a given subset of columns this fact has to be checked only once. It can be easily done with bitmap intersection. + +5. If v1,...,vn never can be a null, then indexes (sorted arrays) can be +created only for rows with nulls. + +6. If v1,...,vn never can be a null and number of rows with nulls is +small do not create indexes and do not create bitmaps. + +7. If you get a row with nulls in all columns stop filling the temporary +table and return UNKNOWN for any tuple <v1,...,vn>. + -=-=(Timour - Fri, 27 Nov 2009, 13:23)=-=- High-Level Specification modified. --- /tmp/wklog.68.old.17140 2009-11-27 11:23:17.000000000 +0000 +++ /tmp/wklog.68.new.17140 2009-11-27 11:23:17.000000000 +0000 @@ -1 +1,72 @@ +This a copy of the initial algorithm proposed by Igor: +====================================================== +For each left side tuple (v_1,...,v_n) we have to find the following set +of rowids for the temp table containing N rows as the result of +materialization of the subquery: + + R= INTERSECT (rowid{a_i=v_i} UNION rowid{a_i is null} where i runs +trough all indexes from [1..n] such that v_i is not null. + +Bear in mind the following specifics of this intersection: + (1) For each i: rowid{a_i=v_i} and rowid{a_i is null} are disjoint + (2) For each i: rowid{a_i is null} is the same for each tuple + +Due to (2) it makes sense to build rowid{a_i is null} only once. +A good representation for such sets would be bitmaps: +- it requires minimum memory: not more than N*n bits in total +- search of an element in a set is extremely cheap + +Taken all above into account I could suggest the following algorithm to +build R: + + Using indexes (read about them below) for each column participating in the + intersection, + merge ordered sets rowid{a_i=v_i} in the following manner. + If a rowid r has been encountered maximum in k sets +rowid{a_i1=v_i1},...,rowid(a_ik=v_ik), + then it has to be checked against all rowid{a_i=v_i} such that i is +not in {i1,...,ik}. + As soon as we fail to find r in one of these sets we discard it. + If r has been found in all of them then r belongs to the set R. + +Here we use the property (1): any r from rowid{a_i=v_i} UNION rowid{a_i +is null} is either +belongs to rowid{a_i=v_i} or to rowid{a_i is null}. From this we can +infer that for any r from R +indexes a_i can be uniquely divided into two groups: one contains +indexes a_i where r belongs to +the sets rowid{a_i=v_i}, the other contains indexes a_j such that r +belongs to rowid{a_j is null}. + +Now let's talk how to get elements from rowid{a_i=v_i} in a sorted order +needed for the merge procedure. We could use BTREE indexes for temp +table. But they are rather expensive and +take a lot of memory as the are implemented with RB trees. +I would suggest creating for each column from the temporary table just +an array of rowids sorted by the value from column a. +Index lookup in such an array is cheap. It's also rather cheap to check +that the next rowid refers to a row with a different value in column a. +The array can be created on demand. + +Other consideration that may be taken into account: +1. If columns a_j1,...,a_jm do not contain null values in the temporary +table at all, create for them only one index array (and of course do not +create any bitmaps for them). +2. Consider the ratio d(a_i)=N'/ V(a_i), where N' is the number of rows +where a_i is not null and V(a_i) is the number of distinct values for +a_i excluding nulls. + If d(a_i) is close to 1 then do not create any index array: check +whether there is a match running through the records that have been +filtered in. Anyway if d(a_i) is close to 1 then a intersection with +rowid{a_i=v_i} would not reduce the number of remaining rowids +significantly. + If additionally N-N' is small do not create a bitmap for this column +either. +3. If for a column a_i d(a_i) is not close to 1, but N-N' is small a +sorted array of rowids from the set rowid{a_i is null} can be used +instead of a bitmap. +4. We always have a match if R0= INTERSECT rowid{a_i is null} is not +empty. Here i runs through all indexes from [1..n] such that v_i is not +null. For a given subset of columns this fact has to be checked only +once. It can be easily done with bitmap intersection. DESCRIPTION: The goal of this task is to implement efficient execution of NOT IN subquery predicates of the form: <oe_1,...,oe_n> NOT IN <non_correlated subquery> when either some oe_i, or some subqury result column contains NULLs. The problem with such predicates is that it is possible to use index lookups only when neither argument of the predicate contains NULLs. If some argument contains a NULL, then due to NULL semantics, it plays the role of a wildcard. If we were to use regular index lookups, then we would get 'no match' for some outer tuple (thus the predicate evaluates to FALSE), while the SQL semantics means 'partial match', and the predicate should evaluate to NULL. This task implements an efficient algorithm to compute such 'parial matches', where a NULL matches any value. HIGH-LEVEL SPECIFICATION: This a copy of the initial algorithm proposed by Igor: ====================================================== For each left side tuple (v_1,...,v_n) we have to find the following set of rowids for the temp table containing N rows as the result of materialization of the subquery: R= INTERSECT (rowid{a_i=v_i} UNION rowid{a_i is null} where i runs trough all indexes from [1..n] such that v_i is not null. Bear in mind the following specifics of this intersection: (1) For each i: rowid{a_i=v_i} and rowid{a_i is null} are disjoint (2) For each i: rowid{a_i is null} is the same for each tuple Due to (2) it makes sense to build rowid{a_i is null} only once. A good representation for such sets would be bitmaps: - it requires minimum memory: not more than N*n bits in total - search of an element in a set is extremely cheap Taken all above into account I could suggest the following algorithm to build R: Using indexes (read about them below) for each column participating in the intersection, merge ordered sets rowid{a_i=v_i} in the following manner. If a rowid r has been encountered maximum in k sets rowid{a_i1=v_i1},...,rowid(a_ik=v_ik), then it has to be checked against all rowid{a_i=v_i} such that i is not in {i1,...,ik}. As soon as we fail to find r in one of these sets we discard it. If r has been found in all of them then r belongs to the set R. Here we use the property (1): any r from rowid{a_i=v_i} UNION rowid{a_i is null} is either belongs to rowid{a_i=v_i} or to rowid{a_i is null}. From this we can infer that for any r from R indexes a_i can be uniquely divided into two groups: one contains indexes a_i where r belongs to the sets rowid{a_i=v_i}, the other contains indexes a_j such that r belongs to rowid{a_j is null}. Now let's talk how to get elements from rowid{a_i=v_i} in a sorted order needed for the merge procedure. We could use BTREE indexes for temp table. But they are rather expensive and take a lot of memory as the are implemented with RB trees. I would suggest creating for each column from the temporary table just an array of rowids sorted by the value from column a. Index lookup in such an array is cheap. It's also rather cheap to check that the next rowid refers to a row with a different value in column a. The array can be created on demand. Other consideration that may be taken into account: 1. If columns a_j1,...,a_jm do not contain null values in the temporary table at all and v_j1,...,v_jm cannot be null, create for these columns only one index array (and of course do not create any bitmaps for them). 2. Consider the ratio d(a_i)=N'(a_i)/V(a_i), where N'(a_i) is the number of rows, where a_i is not null and V(a_i) is the number of distinct values for a_i excluding nulls. If d(a_i) is close to N'(a_i) then do not create any index array: check whether there is a match running through the records that have been filtered in. Anyway if d(a_i) is close to N'(a_i) then the intersection with rowid{a_i=v_i} will not reduce the number of remaining rowids significantly. In other words is V(a_i) exceeds some threshold there is no sense to create an index for a_i. If additionally N-N'(a_i) is small do not create a bitmap for this column either. 3. If for a column a_i d(a_i) is not close to N'(a_i), but N-N'(a_i) is small a sorted array of rowids from the set rowid{a_i is null} can be used instead of a bitmap. 4. We always have a match if R0= INTERSECT rowid{a_i is null} is not empty. Here i runs through all indexes from [1..n] such that v_i is not null. For a given subset of columns this fact has to be checked only once. It can be easily done with bitmap intersection. 5. If v1,...,vn never can be a null, then indexes (sorted arrays) can be created only for rows with nulls. 6. If v1,...,vn never can be a null and number of rows with nulls is small do not create indexes and do not create bitmaps. 7. If you get a row with nulls in all columns stop filling the temporary table and return UNKNOWN for any tuple <v1,...,vn>. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Timour): Subquery optimization: Efficient NOT IN execution with NULLs (68)
by worklog-noreply＠askmonty.org 04 Dec '09

04 Dec '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Subquery optimization: Efficient NOT IN execution with NULLs CREATION DATE..: Fri, 27 Nov 2009, 13:22 SUPERVISOR.....: Monty IMPLEMENTOR....: Timour COPIES TO......: CATEGORY.......: Server-Sprint TASK ID........: 68 (http://askmonty.org/worklog/?tid=68) VERSION........: Benchmarks-3.0 STATUS.........: Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Timour - Fri, 04 Dec 2009, 11:27)=-=- Category updated. --- /tmp/wklog.68.old.5242 2009-12-04 11:27:02.000000000 +0200 +++ /tmp/wklog.68.new.5242 2009-12-04 11:27:02.000000000 +0200 @@ -1 +1 @@ -Server-RawIdeaBin +Server-Sprint -=-=(Timour - Fri, 04 Dec 2009, 11:27)=-=- Status updated. --- /tmp/wklog.68.old.5242 2009-12-04 11:27:02.000000000 +0200 +++ /tmp/wklog.68.new.5242 2009-12-04 11:27:02.000000000 +0200 @@ -1 +1 @@ -Un-Assigned +Assigned -=-=(Timour - Fri, 04 Dec 2009, 11:26)=-=- High-Level Specification modified. --- /tmp/wklog.68.old.5182 2009-12-04 11:26:25.000000000 +0200 +++ /tmp/wklog.68.new.5182 2009-12-04 11:26:25.000000000 +0200 @@ -50,23 +50,39 @@ The array can be created on demand. Other consideration that may be taken into account: + 1. If columns a_j1,...,a_jm do not contain null values in the temporary -table at all, create for them only one index array (and of course do not -create any bitmaps for them). -2. Consider the ratio d(a_i)=N'/ V(a_i), where N' is the number of rows -where a_i is not null and V(a_i) is the number of distinct values for -a_i excluding nulls. - If d(a_i) is close to 1 then do not create any index array: check +table at all and v_j1,...,v_jm cannot be null, create for these columns +only one index array (and of course do not create any bitmaps for them). + +2. Consider the ratio d(a_i)=N'(a_i)/V(a_i), where N'(a_i) is the number +of rows, where a_i is not null and V(a_i) is the number of distinct +values for a_i excluding nulls. +If d(a_i) is close to N'(a_i) then do not create any index array: check whether there is a match running through the records that have been -filtered in. Anyway if d(a_i) is close to 1 then a intersection with -rowid{a_i=v_i} would not reduce the number of remaining rowids +filtered in. Anyway if d(a_i) is close to N'(a_i) then the intersection + with rowid{a_i=v_i} will not reduce the number of remaining rowids significantly. - If additionally N-N' is small do not create a bitmap for this column -either. -3. If for a column a_i d(a_i) is not close to 1, but N-N' is small a -sorted array of rowids from the set rowid{a_i is null} can be used -instead of a bitmap. +In other words is V(a_i) exceeds some threshold there is no sense to +create an index for a_i. +If additionally N-N'(a_i) is small do not create a bitmap for this +column either. + +3. If for a column a_i d(a_i) is not close to N'(a_i), but N-N'(a_i) is +small a sorted array of rowids from the set rowid{a_i is null} can be +used instead of a bitmap. + 4. We always have a match if R0= INTERSECT rowid{a_i is null} is not empty. Here i runs through all indexes from [1..n] such that v_i is not null. For a given subset of columns this fact has to be checked only once. It can be easily done with bitmap intersection. + +5. If v1,...,vn never can be a null, then indexes (sorted arrays) can be +created only for rows with nulls. + +6. If v1,...,vn never can be a null and number of rows with nulls is +small do not create indexes and do not create bitmaps. + +7. If you get a row with nulls in all columns stop filling the temporary +table and return UNKNOWN for any tuple <v1,...,vn>. + -=-=(Timour - Fri, 27 Nov 2009, 13:23)=-=- High-Level Specification modified. --- /tmp/wklog.68.old.17140 2009-11-27 11:23:17.000000000 +0000 +++ /tmp/wklog.68.new.17140 2009-11-27 11:23:17.000000000 +0000 @@ -1 +1,72 @@ +This a copy of the initial algorithm proposed by Igor: +====================================================== +For each left side tuple (v_1,...,v_n) we have to find the following set +of rowids for the temp table containing N rows as the result of +materialization of the subquery: + + R= INTERSECT (rowid{a_i=v_i} UNION rowid{a_i is null} where i runs +trough all indexes from [1..n] such that v_i is not null. + +Bear in mind the following specifics of this intersection: + (1) For each i: rowid{a_i=v_i} and rowid{a_i is null} are disjoint + (2) For each i: rowid{a_i is null} is the same for each tuple + +Due to (2) it makes sense to build rowid{a_i is null} only once. +A good representation for such sets would be bitmaps: +- it requires minimum memory: not more than N*n bits in total +- search of an element in a set is extremely cheap + +Taken all above into account I could suggest the following algorithm to +build R: + + Using indexes (read about them below) for each column participating in the + intersection, + merge ordered sets rowid{a_i=v_i} in the following manner. + If a rowid r has been encountered maximum in k sets +rowid{a_i1=v_i1},...,rowid(a_ik=v_ik), + then it has to be checked against all rowid{a_i=v_i} such that i is +not in {i1,...,ik}. + As soon as we fail to find r in one of these sets we discard it. + If r has been found in all of them then r belongs to the set R. + +Here we use the property (1): any r from rowid{a_i=v_i} UNION rowid{a_i +is null} is either +belongs to rowid{a_i=v_i} or to rowid{a_i is null}. From this we can +infer that for any r from R +indexes a_i can be uniquely divided into two groups: one contains +indexes a_i where r belongs to +the sets rowid{a_i=v_i}, the other contains indexes a_j such that r +belongs to rowid{a_j is null}. + +Now let's talk how to get elements from rowid{a_i=v_i} in a sorted order +needed for the merge procedure. We could use BTREE indexes for temp +table. But they are rather expensive and +take a lot of memory as the are implemented with RB trees. +I would suggest creating for each column from the temporary table just +an array of rowids sorted by the value from column a. +Index lookup in such an array is cheap. It's also rather cheap to check +that the next rowid refers to a row with a different value in column a. +The array can be created on demand. + +Other consideration that may be taken into account: +1. If columns a_j1,...,a_jm do not contain null values in the temporary +table at all, create for them only one index array (and of course do not +create any bitmaps for them). +2. Consider the ratio d(a_i)=N'/ V(a_i), where N' is the number of rows +where a_i is not null and V(a_i) is the number of distinct values for +a_i excluding nulls. + If d(a_i) is close to 1 then do not create any index array: check +whether there is a match running through the records that have been +filtered in. Anyway if d(a_i) is close to 1 then a intersection with +rowid{a_i=v_i} would not reduce the number of remaining rowids +significantly. + If additionally N-N' is small do not create a bitmap for this column +either. +3. If for a column a_i d(a_i) is not close to 1, but N-N' is small a +sorted array of rowids from the set rowid{a_i is null} can be used +instead of a bitmap. +4. We always have a match if R0= INTERSECT rowid{a_i is null} is not +empty. Here i runs through all indexes from [1..n] such that v_i is not +null. For a given subset of columns this fact has to be checked only +once. It can be easily done with bitmap intersection. DESCRIPTION: The goal of this task is to implement efficient execution of NOT IN subquery predicates of the form: <oe_1,...,oe_n> NOT IN <non_correlated subquery> when either some oe_i, or some subqury result column contains NULLs. The problem with such predicates is that it is possible to use index lookups only when neither argument of the predicate contains NULLs. If some argument contains a NULL, then due to NULL semantics, it plays the role of a wildcard. If we were to use regular index lookups, then we would get 'no match' for some outer tuple (thus the predicate evaluates to FALSE), while the SQL semantics means 'partial match', and the predicate should evaluate to NULL. This task implements an efficient algorithm to compute such 'parial matches', where a NULL matches any value. HIGH-LEVEL SPECIFICATION: This a copy of the initial algorithm proposed by Igor: ====================================================== For each left side tuple (v_1,...,v_n) we have to find the following set of rowids for the temp table containing N rows as the result of materialization of the subquery: R= INTERSECT (rowid{a_i=v_i} UNION rowid{a_i is null} where i runs trough all indexes from [1..n] such that v_i is not null. Bear in mind the following specifics of this intersection: (1) For each i: rowid{a_i=v_i} and rowid{a_i is null} are disjoint (2) For each i: rowid{a_i is null} is the same for each tuple Due to (2) it makes sense to build rowid{a_i is null} only once. A good representation for such sets would be bitmaps: - it requires minimum memory: not more than N*n bits in total - search of an element in a set is extremely cheap Taken all above into account I could suggest the following algorithm to build R: Using indexes (read about them below) for each column participating in the intersection, merge ordered sets rowid{a_i=v_i} in the following manner. If a rowid r has been encountered maximum in k sets rowid{a_i1=v_i1},...,rowid(a_ik=v_ik), then it has to be checked against all rowid{a_i=v_i} such that i is not in {i1,...,ik}. As soon as we fail to find r in one of these sets we discard it. If r has been found in all of them then r belongs to the set R. Here we use the property (1): any r from rowid{a_i=v_i} UNION rowid{a_i is null} is either belongs to rowid{a_i=v_i} or to rowid{a_i is null}. From this we can infer that for any r from R indexes a_i can be uniquely divided into two groups: one contains indexes a_i where r belongs to the sets rowid{a_i=v_i}, the other contains indexes a_j such that r belongs to rowid{a_j is null}. Now let's talk how to get elements from rowid{a_i=v_i} in a sorted order needed for the merge procedure. We could use BTREE indexes for temp table. But they are rather expensive and take a lot of memory as the are implemented with RB trees. I would suggest creating for each column from the temporary table just an array of rowids sorted by the value from column a. Index lookup in such an array is cheap. It's also rather cheap to check that the next rowid refers to a row with a different value in column a. The array can be created on demand. Other consideration that may be taken into account: 1. If columns a_j1,...,a_jm do not contain null values in the temporary table at all and v_j1,...,v_jm cannot be null, create for these columns only one index array (and of course do not create any bitmaps for them). 2. Consider the ratio d(a_i)=N'(a_i)/V(a_i), where N'(a_i) is the number of rows, where a_i is not null and V(a_i) is the number of distinct values for a_i excluding nulls. If d(a_i) is close to N'(a_i) then do not create any index array: check whether there is a match running through the records that have been filtered in. Anyway if d(a_i) is close to N'(a_i) then the intersection with rowid{a_i=v_i} will not reduce the number of remaining rowids significantly. In other words is V(a_i) exceeds some threshold there is no sense to create an index for a_i. If additionally N-N'(a_i) is small do not create a bitmap for this column either. 3. If for a column a_i d(a_i) is not close to N'(a_i), but N-N'(a_i) is small a sorted array of rowids from the set rowid{a_i is null} can be used instead of a bitmap. 4. We always have a match if R0= INTERSECT rowid{a_i is null} is not empty. Here i runs through all indexes from [1..n] such that v_i is not null. For a given subset of columns this fact has to be checked only once. It can be easily done with bitmap intersection. 5. If v1,...,vn never can be a null, then indexes (sorted arrays) can be created only for rows with nulls. 6. If v1,...,vn never can be a null and number of rows with nulls is small do not create indexes and do not create bitmaps. 7. If you get a row with nulls in all columns stop filling the temporary table and return UNKNOWN for any tuple <v1,...,vn>. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Timour): Subquery optimization: Efficient NOT IN execution with NULLs (68)
by worklog-noreply＠askmonty.org 04 Dec '09

04 Dec '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Subquery optimization: Efficient NOT IN execution with NULLs CREATION DATE..: Fri, 27 Nov 2009, 13:22 SUPERVISOR.....: Monty IMPLEMENTOR....: Timour COPIES TO......: CATEGORY.......: Server-Sprint TASK ID........: 68 (http://askmonty.org/worklog/?tid=68) VERSION........: Benchmarks-3.0 STATUS.........: Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Timour - Fri, 04 Dec 2009, 11:27)=-=- Category updated. --- /tmp/wklog.68.old.5242 2009-12-04 11:27:02.000000000 +0200 +++ /tmp/wklog.68.new.5242 2009-12-04 11:27:02.000000000 +0200 @@ -1 +1 @@ -Server-RawIdeaBin +Server-Sprint -=-=(Timour - Fri, 04 Dec 2009, 11:27)=-=- Status updated. --- /tmp/wklog.68.old.5242 2009-12-04 11:27:02.000000000 +0200 +++ /tmp/wklog.68.new.5242 2009-12-04 11:27:02.000000000 +0200 @@ -1 +1 @@ -Un-Assigned +Assigned -=-=(Timour - Fri, 04 Dec 2009, 11:26)=-=- High-Level Specification modified. --- /tmp/wklog.68.old.5182 2009-12-04 11:26:25.000000000 +0200 +++ /tmp/wklog.68.new.5182 2009-12-04 11:26:25.000000000 +0200 @@ -50,23 +50,39 @@ The array can be created on demand. Other consideration that may be taken into account: + 1. If columns a_j1,...,a_jm do not contain null values in the temporary -table at all, create for them only one index array (and of course do not -create any bitmaps for them). -2. Consider the ratio d(a_i)=N'/ V(a_i), where N' is the number of rows -where a_i is not null and V(a_i) is the number of distinct values for -a_i excluding nulls. - If d(a_i) is close to 1 then do not create any index array: check +table at all and v_j1,...,v_jm cannot be null, create for these columns +only one index array (and of course do not create any bitmaps for them). + +2. Consider the ratio d(a_i)=N'(a_i)/V(a_i), where N'(a_i) is the number +of rows, where a_i is not null and V(a_i) is the number of distinct +values for a_i excluding nulls. +If d(a_i) is close to N'(a_i) then do not create any index array: check whether there is a match running through the records that have been -filtered in. Anyway if d(a_i) is close to 1 then a intersection with -rowid{a_i=v_i} would not reduce the number of remaining rowids +filtered in. Anyway if d(a_i) is close to N'(a_i) then the intersection + with rowid{a_i=v_i} will not reduce the number of remaining rowids significantly. - If additionally N-N' is small do not create a bitmap for this column -either. -3. If for a column a_i d(a_i) is not close to 1, but N-N' is small a -sorted array of rowids from the set rowid{a_i is null} can be used -instead of a bitmap. +In other words is V(a_i) exceeds some threshold there is no sense to +create an index for a_i. +If additionally N-N'(a_i) is small do not create a bitmap for this +column either. + +3. If for a column a_i d(a_i) is not close to N'(a_i), but N-N'(a_i) is +small a sorted array of rowids from the set rowid{a_i is null} can be +used instead of a bitmap. + 4. We always have a match if R0= INTERSECT rowid{a_i is null} is not empty. Here i runs through all indexes from [1..n] such that v_i is not null. For a given subset of columns this fact has to be checked only once. It can be easily done with bitmap intersection. + +5. If v1,...,vn never can be a null, then indexes (sorted arrays) can be +created only for rows with nulls. + +6. If v1,...,vn never can be a null and number of rows with nulls is +small do not create indexes and do not create bitmaps. + +7. If you get a row with nulls in all columns stop filling the temporary +table and return UNKNOWN for any tuple <v1,...,vn>. + -=-=(Timour - Fri, 27 Nov 2009, 13:23)=-=- High-Level Specification modified. --- /tmp/wklog.68.old.17140 2009-11-27 11:23:17.000000000 +0000 +++ /tmp/wklog.68.new.17140 2009-11-27 11:23:17.000000000 +0000 @@ -1 +1,72 @@ +This a copy of the initial algorithm proposed by Igor: +====================================================== +For each left side tuple (v_1,...,v_n) we have to find the following set +of rowids for the temp table containing N rows as the result of +materialization of the subquery: + + R= INTERSECT (rowid{a_i=v_i} UNION rowid{a_i is null} where i runs +trough all indexes from [1..n] such that v_i is not null. + +Bear in mind the following specifics of this intersection: + (1) For each i: rowid{a_i=v_i} and rowid{a_i is null} are disjoint + (2) For each i: rowid{a_i is null} is the same for each tuple + +Due to (2) it makes sense to build rowid{a_i is null} only once. +A good representation for such sets would be bitmaps: +- it requires minimum memory: not more than N*n bits in total +- search of an element in a set is extremely cheap + +Taken all above into account I could suggest the following algorithm to +build R: + + Using indexes (read about them below) for each column participating in the + intersection, + merge ordered sets rowid{a_i=v_i} in the following manner. + If a rowid r has been encountered maximum in k sets +rowid{a_i1=v_i1},...,rowid(a_ik=v_ik), + then it has to be checked against all rowid{a_i=v_i} such that i is +not in {i1,...,ik}. + As soon as we fail to find r in one of these sets we discard it. + If r has been found in all of them then r belongs to the set R. + +Here we use the property (1): any r from rowid{a_i=v_i} UNION rowid{a_i +is null} is either +belongs to rowid{a_i=v_i} or to rowid{a_i is null}. From this we can +infer that for any r from R +indexes a_i can be uniquely divided into two groups: one contains +indexes a_i where r belongs to +the sets rowid{a_i=v_i}, the other contains indexes a_j such that r +belongs to rowid{a_j is null}. + +Now let's talk how to get elements from rowid{a_i=v_i} in a sorted order +needed for the merge procedure. We could use BTREE indexes for temp +table. But they are rather expensive and +take a lot of memory as the are implemented with RB trees. +I would suggest creating for each column from the temporary table just +an array of rowids sorted by the value from column a. +Index lookup in such an array is cheap. It's also rather cheap to check +that the next rowid refers to a row with a different value in column a. +The array can be created on demand. + +Other consideration that may be taken into account: +1. If columns a_j1,...,a_jm do not contain null values in the temporary +table at all, create for them only one index array (and of course do not +create any bitmaps for them). +2. Consider the ratio d(a_i)=N'/ V(a_i), where N' is the number of rows +where a_i is not null and V(a_i) is the number of distinct values for +a_i excluding nulls. + If d(a_i) is close to 1 then do not create any index array: check +whether there is a match running through the records that have been +filtered in. Anyway if d(a_i) is close to 1 then a intersection with +rowid{a_i=v_i} would not reduce the number of remaining rowids +significantly. + If additionally N-N' is small do not create a bitmap for this column +either. +3. If for a column a_i d(a_i) is not close to 1, but N-N' is small a +sorted array of rowids from the set rowid{a_i is null} can be used +instead of a bitmap. +4. We always have a match if R0= INTERSECT rowid{a_i is null} is not +empty. Here i runs through all indexes from [1..n] such that v_i is not +null. For a given subset of columns this fact has to be checked only +once. It can be easily done with bitmap intersection. DESCRIPTION: The goal of this task is to implement efficient execution of NOT IN subquery predicates of the form: <oe_1,...,oe_n> NOT IN <non_correlated subquery> when either some oe_i, or some subqury result column contains NULLs. The problem with such predicates is that it is possible to use index lookups only when neither argument of the predicate contains NULLs. If some argument contains a NULL, then due to NULL semantics, it plays the role of a wildcard. If we were to use regular index lookups, then we would get 'no match' for some outer tuple (thus the predicate evaluates to FALSE), while the SQL semantics means 'partial match', and the predicate should evaluate to NULL. This task implements an efficient algorithm to compute such 'parial matches', where a NULL matches any value. HIGH-LEVEL SPECIFICATION: This a copy of the initial algorithm proposed by Igor: ====================================================== For each left side tuple (v_1,...,v_n) we have to find the following set of rowids for the temp table containing N rows as the result of materialization of the subquery: R= INTERSECT (rowid{a_i=v_i} UNION rowid{a_i is null} where i runs trough all indexes from [1..n] such that v_i is not null. Bear in mind the following specifics of this intersection: (1) For each i: rowid{a_i=v_i} and rowid{a_i is null} are disjoint (2) For each i: rowid{a_i is null} is the same for each tuple Due to (2) it makes sense to build rowid{a_i is null} only once. A good representation for such sets would be bitmaps: - it requires minimum memory: not more than N*n bits in total - search of an element in a set is extremely cheap Taken all above into account I could suggest the following algorithm to build R: Using indexes (read about them below) for each column participating in the intersection, merge ordered sets rowid{a_i=v_i} in the following manner. If a rowid r has been encountered maximum in k sets rowid{a_i1=v_i1},...,rowid(a_ik=v_ik), then it has to be checked against all rowid{a_i=v_i} such that i is not in {i1,...,ik}. As soon as we fail to find r in one of these sets we discard it. If r has been found in all of them then r belongs to the set R. Here we use the property (1): any r from rowid{a_i=v_i} UNION rowid{a_i is null} is either belongs to rowid{a_i=v_i} or to rowid{a_i is null}. From this we can infer that for any r from R indexes a_i can be uniquely divided into two groups: one contains indexes a_i where r belongs to the sets rowid{a_i=v_i}, the other contains indexes a_j such that r belongs to rowid{a_j is null}. Now let's talk how to get elements from rowid{a_i=v_i} in a sorted order needed for the merge procedure. We could use BTREE indexes for temp table. But they are rather expensive and take a lot of memory as the are implemented with RB trees. I would suggest creating for each column from the temporary table just an array of rowids sorted by the value from column a. Index lookup in such an array is cheap. It's also rather cheap to check that the next rowid refers to a row with a different value in column a. The array can be created on demand. Other consideration that may be taken into account: 1. If columns a_j1,...,a_jm do not contain null values in the temporary table at all and v_j1,...,v_jm cannot be null, create for these columns only one index array (and of course do not create any bitmaps for them). 2. Consider the ratio d(a_i)=N'(a_i)/V(a_i), where N'(a_i) is the number of rows, where a_i is not null and V(a_i) is the number of distinct values for a_i excluding nulls. If d(a_i) is close to N'(a_i) then do not create any index array: check whether there is a match running through the records that have been filtered in. Anyway if d(a_i) is close to N'(a_i) then the intersection with rowid{a_i=v_i} will not reduce the number of remaining rowids significantly. In other words is V(a_i) exceeds some threshold there is no sense to create an index for a_i. If additionally N-N'(a_i) is small do not create a bitmap for this column either. 3. If for a column a_i d(a_i) is not close to N'(a_i), but N-N'(a_i) is small a sorted array of rowids from the set rowid{a_i is null} can be used instead of a bitmap. 4. We always have a match if R0= INTERSECT rowid{a_i is null} is not empty. Here i runs through all indexes from [1..n] such that v_i is not null. For a given subset of columns this fact has to be checked only once. It can be easily done with bitmap intersection. 5. If v1,...,vn never can be a null, then indexes (sorted arrays) can be created only for rows with nulls. 6. If v1,...,vn never can be a null and number of rows with nulls is small do not create indexes and do not create bitmaps. 7. If you get a row with nulls in all columns stop filling the temporary table and return UNKNOWN for any tuple <v1,...,vn>. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Timour): Subquery optimization: Efficient NOT IN execution with NULLs (68)
by worklog-noreply＠askmonty.org 04 Dec '09

04 Dec '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Subquery optimization: Efficient NOT IN execution with NULLs CREATION DATE..: Fri, 27 Nov 2009, 13:22 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: CATEGORY.......: Server-RawIdeaBin TASK ID........: 68 (http://askmonty.org/worklog/?tid=68) VERSION........: Benchmarks-3.0 STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Timour - Fri, 04 Dec 2009, 11:26)=-=- High-Level Specification modified. --- /tmp/wklog.68.old.5182 2009-12-04 11:26:25.000000000 +0200 +++ /tmp/wklog.68.new.5182 2009-12-04 11:26:25.000000000 +0200 @@ -50,23 +50,39 @@ The array can be created on demand. Other consideration that may be taken into account: + 1. If columns a_j1,...,a_jm do not contain null values in the temporary -table at all, create for them only one index array (and of course do not -create any bitmaps for them). -2. Consider the ratio d(a_i)=N'/ V(a_i), where N' is the number of rows -where a_i is not null and V(a_i) is the number of distinct values for -a_i excluding nulls. - If d(a_i) is close to 1 then do not create any index array: check +table at all and v_j1,...,v_jm cannot be null, create for these columns +only one index array (and of course do not create any bitmaps for them). + +2. Consider the ratio d(a_i)=N'(a_i)/V(a_i), where N'(a_i) is the number +of rows, where a_i is not null and V(a_i) is the number of distinct +values for a_i excluding nulls. +If d(a_i) is close to N'(a_i) then do not create any index array: check whether there is a match running through the records that have been -filtered in. Anyway if d(a_i) is close to 1 then a intersection with -rowid{a_i=v_i} would not reduce the number of remaining rowids +filtered in. Anyway if d(a_i) is close to N'(a_i) then the intersection + with rowid{a_i=v_i} will not reduce the number of remaining rowids significantly. - If additionally N-N' is small do not create a bitmap for this column -either. -3. If for a column a_i d(a_i) is not close to 1, but N-N' is small a -sorted array of rowids from the set rowid{a_i is null} can be used -instead of a bitmap. +In other words is V(a_i) exceeds some threshold there is no sense to +create an index for a_i. +If additionally N-N'(a_i) is small do not create a bitmap for this +column either. + +3. If for a column a_i d(a_i) is not close to N'(a_i), but N-N'(a_i) is +small a sorted array of rowids from the set rowid{a_i is null} can be +used instead of a bitmap. + 4. We always have a match if R0= INTERSECT rowid{a_i is null} is not empty. Here i runs through all indexes from [1..n] such that v_i is not null. For a given subset of columns this fact has to be checked only once. It can be easily done with bitmap intersection. + +5. If v1,...,vn never can be a null, then indexes (sorted arrays) can be +created only for rows with nulls. + +6. If v1,...,vn never can be a null and number of rows with nulls is +small do not create indexes and do not create bitmaps. + +7. If you get a row with nulls in all columns stop filling the temporary +table and return UNKNOWN for any tuple <v1,...,vn>. + -=-=(Timour - Fri, 27 Nov 2009, 13:23)=-=- High-Level Specification modified. --- /tmp/wklog.68.old.17140 2009-11-27 11:23:17.000000000 +0000 +++ /tmp/wklog.68.new.17140 2009-11-27 11:23:17.000000000 +0000 @@ -1 +1,72 @@ +This a copy of the initial algorithm proposed by Igor: +====================================================== +For each left side tuple (v_1,...,v_n) we have to find the following set +of rowids for the temp table containing N rows as the result of +materialization of the subquery: + + R= INTERSECT (rowid{a_i=v_i} UNION rowid{a_i is null} where i runs +trough all indexes from [1..n] such that v_i is not null. + +Bear in mind the following specifics of this intersection: + (1) For each i: rowid{a_i=v_i} and rowid{a_i is null} are disjoint + (2) For each i: rowid{a_i is null} is the same for each tuple + +Due to (2) it makes sense to build rowid{a_i is null} only once. +A good representation for such sets would be bitmaps: +- it requires minimum memory: not more than N*n bits in total +- search of an element in a set is extremely cheap + +Taken all above into account I could suggest the following algorithm to +build R: + + Using indexes (read about them below) for each column participating in the + intersection, + merge ordered sets rowid{a_i=v_i} in the following manner. + If a rowid r has been encountered maximum in k sets +rowid{a_i1=v_i1},...,rowid(a_ik=v_ik), + then it has to be checked against all rowid{a_i=v_i} such that i is +not in {i1,...,ik}. + As soon as we fail to find r in one of these sets we discard it. + If r has been found in all of them then r belongs to the set R. + +Here we use the property (1): any r from rowid{a_i=v_i} UNION rowid{a_i +is null} is either +belongs to rowid{a_i=v_i} or to rowid{a_i is null}. From this we can +infer that for any r from R +indexes a_i can be uniquely divided into two groups: one contains +indexes a_i where r belongs to +the sets rowid{a_i=v_i}, the other contains indexes a_j such that r +belongs to rowid{a_j is null}. + +Now let's talk how to get elements from rowid{a_i=v_i} in a sorted order +needed for the merge procedure. We could use BTREE indexes for temp +table. But they are rather expensive and +take a lot of memory as the are implemented with RB trees. +I would suggest creating for each column from the temporary table just +an array of rowids sorted by the value from column a. +Index lookup in such an array is cheap. It's also rather cheap to check +that the next rowid refers to a row with a different value in column a. +The array can be created on demand. + +Other consideration that may be taken into account: +1. If columns a_j1,...,a_jm do not contain null values in the temporary +table at all, create for them only one index array (and of course do not +create any bitmaps for them). +2. Consider the ratio d(a_i)=N'/ V(a_i), where N' is the number of rows +where a_i is not null and V(a_i) is the number of distinct values for +a_i excluding nulls. + If d(a_i) is close to 1 then do not create any index array: check +whether there is a match running through the records that have been +filtered in. Anyway if d(a_i) is close to 1 then a intersection with +rowid{a_i=v_i} would not reduce the number of remaining rowids +significantly. + If additionally N-N' is small do not create a bitmap for this column +either. +3. If for a column a_i d(a_i) is not close to 1, but N-N' is small a +sorted array of rowids from the set rowid{a_i is null} can be used +instead of a bitmap. +4. We always have a match if R0= INTERSECT rowid{a_i is null} is not +empty. Here i runs through all indexes from [1..n] such that v_i is not +null. For a given subset of columns this fact has to be checked only +once. It can be easily done with bitmap intersection. DESCRIPTION: The goal of this task is to implement efficient execution of NOT IN subquery predicates of the form: <oe_1,...,oe_n> NOT IN <non_correlated subquery> when either some oe_i, or some subqury result column contains NULLs. The problem with such predicates is that it is possible to use index lookups only when neither argument of the predicate contains NULLs. If some argument contains a NULL, then due to NULL semantics, it plays the role of a wildcard. If we were to use regular index lookups, then we would get 'no match' for some outer tuple (thus the predicate evaluates to FALSE), while the SQL semantics means 'partial match', and the predicate should evaluate to NULL. This task implements an efficient algorithm to compute such 'parial matches', where a NULL matches any value. HIGH-LEVEL SPECIFICATION: This a copy of the initial algorithm proposed by Igor: ====================================================== For each left side tuple (v_1,...,v_n) we have to find the following set of rowids for the temp table containing N rows as the result of materialization of the subquery: R= INTERSECT (rowid{a_i=v_i} UNION rowid{a_i is null} where i runs trough all indexes from [1..n] such that v_i is not null. Bear in mind the following specifics of this intersection: (1) For each i: rowid{a_i=v_i} and rowid{a_i is null} are disjoint (2) For each i: rowid{a_i is null} is the same for each tuple Due to (2) it makes sense to build rowid{a_i is null} only once. A good representation for such sets would be bitmaps: - it requires minimum memory: not more than N*n bits in total - search of an element in a set is extremely cheap Taken all above into account I could suggest the following algorithm to build R: Using indexes (read about them below) for each column participating in the intersection, merge ordered sets rowid{a_i=v_i} in the following manner. If a rowid r has been encountered maximum in k sets rowid{a_i1=v_i1},...,rowid(a_ik=v_ik), then it has to be checked against all rowid{a_i=v_i} such that i is not in {i1,...,ik}. As soon as we fail to find r in one of these sets we discard it. If r has been found in all of them then r belongs to the set R. Here we use the property (1): any r from rowid{a_i=v_i} UNION rowid{a_i is null} is either belongs to rowid{a_i=v_i} or to rowid{a_i is null}. From this we can infer that for any r from R indexes a_i can be uniquely divided into two groups: one contains indexes a_i where r belongs to the sets rowid{a_i=v_i}, the other contains indexes a_j such that r belongs to rowid{a_j is null}. Now let's talk how to get elements from rowid{a_i=v_i} in a sorted order needed for the merge procedure. We could use BTREE indexes for temp table. But they are rather expensive and take a lot of memory as the are implemented with RB trees. I would suggest creating for each column from the temporary table just an array of rowids sorted by the value from column a. Index lookup in such an array is cheap. It's also rather cheap to check that the next rowid refers to a row with a different value in column a. The array can be created on demand. Other consideration that may be taken into account: 1. If columns a_j1,...,a_jm do not contain null values in the temporary table at all and v_j1,...,v_jm cannot be null, create for these columns only one index array (and of course do not create any bitmaps for them). 2. Consider the ratio d(a_i)=N'(a_i)/V(a_i), where N'(a_i) is the number of rows, where a_i is not null and V(a_i) is the number of distinct values for a_i excluding nulls. If d(a_i) is close to N'(a_i) then do not create any index array: check whether there is a match running through the records that have been filtered in. Anyway if d(a_i) is close to N'(a_i) then the intersection with rowid{a_i=v_i} will not reduce the number of remaining rowids significantly. In other words is V(a_i) exceeds some threshold there is no sense to create an index for a_i. If additionally N-N'(a_i) is small do not create a bitmap for this column either. 3. If for a column a_i d(a_i) is not close to N'(a_i), but N-N'(a_i) is small a sorted array of rowids from the set rowid{a_i is null} can be used instead of a bitmap. 4. We always have a match if R0= INTERSECT rowid{a_i is null} is not empty. Here i runs through all indexes from [1..n] such that v_i is not null. For a given subset of columns this fact has to be checked only once. It can be easily done with bitmap intersection. 5. If v1,...,vn never can be a null, then indexes (sorted arrays) can be created only for rows with nulls. 6. If v1,...,vn never can be a null and number of rows with nulls is small do not create indexes and do not create bitmaps. 7. If you get a row with nulls in all columns stop filling the temporary table and return UNKNOWN for any tuple <v1,...,vn>. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Timour): Subquery optimization: Efficient NOT IN execution with NULLs (68)
by worklog-noreply＠askmonty.org 04 Dec '09

04 Dec '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Subquery optimization: Efficient NOT IN execution with NULLs CREATION DATE..: Fri, 27 Nov 2009, 13:22 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: CATEGORY.......: Server-RawIdeaBin TASK ID........: 68 (http://askmonty.org/worklog/?tid=68) VERSION........: Benchmarks-3.0 STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Timour - Fri, 04 Dec 2009, 11:26)=-=- High-Level Specification modified. --- /tmp/wklog.68.old.5182 2009-12-04 11:26:25.000000000 +0200 +++ /tmp/wklog.68.new.5182 2009-12-04 11:26:25.000000000 +0200 @@ -50,23 +50,39 @@ The array can be created on demand. Other consideration that may be taken into account: + 1. If columns a_j1,...,a_jm do not contain null values in the temporary -table at all, create for them only one index array (and of course do not -create any bitmaps for them). -2. Consider the ratio d(a_i)=N'/ V(a_i), where N' is the number of rows -where a_i is not null and V(a_i) is the number of distinct values for -a_i excluding nulls. - If d(a_i) is close to 1 then do not create any index array: check +table at all and v_j1,...,v_jm cannot be null, create for these columns +only one index array (and of course do not create any bitmaps for them). + +2. Consider the ratio d(a_i)=N'(a_i)/V(a_i), where N'(a_i) is the number +of rows, where a_i is not null and V(a_i) is the number of distinct +values for a_i excluding nulls. +If d(a_i) is close to N'(a_i) then do not create any index array: check whether there is a match running through the records that have been -filtered in. Anyway if d(a_i) is close to 1 then a intersection with -rowid{a_i=v_i} would not reduce the number of remaining rowids +filtered in. Anyway if d(a_i) is close to N'(a_i) then the intersection + with rowid{a_i=v_i} will not reduce the number of remaining rowids significantly. - If additionally N-N' is small do not create a bitmap for this column -either. -3. If for a column a_i d(a_i) is not close to 1, but N-N' is small a -sorted array of rowids from the set rowid{a_i is null} can be used -instead of a bitmap. +In other words is V(a_i) exceeds some threshold there is no sense to +create an index for a_i. +If additionally N-N'(a_i) is small do not create a bitmap for this +column either. + +3. If for a column a_i d(a_i) is not close to N'(a_i), but N-N'(a_i) is +small a sorted array of rowids from the set rowid{a_i is null} can be +used instead of a bitmap. + 4. We always have a match if R0= INTERSECT rowid{a_i is null} is not empty. Here i runs through all indexes from [1..n] such that v_i is not null. For a given subset of columns this fact has to be checked only once. It can be easily done with bitmap intersection. + +5. If v1,...,vn never can be a null, then indexes (sorted arrays) can be +created only for rows with nulls. + +6. If v1,...,vn never can be a null and number of rows with nulls is +small do not create indexes and do not create bitmaps. + +7. If you get a row with nulls in all columns stop filling the temporary +table and return UNKNOWN for any tuple <v1,...,vn>. + -=-=(Timour - Fri, 27 Nov 2009, 13:23)=-=- High-Level Specification modified. --- /tmp/wklog.68.old.17140 2009-11-27 11:23:17.000000000 +0000 +++ /tmp/wklog.68.new.17140 2009-11-27 11:23:17.000000000 +0000 @@ -1 +1,72 @@ +This a copy of the initial algorithm proposed by Igor: +====================================================== +For each left side tuple (v_1,...,v_n) we have to find the following set +of rowids for the temp table containing N rows as the result of +materialization of the subquery: + + R= INTERSECT (rowid{a_i=v_i} UNION rowid{a_i is null} where i runs +trough all indexes from [1..n] such that v_i is not null. + +Bear in mind the following specifics of this intersection: + (1) For each i: rowid{a_i=v_i} and rowid{a_i is null} are disjoint + (2) For each i: rowid{a_i is null} is the same for each tuple + +Due to (2) it makes sense to build rowid{a_i is null} only once. +A good representation for such sets would be bitmaps: +- it requires minimum memory: not more than N*n bits in total +- search of an element in a set is extremely cheap + +Taken all above into account I could suggest the following algorithm to +build R: + + Using indexes (read about them below) for each column participating in the + intersection, + merge ordered sets rowid{a_i=v_i} in the following manner. + If a rowid r has been encountered maximum in k sets +rowid{a_i1=v_i1},...,rowid(a_ik=v_ik), + then it has to be checked against all rowid{a_i=v_i} such that i is +not in {i1,...,ik}. + As soon as we fail to find r in one of these sets we discard it. + If r has been found in all of them then r belongs to the set R. + +Here we use the property (1): any r from rowid{a_i=v_i} UNION rowid{a_i +is null} is either +belongs to rowid{a_i=v_i} or to rowid{a_i is null}. From this we can +infer that for any r from R +indexes a_i can be uniquely divided into two groups: one contains +indexes a_i where r belongs to +the sets rowid{a_i=v_i}, the other contains indexes a_j such that r +belongs to rowid{a_j is null}. + +Now let's talk how to get elements from rowid{a_i=v_i} in a sorted order +needed for the merge procedure. We could use BTREE indexes for temp +table. But they are rather expensive and +take a lot of memory as the are implemented with RB trees. +I would suggest creating for each column from the temporary table just +an array of rowids sorted by the value from column a. +Index lookup in such an array is cheap. It's also rather cheap to check +that the next rowid refers to a row with a different value in column a. +The array can be created on demand. + +Other consideration that may be taken into account: +1. If columns a_j1,...,a_jm do not contain null values in the temporary +table at all, create for them only one index array (and of course do not +create any bitmaps for them). +2. Consider the ratio d(a_i)=N'/ V(a_i), where N' is the number of rows +where a_i is not null and V(a_i) is the number of distinct values for +a_i excluding nulls. + If d(a_i) is close to 1 then do not create any index array: check +whether there is a match running through the records that have been +filtered in. Anyway if d(a_i) is close to 1 then a intersection with +rowid{a_i=v_i} would not reduce the number of remaining rowids +significantly. + If additionally N-N' is small do not create a bitmap for this column +either. +3. If for a column a_i d(a_i) is not close to 1, but N-N' is small a +sorted array of rowids from the set rowid{a_i is null} can be used +instead of a bitmap. +4. We always have a match if R0= INTERSECT rowid{a_i is null} is not +empty. Here i runs through all indexes from [1..n] such that v_i is not +null. For a given subset of columns this fact has to be checked only +once. It can be easily done with bitmap intersection. DESCRIPTION: The goal of this task is to implement efficient execution of NOT IN subquery predicates of the form: <oe_1,...,oe_n> NOT IN <non_correlated subquery> when either some oe_i, or some subqury result column contains NULLs. The problem with such predicates is that it is possible to use index lookups only when neither argument of the predicate contains NULLs. If some argument contains a NULL, then due to NULL semantics, it plays the role of a wildcard. If we were to use regular index lookups, then we would get 'no match' for some outer tuple (thus the predicate evaluates to FALSE), while the SQL semantics means 'partial match', and the predicate should evaluate to NULL. This task implements an efficient algorithm to compute such 'parial matches', where a NULL matches any value. HIGH-LEVEL SPECIFICATION: This a copy of the initial algorithm proposed by Igor: ====================================================== For each left side tuple (v_1,...,v_n) we have to find the following set of rowids for the temp table containing N rows as the result of materialization of the subquery: R= INTERSECT (rowid{a_i=v_i} UNION rowid{a_i is null} where i runs trough all indexes from [1..n] such that v_i is not null. Bear in mind the following specifics of this intersection: (1) For each i: rowid{a_i=v_i} and rowid{a_i is null} are disjoint (2) For each i: rowid{a_i is null} is the same for each tuple Due to (2) it makes sense to build rowid{a_i is null} only once. A good representation for such sets would be bitmaps: - it requires minimum memory: not more than N*n bits in total - search of an element in a set is extremely cheap Taken all above into account I could suggest the following algorithm to build R: Using indexes (read about them below) for each column participating in the intersection, merge ordered sets rowid{a_i=v_i} in the following manner. If a rowid r has been encountered maximum in k sets rowid{a_i1=v_i1},...,rowid(a_ik=v_ik), then it has to be checked against all rowid{a_i=v_i} such that i is not in {i1,...,ik}. As soon as we fail to find r in one of these sets we discard it. If r has been found in all of them then r belongs to the set R. Here we use the property (1): any r from rowid{a_i=v_i} UNION rowid{a_i is null} is either belongs to rowid{a_i=v_i} or to rowid{a_i is null}. From this we can infer that for any r from R indexes a_i can be uniquely divided into two groups: one contains indexes a_i where r belongs to the sets rowid{a_i=v_i}, the other contains indexes a_j such that r belongs to rowid{a_j is null}. Now let's talk how to get elements from rowid{a_i=v_i} in a sorted order needed for the merge procedure. We could use BTREE indexes for temp table. But they are rather expensive and take a lot of memory as the are implemented with RB trees. I would suggest creating for each column from the temporary table just an array of rowids sorted by the value from column a. Index lookup in such an array is cheap. It's also rather cheap to check that the next rowid refers to a row with a different value in column a. The array can be created on demand. Other consideration that may be taken into account: 1. If columns a_j1,...,a_jm do not contain null values in the temporary table at all and v_j1,...,v_jm cannot be null, create for these columns only one index array (and of course do not create any bitmaps for them). 2. Consider the ratio d(a_i)=N'(a_i)/V(a_i), where N'(a_i) is the number of rows, where a_i is not null and V(a_i) is the number of distinct values for a_i excluding nulls. If d(a_i) is close to N'(a_i) then do not create any index array: check whether there is a match running through the records that have been filtered in. Anyway if d(a_i) is close to N'(a_i) then the intersection with rowid{a_i=v_i} will not reduce the number of remaining rowids significantly. In other words is V(a_i) exceeds some threshold there is no sense to create an index for a_i. If additionally N-N'(a_i) is small do not create a bitmap for this column either. 3. If for a column a_i d(a_i) is not close to N'(a_i), but N-N'(a_i) is small a sorted array of rowids from the set rowid{a_i is null} can be used instead of a bitmap. 4. We always have a match if R0= INTERSECT rowid{a_i is null} is not empty. Here i runs through all indexes from [1..n] such that v_i is not null. For a given subset of columns this fact has to be checked only once. It can be easily done with bitmap intersection. 5. If v1,...,vn never can be a null, then indexes (sorted arrays) can be created only for rows with nulls. 6. If v1,...,vn never can be a null and number of rows with nulls is small do not create indexes and do not create bitmaps. 7. If you get a row with nulls in all columns stop filling the temporary table and return UNKNOWN for any tuple <v1,...,vn>. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Rev 2775: Fix for maria BUG#491097. in file:///home/bell/maria/bzr/work-maria-5.1-test/
by sanja＠askmonty.org 03 Dec '09

03 Dec '09

At file:///home/bell/maria/bzr/work-maria-5.1-test/ ------------------------------------------------------------ revno: 2775 revision-id: sanja(a)askmonty.org-20091203210748-x0t8m9pim0x20s7g parent: monty(a)askmonty.org-20091201160902-a3iq6lw82xbto9yp committer: sanja(a)askmonty.org branch nick: work-maria-5.1-test timestamp: Thu 2009-12-03 23:07:48 +0200 message: Fix for maria BUG#491097. Unlock tables before sending OK to client. === modified file 'sql/mysql_priv.h' --- a/sql/mysql_priv.h 2009-11-16 20:49:51 +0000 +++ b/sql/mysql_priv.h 2009-12-03 21:07:48 +0000 @@ -796,6 +796,7 @@ void cleanup_items(Item *item); class THD; void close_thread_tables(THD *thd); +void unlock_thread_tables(THD *thd); #ifndef NO_EMBEDDED_ACCESS_CHECKS bool check_one_table_access(THD *thd, ulong privilege, TABLE_LIST *tables); === modified file 'sql/sql_base.cc' --- a/sql/sql_base.cc 2009-11-30 21:37:27 +0000 +++ b/sql/sql_base.cc 2009-12-03 21:07:48 +0000 @@ -1213,6 +1213,33 @@ } +/** + Unlocks tables in thd->lock() + + + @param thd Thread handler +*/ + +void unlock_thread_tables(THD *thd) +{ + if (thd->lock) + { + /* + For RBR we flush the pending event just before we unlock all the + tables. This means that we are at the end of a topmost + statement, so we ensure that the STMT_END_F flag is set on the + pending event. For statements that are *inside* stored + functions, the pending event will not be flushed: that will be + handled either before writing a query log event (inside + binlog_query()) or when preparing a pending event. + */ + thd->binlog_flush_pending_rows_event(TRUE); + mysql_unlock_tables(thd, thd->lock); + thd->lock=0; + } +} + + /* Close all tables used by the current substatement, or all tables used by this thread if we are on the upper level. @@ -1325,21 +1352,8 @@ /* Fallthrough */ } - if (thd->lock) - { - /* - For RBR we flush the pending event just before we unlock all the - tables. This means that we are at the end of a topmost - statement, so we ensure that the STMT_END_F flag is set on the - pending event. For statements that are *inside* stored - functions, the pending event will not be flushed: that will be - handled either before writing a query log event (inside - binlog_query()) or when preparing a pending event. - */ - thd->binlog_flush_pending_rows_event(TRUE); - mysql_unlock_tables(thd, thd->lock); - thd->lock=0; - } + unlock_thread_tables(thd); + /* Note that we need to hold LOCK_open while changing the open_tables list. Another thread may work on it. === modified file 'sql/sql_parse.cc' --- a/sql/sql_parse.cc 2009-11-30 21:37:27 +0000 +++ b/sql/sql_parse.cc 2009-12-03 21:07:48 +0000 @@ -1252,6 +1252,12 @@ ha_maria::implicit_commit(thd, FALSE); #endif + /* + We should unlock tables before sending OK to client, because + unlocking makes results of insert (for example) visible for other + threads (for some engines like MyISAM). + */ + unlock_thread_tables(thd); net_end_statement(thd); query_cache_end_of_result(thd); /* @@ -1645,6 +1651,12 @@ ha_maria::implicit_commit(thd, FALSE); #endif + /* + We should unlock tables before sending OK to client, because unlocking + makes results of insert (for example) visible for other threads (for + some engines like MyISAM). + */ + unlock_thread_tables(thd); net_end_statement(thd); query_cache_end_of_result(thd);

1 0

[Maria-developers] [Branch ~maria-captains/maria/5.1] Rev 2776: Applied patch from to fix some problems with Croatian character set and LIKE queries
by noreply＠launchpad.net 03 Dec '09

03 Dec '09

------------------------------------------------------------ revno: 2776 committer: Michael Widenius <monty(a)askmonty.org> branch nick: maria-5.1 timestamp: Thu 2009-12-03 14:02:37 +0200 message: Applied patch from to fix some problems with Croatian character set and LIKE queries Author: Alexander Barkov License: GPL modified: mysql-test/r/ctype_ucs.result mysql-test/t/ctype_ucs.test strings/ctype-ucs2.c -- lp:maria https://code.launchpad.net/~maria-captains/maria/5.1 Your team Maria developers is subscribed to branch lp:maria. To unsubscribe from this branch go to https://code.launchpad.net/~maria-captains/maria/5.1/+edit-subscription.

1 0

[Maria-developers] [Branch ~maria-captains/maria/5.1] Rev 2775: Merge
by noreply＠launchpad.net 03 Dec '09

03 Dec '09

Merge authors: Antony T Curtis (atcurtis) hery.ramilison(a)sun.com Igor Babaev (igorb-seattle) Michael Widenius (monty) Michael Widenius (monty) Related merge proposals: https://code.launchpad.net/~atcurtis/maria/maria-5.1-federatedx/+merge/14409 proposed by: Antony T Curtis (atcurtis) review: Needs Fixing - Michael Widenius (monty) ------------------------------------------------------------ revno: 2775 [merge] committer: Michael Widenius <monty(a)askmonty.org> branch nick: maria-5.1 timestamp: Thu 2009-12-03 13:34:11 +0200 message: Merge modified: client/mysql.cc client/mysqlcheck.c client/mysqlslap.c client/mysqltest.cc mysql-test/lib/mtr_report.pm mysql-test/mysql-test-run.pl mysql-test/suite/federated/disabled.def mysql-test/suite/federated/federated_server.result mysql-test/suite/federated/federated_server.test storage/federatedx/ha_federatedx.cc unittest/mysys/Makefile.am -- lp:maria https://code.launchpad.net/~maria-captains/maria/5.1 Your team Maria developers is subscribed to branch lp:maria. To unsubscribe from this branch go to https://code.launchpad.net/~maria-captains/maria/5.1/+edit-subscription.

1 0

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (monty:2776)
by Michael Widenius 03 Dec '09

03 Dec '09

#At lp:maria based on revid:monty@askmonty.org-20091203113411-cmr8g2lcp45n0prv 2776 Michael Widenius 2009-12-03 Applied patch from to fix some problems with Croatian character set and LIKE queries Author: Alexander Barkov License: GPL modified: mysql-test/r/ctype_ucs.result mysql-test/t/ctype_ucs.test strings/ctype-ucs2.c per-file messages: mysql-test/t/ctype_ucs.test Added test case for Croatina character set === modified file 'mysql-test/r/ctype_ucs.result' --- a/mysql-test/r/ctype_ucs.result 2008-12-23 14:21:01 +0000 +++ b/mysql-test/r/ctype_ucs.result 2009-12-03 12:02:37 +0000 @@ -1211,3 +1211,47 @@ HEX(DAYNAME(19700101)) 0427043504420432043504400433 SET character_set_connection=latin1; End of 5.0 tests +Start of 5.1 tests +SET NAMES utf8; +CREATE TABLE t1 ( +a varchar(10) CHARACTER SET ucs2 COLLATE ucs2_czech_ci, +key(a) +); +INSERT INTO t1 VALUES +('aa'),('bb'),('cc'),('dd'),('ee'),('ff'),('gg'),('hh'),('ii'), +('jj'),('kk'),('ll'),('mm'),('nn'),('oo'),('pp'),('rr'),('ss'), +('tt'),('uu'),('vv'),('ww'),('xx'),('yy'),('zz'); +INSERT INTO t1 VALUES ('ca'),('cz'),('ch'); +INSERT INTO t1 VALUES ('da'),('dz'), (X'0064017E'); +EXPLAIN SELECT * FROM t1 WHERE a LIKE 'b%'; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 range a a 23 NULL 1 Using where; Using index +EXPLAIN SELECT * FROM t1 WHERE a LIKE 'c%'; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 range a a 23 NULL 30 Using where; Using index +SELECT * FROM t1 WHERE a LIKE 'c%'; +a +ca +cc +cz +ch +EXPLAIN SELECT * FROM t1 WHERE a LIKE 'ch%'; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 range a a 23 NULL 1 Using where; Using index +SELECT * FROM t1 WHERE a LIKE 'ch%'; +a +ch +ALTER TABLE t1 MODIFY a VARCHAR(10) CHARACTER SET ucs2 COLLATE ucs2_croatian_ci; +EXPLAIN SELECT * FROM t1 WHERE a LIKE 'd%'; +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 range a a 23 NULL 1 Using where; Using index +SELECT hex(concat('d',_ucs2 0x017E,'%')); +hex(concat('d',_ucs2 0x017E,'%')) +0064017E0025 +EXPLAIN SELECT * FROM t1 WHERE a LIKE concat('d',_ucs2 0x017E,'%'); +id select_type table type possible_keys key key_len ref rows Extra +1 SIMPLE t1 range a a 23 NULL 1 Using where; Using index +SELECT hex(a) FROM t1 WHERE a LIKE concat('D',_ucs2 0x017E,'%'); +hex(a) +0064017E +DROP TABLE t1; === modified file 'mysql-test/t/ctype_ucs.test' --- a/mysql-test/t/ctype_ucs.test 2008-12-23 14:21:01 +0000 +++ b/mysql-test/t/ctype_ucs.test 2009-12-03 12:02:37 +0000 @@ -723,3 +723,34 @@ SELECT HEX(DAYNAME(19700101)); SET character_set_connection=latin1; --echo End of 5.0 tests + + +--echo Start of 5.1 tests +# +# Checking my_like_range_ucs2 +# +SET NAMES utf8; +CREATE TABLE t1 ( + a varchar(10) CHARACTER SET ucs2 COLLATE ucs2_czech_ci, + key(a) +); +INSERT INTO t1 VALUES +('aa'),('bb'),('cc'),('dd'),('ee'),('ff'),('gg'),('hh'),('ii'), +('jj'),('kk'),('ll'),('mm'),('nn'),('oo'),('pp'),('rr'),('ss'), +('tt'),('uu'),('vv'),('ww'),('xx'),('yy'),('zz'); +INSERT INTO t1 VALUES ('ca'),('cz'),('ch'); +INSERT INTO t1 VALUES ('da'),('dz'), (X'0064017E'); +# This one should scan only one row +EXPLAIN SELECT * FROM t1 WHERE a LIKE 'b%'; +# This one should scan many rows: 'c' is a contraction head +EXPLAIN SELECT * FROM t1 WHERE a LIKE 'c%'; +SELECT * FROM t1 WHERE a LIKE 'c%'; +EXPLAIN SELECT * FROM t1 WHERE a LIKE 'ch%'; +SELECT * FROM t1 WHERE a LIKE 'ch%'; +ALTER TABLE t1 MODIFY a VARCHAR(10) CHARACTER SET ucs2 COLLATE ucs2_croatian_ci; +EXPLAIN SELECT * FROM t1 WHERE a LIKE 'd%'; +SELECT hex(concat('d',_ucs2 0x017E,'%')); +EXPLAIN SELECT * FROM t1 WHERE a LIKE concat('d',_ucs2 0x017E,'%'); +SELECT hex(a) FROM t1 WHERE a LIKE concat('D',_ucs2 0x017E,'%'); + +DROP TABLE t1; === modified file 'strings/ctype-ucs2.c' --- a/strings/ctype-ucs2.c 2009-11-30 12:42:24 +0000 +++ b/strings/ctype-ucs2.c 2009-12-03 12:02:37 +0000 @@ -1498,6 +1498,14 @@ void my_hash_sort_ucs2_bin(CHARSET_INFO } } + +static inline my_wc_t +ucs2_to_wc(const uchar *ptr) +{ + return (((uint) ptr[0]) << 8) + ptr[1]; +} + + /* ** Calculate min_str and max_str that ranges a LIKE string. ** Arguments: @@ -1531,6 +1539,7 @@ my_bool my_like_range_ucs2(CHARSET_INFO for ( ; ptr + 1 < end && min_str + 1 < min_end && charlen > 0 ; ptr+=2, charlen--) { + my_wc_t wc; if (ptr[0] == '\0' && ptr[1] == escape && ptr + 1 < end) { ptr+=2; /* Skip escape */ @@ -1567,9 +1576,9 @@ fill_max_and_min: } if (have_contractions && ptr + 3 < end && - ptr[0] == '\0' && - my_uca_can_be_contraction_head(cs, (uchar) ptr[1])) + my_uca_can_be_contraction_head(cs, (wc= ucs2_to_wc((uchar*) ptr)))) { + my_wc_t wc2; /* Contraction head found */ if (ptr[2] == '\0' && (ptr[3] == w_one || ptr[3] == w_many)) { @@ -1581,9 +1590,8 @@ fill_max_and_min: Check if the second letter can be contraction part, and if two letters really produce a contraction. */ - if (ptr[2] == '\0' && - my_uca_can_be_contraction_tail(cs, (uchar) ptr[3]) && - my_uca_contraction2_weight(cs,(uchar) ptr[1], (uchar) ptr[3])) + if (my_uca_can_be_contraction_tail(cs, (wc2= ucs2_to_wc((uchar*) ptr + 2))) && + my_uca_contraction2_weight(cs, wc , wc2)) { /* Contraction found */ if (charlen == 1 || min_str + 2 >= min_end)

1 0

[Maria-developers] 5.1 release todo page
by Kristian Nielsen 03 Dec '09

03 Dec '09

Seems we had 3 pages in the wiki for 5.1 release todo with mostly overlapping content. I merged these two: http://askmonty.org/wiki/index.php?title=MariaDB_5.1_TODO http://askmonty.org/wiki/index.php/MariaDB_Fix_Failures into this one: http://askmonty.org/wiki/index.php/MariaDB51_release_plan and added redirects to the former two instead. I also updated some of the information that was outdated. - Kristian.

1 0

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (monty:2775)
by Michael Widenius 03 Dec '09

03 Dec '09

#At lp:maria based on revid:monty@askmonty.org-20091201160902-a3iq6lw82xbto9yp 2775 Michael Widenius 2009-12-03 [merge] Merge modified: client/mysql.cc client/mysqlcheck.c client/mysqlslap.c client/mysqltest.cc mysql-test/lib/mtr_report.pm mysql-test/mysql-test-run.pl mysql-test/suite/federated/disabled.def mysql-test/suite/federated/federated_server.result mysql-test/suite/federated/federated_server.test storage/federatedx/ha_federatedx.cc unittest/mysys/Makefile.am === modified file 'client/mysql.cc' --- a/client/mysql.cc 2009-11-30 21:37:27 +0000 +++ b/client/mysql.cc 2009-12-03 11:34:11 +0000 @@ -1280,7 +1280,6 @@ sig_handler handle_sigint(int sig) MYSQL *kill_mysql= NULL; /* terminate if no query being executed, or we already tried interrupting */ - /* terminate if no query being executed, or we already tried interrupting */ if (!executing_query || (interrupted_query == 2)) { tee_fprintf(stdout, "Ctrl-C -- exit!\n"); @@ -1295,6 +1294,7 @@ sig_handler handle_sigint(int sig) goto err; } + /* First time try to kill the query, second time the connection */ interrupted_query++; /* mysqld < 5 does not understand KILL QUERY, skip to KILL CONNECTION */ @@ -1305,10 +1305,13 @@ sig_handler handle_sigint(int sig) sprintf(kill_buffer, "KILL %s%lu", (interrupted_query == 1) ? "QUERY " : "", mysql_thread_id(&mysql)); - tee_fprintf(stdout, "Ctrl-C -- sending \"%s\" to server ...\n", kill_buffer); + if (verbose) + tee_fprintf(stdout, "Ctrl-C -- sending \"%s\" to server ...\n", + kill_buffer); mysql_real_query(kill_mysql, kill_buffer, (uint) strlen(kill_buffer)); mysql_close(kill_mysql); - tee_fprintf(stdout, "Ctrl-C -- query aborted.\n"); + tee_fprintf(stdout, "Ctrl-C -- query killed. Continuing normally.\n"); + interrupted_query= 0; return; @@ -1321,7 +1324,6 @@ err: handler called mysql_end(). */ mysql_thread_end(); - return; #else mysql_end(sig); #endif @@ -2881,13 +2883,8 @@ com_help(String *buffer __attribute__((u return com_server_help(buffer,line,help_arg); } - put_info("\nFor information about MySQL products and services, visit:\n" - " http://www.mysql.com/\n" - "For developer information, including the MySQL Reference Manual, " - "visit:\n" - " http://dev.mysql.com/\n" - "To buy MySQL Enterprise support, training, or other products, visit:\n" - " https://shop.mysql.com/\n", INFO_INFO); + put_info("\nGeneral information about MariaDB can be found at\n" + "http://askmonty.org/wiki/index.php/Manual:Contents\n", INFO_INFO); put_info("List of all MySQL commands:", INFO_INFO); if (!named_cmds) put_info("Note that all text commands must be first on line and end with ';'",INFO_INFO); === modified file 'client/mysqlcheck.c' --- a/client/mysqlcheck.c 2009-09-28 06:24:19 +0000 +++ b/client/mysqlcheck.c 2009-12-03 11:19:05 +0000 @@ -857,7 +857,8 @@ int main(int argc, char **argv) if (!opt_write_binlog) { - if (disable_binlog()) { + if (disable_binlog()) + { first_error= 1; goto end; } === modified file 'client/mysqlslap.c' --- a/client/mysqlslap.c 2009-11-30 21:37:27 +0000 +++ b/client/mysqlslap.c 2009-12-03 11:34:11 +0000 @@ -472,11 +472,10 @@ void concurrency_loop(MYSQL *mysql, uint if (commit_rate) run_query(mysql, "SET AUTOCOMMIT=0", strlen("SET AUTOCOMMIT=0")); - if (pre_system) - if ((sysret= system(pre_system)) != 0) - fprintf(stderr, - "Warning: Execution of pre_system option returned %d.\n", - sysret); + if (pre_system && (sysret= system(pre_system)) != 0) + fprintf(stderr, + "Warning: Execution of pre_system option returned %d.\n", + sysret); /* Pre statements are always run after all other logic so they can @@ -490,11 +489,10 @@ void concurrency_loop(MYSQL *mysql, uint if (post_statements) run_statements(mysql, post_statements); - if (post_system) - if ((sysret= system(post_system)) != 0) - fprintf(stderr, - "Warning: Execution of post_system option returned %d.\n", - sysret); + if (post_system && (sysret= system(post_system)) != 0) + fprintf(stderr, + "Warning: Execution of post_system option returned %d.\n", + sysret); /* We are finished with this run */ if (auto_generate_sql_autoincrement || auto_generate_sql_guid_primary) drop_primary_key_list(); === modified file 'client/mysqltest.cc' --- a/client/mysqltest.cc 2009-11-16 20:49:51 +0000 +++ b/client/mysqltest.cc 2009-12-03 11:19:05 +0000 @@ -3497,9 +3497,10 @@ void do_diff_files(struct st_command *co if ((error= compare_files(ds_filename.str, ds_filename2.str)) && match_expected_error(command, error, NULL) < 0) { - /* Compare of the two files failed, append them to output - so the failure can be analyzed, but only if it was not - expected to fail. + /* + Compare of the two files failed, append them to output + so the failure can be analyzed, but only if it was not + expected to fail. */ show_diff(&ds_res, ds_filename.str, ds_filename2.str); log_file.write(&ds_res); @@ -5013,7 +5014,8 @@ void do_connect(struct st_command *comma con_options= ds_options.str; while (*con_options) { - char* end; + size_t length; + char *end; /* Step past any spaces in beginning of option*/ while (*con_options && my_isspace(charset_info, *con_options)) con_options++; @@ -5021,13 +5023,14 @@ void do_connect(struct st_command *comma end= con_options; while (*end && !my_isspace(charset_info, *end)) end++; - if (!strncmp(con_options, "SSL", 3)) + length= (size_t) (end - con_options); + if (length == 3 && !strncmp(con_options, "SSL", 3)) con_ssl= 1; - else if (!strncmp(con_options, "COMPRESS", 8)) + else if (length == 8 && !strncmp(con_options, "COMPRESS", 8)) con_compress= 1; - else if (!strncmp(con_options, "PIPE", 4)) + else if (length == 4 && !strncmp(con_options, "PIPE", 4)) con_pipe= 1; - else if (!strncmp(con_options, "SHM", 3)) + else if (length == 3 && !strncmp(con_options, "SHM", 3)) con_shm= 1; else die("Illegal option to connect: %.*s", @@ -5096,14 +5099,13 @@ void do_connect(struct st_command *comma mysql_options(&con_slot->mysql, MYSQL_SHARED_MEMORY_BASE_NAME, ds_shm.str); mysql_options(&con_slot->mysql, MYSQL_OPT_PROTOCOL, &protocol); } - else if(shared_memory_base_name) + else if (shared_memory_base_name) { mysql_options(&con_slot->mysql, MYSQL_SHARED_MEMORY_BASE_NAME, - shared_memory_base_name); + shared_memory_base_name); } #endif - /* Use default db name */ if (ds_database.length == 0) dynstr_set(&ds_database, opt_db); @@ -6879,10 +6881,8 @@ void run_query_stmt(MYSQL *mysql, struct MYSQL_STMT *stmt; DYNAMIC_STRING ds_prepare_warnings; DYNAMIC_STRING ds_execute_warnings; - ulonglong affected_rows; DBUG_ENTER("run_query_stmt"); DBUG_PRINT("query", ("'%-.60s'", query)); - LINT_INIT(affected_rows); /* Init a new stmt if it's not already one created for this connection @@ -6981,6 +6981,9 @@ void run_query_stmt(MYSQL *mysql, struct handle_no_error(command); if (!disable_result_log) { + ulonglong affected_rows; + LINT_INIT(affected_rows); + /* Not all statements creates a result set. If there is one we can now create another normal result set that contains the meta @@ -7026,39 +7029,33 @@ void run_query_stmt(MYSQL *mysql, struct Need to grab affected rows information before getting warnings here */ - { - ulonglong affected_rows; - LINT_INIT(affected_rows); + if (!disable_info) + affected_rows= mysql_affected_rows(mysql); - if (!disable_info) - affected_rows= mysql_affected_rows(mysql); + if (!disable_warnings) + { + /* Get the warnings from execute */ - if (!disable_warnings) + /* Append warnings to ds - if there are any */ + if (append_warnings(&ds_execute_warnings, mysql) || + ds_execute_warnings.length || + ds_prepare_warnings.length || + ds_warnings->length) { - /* Get the warnings from execute */ - - /* Append warnings to ds - if there are any */ - if (append_warnings(&ds_execute_warnings, mysql) || - ds_execute_warnings.length || - ds_prepare_warnings.length || - ds_warnings->length) - { - dynstr_append_mem(ds, "Warnings:\n", 10); - if (ds_warnings->length) - dynstr_append_mem(ds, ds_warnings->str, - ds_warnings->length); - if (ds_prepare_warnings.length) - dynstr_append_mem(ds, ds_prepare_warnings.str, - ds_prepare_warnings.length); - if (ds_execute_warnings.length) - dynstr_append_mem(ds, ds_execute_warnings.str, - ds_execute_warnings.length); - } + dynstr_append_mem(ds, "Warnings:\n", 10); + if (ds_warnings->length) + dynstr_append_mem(ds, ds_warnings->str, + ds_warnings->length); + if (ds_prepare_warnings.length) + dynstr_append_mem(ds, ds_prepare_warnings.str, + ds_prepare_warnings.length); + if (ds_execute_warnings.length) + dynstr_append_mem(ds, ds_execute_warnings.str, + ds_execute_warnings.length); } - - if (!disable_info) - append_info(ds, affected_rows, mysql_info(mysql)); } + if (!disable_info) + append_info(ds, affected_rows, mysql_info(mysql)); } end: @@ -7235,7 +7232,6 @@ void run_query(struct st_connection *cn, } dynstr_free(&query_str); - } if (sp_protocol_enabled && @@ -7662,6 +7658,7 @@ int main(int argc, char **argv) my_bool q_send_flag= 0, abort_flag= 0; uint command_executed= 0, last_command_executed= 0; char save_file[FN_REFLEN]; + bool empty_result= FALSE; MY_INIT(argv[0]); save_file[0]= 0; @@ -7819,6 +7816,7 @@ int main(int argc, char **argv) verbose_msg("Start processing test commands from '%s' ...", cur_file->file_name); while (!read_command(&command) && !abort_flag) { + my_bool ok_to_do; int current_line_inc = 1, processed = 0; if (command->type == Q_UNKNOWN || command->type == Q_COMMENT_WITH_COMMAND) get_command_type(command); @@ -7831,7 +7829,7 @@ int main(int argc, char **argv) command->type= Q_COMMENT; } - my_bool ok_to_do= cur_block->ok; + ok_to_do= cur_block->ok; /* Some commands need to be "done" the first time if they may get re-iterated over in a true context. This can only happen if there's @@ -8167,8 +8165,6 @@ int main(int argc, char **argv) if (parsing_disabled) die("Test ended with parsing disabled"); - my_bool empty_result= FALSE; - /* The whole test has been executed _sucessfully_. Time to compare result or save it to record file. === modified file 'mysql-test/lib/mtr_report.pm' --- a/mysql-test/lib/mtr_report.pm 2009-11-16 20:49:51 +0000 +++ b/mysql-test/lib/mtr_report.pm 2009-12-03 11:19:05 +0000 @@ -388,7 +388,7 @@ MSG } elsif (@$extra_warnings) { - mtr_error("There were errors/warnings in server logs after running test cases."); + mtr_error("There where errors/warnings in server logs after running test cases."); } elsif ($fail) { === modified file 'mysql-test/mysql-test-run.pl' --- a/mysql-test/mysql-test-run.pl 2009-11-30 21:37:27 +0000 +++ b/mysql-test/mysql-test-run.pl 2009-12-03 11:34:11 +0000 @@ -201,10 +201,10 @@ my $opt_mark_progress; my $opt_sleep; -my $opt_testcase_timeout= 15; # minutes -my $opt_suite_timeout = 300; # minutes -my $opt_shutdown_timeout= 10; # seconds -my $opt_start_timeout = 180; # seconds +my $opt_testcase_timeout= 15; # 15 minutes +my $opt_suite_timeout = 360; # 6 hours +my $opt_shutdown_timeout= 10; # 10 seconds +my $opt_start_timeout = 180; # 180 seconds sub testcase_timeout { return $opt_testcase_timeout * 60; }; sub suite_timeout { return $opt_suite_timeout * 60; }; @@ -1319,6 +1319,8 @@ sub command_line_setup { { # Indicate that we are using debugger $glob_debugger= 1; + $opt_testcase_timeout= 60*60*24; # Don't abort debugging with timeout + $opt_suite_timeout= $opt_testcase_timeout; $opt_retry= 1; $opt_retry_failure= 1; @@ -2151,7 +2153,6 @@ sub environment_setup { # Create an environment variable to make it possible # to detect that valgrind is being used from test cases $ENV{'VALGRIND_TEST'}= $opt_valgrind; - } @@ -2908,8 +2909,8 @@ sub mysql_install_db { my $bootstrap_sql_file= "$opt_vardir/tmp/bootstrap.sql"; my $path_sql= my_find_file($install_basedir, - ["mysql", "sql/share", "share/mysql", - "share/mariadb", "share", "scripts"], + ["mysql", "sql/share", "share/mariadb", + "share/mysql", "share", "scripts"], "mysql_system_tables.sql", NOT_REQUIRED); @@ -3861,7 +3862,7 @@ sub extract_server_log ($$) { my ($error_log, $tname) = @_; # Open the servers .err log file and read all lines - # belonging to current tets into @lines + # belonging to current test into @lines my $Ferr = IO::File->new($error_log) or mtr_error("Could not open file '$error_log' for reading: $!"); === modified file 'mysql-test/suite/federated/disabled.def' --- a/mysql-test/suite/federated/disabled.def 2009-10-30 18:50:56 +0000 +++ b/mysql-test/suite/federated/disabled.def 2009-11-14 19:33:59 +0000 @@ -9,5 +9,4 @@ # Do not use any TAB characters for whitespace. # ############################################################################## -federated_server : needs fixup === modified file 'mysql-test/suite/federated/federated_server.result' --- a/mysql-test/suite/federated/federated_server.result 2009-10-30 18:50:56 +0000 +++ b/mysql-test/suite/federated/federated_server.result 2009-11-14 19:33:59 +0000 @@ -175,6 +175,8 @@ CREATE TABLE db_bogus.t1 ( ) ; INSERT INTO db_bogus.t1 VALUES ('2','this is bogus'); +create user test_fed@localhost identified by 'foo'; +grant all on db_legitimate.* to test_fed@localhost; create server 's1' foreign data wrapper 'mysql' options (HOST '127.0.0.1', DATABASE 'db_legitimate', @@ -211,15 +213,14 @@ id name alter server s1 options (database 'db_bogus'); flush tables; select * from federated.t1; -id name -2 this is bogus +ERROR HY000: There was a problem processing the query on the foreign data source. Data source error: : 1044 : Access denied for user 'test_fed'@'localhost' to databa drop server if exists 's1'; ERROR 42000: Access denied; you need the SUPER privilege for this operation create server 's1' foreign data wrapper 'mysql' options (HOST '127.0.0.1', DATABASE 'db_legitimate', -USER 'root', -PASSWORD '', +USER 'test_fed', +PASSWORD 'foo', PORT SLAVE_PORT, SOCKET '', OWNER 'root'); @@ -228,8 +229,8 @@ drop server 's1'; create server 's1' foreign data wrapper 'mysql' options (HOST '127.0.0.1', DATABASE 'db_legitimate', -USER 'root', -PASSWORD '', +USER 'test_fed', +PASSWORD 'foo', PORT SLAVE_PORT, SOCKET '', OWNER 'root'); @@ -237,6 +238,7 @@ flush tables; select * from federated.t1; id name 1 this is legitimate +drop user test_fed@localhost; drop database db_legitimate; drop database db_bogus; drop user guest_super@localhost; @@ -275,6 +277,6 @@ call p1(); drop procedure p1; drop server if exists s; DROP TABLE IF EXISTS federated.t1; -DROP DATABASE federated; +DROP DATABASE IF EXISTS federated; DROP TABLE IF EXISTS federated.t1; -DROP DATABASE federated; +DROP DATABASE IF EXISTS federated; === modified file 'mysql-test/suite/federated/federated_server.test' --- a/mysql-test/suite/federated/federated_server.test 2009-10-30 18:50:56 +0000 +++ b/mysql-test/suite/federated/federated_server.test 2009-11-14 19:33:59 +0000 @@ -239,6 +239,7 @@ alter server s1 options (database 'db_bo connection master; flush tables; +--error ER_QUERY_ON_FOREIGN_DATA_SOURCE select * from federated.t1; connection conn_select; @@ -249,8 +250,8 @@ drop server if exists 's1'; eval create server 's1' foreign data wrapper 'mysql' options (HOST '127.0.0.1', DATABASE 'db_legitimate', - USER 'root', - PASSWORD '', + USER 'test_fed', + PASSWORD 'foo', PORT $SLAVE_MYPORT, SOCKET '', OWNER 'root'); @@ -261,8 +262,8 @@ drop server 's1'; eval create server 's1' foreign data wrapper 'mysql' options (HOST '127.0.0.1', DATABASE 'db_legitimate', - USER 'root', - PASSWORD '', + USER 'test_fed', + PASSWORD 'foo', PORT $SLAVE_MYPORT, SOCKET '', OWNER 'root'); @@ -273,6 +274,7 @@ select * from federated.t1; # clean up test connection slave; +drop user test_fed@localhost; drop database db_legitimate; drop database db_bogus; === modified file 'storage/federatedx/ha_federatedx.cc' --- a/storage/federatedx/ha_federatedx.cc 2009-11-30 21:37:27 +0000 +++ b/storage/federatedx/ha_federatedx.cc 2009-12-03 11:34:11 +0000 @@ -1,5 +1,5 @@ /* -Copyright (c) 2008, Patrick Galbraith +Copyright (c) 2008-2009, Patrick Galbraith & Antony Curtis All rights reserved. Redistribution and use in source and binary forms, with or without @@ -308,7 +308,7 @@ OF THIS SOFTWARE, EVEN IF ADVISED OF THE */ -#define MYSQL_SERVER 1q +#define MYSQL_SERVER 1 #include "mysql_priv.h" #include <mysql/plugin.h> @@ -1627,7 +1627,13 @@ static int free_server(federatedx_txn *t { MEM_ROOT mem_root; - txn->close(server); + if (!txn) + { + federatedx_txn tmp_txn; + tmp_txn.close(server); + } + else + txn->close(server); DBUG_ASSERT(server->io_count == 0); @@ -1785,21 +1791,25 @@ int ha_federatedx::close(void) if (stored_result) retval= free_result(); - /* Disconnect from mysql. thd may be null during refresh */ - txn= thd ? get_txn(thd, true) : new federatedx_txn(); + /* Disconnect from mysql */ + if (!thd || !(txn= get_txn(thd, true))) + { + federatedx_txn tmp_txn; + + tmp_txn.release(&io); - if (txn) + DBUG_ASSERT(io == NULL); + + if ((error= free_share(&tmp_txn, share))) + retval= error; + } + else { txn->release(&io); - DBUG_ASSERT(io == NULL); if ((error= free_share(txn, share))) retval= error; - - if (!thd) - delete txn; - } DBUG_RETURN(retval); } @@ -2793,14 +2803,16 @@ int ha_federatedx::rnd_end() int ha_federatedx::free_result() { int error; + federatedx_io *tmp_io= 0, **iop; DBUG_ASSERT(stored_result); - if ((error= txn->acquire(share, FALSE, &io))) + if (!*(iop= &io) && (error= txn->acquire(share, TRUE, (iop= &tmp_io)))) { DBUG_ASSERT(0); // Fail when testing return error; } - io->free_result(stored_result); + (*iop)->free_result(stored_result); stored_result= 0; + txn->release(&tmp_io); return 0; } @@ -2985,7 +2997,7 @@ int ha_federatedx::info(uint flag) { char error_buffer[FEDERATEDX_QUERY_BUFFER_SIZE]; uint error_code; - federatedx_io *tmp_io= 0; + federatedx_io *tmp_io= 0, **iop= 0; DBUG_ENTER("ha_federatedx::info"); error_code= ER_QUERY_ON_FOREIGN_DATA_SOURCE; @@ -2993,7 +3005,7 @@ int ha_federatedx::info(uint flag) /* we want not to show table status if not needed to do so */ if (flag & (HA_STATUS_VARIABLE | HA_STATUS_CONST | HA_STATUS_AUTO)) { - if ((error_code= txn->acquire(share, TRUE, &tmp_io))) + if (!*(iop= &io) && (error_code= txn->acquire(share, TRUE, (iop= &tmp_io)))) goto fail; } @@ -3006,13 +3018,13 @@ int ha_federatedx::info(uint flag) if (flag & HA_STATUS_CONST) stats.block_size= 4096; - if (tmp_io->table_metadata(&stats, share->table_name, + if ((*iop)->table_metadata(&stats, share->table_name, share->table_name_length, flag)) goto error; } if (flag & HA_STATUS_AUTO) - stats.auto_increment_value= tmp_io->last_insert_id(); + stats.auto_increment_value= (*iop)->last_insert_id(); /* If ::info created it's own transaction, close it. This happens in case @@ -3023,10 +3035,10 @@ int ha_federatedx::info(uint flag) DBUG_RETURN(0); error: - if (tmp_io) + if (iop && *iop) { my_sprintf(error_buffer, (error_buffer, ": %d : %s", - tmp_io->error_code(), tmp_io->error_str())); + (*iop)->error_code(), (*iop)->error_str())); my_error(error_code, MYF(0), error_buffer); } else === modified file 'unittest/mysys/Makefile.am' --- a/unittest/mysys/Makefile.am 2009-11-26 08:44:38 +0000 +++ b/unittest/mysys/Makefile.am 2009-12-03 11:19:05 +0000 @@ -24,7 +24,7 @@ LDADD = $(top_builddir)/unittest/mytap $(top_builddir)/strings/libmystrings.a EXTRA_DIST = CMakeLists.txt -noinst_PROGRAMS = bitmap-t base64-t lf-t waiting_threads-t +noinst_PROGRAMS = bitmap-t base64-t my_atomic-t lf-t waiting_threads-t if NEED_THREAD # my_atomic-t is used to check thread functions, so it is safe to

1 0

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (monty:2778)
by Michael Widenius 03 Dec '09

03 Dec '09

#At lp:maria based on revid:monty@askmonty.org-20091130124224-45xbn8nkmgnu1c2f 2778 Michael Widenius 2009-12-03 [merge] Merge with maria-5.1-federatedx; A patch to fix bugs in federatedx and enable federated_server.test Author: Antony Curtis License: BSD modified: client/mysql.cc client/mysqlcheck.c client/mysqlslap.c client/mysqltest.cc configure.in mysql-test/lib/mtr_report.pm mysql-test/mysql-test-run.pl mysql-test/suite/federated/disabled.def mysql-test/suite/federated/federated_server.result mysql-test/suite/federated/federated_server.test storage/federatedx/ha_federatedx.cc unittest/mysys/Makefile.am per-file messages: client/mysql.cc Reset variable if CTRL-C was used to kill running query, so that the user can do it again client/mysqlcheck.c Indentation fix client/mysqlslap.c Indentation fixes client/mysqltest.cc Make testing of commands safer by also testing length Removed not used variable Fixed indentation to be as it was before last patch mysql-test/lib/mtr_report.pm Fixed typo mysql-test/mysql-test-run.pl Merge (Align code with default mysql-tes-run.pl) mysql-test/suite/federated/disabled.def Removed test case storage/federatedx/ha_federatedx.cc Removed my changes and applied Antony's instead === modified file 'client/mysql.cc' --- a/client/mysql.cc 2009-11-16 20:49:51 +0000 +++ b/client/mysql.cc 2009-12-03 11:19:05 +0000 @@ -1281,7 +1281,6 @@ sig_handler handle_sigint(int sig) MYSQL *kill_mysql= NULL; /* terminate if no query being executed, or we already tried interrupting */ - /* terminate if no query being executed, or we already tried interrupting */ if (!executing_query || (interrupted_query == 2)) { tee_fprintf(stdout, "Ctrl-C -- exit!\n"); @@ -1296,6 +1295,7 @@ sig_handler handle_sigint(int sig) goto err; } + /* First time try to kill the query, second time the connection */ interrupted_query++; /* mysqld < 5 does not understand KILL QUERY, skip to KILL CONNECTION */ @@ -1306,10 +1306,13 @@ sig_handler handle_sigint(int sig) sprintf(kill_buffer, "KILL %s%lu", (interrupted_query == 1) ? "QUERY " : "", mysql_thread_id(&mysql)); - tee_fprintf(stdout, "Ctrl-C -- sending \"%s\" to server ...\n", kill_buffer); + if (verbose) + tee_fprintf(stdout, "Ctrl-C -- sending \"%s\" to server ...\n", + kill_buffer); mysql_real_query(kill_mysql, kill_buffer, (uint) strlen(kill_buffer)); mysql_close(kill_mysql); - tee_fprintf(stdout, "Ctrl-C -- query aborted.\n"); + tee_fprintf(stdout, "Ctrl-C -- query killed. Continuing normally.\n"); + interrupted_query= 0; return; @@ -1322,7 +1325,6 @@ err: handler called mysql_end(). */ mysql_thread_end(); - return; #else mysql_end(sig); #endif @@ -2882,13 +2884,8 @@ com_help(String *buffer __attribute__((u return com_server_help(buffer,line,help_arg); } - put_info("\nFor information about MySQL products and services, visit:\n" - " http://www.mysql.com/\n" - "For developer information, including the MySQL Reference Manual, " - "visit:\n" - " http://dev.mysql.com/\n" - "To buy MySQL Enterprise support, training, or other products, visit:\n" - " https://shop.mysql.com/\n", INFO_INFO); + put_info("\nGeneral information about MariaDB can be found at\n" + "http://askmonty.org/wiki/index.php/Manual:Contents\n", INFO_INFO); put_info("List of all MySQL commands:", INFO_INFO); if (!named_cmds) put_info("Note that all text commands must be first on line and end with ';'",INFO_INFO); === modified file 'client/mysqlcheck.c' --- a/client/mysqlcheck.c 2009-09-28 06:24:19 +0000 +++ b/client/mysqlcheck.c 2009-12-03 11:19:05 +0000 @@ -857,7 +857,8 @@ int main(int argc, char **argv) if (!opt_write_binlog) { - if (disable_binlog()) { + if (disable_binlog()) + { first_error= 1; goto end; } === modified file 'client/mysqlslap.c' --- a/client/mysqlslap.c 2009-11-16 20:49:51 +0000 +++ b/client/mysqlslap.c 2009-12-03 11:19:05 +0000 @@ -472,11 +472,10 @@ void concurrency_loop(MYSQL *mysql, uint if (commit_rate) run_query(mysql, "SET AUTOCOMMIT=0", strlen("SET AUTOCOMMIT=0")); - if (pre_system) - if ((sysret= system(pre_system)) != 0) - fprintf(stderr, - "Warning: Execution of pre_system option returned %d.\n", - sysret); + if (pre_system && (sysret= system(pre_system)) != 0) + fprintf(stderr, + "Warning: Execution of pre_system option returned %d.\n", + sysret); /* Pre statements are always run after all other logic so they can @@ -490,11 +489,10 @@ void concurrency_loop(MYSQL *mysql, uint if (post_statements) run_statements(mysql, post_statements); - if (post_system) - if ((sysret= system(post_system)) != 0) - fprintf(stderr, - "Warning: Execution of post_system option returned %d.\n", - sysret); + if (post_system && (sysret= system(post_system)) != 0) + fprintf(stderr, + "Warning: Execution of post_system option returned %d.\n", + sysret); /* We are finished with this run */ if (auto_generate_sql_autoincrement || auto_generate_sql_guid_primary) drop_primary_key_list(); === modified file 'client/mysqltest.cc' --- a/client/mysqltest.cc 2009-11-16 20:49:51 +0000 +++ b/client/mysqltest.cc 2009-12-03 11:19:05 +0000 @@ -3497,9 +3497,10 @@ void do_diff_files(struct st_command *co if ((error= compare_files(ds_filename.str, ds_filename2.str)) && match_expected_error(command, error, NULL) < 0) { - /* Compare of the two files failed, append them to output - so the failure can be analyzed, but only if it was not - expected to fail. + /* + Compare of the two files failed, append them to output + so the failure can be analyzed, but only if it was not + expected to fail. */ show_diff(&ds_res, ds_filename.str, ds_filename2.str); log_file.write(&ds_res); @@ -5013,7 +5014,8 @@ void do_connect(struct st_command *comma con_options= ds_options.str; while (*con_options) { - char* end; + size_t length; + char *end; /* Step past any spaces in beginning of option*/ while (*con_options && my_isspace(charset_info, *con_options)) con_options++; @@ -5021,13 +5023,14 @@ void do_connect(struct st_command *comma end= con_options; while (*end && !my_isspace(charset_info, *end)) end++; - if (!strncmp(con_options, "SSL", 3)) + length= (size_t) (end - con_options); + if (length == 3 && !strncmp(con_options, "SSL", 3)) con_ssl= 1; - else if (!strncmp(con_options, "COMPRESS", 8)) + else if (length == 8 && !strncmp(con_options, "COMPRESS", 8)) con_compress= 1; - else if (!strncmp(con_options, "PIPE", 4)) + else if (length == 4 && !strncmp(con_options, "PIPE", 4)) con_pipe= 1; - else if (!strncmp(con_options, "SHM", 3)) + else if (length == 3 && !strncmp(con_options, "SHM", 3)) con_shm= 1; else die("Illegal option to connect: %.*s", @@ -5096,14 +5099,13 @@ void do_connect(struct st_command *comma mysql_options(&con_slot->mysql, MYSQL_SHARED_MEMORY_BASE_NAME, ds_shm.str); mysql_options(&con_slot->mysql, MYSQL_OPT_PROTOCOL, &protocol); } - else if(shared_memory_base_name) + else if (shared_memory_base_name) { mysql_options(&con_slot->mysql, MYSQL_SHARED_MEMORY_BASE_NAME, - shared_memory_base_name); + shared_memory_base_name); } #endif - /* Use default db name */ if (ds_database.length == 0) dynstr_set(&ds_database, opt_db); @@ -6879,10 +6881,8 @@ void run_query_stmt(MYSQL *mysql, struct MYSQL_STMT *stmt; DYNAMIC_STRING ds_prepare_warnings; DYNAMIC_STRING ds_execute_warnings; - ulonglong affected_rows; DBUG_ENTER("run_query_stmt"); DBUG_PRINT("query", ("'%-.60s'", query)); - LINT_INIT(affected_rows); /* Init a new stmt if it's not already one created for this connection @@ -6981,6 +6981,9 @@ void run_query_stmt(MYSQL *mysql, struct handle_no_error(command); if (!disable_result_log) { + ulonglong affected_rows; + LINT_INIT(affected_rows); + /* Not all statements creates a result set. If there is one we can now create another normal result set that contains the meta @@ -7026,39 +7029,33 @@ void run_query_stmt(MYSQL *mysql, struct Need to grab affected rows information before getting warnings here */ - { - ulonglong affected_rows; - LINT_INIT(affected_rows); + if (!disable_info) + affected_rows= mysql_affected_rows(mysql); - if (!disable_info) - affected_rows= mysql_affected_rows(mysql); + if (!disable_warnings) + { + /* Get the warnings from execute */ - if (!disable_warnings) + /* Append warnings to ds - if there are any */ + if (append_warnings(&ds_execute_warnings, mysql) || + ds_execute_warnings.length || + ds_prepare_warnings.length || + ds_warnings->length) { - /* Get the warnings from execute */ - - /* Append warnings to ds - if there are any */ - if (append_warnings(&ds_execute_warnings, mysql) || - ds_execute_warnings.length || - ds_prepare_warnings.length || - ds_warnings->length) - { - dynstr_append_mem(ds, "Warnings:\n", 10); - if (ds_warnings->length) - dynstr_append_mem(ds, ds_warnings->str, - ds_warnings->length); - if (ds_prepare_warnings.length) - dynstr_append_mem(ds, ds_prepare_warnings.str, - ds_prepare_warnings.length); - if (ds_execute_warnings.length) - dynstr_append_mem(ds, ds_execute_warnings.str, - ds_execute_warnings.length); - } + dynstr_append_mem(ds, "Warnings:\n", 10); + if (ds_warnings->length) + dynstr_append_mem(ds, ds_warnings->str, + ds_warnings->length); + if (ds_prepare_warnings.length) + dynstr_append_mem(ds, ds_prepare_warnings.str, + ds_prepare_warnings.length); + if (ds_execute_warnings.length) + dynstr_append_mem(ds, ds_execute_warnings.str, + ds_execute_warnings.length); } - - if (!disable_info) - append_info(ds, affected_rows, mysql_info(mysql)); } + if (!disable_info) + append_info(ds, affected_rows, mysql_info(mysql)); } end: @@ -7235,7 +7232,6 @@ void run_query(struct st_connection *cn, } dynstr_free(&query_str); - } if (sp_protocol_enabled && @@ -7662,6 +7658,7 @@ int main(int argc, char **argv) my_bool q_send_flag= 0, abort_flag= 0; uint command_executed= 0, last_command_executed= 0; char save_file[FN_REFLEN]; + bool empty_result= FALSE; MY_INIT(argv[0]); save_file[0]= 0; @@ -7819,6 +7816,7 @@ int main(int argc, char **argv) verbose_msg("Start processing test commands from '%s' ...", cur_file->file_name); while (!read_command(&command) && !abort_flag) { + my_bool ok_to_do; int current_line_inc = 1, processed = 0; if (command->type == Q_UNKNOWN || command->type == Q_COMMENT_WITH_COMMAND) get_command_type(command); @@ -7831,7 +7829,7 @@ int main(int argc, char **argv) command->type= Q_COMMENT; } - my_bool ok_to_do= cur_block->ok; + ok_to_do= cur_block->ok; /* Some commands need to be "done" the first time if they may get re-iterated over in a true context. This can only happen if there's @@ -8167,8 +8165,6 @@ int main(int argc, char **argv) if (parsing_disabled) die("Test ended with parsing disabled"); - my_bool empty_result= FALSE; - /* The whole test has been executed _sucessfully_. Time to compare result or save it to record file. === modified file 'configure.in' --- a/configure.in 2009-11-16 20:49:51 +0000 +++ b/configure.in 2009-12-03 11:19:05 +0000 @@ -15,7 +15,7 @@ AC_CANONICAL_SYSTEM # MySQL version number. # # Note: the following line must be parseable by win/configure.js:GetVersion() -AM_INIT_AUTOMAKE(mysql, 5.1.41-mariadb-beta) +AM_INIT_AUTOMAKE(mysql, 5.1.41-MariaDB-beta) AM_CONFIG_HEADER([include/config.h:config.h.in]) PROTOCOL_VERSION=10 === modified file 'mysql-test/lib/mtr_report.pm' --- a/mysql-test/lib/mtr_report.pm 2009-11-16 20:49:51 +0000 +++ b/mysql-test/lib/mtr_report.pm 2009-12-03 11:19:05 +0000 @@ -388,7 +388,7 @@ MSG } elsif (@$extra_warnings) { - mtr_error("There were errors/warnings in server logs after running test cases."); + mtr_error("There where errors/warnings in server logs after running test cases."); } elsif ($fail) { === modified file 'mysql-test/mysql-test-run.pl' --- a/mysql-test/mysql-test-run.pl 2009-11-27 13:20:59 +0000 +++ b/mysql-test/mysql-test-run.pl 2009-12-03 11:19:05 +0000 @@ -201,10 +201,10 @@ my $opt_mark_progress; my $opt_sleep; -my $opt_testcase_timeout= 15; # minutes -my $opt_suite_timeout = 300; # minutes -my $opt_shutdown_timeout= 10; # seconds -my $opt_start_timeout = 180; # seconds +my $opt_testcase_timeout= 15; # 15 minutes +my $opt_suite_timeout = 360; # 6 hours +my $opt_shutdown_timeout= 10; # 10 seconds +my $opt_start_timeout = 180; # 180 seconds sub testcase_timeout { return $opt_testcase_timeout * 60; }; sub suite_timeout { return $opt_suite_timeout * 60; }; @@ -215,7 +215,7 @@ my $opt_start_dirty; my $start_only; my $opt_wait_all; my $opt_repeat= 1; -my $opt_retry= 3; +my $opt_retry= 1; my $opt_retry_failure= 2; my $opt_strace_client; @@ -612,8 +612,9 @@ sub run_test_server ($$$) { my $fake_test= My::Test::read_test($sock); my $test_list= join (" ", @{$fake_test->{testnames}}); push @$extra_warnings, $test_list; + my $report= $fake_test->{'warnings'}; mtr_report("***Warnings generated in error logs during shutdown ". - "after running tests: $test_list"); + "after running tests: $test_list\n\n$report"); $test_failure= 1; if ( !$opt_force ) { # Test failure due to warnings, force is off @@ -1318,6 +1319,8 @@ sub command_line_setup { { # Indicate that we are using debugger $glob_debugger= 1; + $opt_testcase_timeout= 60*60*24; # Don't abort debugging with timeout + $opt_suite_timeout= $opt_testcase_timeout; $opt_retry= 1; $opt_retry_failure= 1; @@ -2150,7 +2153,6 @@ sub environment_setup { # Create an environment variable to make it possible # to detect that valgrind is being used from test cases $ENV{'VALGRIND_TEST'}= $opt_valgrind; - } @@ -2907,8 +2909,8 @@ sub mysql_install_db { my $bootstrap_sql_file= "$opt_vardir/tmp/bootstrap.sql"; my $path_sql= my_find_file($install_basedir, - ["mysql", "sql/share", "share/mysql", - "share/mariadb", "share", "scripts"], + ["mysql", "sql/share", "share/mariadb", + "share/mysql", "share", "scripts"], "mysql_system_tables.sql", NOT_REQUIRED); @@ -3860,7 +3862,7 @@ sub extract_server_log ($$) { my ($error_log, $tname) = @_; # Open the servers .err log file and read all lines - # belonging to current tets into @lines + # belonging to current test into @lines my $Ferr = IO::File->new($error_log) or mtr_error("Could not open file '$error_log' for reading: $!"); @@ -3999,9 +4001,11 @@ sub extract_warning_lines ($) { qr/Slave I\/O: Get master COLLATION_SERVER failed with error:.*/, qr/Slave I\/O: Get master TIME_ZONE failed with error:.*/, qr/Slave I\/O: error reconnecting to master '.*' - retry-time: [1-3] retries/, + qr/Error reading packet/, + qr/Slave: Can't drop database.* database doesn't exist/, ); - my $match_count= 0; + my $matched_lines= []; LINE: foreach my $line ( @lines ) { PAT: foreach my $pat ( @patterns ) @@ -4013,18 +4017,18 @@ sub extract_warning_lines ($) { next LINE if $line =~ $apat; } print $Fwarn $line; - ++$match_count; + push @$matched_lines, $line; last PAT; } } } $Fwarn = undef; # Close file - if ($match_count > 0 && + if (scalar(@$matched_lines) > 0 && defined($last_warning_position->{$error_log}{test_names})) { - return $last_warning_position->{$error_log}{test_names}; + return ($last_warning_position->{$error_log}{test_names}, $matched_lines); } else { - return []; + return ([], $matched_lines); } } @@ -4201,14 +4205,18 @@ sub check_warnings ($) { sub check_warnings_post_shutdown { my ($server_socket)= @_; my $testname_hash= { }; + my $report= ''; foreach my $mysqld ( mysqlds()) { - my $testlist= extract_warning_lines($mysqld->value('#log-error')); + my ($testlist, $match_lines)= + extract_warning_lines($mysqld->value('#log-error')); $testname_hash->{$_}= 1 for @$testlist; + $report.= join('', @$match_lines); } my @warning_tests= keys(%$testname_hash); if (@warning_tests) { my $fake_test= My::Test->new(testnames => \@warning_tests); + $fake_test->{'warnings'}= $report; $fake_test->write_test($server_socket, 'WARNINGS'); } } === modified file 'mysql-test/suite/federated/disabled.def' --- a/mysql-test/suite/federated/disabled.def 2009-10-30 18:50:56 +0000 +++ b/mysql-test/suite/federated/disabled.def 2009-11-14 19:33:59 +0000 @@ -9,5 +9,4 @@ # Do not use any TAB characters for whitespace. # ############################################################################## -federated_server : needs fixup === modified file 'mysql-test/suite/federated/federated_server.result' --- a/mysql-test/suite/federated/federated_server.result 2009-10-30 18:50:56 +0000 +++ b/mysql-test/suite/federated/federated_server.result 2009-11-14 19:33:59 +0000 @@ -175,6 +175,8 @@ CREATE TABLE db_bogus.t1 ( ) ; INSERT INTO db_bogus.t1 VALUES ('2','this is bogus'); +create user test_fed@localhost identified by 'foo'; +grant all on db_legitimate.* to test_fed@localhost; create server 's1' foreign data wrapper 'mysql' options (HOST '127.0.0.1', DATABASE 'db_legitimate', @@ -211,15 +213,14 @@ id name alter server s1 options (database 'db_bogus'); flush tables; select * from federated.t1; -id name -2 this is bogus +ERROR HY000: There was a problem processing the query on the foreign data source. Data source error: : 1044 : Access denied for user 'test_fed'@'localhost' to databa drop server if exists 's1'; ERROR 42000: Access denied; you need the SUPER privilege for this operation create server 's1' foreign data wrapper 'mysql' options (HOST '127.0.0.1', DATABASE 'db_legitimate', -USER 'root', -PASSWORD '', +USER 'test_fed', +PASSWORD 'foo', PORT SLAVE_PORT, SOCKET '', OWNER 'root'); @@ -228,8 +229,8 @@ drop server 's1'; create server 's1' foreign data wrapper 'mysql' options (HOST '127.0.0.1', DATABASE 'db_legitimate', -USER 'root', -PASSWORD '', +USER 'test_fed', +PASSWORD 'foo', PORT SLAVE_PORT, SOCKET '', OWNER 'root'); @@ -237,6 +238,7 @@ flush tables; select * from federated.t1; id name 1 this is legitimate +drop user test_fed@localhost; drop database db_legitimate; drop database db_bogus; drop user guest_super@localhost; @@ -275,6 +277,6 @@ call p1(); drop procedure p1; drop server if exists s; DROP TABLE IF EXISTS federated.t1; -DROP DATABASE federated; +DROP DATABASE IF EXISTS federated; DROP TABLE IF EXISTS federated.t1; -DROP DATABASE federated; +DROP DATABASE IF EXISTS federated; === modified file 'mysql-test/suite/federated/federated_server.test' --- a/mysql-test/suite/federated/federated_server.test 2009-10-30 18:50:56 +0000 +++ b/mysql-test/suite/federated/federated_server.test 2009-11-14 19:33:59 +0000 @@ -239,6 +239,7 @@ alter server s1 options (database 'db_bo connection master; flush tables; +--error ER_QUERY_ON_FOREIGN_DATA_SOURCE select * from federated.t1; connection conn_select; @@ -249,8 +250,8 @@ drop server if exists 's1'; eval create server 's1' foreign data wrapper 'mysql' options (HOST '127.0.0.1', DATABASE 'db_legitimate', - USER 'root', - PASSWORD '', + USER 'test_fed', + PASSWORD 'foo', PORT $SLAVE_MYPORT, SOCKET '', OWNER 'root'); @@ -261,8 +262,8 @@ drop server 's1'; eval create server 's1' foreign data wrapper 'mysql' options (HOST '127.0.0.1', DATABASE 'db_legitimate', - USER 'root', - PASSWORD '', + USER 'test_fed', + PASSWORD 'foo', PORT $SLAVE_MYPORT, SOCKET '', OWNER 'root'); @@ -273,6 +274,7 @@ select * from federated.t1; # clean up test connection slave; +drop user test_fed@localhost; drop database db_legitimate; drop database db_bogus; === modified file 'storage/federatedx/ha_federatedx.cc' --- a/storage/federatedx/ha_federatedx.cc 2009-11-16 20:49:51 +0000 +++ b/storage/federatedx/ha_federatedx.cc 2009-12-03 11:19:05 +0000 @@ -1,5 +1,5 @@ /* -Copyright (c) 2008, Patrick Galbraith +Copyright (c) 2008-2009, Patrick Galbraith & Antony Curtis All rights reserved. Redistribution and use in source and binary forms, with or without @@ -308,7 +308,7 @@ OF THIS SOFTWARE, EVEN IF ADVISED OF THE */ -#define MYSQL_SERVER 1q +#define MYSQL_SERVER 1 #include "mysql_priv.h" #include <mysql/plugin.h> @@ -1627,7 +1627,13 @@ static int free_server(federatedx_txn *t { MEM_ROOT mem_root; - txn->close(server); + if (!txn) + { + federatedx_txn tmp_txn; + tmp_txn.close(server); + } + else + txn->close(server); DBUG_ASSERT(server->io_count == 0); @@ -1785,21 +1791,25 @@ int ha_federatedx::close(void) if (stored_result) retval= free_result(); - /* Disconnect from mysql. thd may be null during refresh */ - txn= thd ? get_txn(thd, true) : new federatedx_txn(); + /* Disconnect from mysql */ + if (!thd || !(txn= get_txn(thd, true))) + { + federatedx_txn tmp_txn; + + tmp_txn.release(&io); - if (txn) + DBUG_ASSERT(io == NULL); + + if ((error= free_share(&tmp_txn, share))) + retval= error; + } + else { txn->release(&io); - DBUG_ASSERT(io == NULL); if ((error= free_share(txn, share))) retval= error; - - if (!thd) - delete txn; - } DBUG_RETURN(retval); } @@ -2793,14 +2803,16 @@ int ha_federatedx::rnd_end() int ha_federatedx::free_result() { int error; + federatedx_io *tmp_io= 0, **iop; DBUG_ASSERT(stored_result); - if ((error= txn->acquire(share, FALSE, &io))) + if (!*(iop= &io) && (error= txn->acquire(share, TRUE, (iop= &tmp_io)))) { DBUG_ASSERT(0); // Fail when testing return error; } - io->free_result(stored_result); + (*iop)->free_result(stored_result); stored_result= 0; + txn->release(&tmp_io); return 0; } @@ -2985,7 +2997,7 @@ int ha_federatedx::info(uint flag) { char error_buffer[FEDERATEDX_QUERY_BUFFER_SIZE]; uint error_code; - federatedx_io *tmp_io= 0; + federatedx_io *tmp_io= 0, **iop= 0; DBUG_ENTER("ha_federatedx::info"); error_code= ER_QUERY_ON_FOREIGN_DATA_SOURCE; @@ -2993,7 +3005,7 @@ int ha_federatedx::info(uint flag) /* we want not to show table status if not needed to do so */ if (flag & (HA_STATUS_VARIABLE | HA_STATUS_CONST | HA_STATUS_AUTO)) { - if ((error_code= txn->acquire(share, TRUE, &tmp_io))) + if (!*(iop= &io) && (error_code= txn->acquire(share, TRUE, (iop= &tmp_io)))) goto fail; } @@ -3006,13 +3018,13 @@ int ha_federatedx::info(uint flag) if (flag & HA_STATUS_CONST) stats.block_size= 4096; - if (tmp_io->table_metadata(&stats, share->table_name, + if ((*iop)->table_metadata(&stats, share->table_name, share->table_name_length, flag)) goto error; } if (flag & HA_STATUS_AUTO) - stats.auto_increment_value= tmp_io->last_insert_id(); + stats.auto_increment_value= (*iop)->last_insert_id(); /* If ::info created it's own transaction, close it. This happens in case @@ -3023,10 +3035,10 @@ int ha_federatedx::info(uint flag) DBUG_RETURN(0); error: - if (tmp_io) + if (iop && *iop) { my_sprintf(error_buffer, (error_buffer, ": %d : %s", - tmp_io->error_code(), tmp_io->error_str())); + (*iop)->error_code(), (*iop)->error_str())); my_error(error_code, MYF(0), error_buffer); } else === modified file 'unittest/mysys/Makefile.am' --- a/unittest/mysys/Makefile.am 2009-11-26 08:44:38 +0000 +++ b/unittest/mysys/Makefile.am 2009-12-03 11:19:05 +0000 @@ -24,7 +24,7 @@ LDADD = $(top_builddir)/unittest/mytap $(top_builddir)/strings/libmystrings.a EXTRA_DIST = CMakeLists.txt -noinst_PROGRAMS = bitmap-t base64-t lf-t waiting_threads-t +noinst_PROGRAMS = bitmap-t base64-t my_atomic-t lf-t waiting_threads-t if NEED_THREAD # my_atomic-t is used to check thread functions, so it is safe to

1 0

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (monty:2778)
by Michael Widenius 03 Dec '09

03 Dec '09

#At lp:maria based on revid:monty@askmonty.org-20091130124224-45xbn8nkmgnu1c2f 2778 Michael Widenius 2009-12-03 [merge] Merge with maria-5.1-federatedx; Antony's patch for federatedx modified: client/mysql.cc client/mysqlcheck.c client/mysqlslap.c client/mysqltest.cc configure.in mysql-test/lib/mtr_report.pm mysql-test/mysql-test-run.pl mysql-test/suite/federated/disabled.def mysql-test/suite/federated/federated_server.result mysql-test/suite/federated/federated_server.test storage/federatedx/ha_federatedx.cc unittest/mysys/Makefile.am per-file messages: client/mysql.cc Reset variable if CTRL-C was used to kill running query, so that the user can do it again client/mysqlcheck.c Indentation fix client/mysqlslap.c Indentation fixes client/mysqltest.cc Make testing of commands safer by also testing length Removed not used variable Fixed indentation to be as it was before last patch mysql-test/lib/mtr_report.pm Fixed typo mysql-test/mysql-test-run.pl Merge (Align code with default mysql-tes-run.pl) mysql-test/suite/federated/disabled.def Removed test case storage/federatedx/ha_federatedx.cc Removed my changes and applied Antony's instead === modified file 'client/mysql.cc' --- a/client/mysql.cc 2009-11-16 20:49:51 +0000 +++ b/client/mysql.cc 2009-12-03 11:07:55 +0000 @@ -1281,7 +1281,6 @@ sig_handler handle_sigint(int sig) MYSQL *kill_mysql= NULL; /* terminate if no query being executed, or we already tried interrupting */ - /* terminate if no query being executed, or we already tried interrupting */ if (!executing_query || (interrupted_query == 2)) { tee_fprintf(stdout, "Ctrl-C -- exit!\n"); @@ -1296,6 +1295,7 @@ sig_handler handle_sigint(int sig) goto err; } + /* First time try to kill the query, second time the connection */ interrupted_query++; /* mysqld < 5 does not understand KILL QUERY, skip to KILL CONNECTION */ @@ -1306,10 +1306,13 @@ sig_handler handle_sigint(int sig) sprintf(kill_buffer, "KILL %s%lu", (interrupted_query == 1) ? "QUERY " : "", mysql_thread_id(&mysql)); - tee_fprintf(stdout, "Ctrl-C -- sending \"%s\" to server ...\n", kill_buffer); + if (verbose) + tee_fprintf(stdout, "Ctrl-C -- sending \"%s\" to server ...\n", + kill_buffer); mysql_real_query(kill_mysql, kill_buffer, (uint) strlen(kill_buffer)); mysql_close(kill_mysql); - tee_fprintf(stdout, "Ctrl-C -- query aborted.\n"); + tee_fprintf(stdout, "Ctrl-C -- query killed. Continuing normally.\n"); + interrupted_query= 0; return; @@ -1322,7 +1325,6 @@ err: handler called mysql_end(). */ mysql_thread_end(); - return; #else mysql_end(sig); #endif @@ -2882,13 +2884,8 @@ com_help(String *buffer __attribute__((u return com_server_help(buffer,line,help_arg); } - put_info("\nFor information about MySQL products and services, visit:\n" - " http://www.mysql.com/\n" - "For developer information, including the MySQL Reference Manual, " - "visit:\n" - " http://dev.mysql.com/\n" - "To buy MySQL Enterprise support, training, or other products, visit:\n" - " https://shop.mysql.com/\n", INFO_INFO); + put_info("\nGeneral information about MariaDB can be found at\n" + "http://askmonty.org/wiki/index.php/Manual:Contents\n", INFO_INFO); put_info("List of all MySQL commands:", INFO_INFO); if (!named_cmds) put_info("Note that all text commands must be first on line and end with ';'",INFO_INFO); === modified file 'client/mysqlcheck.c' --- a/client/mysqlcheck.c 2009-09-28 06:24:19 +0000 +++ b/client/mysqlcheck.c 2009-12-03 11:07:55 +0000 @@ -857,7 +857,8 @@ int main(int argc, char **argv) if (!opt_write_binlog) { - if (disable_binlog()) { + if (disable_binlog()) + { first_error= 1; goto end; } === modified file 'client/mysqlslap.c' --- a/client/mysqlslap.c 2009-11-16 20:49:51 +0000 +++ b/client/mysqlslap.c 2009-12-03 11:07:55 +0000 @@ -472,11 +472,10 @@ void concurrency_loop(MYSQL *mysql, uint if (commit_rate) run_query(mysql, "SET AUTOCOMMIT=0", strlen("SET AUTOCOMMIT=0")); - if (pre_system) - if ((sysret= system(pre_system)) != 0) - fprintf(stderr, - "Warning: Execution of pre_system option returned %d.\n", - sysret); + if (pre_system && (sysret= system(pre_system)) != 0) + fprintf(stderr, + "Warning: Execution of pre_system option returned %d.\n", + sysret); /* Pre statements are always run after all other logic so they can @@ -490,11 +489,10 @@ void concurrency_loop(MYSQL *mysql, uint if (post_statements) run_statements(mysql, post_statements); - if (post_system) - if ((sysret= system(post_system)) != 0) - fprintf(stderr, - "Warning: Execution of post_system option returned %d.\n", - sysret); + if (post_system && (sysret= system(post_system)) != 0) + fprintf(stderr, + "Warning: Execution of post_system option returned %d.\n", + sysret); /* We are finished with this run */ if (auto_generate_sql_autoincrement || auto_generate_sql_guid_primary) drop_primary_key_list(); === modified file 'client/mysqltest.cc' --- a/client/mysqltest.cc 2009-11-16 20:49:51 +0000 +++ b/client/mysqltest.cc 2009-12-03 11:07:55 +0000 @@ -3497,9 +3497,10 @@ void do_diff_files(struct st_command *co if ((error= compare_files(ds_filename.str, ds_filename2.str)) && match_expected_error(command, error, NULL) < 0) { - /* Compare of the two files failed, append them to output - so the failure can be analyzed, but only if it was not - expected to fail. + /* + Compare of the two files failed, append them to output + so the failure can be analyzed, but only if it was not + expected to fail. */ show_diff(&ds_res, ds_filename.str, ds_filename2.str); log_file.write(&ds_res); @@ -5013,7 +5014,8 @@ void do_connect(struct st_command *comma con_options= ds_options.str; while (*con_options) { - char* end; + size_t length; + char *end; /* Step past any spaces in beginning of option*/ while (*con_options && my_isspace(charset_info, *con_options)) con_options++; @@ -5021,13 +5023,14 @@ void do_connect(struct st_command *comma end= con_options; while (*end && !my_isspace(charset_info, *end)) end++; - if (!strncmp(con_options, "SSL", 3)) + length= (size_t) (end - con_options); + if (length == 3 && !strncmp(con_options, "SSL", 3)) con_ssl= 1; - else if (!strncmp(con_options, "COMPRESS", 8)) + else if (length == 8 && !strncmp(con_options, "COMPRESS", 8)) con_compress= 1; - else if (!strncmp(con_options, "PIPE", 4)) + else if (length == 4 && !strncmp(con_options, "PIPE", 4)) con_pipe= 1; - else if (!strncmp(con_options, "SHM", 3)) + else if (length == 3 && !strncmp(con_options, "SHM", 3)) con_shm= 1; else die("Illegal option to connect: %.*s", @@ -5096,14 +5099,13 @@ void do_connect(struct st_command *comma mysql_options(&con_slot->mysql, MYSQL_SHARED_MEMORY_BASE_NAME, ds_shm.str); mysql_options(&con_slot->mysql, MYSQL_OPT_PROTOCOL, &protocol); } - else if(shared_memory_base_name) + else if (shared_memory_base_name) { mysql_options(&con_slot->mysql, MYSQL_SHARED_MEMORY_BASE_NAME, - shared_memory_base_name); + shared_memory_base_name); } #endif - /* Use default db name */ if (ds_database.length == 0) dynstr_set(&ds_database, opt_db); @@ -6879,10 +6881,8 @@ void run_query_stmt(MYSQL *mysql, struct MYSQL_STMT *stmt; DYNAMIC_STRING ds_prepare_warnings; DYNAMIC_STRING ds_execute_warnings; - ulonglong affected_rows; DBUG_ENTER("run_query_stmt"); DBUG_PRINT("query", ("'%-.60s'", query)); - LINT_INIT(affected_rows); /* Init a new stmt if it's not already one created for this connection @@ -6981,6 +6981,9 @@ void run_query_stmt(MYSQL *mysql, struct handle_no_error(command); if (!disable_result_log) { + ulonglong affected_rows; + LINT_INIT(affected_rows); + /* Not all statements creates a result set. If there is one we can now create another normal result set that contains the meta @@ -7026,39 +7029,33 @@ void run_query_stmt(MYSQL *mysql, struct Need to grab affected rows information before getting warnings here */ - { - ulonglong affected_rows; - LINT_INIT(affected_rows); + if (!disable_info) + affected_rows= mysql_affected_rows(mysql); - if (!disable_info) - affected_rows= mysql_affected_rows(mysql); + if (!disable_warnings) + { + /* Get the warnings from execute */ - if (!disable_warnings) + /* Append warnings to ds - if there are any */ + if (append_warnings(&ds_execute_warnings, mysql) || + ds_execute_warnings.length || + ds_prepare_warnings.length || + ds_warnings->length) { - /* Get the warnings from execute */ - - /* Append warnings to ds - if there are any */ - if (append_warnings(&ds_execute_warnings, mysql) || - ds_execute_warnings.length || - ds_prepare_warnings.length || - ds_warnings->length) - { - dynstr_append_mem(ds, "Warnings:\n", 10); - if (ds_warnings->length) - dynstr_append_mem(ds, ds_warnings->str, - ds_warnings->length); - if (ds_prepare_warnings.length) - dynstr_append_mem(ds, ds_prepare_warnings.str, - ds_prepare_warnings.length); - if (ds_execute_warnings.length) - dynstr_append_mem(ds, ds_execute_warnings.str, - ds_execute_warnings.length); - } + dynstr_append_mem(ds, "Warnings:\n", 10); + if (ds_warnings->length) + dynstr_append_mem(ds, ds_warnings->str, + ds_warnings->length); + if (ds_prepare_warnings.length) + dynstr_append_mem(ds, ds_prepare_warnings.str, + ds_prepare_warnings.length); + if (ds_execute_warnings.length) + dynstr_append_mem(ds, ds_execute_warnings.str, + ds_execute_warnings.length); } - - if (!disable_info) - append_info(ds, affected_rows, mysql_info(mysql)); } + if (!disable_info) + append_info(ds, affected_rows, mysql_info(mysql)); } end: @@ -7235,7 +7232,6 @@ void run_query(struct st_connection *cn, } dynstr_free(&query_str); - } if (sp_protocol_enabled && @@ -7662,6 +7658,7 @@ int main(int argc, char **argv) my_bool q_send_flag= 0, abort_flag= 0; uint command_executed= 0, last_command_executed= 0; char save_file[FN_REFLEN]; + bool empty_result= FALSE; MY_INIT(argv[0]); save_file[0]= 0; @@ -7819,6 +7816,7 @@ int main(int argc, char **argv) verbose_msg("Start processing test commands from '%s' ...", cur_file->file_name); while (!read_command(&command) && !abort_flag) { + my_bool ok_to_do; int current_line_inc = 1, processed = 0; if (command->type == Q_UNKNOWN || command->type == Q_COMMENT_WITH_COMMAND) get_command_type(command); @@ -7831,7 +7829,7 @@ int main(int argc, char **argv) command->type= Q_COMMENT; } - my_bool ok_to_do= cur_block->ok; + ok_to_do= cur_block->ok; /* Some commands need to be "done" the first time if they may get re-iterated over in a true context. This can only happen if there's @@ -8167,8 +8165,6 @@ int main(int argc, char **argv) if (parsing_disabled) die("Test ended with parsing disabled"); - my_bool empty_result= FALSE; - /* The whole test has been executed _sucessfully_. Time to compare result or save it to record file. === modified file 'configure.in' --- a/configure.in 2009-11-16 20:49:51 +0000 +++ b/configure.in 2009-12-03 11:07:55 +0000 @@ -15,7 +15,7 @@ AC_CANONICAL_SYSTEM # MySQL version number. # # Note: the following line must be parseable by win/configure.js:GetVersion() -AM_INIT_AUTOMAKE(mysql, 5.1.41-mariadb-beta) +AM_INIT_AUTOMAKE(mysql, 5.1.41-MariaDB-beta) AM_CONFIG_HEADER([include/config.h:config.h.in]) PROTOCOL_VERSION=10 === modified file 'mysql-test/lib/mtr_report.pm' --- a/mysql-test/lib/mtr_report.pm 2009-11-16 20:49:51 +0000 +++ b/mysql-test/lib/mtr_report.pm 2009-12-03 11:07:55 +0000 @@ -388,7 +388,7 @@ MSG } elsif (@$extra_warnings) { - mtr_error("There were errors/warnings in server logs after running test cases."); + mtr_error("There where errors/warnings in server logs after running test cases."); } elsif ($fail) { === modified file 'mysql-test/mysql-test-run.pl' --- a/mysql-test/mysql-test-run.pl 2009-11-27 13:20:59 +0000 +++ b/mysql-test/mysql-test-run.pl 2009-12-03 11:07:55 +0000 @@ -201,10 +201,10 @@ my $opt_mark_progress; my $opt_sleep; -my $opt_testcase_timeout= 15; # minutes -my $opt_suite_timeout = 300; # minutes -my $opt_shutdown_timeout= 10; # seconds -my $opt_start_timeout = 180; # seconds +my $opt_testcase_timeout= 15; # 15 minutes +my $opt_suite_timeout = 360; # 6 hours +my $opt_shutdown_timeout= 10; # 10 seconds +my $opt_start_timeout = 180; # 180 seconds sub testcase_timeout { return $opt_testcase_timeout * 60; }; sub suite_timeout { return $opt_suite_timeout * 60; }; @@ -215,7 +215,7 @@ my $opt_start_dirty; my $start_only; my $opt_wait_all; my $opt_repeat= 1; -my $opt_retry= 3; +my $opt_retry= 1; my $opt_retry_failure= 2; my $opt_strace_client; @@ -612,8 +612,9 @@ sub run_test_server ($$$) { my $fake_test= My::Test::read_test($sock); my $test_list= join (" ", @{$fake_test->{testnames}}); push @$extra_warnings, $test_list; + my $report= $fake_test->{'warnings'}; mtr_report("***Warnings generated in error logs during shutdown ". - "after running tests: $test_list"); + "after running tests: $test_list\n\n$report"); $test_failure= 1; if ( !$opt_force ) { # Test failure due to warnings, force is off @@ -1318,6 +1319,8 @@ sub command_line_setup { { # Indicate that we are using debugger $glob_debugger= 1; + $opt_testcase_timeout= 60*60*24; # Don't abort debugging with timeout + $opt_suite_timeout= $opt_testcase_timeout; $opt_retry= 1; $opt_retry_failure= 1; @@ -2150,7 +2153,6 @@ sub environment_setup { # Create an environment variable to make it possible # to detect that valgrind is being used from test cases $ENV{'VALGRIND_TEST'}= $opt_valgrind; - } @@ -2907,8 +2909,8 @@ sub mysql_install_db { my $bootstrap_sql_file= "$opt_vardir/tmp/bootstrap.sql"; my $path_sql= my_find_file($install_basedir, - ["mysql", "sql/share", "share/mysql", - "share/mariadb", "share", "scripts"], + ["mysql", "sql/share", "share/mariadb", + "share/mysql", "share", "scripts"], "mysql_system_tables.sql", NOT_REQUIRED); @@ -3860,7 +3862,7 @@ sub extract_server_log ($$) { my ($error_log, $tname) = @_; # Open the servers .err log file and read all lines - # belonging to current tets into @lines + # belonging to current test into @lines my $Ferr = IO::File->new($error_log) or mtr_error("Could not open file '$error_log' for reading: $!"); @@ -3999,9 +4001,11 @@ sub extract_warning_lines ($) { qr/Slave I\/O: Get master COLLATION_SERVER failed with error:.*/, qr/Slave I\/O: Get master TIME_ZONE failed with error:.*/, qr/Slave I\/O: error reconnecting to master '.*' - retry-time: [1-3] retries/, + qr/Error reading packet/, + qr/Slave: Can't drop database.* database doesn't exist/, ); - my $match_count= 0; + my $matched_lines= []; LINE: foreach my $line ( @lines ) { PAT: foreach my $pat ( @patterns ) @@ -4013,18 +4017,18 @@ sub extract_warning_lines ($) { next LINE if $line =~ $apat; } print $Fwarn $line; - ++$match_count; + push @$matched_lines, $line; last PAT; } } } $Fwarn = undef; # Close file - if ($match_count > 0 && + if (scalar(@$matched_lines) > 0 && defined($last_warning_position->{$error_log}{test_names})) { - return $last_warning_position->{$error_log}{test_names}; + return ($last_warning_position->{$error_log}{test_names}, $matched_lines); } else { - return []; + return ([], $matched_lines); } } @@ -4201,14 +4205,18 @@ sub check_warnings ($) { sub check_warnings_post_shutdown { my ($server_socket)= @_; my $testname_hash= { }; + my $report= ''; foreach my $mysqld ( mysqlds()) { - my $testlist= extract_warning_lines($mysqld->value('#log-error')); + my ($testlist, $match_lines)= + extract_warning_lines($mysqld->value('#log-error')); $testname_hash->{$_}= 1 for @$testlist; + $report.= join('', @$match_lines); } my @warning_tests= keys(%$testname_hash); if (@warning_tests) { my $fake_test= My::Test->new(testnames => \@warning_tests); + $fake_test->{'warnings'}= $report; $fake_test->write_test($server_socket, 'WARNINGS'); } } === modified file 'mysql-test/suite/federated/disabled.def' --- a/mysql-test/suite/federated/disabled.def 2009-10-30 18:50:56 +0000 +++ b/mysql-test/suite/federated/disabled.def 2009-11-14 19:33:59 +0000 @@ -9,5 +9,4 @@ # Do not use any TAB characters for whitespace. # ############################################################################## -federated_server : needs fixup === modified file 'mysql-test/suite/federated/federated_server.result' --- a/mysql-test/suite/federated/federated_server.result 2009-10-30 18:50:56 +0000 +++ b/mysql-test/suite/federated/federated_server.result 2009-11-14 19:33:59 +0000 @@ -175,6 +175,8 @@ CREATE TABLE db_bogus.t1 ( ) ; INSERT INTO db_bogus.t1 VALUES ('2','this is bogus'); +create user test_fed@localhost identified by 'foo'; +grant all on db_legitimate.* to test_fed@localhost; create server 's1' foreign data wrapper 'mysql' options (HOST '127.0.0.1', DATABASE 'db_legitimate', @@ -211,15 +213,14 @@ id name alter server s1 options (database 'db_bogus'); flush tables; select * from federated.t1; -id name -2 this is bogus +ERROR HY000: There was a problem processing the query on the foreign data source. Data source error: : 1044 : Access denied for user 'test_fed'@'localhost' to databa drop server if exists 's1'; ERROR 42000: Access denied; you need the SUPER privilege for this operation create server 's1' foreign data wrapper 'mysql' options (HOST '127.0.0.1', DATABASE 'db_legitimate', -USER 'root', -PASSWORD '', +USER 'test_fed', +PASSWORD 'foo', PORT SLAVE_PORT, SOCKET '', OWNER 'root'); @@ -228,8 +229,8 @@ drop server 's1'; create server 's1' foreign data wrapper 'mysql' options (HOST '127.0.0.1', DATABASE 'db_legitimate', -USER 'root', -PASSWORD '', +USER 'test_fed', +PASSWORD 'foo', PORT SLAVE_PORT, SOCKET '', OWNER 'root'); @@ -237,6 +238,7 @@ flush tables; select * from federated.t1; id name 1 this is legitimate +drop user test_fed@localhost; drop database db_legitimate; drop database db_bogus; drop user guest_super@localhost; @@ -275,6 +277,6 @@ call p1(); drop procedure p1; drop server if exists s; DROP TABLE IF EXISTS federated.t1; -DROP DATABASE federated; +DROP DATABASE IF EXISTS federated; DROP TABLE IF EXISTS federated.t1; -DROP DATABASE federated; +DROP DATABASE IF EXISTS federated; === modified file 'mysql-test/suite/federated/federated_server.test' --- a/mysql-test/suite/federated/federated_server.test 2009-10-30 18:50:56 +0000 +++ b/mysql-test/suite/federated/federated_server.test 2009-11-14 19:33:59 +0000 @@ -239,6 +239,7 @@ alter server s1 options (database 'db_bo connection master; flush tables; +--error ER_QUERY_ON_FOREIGN_DATA_SOURCE select * from federated.t1; connection conn_select; @@ -249,8 +250,8 @@ drop server if exists 's1'; eval create server 's1' foreign data wrapper 'mysql' options (HOST '127.0.0.1', DATABASE 'db_legitimate', - USER 'root', - PASSWORD '', + USER 'test_fed', + PASSWORD 'foo', PORT $SLAVE_MYPORT, SOCKET '', OWNER 'root'); @@ -261,8 +262,8 @@ drop server 's1'; eval create server 's1' foreign data wrapper 'mysql' options (HOST '127.0.0.1', DATABASE 'db_legitimate', - USER 'root', - PASSWORD '', + USER 'test_fed', + PASSWORD 'foo', PORT $SLAVE_MYPORT, SOCKET '', OWNER 'root'); @@ -273,6 +274,7 @@ select * from federated.t1; # clean up test connection slave; +drop user test_fed@localhost; drop database db_legitimate; drop database db_bogus; === modified file 'storage/federatedx/ha_federatedx.cc' --- a/storage/federatedx/ha_federatedx.cc 2009-11-16 20:49:51 +0000 +++ b/storage/federatedx/ha_federatedx.cc 2009-12-03 11:07:55 +0000 @@ -1,5 +1,5 @@ /* -Copyright (c) 2008, Patrick Galbraith +Copyright (c) 2008-2009, Patrick Galbraith & Antony Curtis All rights reserved. Redistribution and use in source and binary forms, with or without @@ -308,7 +308,7 @@ OF THIS SOFTWARE, EVEN IF ADVISED OF THE */ -#define MYSQL_SERVER 1q +#define MYSQL_SERVER 1 #include "mysql_priv.h" #include <mysql/plugin.h> @@ -1627,7 +1627,13 @@ static int free_server(federatedx_txn *t { MEM_ROOT mem_root; - txn->close(server); + if (!txn) + { + federatedx_txn tmp_txn; + tmp_txn.close(server); + } + else + txn->close(server); DBUG_ASSERT(server->io_count == 0); @@ -1785,21 +1791,25 @@ int ha_federatedx::close(void) if (stored_result) retval= free_result(); - /* Disconnect from mysql. thd may be null during refresh */ - txn= thd ? get_txn(thd, true) : new federatedx_txn(); + /* Disconnect from mysql */ + if (!thd || !(txn= get_txn(thd, true))) + { + federatedx_txn tmp_txn; + + tmp_txn.release(&io); - if (txn) + DBUG_ASSERT(io == NULL); + + if ((error= free_share(&tmp_txn, share))) + retval= error; + } + else { txn->release(&io); - DBUG_ASSERT(io == NULL); if ((error= free_share(txn, share))) retval= error; - - if (!thd) - delete txn; - } DBUG_RETURN(retval); } @@ -2793,14 +2803,16 @@ int ha_federatedx::rnd_end() int ha_federatedx::free_result() { int error; + federatedx_io *tmp_io= 0, **iop; DBUG_ASSERT(stored_result); - if ((error= txn->acquire(share, FALSE, &io))) + if (!*(iop= &io) && (error= txn->acquire(share, TRUE, (iop= &tmp_io)))) { DBUG_ASSERT(0); // Fail when testing return error; } - io->free_result(stored_result); + (*iop)->free_result(stored_result); stored_result= 0; + txn->release(&tmp_io); return 0; } @@ -2985,7 +2997,7 @@ int ha_federatedx::info(uint flag) { char error_buffer[FEDERATEDX_QUERY_BUFFER_SIZE]; uint error_code; - federatedx_io *tmp_io= 0; + federatedx_io *tmp_io= 0, **iop= 0; DBUG_ENTER("ha_federatedx::info"); error_code= ER_QUERY_ON_FOREIGN_DATA_SOURCE; @@ -2993,7 +3005,7 @@ int ha_federatedx::info(uint flag) /* we want not to show table status if not needed to do so */ if (flag & (HA_STATUS_VARIABLE | HA_STATUS_CONST | HA_STATUS_AUTO)) { - if ((error_code= txn->acquire(share, TRUE, &tmp_io))) + if (!*(iop= &io) && (error_code= txn->acquire(share, TRUE, (iop= &tmp_io)))) goto fail; } @@ -3006,13 +3018,13 @@ int ha_federatedx::info(uint flag) if (flag & HA_STATUS_CONST) stats.block_size= 4096; - if (tmp_io->table_metadata(&stats, share->table_name, + if ((*iop)->table_metadata(&stats, share->table_name, share->table_name_length, flag)) goto error; } if (flag & HA_STATUS_AUTO) - stats.auto_increment_value= tmp_io->last_insert_id(); + stats.auto_increment_value= (*iop)->last_insert_id(); /* If ::info created it's own transaction, close it. This happens in case @@ -3023,10 +3035,10 @@ int ha_federatedx::info(uint flag) DBUG_RETURN(0); error: - if (tmp_io) + if (iop && *iop) { my_sprintf(error_buffer, (error_buffer, ": %d : %s", - tmp_io->error_code(), tmp_io->error_str())); + (*iop)->error_code(), (*iop)->error_str())); my_error(error_code, MYF(0), error_buffer); } else === modified file 'unittest/mysys/Makefile.am' --- a/unittest/mysys/Makefile.am 2009-11-26 08:44:38 +0000 +++ b/unittest/mysys/Makefile.am 2009-12-03 11:07:55 +0000 @@ -24,7 +24,7 @@ LDADD = $(top_builddir)/unittest/mytap $(top_builddir)/strings/libmystrings.a EXTRA_DIST = CMakeLists.txt -noinst_PROGRAMS = bitmap-t base64-t lf-t waiting_threads-t +noinst_PROGRAMS = bitmap-t base64-t my_atomic-t lf-t waiting_threads-t if NEED_THREAD # my_atomic-t is used to check thread functions, so it is safe to

1 0

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (knielsen:2776)
by knielsen＠knielsen-hq.org 02 Dec '09

02 Dec '09

#At lp:maria 2776 knielsen(a)knielsen-hq.org 2009-12-02 Fix PBXT crash during shutdown if shutdown happens early (like wrong command args). modified: storage/pbxt/src/myxt_xt.cc === modified file 'storage/pbxt/src/myxt_xt.cc' --- a/storage/pbxt/src/myxt_xt.cc 2009-12-01 09:50:46 +0000 +++ b/storage/pbxt/src/myxt_xt.cc 2009-12-02 18:07:59 +0000 @@ -3112,10 +3112,19 @@ xtPublic void *myxt_create_thread() * references to the PBXT plugin object and will effectively deadlock the plugin so * that server will have to force plugin shutdown. To avoid deadlocking and forced shutdown * we must dereference the plugin after creating THD objects. + * Similarly to global_system_variables.table_plugin as described above, + * new_thd->valriables.table_plugin can also become NULL due to shutdown. */ - LEX_STRING& plugin_name = REF_MYSQL_PLUGIN(new_thd->variables.table_plugin)->name; + plugin_ref table_plugin = new_thd->variables.table_plugin; + if (!table_plugin) { + delete new_thd; + xt_register_xterr(XT_REG_CONTEXT, XT_ERR_MYSQL_NO_THREAD); + return NULL; + } + + LEX_STRING& plugin_name = REF_MYSQL_PLUGIN(table_plugin)->name; if ((plugin_name.length == 4) && (strncmp(plugin_name.str, "PBXT", plugin_name.length) == 0)) { - REF_MYSQL_PLUGIN(new_thd->variables.table_plugin)->ref_count--; + REF_MYSQL_PLUGIN(table_plugin)->ref_count--; } new_thd->thread_stack = (char *) &new_thd; new_thd->store_globals();

1 0

Re: [Maria-developers] [Merge] lp:~paul-mccullagh/maria/maria-pbxt-rc3 into lp:maria
by Sergei Golubchik 02 Dec '09

02 Dec '09

Hi, Paul! On Dec 02, Paul McCullagh wrote: > On Dec 2, 2009, at 2:46 PM, Paul McCullagh wrote: >> >> I have tested this, and it all works as you specified, but now I have >> found a problem with PBXT. >> >> The problem stems from code in PBXT (in functions ha_pbxt::write_row, >> ha_pbxt::update_row, ha_pbxt::delete_row) which I have pasted below. >> The code is a hack! ... >> I noticed that InnoDB and NDB call trans_register_ha() >> unconditionally at the start of a statement. >> >> Could it be that the problem that lead to my "hack" has been fixed? > > Just doing some testing, and it looks like the problem has been fixed, > so I can remove my hack ... :) Great! Yes, I vaguely remember that there were changes that could've had fixed that. I'd need to check the code to be sure, but as you say it works... :) Regards / Mit vielen Grüßen, Sergei -- __ ___ ___ ____ __ / |/ /_ __/ __/ __ \/ / Sergei Golubchik <serg(a)sun.com> / /|_/ / // /\ \/ /_/ / /__ Principal Software Engineer/Server Architect /_/ /_/\_, /___/\___\_\___/ Sun Microsystems GmbH, HRB München 161028 <___/ Sonnenallee 1, 85551 Kirchheim-Heimstetten Geschäftsführer: Thomas Schroeder, Wolfgang Engels, Wolf Frenkel Vorsitzender des Aufsichtsrates: Martin Häring

1 0

Re: [Maria-developers] [Merge] lp:~paul-mccullagh/maria/maria-pbxt-rc3 into lp:maria
by Sergei Golubchik 02 Dec '09

02 Dec '09

Hi, Paul! On Dec 01, Paul McCullagh wrote: > > This patch changes MariaDB code in one case: > > File handler.cc, Line 1591 > > DBUG_ASSERT(total_ha_2pc == (ulong) opt_bin_log+1); // only InnoDB and binlog > > This line has been commented out, because the assertions fails now > that PBXT supports XA. > > Although this assertion implies that not everything will work, I have > not found any problem with the stability of MySQL due to the addition > of PBXT XA. Eh, no, not exactly. This assertion was to make sure that the ifdef'ed code only runs when InnoDB is the only XA-capable engine. As it's no longer the case, you need to remove the whole ifdef block, not only the assert. > PBXT XA can be disabled by setting pbxt_support_xa=0, it is enabled by > default. > Regards / Mit vielen Grüßen, Sergei -- __ ___ ___ ____ __ / |/ /_ __/ __/ __ \/ / Sergei Golubchik <serg(a)sun.com> / /|_/ / // /\ \/ /_/ / /__ Principal Software Engineer/Server Architect /_/ /_/\_, /___/\___\_\___/ Sun Microsystems GmbH, HRB München 161028 <___/ Sonnenallee 1, 85551 Kirchheim-Heimstetten Geschäftsführer: Thomas Schroeder, Wolfgang Engels, Wolf Frenkel Vorsitzender des Aufsichtsrates: Martin Häring

3 7

[Maria-developers] Rev 2733: Remove compiler warning: always use literal string for printf-format line. in file:///home/psergey/dev/maria-5.2/
by Sergey Petrunya 02 Dec '09

02 Dec '09

At file:///home/psergey/dev/maria-5.2/ ------------------------------------------------------------ revno: 2733 revision-id: psergey(a)askmonty.org-20091202142609-18bp41q8mejxl47t parent: igor(a)askmonty.org-20091112043128-yd6odg8zr5ribjal committer: Sergey Petrunya <psergey(a)askmonty.org> branch nick: maria-5.2 timestamp: Wed 2009-12-02 16:26:09 +0200 message: Remove compiler warning: always use literal string for printf-format line. === modified file 'sql/item.h' --- a/sql/item.h 2009-11-10 02:32:39 +0000 +++ b/sql/item.h 2009-12-02 14:26:09 +0000 @@ -25,7 +25,7 @@ sprintf(buff, "%s::%s", where, processor_name); DBUG_ENTER(buff); sprintf(buff, "%s returns TRUE: unsupported function", processor_name); - DBUG_PRINT("info", (buff)); + DBUG_PRINT("info", ("%s", buff)); DBUG_RETURN(TRUE); }

1 0

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (knielsen:2776)
by knielsen＠knielsen-hq.org 02 Dec '09

02 Dec '09

#At lp:maria 2776 knielsen(a)knielsen-hq.org 2009-12-02 Fix PBXT crash during shutdown if shutdown happens early (like wrong command args). modified: storage/pbxt/src/myxt_xt.cc === modified file 'storage/pbxt/src/myxt_xt.cc' --- a/storage/pbxt/src/myxt_xt.cc 2009-12-01 09:50:46 +0000 +++ b/storage/pbxt/src/myxt_xt.cc 2009-12-02 14:19:34 +0000 @@ -3112,8 +3112,17 @@ xtPublic void *myxt_create_thread() * references to the PBXT plugin object and will effectively deadlock the plugin so * that server will have to force plugin shutdown. To avoid deadlocking and forced shutdown * we must dereference the plugin after creating THD objects. + * Similarly to global_system_variables.table_plugin as described above, + * new_thd->valriables.table_plugin can also become NULL due to shutdown. */ - LEX_STRING& plugin_name = REF_MYSQL_PLUGIN(new_thd->variables.table_plugin)->name; + plugin_ref table_plugin = REF_MYSQL_PLUGIN(new_thd->variables.table_plugin); + if (!table_plugin) { + delete new_thd; + xt_register_xterr(XT_REG_CONTEXT, XT_ERR_MYSQL_NO_THREAD); + return NULL; + } + + LEX_STRING& plugin_name = table_plugin->name; if ((plugin_name.length == 4) && (strncmp(plugin_name.str, "PBXT", plugin_name.length) == 0)) { REF_MYSQL_PLUGIN(new_thd->variables.table_plugin)->ref_count--; }

1 0

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (knielsen:2775)
by knielsen＠knielsen-hq.org 02 Dec '09

02 Dec '09

#At lp:maria 2775 knielsen(a)knielsen-hq.org 2009-12-02 [merge] Merge PBXT 1.0.09f RC3 into MariaDB. added: storage/pbxt/src/backup_xt.cc storage/pbxt/src/backup_xt.h modified: sql/handler.cc storage/pbxt/ChangeLog storage/pbxt/src/Makefile.am storage/pbxt/src/cache_xt.cc storage/pbxt/src/cache_xt.h storage/pbxt/src/database_xt.cc storage/pbxt/src/database_xt.h storage/pbxt/src/datadic_xt.cc storage/pbxt/src/datalog_xt.cc storage/pbxt/src/discover_xt.cc storage/pbxt/src/filesys_xt.cc storage/pbxt/src/filesys_xt.h storage/pbxt/src/ha_pbxt.cc storage/pbxt/src/ha_pbxt.h storage/pbxt/src/ha_xtsys.h storage/pbxt/src/heap_xt.cc storage/pbxt/src/index_xt.cc storage/pbxt/src/lock_xt.cc storage/pbxt/src/locklist_xt.h storage/pbxt/src/memory_xt.cc storage/pbxt/src/myxt_xt.cc storage/pbxt/src/myxt_xt.h storage/pbxt/src/pbms.h storage/pbxt/src/pbms_enabled.cc storage/pbxt/src/pbms_enabled.h storage/pbxt/src/pthread_xt.cc storage/pbxt/src/restart_xt.cc storage/pbxt/src/restart_xt.h storage/pbxt/src/strutil_xt.cc storage/pbxt/src/systab_xt.cc storage/pbxt/src/tabcache_xt.cc storage/pbxt/src/tabcache_xt.h storage/pbxt/src/table_xt.cc storage/pbxt/src/table_xt.h storage/pbxt/src/thread_xt.cc storage/pbxt/src/thread_xt.h storage/pbxt/src/util_xt.cc storage/pbxt/src/util_xt.h storage/pbxt/src/xaction_xt.cc storage/pbxt/src/xaction_xt.h storage/pbxt/src/xactlog_xt.cc storage/pbxt/src/xactlog_xt.h storage/pbxt/src/xt_config.h storage/pbxt/src/xt_defs.h storage/pbxt/src/xt_errno.h === modified file 'sql/handler.cc' --- a/sql/handler.cc 2009-11-16 20:49:51 +0000 +++ b/sql/handler.cc 2009-12-02 11:50:40 +0000 @@ -1584,16 +1584,6 @@ int ha_recover(HASH *commit_list) if (info.commit_list) sql_print_information("Starting crash recovery..."); -#ifndef WILL_BE_DELETED_LATER - /* - for now, only InnoDB supports 2pc. It means we can always safely - rollback all pending transactions, without risking inconsistent data - */ - DBUG_ASSERT(total_ha_2pc == (ulong) opt_bin_log+1); // only InnoDB and binlog - tc_heuristic_recover= TC_HEURISTIC_RECOVER_ROLLBACK; // forcing ROLLBACK - info.dry_run=FALSE; -#endif - for (info.len= MAX_XID_LIST_SIZE ; info.list==0 && info.len > MIN_XID_LIST_SIZE; info.len/=2) { === modified file 'storage/pbxt/ChangeLog' --- a/storage/pbxt/ChangeLog 2009-09-03 06:15:03 +0000 +++ b/storage/pbxt/ChangeLog 2009-12-01 09:50:46 +0000 @@ -1,6 +1,58 @@ PBXT Release Notes ================== +------- 1.0.09f RC3 - 2009-11-30 + +RN291: Fixed bug #489088: On shutdown MySQL reports: [Warning] Plugin 'PBXT' will be forced to shutdown. + +RN290: Fixed bug #345524: pbxt does not compile on 64 bit windows. Currently atomic operations are not supported on this platform. + +RN286: Fixed a bug introduced in RN281, which could cause an index scan to hang. The original change was to prevent a warning in Valgrind. + +RN285: Merged changes required to compile with Drizzle. + +RN284: Fixed bug that cause the error "[ERROR] Invalid (old?) table or database name 'mysqld.1'", when running temp_table.test under MariaDB (thanks to Monty for his initial bug fix). Added a fix for partition table names as well. + +RN283: Added win_inttypes.h to the distribution. This file is only required for the Windows build. + +RN282: Fixed bug #451101: jump or move depends on uninitialised value in myxt_get_key_length + +RN281: Fixed bug #451080: Uninitialised memory write in XTDatabaseLog::xlog_append + +RN280: Fixed bug #451085: jump or move depends on uninitialised value in my_type_to_string + +RN279: Fixed bug #441000: xtstat crashes with segmentation fault on startup if max_pbxt_threads exceeded. + +------- 1.0.09e RC3 - 2009-11-20 + +RN278: Fixed compile error with MySQL 5.1.41. + +------- 1.0.09d RC3 - 2009-09-30 + +RN277: Added r/o flag to pbxt_max_threads server variable (this fix is related to bug #430637) + +RN276: Added test case for replication on tables w/o PKs (see bug #430716) + +RN275: Fixed bug #430600: 'Failed to read auto-increment value from storage engine' error. + +RN274: Fixed bug #431240: This report is public edit xtstat fails if no PBXT table has been created. xtstat now accepts --database=information_schema or --database=pbxt. Depending on this setting PBXT will either use the information_schema.pbxt_statistics or the pbxt.statistics table. If information_schema is used, then the statistics are displayed even when no PBXT table exists. Recovery activity is also displayed, unless pbxt_support_xa=1, in which case MySQL will wait for PBXT recovery to complete before allowing connections. + +RN273: Fixed bug #430633: XA_RBDEADLOCK is not returned on XA END after the transacting ended with a deadlock. + +RN272: Fixed bug #430596: Backup/restore does not work well even on a basic PBXT table with auto-increment. + +------- 1.0.09c RC3 - 2009-09-16 + +RN271: Windows build update: now you can simply put the pbxt directory under <mysql-root>/storage and build the PBXT engine as a part of the source tree. The engine will be linked statically. Be sure to specify the WITH_PBXT_STORAGE_ENGINE option when running win\configure.js + +RN270: Correctly disabled PBMS so that this version now compiles under Windows. If PBMS_ENABLED is defined, PBXT will not compile under Windows becaause of a getpid() call in pbms.h. + +------- 1.0.09 RC3 - 2009-09-09 + +RN269: Implemented online backup. A native online backup driver now performs BACKUP and RESTORE DATABASE operations for PBXT. NOTE: This feature is only supported by MySQL 6.0.9 or later. + +RN268: Implemented XA support. PBXT now supports all XA related MySQL statements. The variable pbxt_support_xa determines if XA support is enabled. Note: due to MySQL bug #47134, enabling XA support could lead to a crash. + ------- 1.0.08d RC2 - 2009-09-02 RN267: Fixed a bug that caused MySQL to crash on shutdown, after an incorrect command line parameter was given. The crash occurred because the background recovery task was not cleaned up before the PBXT engine was de-initialized. === modified file 'storage/pbxt/src/Makefile.am' --- a/storage/pbxt/src/Makefile.am 2009-10-07 07:40:56 +0000 +++ b/storage/pbxt/src/Makefile.am 2009-11-24 10:55:06 +0000 @@ -22,7 +22,7 @@ noinst_HEADERS = bsearch_xt.h cache_xt. pbms_enabled.h sortedlist_xt.h strutil_xt.h \ tabcache_xt.h table_xt.h trace_xt.h thread_xt.h \ util_xt.h xaction_xt.h xactlog_xt.h lock_xt.h \ - systab_xt.h ha_xtsys.h discover_xt.h \ + systab_xt.h ha_xtsys.h discover_xt.h backup_xt.h \ pbms.h xt_config.h xt_defs.h xt_errno.h locklist_xt.h EXTRA_LTLIBRARIES = libpbxt.la @@ -32,7 +32,7 @@ libpbxt_la_SOURCES = bsearch_xt.cc cache memory_xt.cc myxt_xt.cc pthread_xt.cc restart_xt.cc \ sortedlist_xt.cc strutil_xt.cc \ tabcache_xt.cc table_xt.cc trace_xt.cc thread_xt.cc \ - systab_xt.cc ha_xtsys.cc discover_xt.cc \ + systab_xt.cc ha_xtsys.cc discover_xt.cc backup_xt.cc \ util_xt.cc xaction_xt.cc xactlog_xt.cc lock_xt.cc locklist_xt.cc libpbxt_la_LDFLAGS = -module === added file 'storage/pbxt/src/backup_xt.cc' --- a/storage/pbxt/src/backup_xt.cc 1970-01-01 00:00:00 +0000 +++ b/storage/pbxt/src/backup_xt.cc 2009-11-24 10:55:06 +0000 @@ -0,0 +1,802 @@ +/* Copyright (c) 2009 PrimeBase Technologies GmbH + * + * PrimeBase XT + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * + * 2009-09-07 Paul McCullagh + * + * H&G2JCtL + */ + +#include "xt_config.h" + +#ifdef MYSQL_SUPPORTS_BACKUP + +#include <string.h> +#include <stdio.h> +#include <stdlib.h> +#include <time.h> +#include <ctype.h> + +#include "mysql_priv.h" +#include <backup/api_types.h> +#include <backup/backup_engine.h> +#include <backup/backup_aux.h> // for build_table_list() +#include <hash.h> + +#include "ha_pbxt.h" + +#include "backup_xt.h" +#include "pthread_xt.h" +#include "filesys_xt.h" +#include "database_xt.h" +#include "strutil_xt.h" +#include "memory_xt.h" +#include "trace_xt.h" +#include "myxt_xt.h" + +#ifdef OK +#undef OK +#endif + +#ifdef byte +#undef byte +#endif + +#ifdef DEBUG +//#define TRACE_BACKUP_CALLS +//#define TEST_SMALL_BLOCK 100000 +#endif + +using backup::byte; +using backup::result_t; +using backup::version_t; +using backup::Table_list; +using backup::Table_ref; +using backup::Buffer; + +#ifdef TRACE_BACKUP_CALLS +#define XT_TRACE_CALL() ha_trace_function(__FUNC__, NULL) +#else +#define XT_TRACE_CALL() +#endif + +#define XT_RESTORE_BATCH_SIZE 10000 + +#define BUP_STATE_BEFORE_LOCK 0 +#define BUP_STATE_AFTER_LOCK 1 + +#define BUP_STANDARD_VAR_RECORD 1 +#define BUP_RECORD_BLOCK_4_START 2 // Part of a record, with a 4 byte total length, and 4 byte data length +#define BUP_RECORD_BLOCK_4 3 // Part of a record, with a 4 byte length +#define BUP_RECORD_BLOCK_4_END 4 // Last part of a record with a 4 byte length + +/* + * ----------------------------------------------------------------------- + * UTILITIES + */ + +#ifdef TRACE_BACKUP_CALLS +static void ha_trace_function(const char *function, char *table) +{ + char func_buf[50], *ptr; + XTThreadPtr thread = xt_get_self(); + + if ((ptr = strchr(function, '('))) { + ptr--; + while (ptr > function) { + if (!(isalnum(*ptr) || *ptr == '_')) + break; + ptr--; + } + ptr++; + xt_strcpy(50, func_buf, ptr); + if ((ptr = strchr(func_buf, '('))) + *ptr = 0; + } + else + xt_strcpy(50, func_buf, function); + if (table) + printf("%s %s (%s)\n", thread ? thread->t_name : "-unknown-", func_buf, table); + else + printf("%s %s\n", thread ? thread->t_name : "-unknown-", func_buf); +} +#endif + +/* + * ----------------------------------------------------------------------- + * BACKUP DRIVER + */ + +class PBXTBackupDriver: public Backup_driver +{ + public: + PBXTBackupDriver(const Table_list &); + virtual ~PBXTBackupDriver(); + + virtual size_t size(); + virtual size_t init_size(); + virtual result_t begin(const size_t); + virtual result_t end(); + virtual result_t get_data(Buffer &); + virtual result_t prelock(); + virtual result_t lock(); + virtual result_t unlock(); + virtual result_t cancel(); + virtual void free(); + void lock_tables_TL_READ_NO_INSERT(); + + private: + XTThreadPtr bd_thread; + int bd_state; + u_int bd_table_no; + XTOpenTablePtr bd_ot; + xtWord1 *bd_row_buf; + + /* Non-zero if we last returned only part of + * a row. + */ + xtWord1 *db_write_block(xtWord1 *buffer, xtWord1 bup_type, size_t *size, xtWord4 row_len); + xtWord1 *db_write_block(xtWord1 *buffer, xtWord1 bup_type, size_t *size, xtWord4 total_len, xtWord4 row_len); + + xtWord4 bd_row_offset; + xtWord4 bd_row_size; +}; + + +PBXTBackupDriver::PBXTBackupDriver(const Table_list &tables): +Backup_driver(tables), +bd_state(BUP_STATE_BEFORE_LOCK), +bd_table_no(0), +bd_ot(NULL), +bd_row_buf(NULL), +bd_row_offset(0), +bd_row_size(0) +{ +} + +PBXTBackupDriver::~PBXTBackupDriver() +{ +} + +/** Estimates total size of backup. @todo improve it */ +size_t PBXTBackupDriver::size() +{ + XT_TRACE_CALL(); + return UNKNOWN_SIZE; +} + +/** Estimates size of backup before lock. @todo improve it */ +size_t PBXTBackupDriver::init_size() +{ + XT_TRACE_CALL(); + return 0; +} + +result_t PBXTBackupDriver::begin(const size_t) +{ + THD *thd = current_thd; + XTExceptionRec e; + + XT_TRACE_CALL(); + + if (!(bd_thread = xt_ha_set_current_thread(thd, &e))) { + xt_log_exception(NULL, &e, XT_LOG_DEFAULT); + return backup::ERROR; + } + + return backup::OK; +} + +result_t PBXTBackupDriver::end() +{ + XT_TRACE_CALL(); + if (bd_ot) { + xt_tab_seq_exit(bd_ot); + xt_db_return_table_to_pool_ns(bd_ot); + bd_ot = NULL; + } + if (bd_thread->st_xact_data) { + if (!xt_xn_commit(bd_thread)) + return backup::ERROR; + } + return backup::OK; +} + +xtWord1 *PBXTBackupDriver::db_write_block(xtWord1 *buffer, xtWord1 bup_type, size_t *ret_size, xtWord4 row_len) +{ + register size_t size = *ret_size; + + *buffer = bup_type; // Record type identifier. + buffer++; + size--; + memcpy(buffer, bd_ot->ot_row_wbuffer, row_len); + buffer += row_len; + size -= row_len; + *ret_size = size; + return buffer; +} + +xtWord1 *PBXTBackupDriver::db_write_block(xtWord1 *buffer, xtWord1 bup_type, size_t *ret_size, xtWord4 total_len, xtWord4 row_len) +{ + register size_t size = *ret_size; + + *buffer = bup_type; // Record type identifier. + buffer++; + size--; + if (bup_type == BUP_RECORD_BLOCK_4_START) { + XT_SET_DISK_4(buffer, total_len); + buffer += 4; + size -= 4; + } + XT_SET_DISK_4(buffer, row_len); + buffer += 4; + size -= 4; + memcpy(buffer, bd_ot->ot_row_wbuffer+bd_row_offset, row_len); + buffer += row_len; + size -= row_len; + bd_row_size -= row_len; + bd_row_offset += row_len; + *ret_size = size; + return buffer; +} + +result_t PBXTBackupDriver::get_data(Buffer &buf) +{ + xtBool eof = FALSE; + size_t size; + xtWord4 row_len; + xtWord1 *buffer; + + XT_TRACE_CALL(); + + if (bd_state == BUP_STATE_BEFORE_LOCK) { + buf.table_num = 0; + buf.size = 0; + buf.last = FALSE; + return backup::READY; + } + + /* Open the backup table: */ + if (!bd_ot) { + XTThreadPtr self = bd_thread; + XTTableHPtr tab; + char path[PATH_MAX]; + + if (bd_table_no == m_tables.count()) { + buf.size = 0; + buf.table_num = 0; + buf.last = TRUE; + return backup::DONE; + } + + m_tables[bd_table_no].internal_name(path, sizeof(path)); + bd_table_no++; + try_(a) { + xt_ha_open_database_of_table(self, (XTPathStrPtr) path); + tab = xt_use_table(self, (XTPathStrPtr) path, FALSE, FALSE, NULL); + pushr_(xt_heap_release, tab); + if (!(bd_ot = xt_db_open_table_using_tab(tab, bd_thread))) + xt_throw(self); + freer_(); // xt_heap_release(tab) + + /* Prepare the seqential scan: */ + xt_tab_seq_exit(bd_ot); + if (!xt_tab_seq_init(bd_ot)) + xt_throw(self); + + if (bd_row_buf) { + xt_free(self, bd_row_buf); + bd_row_buf = NULL; + } + bd_row_buf = (xtWord1 *) xt_malloc(self, bd_ot->ot_table->tab_dic.dic_mysql_buf_size); + bd_ot->ot_cols_req = bd_ot->ot_table->tab_dic.dic_no_of_cols; + } + catch_(a) { + ; + } + cont_(a); + + if (!bd_ot) + goto failed; + } + + buf.table_num = bd_table_no; +#ifdef TEST_SMALL_BLOCK + buf.size = TEST_SMALL_BLOCK; +#endif + size = buf.size; + buffer = (xtWord1 *) buf.data; + ASSERT_NS(size > 9); + + /* First check of a record was partically written + * last time. + */ + write_row: + if (bd_row_size > 0) { + row_len = bd_row_size; + if (bd_row_offset == 0) { + if (row_len+1 > size) { + ASSERT_NS(size > 9); + row_len = size - 9; + buffer = db_write_block(buffer, BUP_RECORD_BLOCK_4_START, &size, bd_row_size, row_len); + goto done; + } + buffer = db_write_block(buffer, BUP_STANDARD_VAR_RECORD, &size, row_len); + bd_row_size = 0; + } + else { + if (row_len+5 > size) { + row_len = size - 5; + buffer = db_write_block(buffer, BUP_RECORD_BLOCK_4, &size, 0, row_len); + goto done; + } + buffer = db_write_block(buffer, BUP_RECORD_BLOCK_4_END, &size, 0, row_len); + } + } + + /* Now continue with the sequential scan. */ + while (size > 1) { + if (!xt_tab_seq_next(bd_ot, bd_row_buf, &eof)) + goto failed; + if (eof) { + /* We will go the next table, on the next call. */ + xt_tab_seq_exit(bd_ot); + xt_db_return_table_to_pool_ns(bd_ot); + bd_ot = NULL; + break; + } + if (!(row_len = myxt_store_row_data(bd_ot, 0, (char *) bd_row_buf))) + goto failed; + if (row_len+1 > size) { + /* Does not fit: */ + bd_row_offset = 0; + bd_row_size = row_len; + /* Only add part of the row, if there is still + * quite a bit of space left: + */ + if (size >= (32 * 1024)) + goto write_row; + break; + } + buffer = db_write_block(buffer, BUP_STANDARD_VAR_RECORD, &size, row_len); + } + + done: + buf.size = buf.size - size; + /* This indicates wnd of data for a table! */ + buf.last = eof; + + return backup::OK; + + failed: + xt_log_and_clear_exception(bd_thread); + return backup::ERROR; +} + +result_t PBXTBackupDriver::prelock() +{ + XT_TRACE_CALL(); + return backup::READY; +} + +result_t PBXTBackupDriver::lock() +{ + XT_TRACE_CALL(); + bd_thread->st_xact_mode = XT_XACT_COMMITTED_READ; + bd_thread->st_ignore_fkeys = FALSE; + bd_thread->st_auto_commit = FALSE; + bd_thread->st_table_trans = FALSE; + bd_thread->st_abort_trans = FALSE; + bd_thread->st_stat_ended = FALSE; + bd_thread->st_stat_trans = FALSE; + bd_thread->st_is_update = FALSE; + if (!xt_xn_begin(bd_thread)) + return backup::ERROR; + bd_state = BUP_STATE_AFTER_LOCK; + return backup::OK; +} + +result_t PBXTBackupDriver::unlock() +{ + XT_TRACE_CALL(); + return backup::OK; +} + +result_t PBXTBackupDriver::cancel() +{ + XT_TRACE_CALL(); + return backup::OK; // free() will be called and suffice +} + +void PBXTBackupDriver::free() +{ + XT_TRACE_CALL(); + if (bd_ot) { + xt_tab_seq_exit(bd_ot); + xt_db_return_table_to_pool_ns(bd_ot); + bd_ot = NULL; + } + if (bd_row_buf) { + xt_free_ns(bd_row_buf); + bd_row_buf = NULL; + } + if (bd_thread->st_xact_data) + xt_xn_rollback(bd_thread); + delete this; +} + +void PBXTBackupDriver::lock_tables_TL_READ_NO_INSERT() +{ + XT_TRACE_CALL(); +} + +/* + * ----------------------------------------------------------------------- + * BACKUP DRIVER + */ + +class PBXTRestoreDriver: public Restore_driver +{ + public: + PBXTRestoreDriver(const Table_list &tables); + virtual ~PBXTRestoreDriver(); + + virtual result_t begin(const size_t); + virtual result_t end(); + virtual result_t send_data(Buffer &buf); + virtual result_t cancel(); + virtual void free(); + + private: + XTThreadPtr rd_thread; + u_int rd_table_no; + XTOpenTablePtr rd_ot; + STRUCT_TABLE *rd_my_table; + xtWord1 *rb_row_buf; + u_int rb_col_cnt; + u_int rb_insert_count; + + /* Long rows are accumulated here: */ + xtWord4 rb_row_len; + xtWord4 rb_data_size; + xtWord1 *rb_row_data; +}; + +PBXTRestoreDriver::PBXTRestoreDriver(const Table_list &tables): +Restore_driver(tables), +rd_thread(NULL), +rd_table_no(0), +rd_ot(NULL), +rb_row_buf(NULL), +rb_row_len(0), +rb_data_size(0), +rb_row_data(NULL) +{ +} + +PBXTRestoreDriver::~PBXTRestoreDriver() +{ +} + +result_t PBXTRestoreDriver::begin(const size_t) +{ + THD *thd = current_thd; + XTExceptionRec e; + + XT_TRACE_CALL(); + + if (!(rd_thread = xt_ha_set_current_thread(thd, &e))) { + xt_log_exception(NULL, &e, XT_LOG_DEFAULT); + return backup::ERROR; + } + + return backup::OK; +} + +result_t PBXTRestoreDriver::end() +{ + XT_TRACE_CALL(); + if (rd_ot) { + xt_db_return_table_to_pool_ns(rd_ot); + rd_ot = NULL; + } + //if (rb_row_buf) { + // xt_free_ns(rb_row_buf); + // rb_row_buf = NULL; + //} + if (rb_row_data) { + xt_free_ns(rb_row_data); + rb_row_data = NULL; + } + if (rd_thread->st_xact_data) { + if (!xt_xn_commit(rd_thread)) + return backup::ERROR; + } + return backup::OK; +} + + +result_t PBXTRestoreDriver::send_data(Buffer &buf) +{ + size_t size; + xtWord1 type; + xtWord1 *buffer; + xtWord4 row_len; + xtWord1 *rec_data; + + XT_TRACE_CALL(); + + if (buf.table_num != rd_table_no) { + XTThreadPtr self = rd_thread; + XTTableHPtr tab; + char path[PATH_MAX]; + + if (rd_ot) { + xt_db_return_table_to_pool_ns(rd_ot); + rd_ot = NULL; + } + + if (rd_thread->st_xact_data) { + if (!xt_xn_commit(rd_thread)) + goto failed; + } + if (!xt_xn_begin(rd_thread)) + goto failed; + rb_insert_count = 0; + + rd_table_no = buf.table_num; + m_tables[rd_table_no-1].internal_name(path, sizeof(path)); + try_(a) { + xt_ha_open_database_of_table(self, (XTPathStrPtr) path); + tab = xt_use_table(self, (XTPathStrPtr) path, FALSE, FALSE, NULL); + pushr_(xt_heap_release, tab); + if (!(rd_ot = xt_db_open_table_using_tab(tab, rd_thread))) + xt_throw(self); + freer_(); // xt_heap_release(tab) + + rd_my_table = rd_ot->ot_table->tab_dic.dic_my_table; + if (rd_my_table->found_next_number_field) { + rd_my_table->in_use = current_thd; + rd_my_table->next_number_field = rd_my_table->found_next_number_field; + rd_my_table->mark_columns_used_by_index_no_reset(rd_my_table->s->next_number_index, rd_my_table->read_set); + } + + /* This is safe because only one thread can restore a table at + * a time! + */ + rb_row_buf = (xtWord1 *) rd_my_table->record[0]; + //if (rb_row_buf) { + // xt_free(self, rb_row_buf); + // rb_row_buf = NULL; + //} + //rb_row_buf = (xtWord1 *) xt_malloc(self, rd_ot->ot_table->tab_dic.dic_mysql_buf_size); + + rb_col_cnt = rd_ot->ot_table->tab_dic.dic_no_of_cols; + + } + catch_(a) { + ; + } + cont_(a); + + if (!rd_ot) + goto failed; + } + + buffer = (xtWord1 *) buf.data; + size = buf.size; + + while (size > 0) { + type = *buffer; + switch (type) { + case BUP_STANDARD_VAR_RECORD: + rec_data = buffer + 1; + break; + case BUP_RECORD_BLOCK_4_START: + buffer++; + row_len = XT_GET_DISK_4(buffer); + buffer += 4; + if (rb_data_size < row_len) { + if (!xt_realloc_ns((void **) &rb_row_data, row_len)) + goto failed; + rb_data_size = row_len; + } + row_len = XT_GET_DISK_4(buffer); + buffer += 4; + ASSERT_NS(row_len <= rb_data_size); + if (row_len > rb_data_size) { + xt_register_xterr(XT_REG_CONTEXT, XT_ERR_BAD_BACKUP_FORMAT); + goto failed; + } + memcpy(rb_row_data, buffer, row_len); + rb_row_len = row_len; + buffer += row_len; + if (row_len + 9 > size) { + xt_register_xterr(XT_REG_CONTEXT, XT_ERR_BAD_BACKUP_FORMAT); + goto failed; + } + size -= row_len + 9; + continue; + case BUP_RECORD_BLOCK_4: + buffer++; + row_len = XT_GET_DISK_4(buffer); + buffer += 4; + ASSERT_NS(rb_row_len + row_len <= rb_data_size); + if (rb_row_len + row_len > rb_data_size) { + xt_register_xterr(XT_REG_CONTEXT, XT_ERR_BAD_BACKUP_FORMAT); + goto failed; + } + memcpy(rb_row_data + rb_row_len, buffer, row_len); + rb_row_len += row_len; + buffer += row_len; + if (row_len + 5 > size) { + xt_register_xterr(XT_REG_CONTEXT, XT_ERR_BAD_BACKUP_FORMAT); + goto failed; + } + size -= row_len + 5; + continue; + case BUP_RECORD_BLOCK_4_END: + buffer++; + row_len = XT_GET_DISK_4(buffer); + buffer += 4; + ASSERT_NS(rb_row_len + row_len <= rb_data_size); + if (rb_row_len + row_len > rb_data_size) { + xt_register_xterr(XT_REG_CONTEXT, XT_ERR_BAD_BACKUP_FORMAT); + goto failed; + } + memcpy(rb_row_data + rb_row_len, buffer, row_len); + buffer += row_len; + if (row_len + 5 > size) { + xt_register_xterr(XT_REG_CONTEXT, XT_ERR_BAD_BACKUP_FORMAT); + goto failed; + } + size -= row_len + 5; + rec_data = rb_row_data; + break; + default: + xt_register_xterr(XT_REG_CONTEXT, XT_ERR_BAD_BACKUP_FORMAT); + goto failed; + } + + if (!(row_len = myxt_load_row_data(rd_ot, rec_data, rb_row_buf, rb_col_cnt))) + goto failed; + + if (rd_ot->ot_table->tab_dic.dic_my_table->found_next_number_field) + ha_set_auto_increment(rd_ot, rd_ot->ot_table->tab_dic.dic_my_table->found_next_number_field); + + if (!xt_tab_new_record(rd_ot, rb_row_buf)) + goto failed; + + if (type == BUP_STANDARD_VAR_RECORD) { + buffer += row_len+1; + if (row_len + 1 > size) { + xt_register_xterr(XT_REG_CONTEXT, XT_ERR_BAD_BACKUP_FORMAT); + goto failed; + } + size -= row_len + 1; + } + + rb_insert_count++; + if (rb_insert_count == XT_RESTORE_BATCH_SIZE) { + if (!xt_xn_commit(rd_thread)) + goto failed; + if (!xt_xn_begin(rd_thread)) + goto failed; + rb_insert_count = 0; + } + } + + return backup::OK; + + failed: + xt_log_and_clear_exception(rd_thread); + return backup::ERROR; +} + + +result_t PBXTRestoreDriver::cancel() +{ + XT_TRACE_CALL(); + /* Nothing to do in cancel(); free() will suffice */ + return backup::OK; +} + +void PBXTRestoreDriver::free() +{ + XT_TRACE_CALL(); + if (rd_ot) { + xt_db_return_table_to_pool_ns(rd_ot); + rd_ot = NULL; + } + //if (rb_row_buf) { + // xt_free_ns(rb_row_buf); + // rb_row_buf = NULL; + //} + if (rb_row_data) { + xt_free_ns(rb_row_data); + rb_row_data = NULL; + } + if (rd_thread->st_xact_data) + xt_xn_rollback(rd_thread); + delete this; +} + +/* + * ----------------------------------------------------------------------- + * BACKUP ENGINE FACTORY + */ + +#define PBXT_BACKUP_VERSION 1 + + +class PBXTBackupEngine: public Backup_engine +{ + public: + PBXTBackupEngine() { }; + + virtual version_t version() const { + return PBXT_BACKUP_VERSION; + }; + + virtual result_t get_backup(const uint32, const Table_list &, Backup_driver* &); + + virtual result_t get_restore(const version_t, const uint32, const Table_list &,Restore_driver* &); + + virtual void free() + { + delete this; + } +}; + +result_t PBXTBackupEngine::get_backup(const u_int count, const Table_list &tables, Backup_driver* &drv) +{ + PBXTBackupDriver *ptr = new PBXTBackupDriver(tables); + + if (!ptr) + return backup::ERROR; + drv = ptr; + return backup::OK; +} + +result_t PBXTBackupEngine::get_restore(const version_t ver, const uint32, + const Table_list &tables, Restore_driver* &drv) +{ + if (ver > PBXT_BACKUP_VERSION) + { + return backup::ERROR; + } + + PBXTRestoreDriver *ptr = new PBXTRestoreDriver(tables); + + if (!ptr) + return backup::ERROR; + drv = (Restore_driver *) ptr; + return backup::OK; +} + + +Backup_result_t pbxt_backup_engine(handlerton *self, Backup_engine* &be) +{ + be = new PBXTBackupEngine(); + + if (!be) + return backup::ERROR; + + return backup::OK; +} + +#endif === added file 'storage/pbxt/src/backup_xt.h' --- a/storage/pbxt/src/backup_xt.h 1970-01-01 00:00:00 +0000 +++ b/storage/pbxt/src/backup_xt.h 2009-11-24 10:55:06 +0000 @@ -0,0 +1,34 @@ +/* Copyright (c) 2009 PrimeBase Technologies GmbH + * + * PrimeBase XT + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * + * 2009-09-07 Paul McCullagh + * + * H&G2JCtL + */ + +#ifndef __backup_xt_h__ +#define __backup_xt_h__ + +#include "xt_defs.h" + +#ifdef MYSQL_SUPPORTS_BACKUP + +Backup_result_t pbxt_backup_engine(handlerton *self, Backup_engine* &be); + +#endif +#endif === modified file 'storage/pbxt/src/cache_xt.cc' --- a/storage/pbxt/src/cache_xt.cc 2009-10-30 18:50:56 +0000 +++ b/storage/pbxt/src/cache_xt.cc 2009-11-24 10:55:06 +0000 @@ -73,7 +73,7 @@ #define IDX_CAC_UNLOCK(i, o) xt_xsmutex_unlock(&(i)->cs_lock, (o)->t_id) #elif defined(IDX_CAC_USE_PTHREAD_RW) #define IDX_CAC_LOCK_TYPE xt_rwlock_type -#define IDX_CAC_INIT_LOCK(s, i) xt_init_rwlock(s, &(i)->cs_lock) +#define IDX_CAC_INIT_LOCK(s, i) xt_init_rwlock_with_autoname(s, &(i)->cs_lock) #define IDX_CAC_FREE_LOCK(s, i) xt_free_rwlock(&(i)->cs_lock) #define IDX_CAC_READ_LOCK(i, o) xt_slock_rwlock_ns(&(i)->cs_lock) #define IDX_CAC_WRITE_LOCK(i, o) xt_xlock_rwlock_ns(&(i)->cs_lock) @@ -94,8 +94,12 @@ #define IDX_CAC_UNLOCK(i, s) xt_spinxslock_unlock(&(i)->cs_lock, (s)->t_id) #endif +#ifdef XT_NO_ATOMICS +#define ID_HANDLE_USE_PTHREAD_RW +#else #define ID_HANDLE_USE_SPINLOCK //#define ID_HANDLE_USE_PTHREAD_RW +#endif #if defined(ID_HANDLE_USE_PTHREAD_RW) #define ID_HANDLE_LOCK_TYPE xt_mutex_type @@ -374,7 +378,7 @@ xtPublic void xt_ind_release_handle(XTIn { DcHandleSlotPtr hs; XTIndBlockPtr block = NULL; - u_int hash_idx = 0; + u_int hash_idx = 0; DcSegmentPtr seg = NULL; XTIndBlockPtr xblock; @@ -1379,7 +1383,7 @@ xtPublic xtBool xt_ind_fetch(XTOpenTable ASSERT_NS(iref->ir_xlock == 2); #endif if (!(block = ind_cac_fetch(ot, ind, address, &seg, TRUE))) - return 0; + return FAILED; branch_size = XT_GET_DISK_2(((XTIdxBranchDPtr) block->cb_data)->tb_size_2); if (XT_GET_INDEX_BLOCK_LEN(branch_size) < 2 || XT_GET_INDEX_BLOCK_LEN(branch_size) > XT_INDEX_PAGE_SIZE) { === modified file 'storage/pbxt/src/cache_xt.h' --- a/storage/pbxt/src/cache_xt.h 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/cache_xt.h 2009-11-24 10:55:06 +0000 @@ -62,7 +62,7 @@ struct XTIdxReadBuffer; #define XT_IPAGE_UNLOCK(i, x) xt_atomicrwlock_unlock(i, x) #elif defined(XT_IPAGE_USE_PTHREAD_RW) #define XT_IPAGE_LOCK_TYPE xt_rwlock_type -#define XT_IPAGE_INIT_LOCK(s, i) xt_init_rwlock(s, i) +#define XT_IPAGE_INIT_LOCK(s, i) xt_init_rwlock_with_autoname(s, i) #define XT_IPAGE_FREE_LOCK(s, i) xt_free_rwlock(i) #define XT_IPAGE_READ_LOCK(i) xt_slock_rwlock_ns(i) #define XT_IPAGE_WRITE_LOCK(i, s) xt_xlock_rwlock_ns(i) === modified file 'storage/pbxt/src/database_xt.cc' --- a/storage/pbxt/src/database_xt.cc 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/database_xt.cc 2009-11-24 10:55:06 +0000 @@ -54,6 +54,8 @@ * GLOBALS */ +xtPublic XTDatabaseHPtr pbxt_database = NULL; // The global open database + xtPublic xtLogOffset xt_db_log_file_threshold; xtPublic size_t xt_db_log_buffer_size; xtPublic size_t xt_db_transaction_buffer_size; @@ -505,6 +507,15 @@ xtPublic XTDatabaseHPtr xt_get_database( * all index entries that are not visible have * been removed. * + * REASON WHY WE SET ROWID ON RECOVERY: + * The row ID is set on recovery because the + * change to the index may be lost after a crash. + * The change to the index is done by the sweeper, and + * there is no record of this change in the log. + * The sweeper will not "re-sweep" all transations + * that are recovered. As a result, this upadte + * of the index by the sweeper may be lost. + * * {OPEN-DB-SWEEPER-WAIT} * This has been moved to after the release of the open * database lock because: @@ -518,9 +529,12 @@ xtPublic XTDatabaseHPtr xt_get_database( * - To open the database it needs the open database * lock. */ + /* + * This has been moved, see: {WAIT-FOR-SW-AFTER-RECOV} pushr_(xt_heap_release, db); xt_wait_for_sweeper(self, db, 0); popr_(); + */ return db; } === modified file 'storage/pbxt/src/database_xt.h' --- a/storage/pbxt/src/database_xt.h 2009-04-02 10:03:14 +0000 +++ b/storage/pbxt/src/database_xt.h 2009-11-24 10:55:06 +0000 @@ -105,6 +105,8 @@ typedef struct XTTablePath { #define XT_THREAD_IDLE 1 #define XT_THREAD_INERR 2 +#define XT_XA_HASH_TAB_SIZE 223 + typedef struct XTDatabase : public XTHeap { char *db_name; /* The name of the database, last component of the path! */ char *db_main_path; @@ -131,6 +133,9 @@ typedef struct XTDatabase : public XTHea u_int db_stat_sweep_waits; /* STATISTICS: count the sweeper waits. */ XTDatabaseLogRec db_xlog; /* The transaction log for this database. */ XTXactRestartRec db_restart; /* Database recovery stuff. */ + xt_mutex_type db_xn_xa_lock; + XTXactPreparePtr db_xn_xa_table[XT_XA_HASH_TAB_SIZE]; + XTSortedListPtr db_xn_xa_list; /* The "wait-for" list, of transactions waiting for other transactions. */ XTSortedListPtr db_xn_wait_for; /* The "wait-for" list, of transactions waiting for other transactions. */ u_int db_xn_call_start; /* Start of the post wait calls. */ @@ -198,6 +203,7 @@ void xt_check_database(XTThreadPtr se void xt_add_pbxt_file(size_t size, char *path, const char *file); void xt_add_location_file(size_t size, char *path); +void xt_add_pbxt_dir(size_t size, char *path); void xt_add_system_dir(size_t size, char *path); void xt_add_data_dir(size_t size, char *path); @@ -244,4 +250,6 @@ inline void xt_xlog_check_long_writer(XT } } +extern XTDatabaseHPtr pbxt_database; // The global open database + #endif === modified file 'storage/pbxt/src/datadic_xt.cc' --- a/storage/pbxt/src/datadic_xt.cc 2009-08-18 07:46:53 +0000 +++ b/storage/pbxt/src/datadic_xt.cc 2009-11-24 10:55:06 +0000 @@ -35,7 +35,7 @@ #ifdef DEBUG #ifdef DRIZZLED -#include <drizzled/common_includes.h> +//#include <drizzled/common_includes.h> #else #include "mysql_priv.h" #endif @@ -437,11 +437,6 @@ class XTTokenizer { XTToken *nextToken(XTThreadPtr self, c_char *keyword, XTToken *tk); }; -void ri_free_token(XTThreadPtr XT_UNUSED(self), XTToken *tk) -{ - delete tk; -} - XTToken *XTTokenizer::newToken(XTThreadPtr self, u_int type, char *start, char *end) { if (!tkn_current) { === modified file 'storage/pbxt/src/datalog_xt.cc' --- a/storage/pbxt/src/datalog_xt.cc 2009-09-04 07:29:34 +0000 +++ b/storage/pbxt/src/datalog_xt.cc 2009-11-24 10:55:06 +0000 @@ -410,7 +410,7 @@ static void dl_recover_log(XTThreadPtr s ASSERT_NS(seq_read.sl_log_eof == seq_read.sl_rec_log_offset); data_log->dlf_log_eof = seq_read.sl_rec_log_offset; - if ((size_t) data_log->dlf_log_eof < sizeof(XTXactLogHeaderDRec)) { + if (data_log->dlf_log_eof < (off_t) sizeof(XTXactLogHeaderDRec)) { data_log->dlf_log_eof = sizeof(XTXactLogHeaderDRec); if (!dl_create_log_header(data_log, seq_read.sl_log_file, self)) xt_throw(self); @@ -1162,7 +1162,7 @@ xtBool XTDataLogBuffer::dlb_close_log(XT /* When I use 'thread' instead of 'self', this means * that I will not throw an error. */ -xtBool XTDataLogBuffer::dlb_get_log_offset(xtLogID *log_id, xtLogOffset *out_offset, size_t req_size, struct XTThread *thread) +xtBool XTDataLogBuffer::dlb_get_log_offset(xtLogID *log_id, xtLogOffset *out_offset, size_t XT_UNUSED(req_size), struct XTThread *thread) { /* Note, I am allowing a log to grow beyond the threshold. * The amount depends on the maximum extended record size. @@ -1757,7 +1757,7 @@ static xtBool dl_collect_garbage(XTThrea freer_(); // xt_unlock_mutex(&db->db_co_dlog_lock) /* Flush the transaction log. */ - if (!xt_xlog_flush_log(self)) + if (!xt_xlog_flush_log(db, self)) xt_throw(self); xt_lock_mutex_ns(&db->db_datalogs.dlc_head_lock); @@ -1891,7 +1891,7 @@ static xtBool dl_collect_garbage(XTThrea freer_(); // xt_unlock_mutex(&db->db_co_dlog_lock) /* Flush the transaction log. */ - if (!xt_xlog_flush_log(self)) + if (!xt_xlog_flush_log(db, self)) xt_throw(self); /* Save state in source log header. */ === modified file 'storage/pbxt/src/discover_xt.cc' --- a/storage/pbxt/src/discover_xt.cc 2009-11-16 20:49:51 +0000 +++ b/storage/pbxt/src/discover_xt.cc 2009-12-02 11:50:40 +0000 @@ -31,6 +31,9 @@ #include <drizzled/session.h> #include <drizzled/server_includes.h> #include <drizzled/sql_base.h> +#include <drizzled/statement/alter_table.h> +#include <algorithm> +#include <sstream> #endif #include "strutil_xt.h" @@ -39,18 +42,273 @@ #include "ha_xtsys.h" #ifndef DRIZZLED -#if MYSQL_VERSION_ID > 60005 +#if MYSQL_VERSION_ID >= 50404 #define DOT_STR(x) x.str #else #define DOT_STR(x) x #endif #endif -#ifndef DRIZZLED +//#ifndef DRIZZLED #define LOCK_OPEN_HACK_REQUIRED -#endif // DRIZZLED +//#endif // DRIZZLED #ifdef LOCK_OPEN_HACK_REQUIRED +#ifdef DRIZZLED + +using namespace drizzled; +using namespace std; + +#define mysql_create_table_no_lock hacked_mysql_create_table_no_lock + +namespace drizzled { + +int rea_create_table(Session *session, const char *path, + const char *db, const char *table_name, + message::Table *table_proto, + HA_CREATE_INFO *create_info, + List<CreateField> &create_field, + uint32_t key_count,KEY *key_info); +} + +static uint32_t build_tmptable_filename(Session* session, + char *buff, size_t bufflen) +{ + uint32_t length; + ostringstream path_str, post_tmpdir_str; + string tmp; + + path_str << drizzle_tmpdir; + post_tmpdir_str << "/" << TMP_FILE_PREFIX << current_pid; + post_tmpdir_str << session->thread_id << session->tmp_table++; + tmp= post_tmpdir_str.str(); + + transform(tmp.begin(), tmp.end(), tmp.begin(), ::tolower); + + path_str << tmp; + + if (bufflen < path_str.str().length()) + length= 0; + else + length= unpack_filename(buff, path_str.str().c_str()); + + return length; +} + +static bool mysql_create_table_no_lock(Session *session, + const char *db, const char *table_name, + HA_CREATE_INFO *create_info, + message::Table *table_proto, + AlterInfo *alter_info, + bool internal_tmp_table, + uint32_t select_field_count) +{ + char path[FN_REFLEN]; + uint32_t path_length; + uint db_options, key_count; + KEY *key_info_buffer; + Cursor *file; + bool error= true; + /* Check for duplicate fields and check type of table to create */ + if (!alter_info->create_list.elements) + { + my_message(ER_TABLE_MUST_HAVE_COLUMNS, ER(ER_TABLE_MUST_HAVE_COLUMNS), + MYF(0)); + return true; + } + assert(strcmp(table_name,table_proto->name().c_str())==0); + if (check_engine(session, table_name, create_info)) + return true; + db_options= create_info->table_options; + if (create_info->row_type == ROW_TYPE_DYNAMIC) + db_options|=HA_OPTION_PACK_RECORD; + + /*if (!(file= create_info->db_type->getCursor((TableShare*) 0, session->mem_root))) + { + my_error(ER_OUTOFMEMORY, MYF(0), sizeof(Cursor)); + return true; + }*/ + + /* PMC - Done to avoid getting the partition handler by mistake! */ + if (!(file= new (session->mem_root) ha_xtsys(pbxt_hton, NULL))) + { + my_error(ER_OUTOFMEMORY, MYF(0), sizeof(Cursor)); + return true; + } + + set_table_default_charset(create_info, (char*) db); + + if (mysql_prepare_create_table(session, + create_info, + table_proto, + alter_info, + internal_tmp_table, + &db_options, file, + &key_info_buffer, &key_count, + select_field_count)) + goto err; + + /* Check if table exists */ + if (create_info->options & HA_LEX_CREATE_TMP_TABLE) + { + path_length= build_tmptable_filename(session, path, sizeof(path)); + } + else + { + #ifdef FN_DEVCHAR + /* check if the table name contains FN_DEVCHAR when defined */ + if (strchr(table_name, FN_DEVCHAR)) + { + my_error(ER_WRONG_TABLE_NAME, MYF(0), table_name); + return true; + } +#endif + path_length= build_table_filename(path, sizeof(path), db, table_name, internal_tmp_table); + } + + /* Check if table already exists */ + if ((create_info->options & HA_LEX_CREATE_TMP_TABLE) && + session->find_temporary_table(db, table_name)) + { + if (create_info->options & HA_LEX_CREATE_IF_NOT_EXISTS) + { + create_info->table_existed= 1; // Mark that table existed + push_warning_printf(session, DRIZZLE_ERROR::WARN_LEVEL_NOTE, + ER_TABLE_EXISTS_ERROR, ER(ER_TABLE_EXISTS_ERROR), + table_name); + error= 0; + goto err; + } + my_error(ER_TABLE_EXISTS_ERROR, MYF(0), table_name); + goto err; + } + + //pthread_mutex_lock(&LOCK_open); /* CREATE TABLE (some confussion on naming, double check) */ + if (!internal_tmp_table && !(create_info->options & HA_LEX_CREATE_TMP_TABLE)) + { + if (plugin::StorageEngine::getTableDefinition(*session, + path, + db, + table_name, + internal_tmp_table) == EEXIST) + { + if (create_info->options & HA_LEX_CREATE_IF_NOT_EXISTS) + { + error= false; + push_warning_printf(session, DRIZZLE_ERROR::WARN_LEVEL_NOTE, + ER_TABLE_EXISTS_ERROR, ER(ER_TABLE_EXISTS_ERROR), + table_name); + create_info->table_existed= 1; // Mark that table existed + } + else + my_error(ER_TABLE_EXISTS_ERROR,MYF(0),table_name); + + goto unlock_and_end; + } + /* + * We don't assert here, but check the result, because the table could be + * in the table definition cache and in the same time the .frm could be + * missing from the disk, in case of manual intervention which deletes + * the .frm file. The user has to use FLUSH TABLES; to clear the cache. + * Then she could create the table. This case is pretty obscure and + * therefore we don't introduce a new error message only for it. + * */ + if (TableShare::getShare(db, table_name)) + { + my_error(ER_TABLE_EXISTS_ERROR, MYF(0), table_name); + goto unlock_and_end; + } + } + /* + * Check that table with given name does not already + * exist in any storage engine. In such a case it should + * be discovered and the error ER_TABLE_EXISTS_ERROR be returned + * unless user specified CREATE TABLE IF EXISTS + * The LOCK_open mutex has been locked to make sure no + * one else is attempting to discover the table. Since + * it's not on disk as a frm file, no one could be using it! + * */ + if (!(create_info->options & HA_LEX_CREATE_TMP_TABLE)) + { + bool create_if_not_exists = + create_info->options & HA_LEX_CREATE_IF_NOT_EXISTS; + + char table_path[FN_REFLEN]; + uint32_t table_path_length; + + table_path_length= build_table_filename(table_path, sizeof(table_path), + db, table_name, false); + + int retcode= plugin::StorageEngine::getTableDefinition(*session, + table_path, + db, + table_name, + false); + switch (retcode) + { + case ENOENT: + /* Normal case, no table exists. we can go and create it */ + break; + case EEXIST: + if (create_if_not_exists) + { + error= false; + push_warning_printf(session, DRIZZLE_ERROR::WARN_LEVEL_NOTE, + ER_TABLE_EXISTS_ERROR, ER(ER_TABLE_EXISTS_ERROR), + table_name); + create_info->table_existed= 1; // Mark that table existed + goto unlock_and_end; + } + my_error(ER_TABLE_EXISTS_ERROR,MYF(0),table_name); + goto unlock_and_end; + default: + my_error(retcode, MYF(0),table_name); + goto unlock_and_end; + } + } + + session->set_proc_info("creating table"); + create_info->table_existed= 0; // Mark that table is created + + create_info->table_options=db_options; + + if (rea_create_table(session, path, db, table_name, + table_proto, + create_info, alter_info->create_list, + key_count, key_info_buffer)) + goto unlock_and_end; + + if (create_info->options & HA_LEX_CREATE_TMP_TABLE) + { + /* Open table and put in temporary table list */ + if (!(session->open_temporary_table(path, db, table_name, 1, OTM_OPEN))) + { + (void) session->rm_temporary_table(create_info->db_type, path); + goto unlock_and_end; + } + } + + /* + * Don't write statement if: + * - It is an internal temporary table, + * - Row-based logging is used and it we are creating a temporary table, or + * - The binary log is not open. + * Otherwise, the statement shall be binlogged. + * */ + if (!internal_tmp_table && + ((!(create_info->options & HA_LEX_CREATE_TMP_TABLE)))) + write_bin_log(session, session->query, session->query_length); + error= false; +unlock_and_end: + //pthread_mutex_unlock(&LOCK_open); + +err: + session->set_proc_info("After create"); + delete file; + return(error); +} + +#else // MySQL case /////////////////////////////// /* * Unfortunately I cannot use the standard mysql_create_table_no_lock() because it will lock "LOCK_open" @@ -1229,13 +1487,13 @@ static bool mysql_create_table_no_lock(T if (create_info->options & HA_LEX_CREATE_TMP_TABLE) { /* Open table and put in temporary table list */ -#if MYSQL_VERSION_ID > 60005 +#if MYSQL_VERSION_ID >= 50404 if (!(open_temporary_table(thd, path, db, table_name, 1, OTM_OPEN))) #else if (!(open_temporary_table(thd, path, db, table_name, 1))) #endif { -#if MYSQL_VERSION_ID > 60005 +#if MYSQL_VERSION_ID >= 50404 (void) rm_temporary_table(create_info->db_type, path, false); #else (void) rm_temporary_table(create_info->db_type, path); @@ -1252,11 +1510,21 @@ static bool mysql_create_table_no_lock(T - The binary log is not open. Otherwise, the statement shall be binlogged. */ + /* PBXT 1.0.09e + * Firstly we had a compile problem with MySQL 5.1.42 and + * the write_bin_log() call below: + * discover_xt.cc:1259: error: argument of type 'char* (Statement::)()' does not match 'const char*' + * + * And secondly, we should no write the BINLOG anyway because this is + * an internal PBXT system table. + * + * So I am just commenting out the code altogether. if (!internal_tmp_table && (!thd->current_stmt_binlog_row_based || (thd->current_stmt_binlog_row_based && !(create_info->options & HA_LEX_CREATE_TMP_TABLE)))) - write_bin_log(thd, TRUE, thd->query(), thd->query_length()); + write_bin_log(thd, TRUE, thd->query, thd->query_length); + */ error= FALSE; unlock_and_end: pthread_mutex_unlock(&LOCK_open); @@ -1279,37 +1547,51 @@ warn: ////// END OF CUT AND PASTES FROM sql_table.cc //////// //////////////////////////////////////////////////////// +#endif // DRIZZLED #endif // LOCK_OPEN_HACK_REQUIRED //------------------------------ int xt_create_table_frm(handlerton *hton, THD* thd, const char *db, const char *name, DT_FIELD_INFO *info, DT_KEY_INFO *XT_UNUSED(keys), xtBool skip_existing) { #ifdef DRIZZLED - drizzled::message::Table table_proto; +#define MYLEX_CREATE_INFO create_info +#else +#define MYLEX_CREATE_INFO mylex.create_info +#endif + +#ifdef DRIZZLED + drizzled::statement::AlterTable *stmt = new drizzled::statement::AlterTable(thd); + HA_CREATE_INFO create_info; + //AlterInfo alter_info; + drizzled::message::Table table_proto; static const char *ext = ".dfe"; static const int ext_len = 4; + + table_proto.mutable_engine()->mutable_name()->assign("PBXT"); #else static const char *ext = ".frm"; static const int ext_len = 4; #endif int err = 1; - //HA_CREATE_INFO create_info = {0}; - //Alter_info alter_info; char field_length_buffer[12], *field_length_ptr; LEX *save_lex= thd->lex, mylex; - - memset(&mylex.create_info, 0, sizeof(HA_CREATE_INFO)); + + memset(&MYLEX_CREATE_INFO, 0, sizeof(HA_CREATE_INFO)); thd->lex = &mylex; - lex_start(thd); + lex_start(thd); +#ifdef DRIZZLED + mylex.statement = stmt; +#endif /* setup the create info */ - mylex.create_info.db_type = hton; + MYLEX_CREATE_INFO.db_type = hton; + #ifndef DRIZZLED mylex.create_info.frm_only = 1; #endif - mylex.create_info.default_table_charset = system_charset_info; + MYLEX_CREATE_INFO.default_table_charset = system_charset_info; /* setup the column info. */ while (info->field_name) { @@ -1335,7 +1617,7 @@ int xt_create_table_frm(handlerton *hton #else if (add_field_to_list(thd, &field_name, info->field_type, field_length_ptr, info->field_decimal_length, info->field_flags, -#if MYSQL_VERSION_ID > 60005 +#if MYSQL_VERSION_ID >= 50404 HA_SM_DISK, COLUMN_FORMAT_TYPE_FIXED, #endif @@ -1369,7 +1651,7 @@ int xt_create_table_frm(handlerton *hton table_proto.set_name(name); table_proto.set_type(drizzled::message::Table::STANDARD); - if (mysql_create_table_no_lock(thd, db, name, &mylex.create_info, &table_proto, &mylex.alter_info, 1, 0, false)) + if (mysql_create_table_no_lock(thd, db, name, &create_info, &table_proto, &stmt->alter_info, 1, 0)) goto error; #else if (mysql_create_table_no_lock(thd, db, name, &mylex.create_info, &mylex.alter_info, 1, 0)) === modified file 'storage/pbxt/src/filesys_xt.cc' --- a/storage/pbxt/src/filesys_xt.cc 2009-09-03 06:15:03 +0000 +++ b/storage/pbxt/src/filesys_xt.cc 2009-11-24 10:55:06 +0000 @@ -56,6 +56,8 @@ //#define DEBUG_TRACE_FILES //#define INJECT_WRITE_REMAP_ERROR /* This is required to make testing on the Mac faster: */ +/* It turns of full file sync. */ +#define DEBUG_FAST_MAC #endif #ifdef DEBUG_TRACE_FILES @@ -63,10 +65,6 @@ #define PRINTF xt_trace #endif -#if defined(XT_MAC) && defined(F_FULLFSYNC) -#undef F_FULLFSYNC -#endif - #ifdef INJECT_WRITE_REMAP_ERROR #define INJECT_REMAP_FILE_SIZE 1000000 #define INJECT_REMAP_FILE_TYPE "xtd" @@ -883,7 +881,7 @@ xtPublic xtBool xt_flush_file(XTOpenFile * fsync didn't really flush index pages to disk. fcntl(F_FULLFSYNC) is considered more effective * in such case. */ -#ifdef F_FULLFSYNC +#if defined(F_FULLFSYNC) && !defined(DEBUG_FAST_MAC) if (fcntl(of->of_filedes, F_FULLFSYNC, 0) == -1) { xt_register_ferrno(XT_REG_CONTEXT, errno, xt_file_path(of)); goto failed; === modified file 'storage/pbxt/src/filesys_xt.h' --- a/storage/pbxt/src/filesys_xt.h 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/filesys_xt.h 2009-11-24 10:55:06 +0000 @@ -102,7 +102,7 @@ xtBool xt_fs_rename(struct XTThread *s #define FILE_MAP_UNLOCK(i, o) xt_xsmutex_unlock(i, o) #elif defined(FILE_MAP_USE_PTHREAD_RW) #define FILE_MAP_LOCK_TYPE xt_rwlock_type -#define FILE_MAP_INIT_LOCK(s, i) xt_init_rwlock(s, i) +#define FILE_MAP_INIT_LOCK(s, i) xt_init_rwlock_with_autoname(s, i) #define FILE_MAP_FREE_LOCK(s, i) xt_free_rwlock(i) #define FILE_MAP_READ_LOCK(i, o) xt_slock_rwlock_ns(i) #define FILE_MAP_WRITE_LOCK(i, o) xt_xlock_rwlock_ns(i) === modified file 'storage/pbxt/src/ha_pbxt.cc' --- a/storage/pbxt/src/ha_pbxt.cc 2009-09-03 06:15:03 +0000 +++ b/storage/pbxt/src/ha_pbxt.cc 2009-11-27 15:37:02 +0000 @@ -35,6 +35,7 @@ #include <stdlib.h> #include <time.h> +#include <ctype.h> #ifdef DRIZZLED #include <drizzled/common.h> @@ -42,14 +43,21 @@ #include <mysys/my_alloc.h> #include <mysys/hash.h> #include <drizzled/field.h> -#include <drizzled/current_session.h> +#include <drizzled/session.h> #include <drizzled/data_home.h> #include <drizzled/error.h> #include <drizzled/table.h> #include <drizzled/field/timestamp.h> #include <drizzled/server_includes.h> +#include <drizzled/plugin/info_schema_table.h> extern "C" char **session_query(Session *session); #define my_strdup(a,b) strdup(a) + +using drizzled::plugin::Registry; +using drizzled::plugin::ColumnInfo; +using drizzled::plugin::InfoSchemaTable; +using drizzled::plugin::InfoSchemaMethods; + #else #include "mysql_priv.h" #include <mysql/plugin.h> @@ -71,7 +79,7 @@ extern "C" char **session_query(Session #include "tabcache_xt.h" #include "systab_xt.h" #include "xaction_xt.h" -#include "restart_xt.h" +#include "backup_xt.h" #ifdef DEBUG //#define XT_USE_SYS_PAR_DEBUG_SIZES @@ -101,6 +109,10 @@ static void pbxt_drop_database(handlert static int pbxt_close_connection(handlerton *hton, THD* thd); static int pbxt_commit(handlerton *hton, THD *thd, bool all); static int pbxt_rollback(handlerton *hton, THD *thd, bool all); +static int pbxt_prepare(handlerton *hton, THD *thd, bool all); +static int pbxt_recover(handlerton *hton, XID *xid_list, uint len); +static int pbxt_commit_by_xid(handlerton *hton, XID *xid); +static int pbxt_rollback_by_xid(handlerton *hton, XID *xid); #endif static void ha_aquire_exclusive_use(XTThreadPtr self, XTSharePtr share, ha_pbxt *mine); static void ha_release_exclusive_use(XTThreadPtr self, XTSharePtr share); @@ -123,7 +135,9 @@ static void ha_close_open_tables(XTThre #ifdef PBXT_HANDLER_TRACE #define PBXT_ALLOW_PRINTING -#define XT_TRACE_CALL() do { XTThreadPtr s = xt_get_self(); printf("%s %s\n", s ? s->t_name : "-unknown-", __FUNC__); } while (0) +#define XT_TRACE_CALL() ha_trace_function(__FUNC__, NULL) +#define XT_TRACE_METHOD() ha_trace_function(__FUNC__, pb_share->sh_table_path->ps_path) + #ifdef PBXT_TRACE_RETURN #define XT_RETURN(x) do { printf("%d\n", (int) (x)); return (x); } while (0) #define XT_RETURN_VOID do { printf("out\n"); return; } while (0) @@ -135,6 +149,7 @@ static void ha_close_open_tables(XTThre #else #define XT_TRACE_CALL() +#define XT_TRACE_METHOD() #define XT_RETURN(x) return (x) #define XT_RETURN_VOID return @@ -165,10 +180,10 @@ xtBool pbxt_crash_debug = TRUE; xtBool pbxt_crash_debug = FALSE; #endif + /* Variables for pbxt share methods */ static xt_mutex_type pbxt_database_mutex; // Prevent a database from being opened while it is being dropped static XTHashTabPtr pbxt_share_tables; // Hash used to track open tables -XTDatabaseHPtr pbxt_database = NULL; // The global open database static char *pbxt_index_cache_size; static char *pbxt_record_cache_size; static char *pbxt_log_cache_size; @@ -180,6 +195,12 @@ static char *pbxt_data_log_threshold; static char *pbxt_data_file_grow_size; static char *pbxt_row_file_grow_size; static int pbxt_max_threads; +static my_bool pbxt_support_xa; + +#ifndef DRIZZLED +// drizzle complains it's not used +static XTXactEnumXARec pbxt_xa_enum; +#endif #ifdef DEBUG #define XT_SHARE_LOCK_WAIT 5000 @@ -259,6 +280,33 @@ static HAVarParamsRec vp_row_file_grow_s //#define XT_AUTO_INCREMENT_DEF 1 #endif +#ifdef PBXT_HANDLER_TRACE +static void ha_trace_function(const char *function, char *table) +{ + char func_buf[50], *ptr; + XTThreadPtr thread = xt_get_self(); + + if ((ptr = strchr(function, '('))) { + ptr--; + while (ptr > function) { + if (!(isalnum(*ptr) || *ptr == '_')) + break; + ptr--; + } + ptr++; + xt_strcpy(50, func_buf, ptr); + if ((ptr = strchr(func_buf, '('))) + *ptr = 0; + } + else + xt_strcpy(50, func_buf, function); + if (table) + printf("%s %s (%s)\n", thread ? thread->t_name : "-unknown-", func_buf, table); + else + printf("%s %s\n", thread ? thread->t_name : "-unknown-", func_buf); +} +#endif + /* * ----------------------------------------------------------------------- * SHARED TABLE DATA @@ -584,6 +632,9 @@ xtPublic XTThreadPtr xt_ha_thd_to_self(T /* The first bit is 1. */ static u_int ha_get_max_bit(MX_BITMAP *map) { +#ifdef DRIZZLED + return map->getFirstSet(); +#else my_bitmap_map *data_ptr = map->bitmap; my_bitmap_map *end_ptr = map->last_word_ptr; my_bitmap_map b; @@ -612,6 +663,7 @@ static u_int ha_get_max_bit(MX_BITMAP *m cnt -= 32; } return 0; +#endif } /* @@ -684,9 +736,10 @@ xtPublic int xt_ha_pbxt_to_mysql_error(i return(-1); // Unknown error } -xtPublic int xt_ha_pbxt_thread_error_for_mysql(THD *XT_UNUSED(thd), const XTThreadPtr self, int ignore_dup_key) +xtPublic int xt_ha_pbxt_thread_error_for_mysql(THD *thd, const XTThreadPtr self, int ignore_dup_key) { - int xt_err = self->t_exception.e_xt_err; + int xt_err = self->t_exception.e_xt_err; + xtBool dup_key = FALSE; XT_PRINT2(self, "xt_ha_pbxt_thread_error_for_mysql xt_err=%d auto commit=%d\n", (int) xt_err, (int) self->st_auto_commit); switch (xt_err) { @@ -725,6 +778,7 @@ xtPublic int xt_ha_pbxt_thread_error_for /* If we are in auto-commit mode (and we are not ignoring * duplicate keys) then rollback the transaction automatically. */ + dup_key = TRUE; if (!ignore_dup_key && self->st_auto_commit) goto abort_transaction; break; @@ -790,26 +844,20 @@ xtPublic int xt_ha_pbxt_thread_error_for /* Locks are held on tables. * Only rollback after locks are released. */ - self->st_auto_commit = TRUE; + /* I do not think this is required, because + * I tell mysql to rollback below, + * besides it is a hack! + self->st_auto_commit = TRUE; + */ self->st_abort_trans = TRUE; } -#ifdef xxxx -/* DBUG_ASSERT(thd->transaction.stmt.ha_list == NULL || - trans == &thd->transaction.stmt); in handler.cc now - * fails, and I don't know if this function can be called anymore! */ - /* Cause any other DBs to do a rollback as well... */ - if (thd) { - /* - * GOTCHA: - * This is a BUG in MySQL. I cannot rollback a transaction if - * pb_mysql_thd->in_sub_stmt! But I must....?! - */ -#ifdef MYSQL_SERVER - if (!thd->in_sub_stmt) - ha_rollback(thd); -#endif + /* Only tell MySQL to rollback if we automatically rollback. + * Note: calling this with (thd, FALSE), cause sp.test to fail. + */ + if (!dup_key) { + if (thd) + thd_mark_transaction_to_rollback(thd, TRUE); } -#endif } break; } @@ -908,7 +956,11 @@ static void pbxt_call_init(XTThreadPtr s xt_db_data_file_grow_size = (size_t) data_file_grow_size; xt_db_row_file_grow_size = (size_t) row_file_grow_size; +#ifdef DRIZZLED + pbxt_ignore_case = TRUE; +#else pbxt_ignore_case = lower_case_table_names != 0; +#endif if (pbxt_ignore_case) pbxt_share_tables = xt_new_hashtable(self, ha_hash_comp_ci, ha_hash_ci, ha_hash_free, TRUE, FALSE); else @@ -968,7 +1020,7 @@ static void pbxt_call_exit(XTThreadPtr s */ static void ha_exit(XTThreadPtr self) { - xt_xres_wait_for_recovery(self); + xt_xres_terminate_recovery(self); /* Wrap things up... */ xt_unuse_database(self, self); /* Just in case the main thread has a database in use (for testing)? */ @@ -1024,7 +1076,7 @@ static bool pbxt_show_status(handlerton cont_(a); if (!not_ok) { - if (stat_print(thd, "PBXT", 4, "", 0, strbuf.sb_cstring, strbuf.sb_len)) + if (stat_print(thd, "PBXT", 4, "", 0, strbuf.sb_cstring, (uint) strbuf.sb_len)) not_ok = TRUE; } xt_sb_set_size(self, &strbuf, 0); @@ -1038,14 +1090,14 @@ static bool pbxt_show_status(handlerton * return 1 on error, else 0. */ #ifdef DRIZZLED -static int pbxt_init(PluginRegistry &registry) +static int pbxt_init(Registry &registry) #else static int pbxt_init(void *p) #endif { int init_err = 0; - XT_TRACE_CALL(); + XT_PRINT0(NULL, "pbxt_init\n"); if (sizeof(xtWordPS) != sizeof(void *)) { printf("PBXT: This won't work, I require that sizeof(xtWordPS) == sizeof(void *)!\n"); @@ -1076,11 +1128,27 @@ static int pbxt_init(void *p) pbxt_hton->close_connection = pbxt_close_connection; /* close_connection, cleanup thread related data. */ pbxt_hton->commit = pbxt_commit; /* commit */ pbxt_hton->rollback = pbxt_rollback; /* rollback */ + if (pbxt_support_xa) { + pbxt_hton->prepare = pbxt_prepare; + pbxt_hton->recover = pbxt_recover; + pbxt_hton->commit_by_xid = pbxt_commit_by_xid; + pbxt_hton->rollback_by_xid = pbxt_rollback_by_xid; + } + else { + pbxt_hton->prepare = NULL; + pbxt_hton->recover = NULL; + pbxt_hton->commit_by_xid = NULL; + pbxt_hton->rollback_by_xid = NULL; + } pbxt_hton->create = pbxt_create_handler; /* Create a new handler */ pbxt_hton->drop_database = pbxt_drop_database; /* Drop a database */ pbxt_hton->panic = pbxt_panic; /* Panic call */ pbxt_hton->show_status = pbxt_show_status; pbxt_hton->flags = HTON_NO_FLAGS; /* HTON_CAN_RECREATE - Without this flags TRUNCATE uses delete_all_rows() */ + pbxt_hton->slot = (uint)-1; /* assign invald value, so we know when it's inited later */ +#if defined(MYSQL_SUPPORTS_BACKUP) && defined(XT_ENABLE_ONLINE_BACKUP) + pbxt_hton->get_backup_engine = pbxt_backup_engine; +#endif #endif if (!xt_init_logging()) /* Initialize logging */ goto error_1; @@ -1160,8 +1228,10 @@ static int pbxt_init(void *p) * Only real problem, 2 threads try to load the same * plugin at the same time. */ +#if MYSQL_VERSION_ID < 60014 myxt_mutex_unlock(&LOCK_plugin); #endif +#endif /* Can't do this here yet, because I need a THD! */ try_(b) { @@ -1195,8 +1265,10 @@ static int pbxt_init(void *p) if (thd) myxt_destroy_thread(thd, FALSE); #ifndef DRIZZLED +#if MYSQL_VERSION_ID < 60014 myxt_mutex_lock(&LOCK_plugin); #endif +#endif } #endif } @@ -1262,7 +1334,7 @@ static int pbxt_init(void *p) } #ifdef DRIZZLED -static int pbxt_end(PluginRegistry &registry) +static int pbxt_end(Registry &registry) #else static int pbxt_end(void *) #endif @@ -1378,7 +1450,7 @@ static int pbxt_commit(handlerton *hton, * transaction (!all && !self->st_auto_commit). */ if (all || self->st_auto_commit) { - XT_PRINT0(self, "xt_xn_commit\n"); + XT_PRINT0(self, "xt_xn_commit in pbxt_commit\n"); if (!xt_xn_commit(self)) err = xt_ha_pbxt_thread_error_for_mysql(thd, self, FALSE); @@ -1402,7 +1474,7 @@ static int pbxt_rollback(handlerton *hto XTThreadPtr self; if ((self = (XTThreadPtr) *thd_ha_data(thd, hton))) { - XT_PRINT1(self, "pbxt_rollback all=%d\n", all); + XT_PRINT1(self, "pbxt_rollback all=%d in pbxt_commit\n", all); if (self->st_xact_data) { /* There are no table locks, rollback immediately in all cases @@ -1431,7 +1503,7 @@ static int pbxt_rollback(handlerton *hto } #ifdef DRIZZLED -handler *PBXTStorageEngine::create(TABLE_SHARE *table, MEM_ROOT *mem_root) +Cursor *PBXTStorageEngine::create(TABLE_SHARE *table, MEM_ROOT *mem_root) { PBXTStorageEngine * const hton = this; #else @@ -1446,6 +1518,182 @@ static handler *pbxt_create_handler(hand /* * ----------------------------------------------------------------------- + * 2-PHASE COMMIT + * + */ + +#ifndef DRIZZLED + +static int pbxt_prepare(handlerton *hton, THD *thd, bool all) +{ + int err = 0; + XTThreadPtr self; + + XT_TRACE_CALL(); + if ((self = (XTThreadPtr) *thd_ha_data(thd, hton))) { + XT_PRINT1(self, "pbxt_commit all=%d\n", all); + + if (self->st_xact_data) { + /* There are no table locks, commit immediately in all cases + * except when this is a statement commit with an explicit + * transaction (!all && !self->st_auto_commit). + */ + if (all) { + XID xid; + + XT_PRINT0(self, "xt_xn_prepare in pbxt_prepare\n"); + thd_get_xid(thd, (MYSQL_XID*) &xid); + + if (!xt_xn_prepare(xid.length(), (xtWord1 *) &xid, self)) + err = xt_ha_pbxt_thread_error_for_mysql(thd, self, FALSE); + } + } + } + return err; +} + +static XTThreadPtr ha_temp_open_global_database(handlerton *hton, THD **ret_thd, int *temp_thread, char *thread_name, int *err) +{ + THD *thd; + XTThreadPtr self = NULL; + + *temp_thread = 0; + if ((thd = current_thd)) + self = (XTThreadPtr) *thd_ha_data(thd, hton); + else { + //thd = (THD *) myxt_create_thread(); + //*temp_thread |= 2; + } + + if (!self) { + XTExceptionRec e; + + if (!(self = xt_create_thread(thread_name, FALSE, TRUE, &e))) { + *err = xt_ha_pbxt_to_mysql_error(e.e_xt_err); + xt_log_exception(NULL, &e, XT_LOG_DEFAULT); + return NULL; + } + *temp_thread |= 1; + } + + xt_xres_wait_for_recovery(self, XT_RECOVER_DONE); + + try_(a) { + xt_open_database(self, mysql_real_data_home, TRUE); + } + catch_(a) { + *err = xt_ha_pbxt_thread_error_for_mysql(thd, self, FALSE); + if ((*temp_thread & 1)) + xt_free_thread(self); + if (*temp_thread & 2) + myxt_destroy_thread(thd, FALSE); + self = NULL; + } + cont_(a); + + *ret_thd = thd; + return self; +} + +static void ha_temp_close_database(XTThreadPtr self, THD *thd, int temp_thread) +{ + xt_unuse_database(self, self); + if (temp_thread & 1) + xt_free_thread(self); + if (temp_thread & 2) + myxt_destroy_thread(thd, TRUE); +} + +/* Return all prepared transactions, found during recovery. + * This function returns a count. If len is returned, the + * function will be called again. + */ +static int pbxt_recover(handlerton *hton, XID *xid_list, uint len) +{ + xtBool temp_thread; + XTThreadPtr self; + XTDatabaseHPtr db; + uint count = 0; + XTXactPreparePtr xap; + int err; + THD *thd; + + if (!(self = ha_temp_open_global_database(hton, &thd, &temp_thread, "TempForRecover", &err))) + return 0; + + db = self->st_database; + + for (count=0; count<len; count++) { + xap = xt_xn_enum_xa_data(db, &pbxt_xa_enum); + if (!xap) + break; + memcpy(&xid_list[count], xap->xp_xa_data, xap->xp_data_len); + } + + ha_temp_close_database(self, thd, temp_thread); + return (int) count; +} + +static int pbxt_commit_by_xid(handlerton *hton, XID *xid) +{ + xtBool temp_thread; + XTThreadPtr self; + XTDatabaseHPtr db; + int err = 0; + XTXactPreparePtr xap; + THD *thd; + + XT_TRACE_CALL(); + + if (!(self = ha_temp_open_global_database(hton, &thd, &temp_thread, "TempForCommitXA", &err))) + return err; + db = self->st_database; + + if ((xap = xt_xn_find_xa_data(db, xid->length(), (xtWord1 *) xid, TRUE, self))) { + if ((self->st_xact_data = xt_xn_get_xact(db, xap->xp_xact_id, self))) { + self->st_xact_data->xd_flags &= ~XT_XN_XAC_PREPARED; // Prepared transactions cannot be swept! + if (!xt_xn_commit(self)) + err = xt_ha_pbxt_thread_error_for_mysql(thd, self, FALSE); + } + xt_xn_delete_xa_data(db, xap, TRUE, self); + } + + ha_temp_close_database(self, thd, temp_thread); + return 0; +} + +static int pbxt_rollback_by_xid(handlerton *hton, XID *xid) +{ + int temp_thread; + XTThreadPtr self; + XTDatabaseHPtr db; + int err = 0; + XTXactPreparePtr xap; + THD *thd; + + XT_TRACE_CALL(); + + if (!(self = ha_temp_open_global_database(hton, &thd, &temp_thread, "TempForRollbackXA", &err))) + return err; + db = self->st_database; + + if ((xap = xt_xn_find_xa_data(db, xid->length(), (xtWord1 *) xid, TRUE, self))) { + if ((self->st_xact_data = xt_xn_get_xact(db, xap->xp_xact_id, self))) { + self->st_xact_data->xd_flags &= ~XT_XN_XAC_PREPARED; // Prepared transactions cannot be swept! + if (!xt_xn_rollback(self)) + err = xt_ha_pbxt_thread_error_for_mysql(thd, self, FALSE); + } + xt_xn_delete_xa_data(db, xap, TRUE, self); + } + + ha_temp_close_database(self, thd, temp_thread); + return 0; +} + +#endif + +/* + * ----------------------------------------------------------------------- * HANDLER LOCKING FUNCTIONS * * These functions are used get a lock on all handles of a particular table. @@ -1497,7 +1745,7 @@ static void ha_aquire_exclusive_use(XTTh ha_pbxt *handler; time_t end_time = time(NULL) + XT_SHARE_LOCK_TIMEOUT / 1000; - XT_PRINT1(self, "ha_aquire_exclusive_use %s PBXT X lock\n", share->sh_table_path->ps_path); + XT_PRINT1(self, "ha_aquire_exclusive_use (%s) PBXT X lock\n", share->sh_table_path->ps_path); /* GOTCHA: It is possible to hang here, if you hold * onto the sh_ex_mutex lock, before we really * have the exclusive lock (i.e. before all @@ -1578,7 +1826,7 @@ static void ha_release_exclusive_use(XTT static void ha_release_exclusive_use(XTThreadPtr XT_UNUSED(self), XTSharePtr share) #endif { - XT_PRINT1(self, "ha_release_exclusive_use %s PBXT X UNLOCK\n", share->sh_table_path->ps_path); + XT_PRINT1(self, "ha_release_exclusive_use (%s) PBXT X UNLOCK\n", share->sh_table_path->ps_path); xt_lock_mutex_ns((xt_mutex_type *) share->sh_ex_mutex); share->sh_table_lock = FALSE; xt_broadcast_cond_ns((xt_cond_type *) share->sh_ex_cond); @@ -1589,7 +1837,7 @@ static xtBool ha_wait_for_shared_use(ha_ { time_t end_time = time(NULL) + XT_SHARE_LOCK_TIMEOUT / 1000; - XT_PRINT1(xt_get_self(), "ha_wait_for_shared_use %s share lock wait...\n", share->sh_table_path->ps_path); + XT_PRINT1(xt_get_self(), "ha_wait_for_shared_use (%s) share lock wait...\n", share->sh_table_path->ps_path); mine->pb_ex_in_use = 0; xt_lock_mutex_ns((xt_mutex_type *) share->sh_ex_mutex); while (share->sh_table_lock) { @@ -1667,14 +1915,38 @@ xtPublic int ha_pbxt::reopen() * */ -int pbxt_statistics_fill_table(THD *thd, TABLE_LIST *tables, COND *cond) +static int pbxt_statistics_fill_table(THD *thd, TABLE_LIST *tables, COND *cond) { - XTThreadPtr self; + XTThreadPtr self = NULL; int err = 0; + if (!pbxt_hton) { + /* Can't do if PBXT is not loaded! */ + XTExceptionRec e; + + xt_exception_xterr(&e, XT_CONTEXT, XT_ERR_PBXT_NOT_INSTALLED); + xt_log_exception(NULL, &e, XT_LOG_DEFAULT); + /* Just return an empty set: */ + return 0; + } + if (!(self = ha_set_current_thread(thd, &err))) return xt_ha_pbxt_to_mysql_error(err); + + try_(a) { + /* If the thread has no open database, and the global + * database is already open, then open + * the database. Otherwise the statement will be + * executed without an open database, which means + * that the related statistics will be missing. + * + * This includes all background threads. + */ + if (!self->st_database && pbxt_database) { + xt_ha_open_database_of_table(self, (XTPathStrPtr) NULL); + } + err = myxt_statistics_fill_table(self, thd, tables, cond, system_charset_info); } catch_(a) { @@ -1684,6 +1956,24 @@ int pbxt_statistics_fill_table(THD *thd, return err; } +#ifdef DRIZZLED +ColumnInfo pbxt_statistics_fields_info[]= +{ + ColumnInfo("ID", 4, MYSQL_TYPE_LONG, 0, 0, "The ID of the statistic", SKIP_OPEN_TABLE), + ColumnInfo("Name", 40, MYSQL_TYPE_STRING, 0, 0, "The name of the statistic", SKIP_OPEN_TABLE), + ColumnInfo("Value", 8, MYSQL_TYPE_LONGLONG, 0, 0, "The accumulated value", SKIP_OPEN_TABLE), + ColumnInfo() +}; + +class PBXTStatisticsMethods : public InfoSchemaMethods +{ +public: + int fillTable(Session *session, TableList *tables, COND *cond) + { + return pbxt_statistics_fill_table(session, tables, cond); + } +}; +#else ST_FIELD_INFO pbxt_statistics_fields_info[]= { { "ID", 4, MYSQL_TYPE_LONG, 0, 0, "The ID of the statistic", SKIP_OPEN_TABLE}, @@ -1691,24 +1981,28 @@ ST_FIELD_INFO pbxt_statistics_fields_inf { "Value", 8, MYSQL_TYPE_LONGLONG, 0, 0, "The accumulated value", SKIP_OPEN_TABLE}, { 0, 0, MYSQL_TYPE_STRING, 0, 0, 0, SKIP_OPEN_TABLE} }; +#endif #ifdef DRIZZLED static InfoSchemaTable *pbxt_statistics_table; - -int pbxt_init_statitics(PluginRegistry &registry) +static PBXTStatisticsMethods pbxt_statistics_methods; +static int pbxt_init_statistics(Registry &registry) #else -int pbxt_init_statitics(void *p) +static int pbxt_init_statistics(void *p) #endif { #ifdef DRIZZLED - pbxt_statistics_table = (InfoSchemaTable *)xt_calloc_ns(sizeof(InfoSchemaTable)); - pbxt_statistics_table->table_name= "PBXT_STATISTICS"; + //pbxt_statistics_table = (InfoSchemaTable *)xt_calloc_ns(sizeof(InfoSchemaTable)); + //pbxt_statistics_table->table_name= "PBXT_STATISTICS"; + pbxt_statistics_table = new InfoSchemaTable("PBXT_STATISTICS"); + pbxt_statistics_table->setColumnInfo(pbxt_statistics_fields_info); + pbxt_statistics_table->setInfoSchemaMethods(&pbxt_statistics_methods); registry.add(pbxt_statistics_table); #else ST_SCHEMA_TABLE *pbxt_statistics_table = (ST_SCHEMA_TABLE *) p; -#endif pbxt_statistics_table->fields_info = pbxt_statistics_fields_info; pbxt_statistics_table->fill_table = pbxt_statistics_fill_table; +#endif #if defined(XT_WIN) && defined(XT_COREDUMP) void register_crash_filter(); @@ -1721,14 +2015,14 @@ int pbxt_init_statitics(void *p) } #ifdef DRIZZLED -int pbxt_exit_statitics(PluginRegistry &registry) +static int pbxt_exit_statistics(Registry &registry) #else -int pbxt_exit_statitics(void *XT_UNUSED(p)) +static int pbxt_exit_statistics(void *XT_UNUSED(p)) #endif { #ifdef DRIZZLED registry.remove(pbxt_statistics_table); - xt_free_ns(pbxt_statistics_table); + delete pbxt_statistics_table; #endif return(0); } @@ -1758,7 +2052,11 @@ ha_pbxt::ha_pbxt(handlerton *hton, TABLE * exist for the storage engine. This is also used by the default rename_table and * delete_table method in handler.cc. */ +#ifdef DRIZZLED +const char **PBXTStorageEngine::bas_ext() const +#else const char **ha_pbxt::bas_ext() const +#endif { return pbxt_extensions; } @@ -1800,11 +2098,13 @@ MX_TABLE_TYPES_T ha_pbxt::table_flags() * purposes! HA_NOT_EXACT_COUNT | */ +#ifndef DRIZZLED /* * This basically means we have a file with the name of * database table (which we do). */ HA_FILE_BASED | +#endif /* * Not sure what this does (but MyISAM and InnoDB have it)?! * Could it mean that we support the handler functions. @@ -1971,7 +2271,7 @@ int ha_pbxt::open(const char *table_path if (!(self = ha_set_current_thread(thd, &err))) return xt_ha_pbxt_to_mysql_error(err); - XT_PRINT1(self, "ha_pbxt::open %s\n", table_path); + XT_PRINT1(self, "open (%s)\n", table_path); pb_ex_in_use = 1; try_(a) { @@ -2049,7 +2349,7 @@ int ha_pbxt::close(void) } } - XT_PRINT1(self, "ha_pbxt::close %s\n", pb_share && pb_share->sh_table_path->ps_path ? pb_share->sh_table_path->ps_path : "unknown"); + XT_PRINT1(self, "close (%s)\n", pb_share && pb_share->sh_table_path->ps_path ? pb_share->sh_table_path->ps_path : "unknown"); if (self) { try_(a) { @@ -2125,9 +2425,10 @@ void ha_pbxt::init_auto_increment(xtWord if (!TS(table)->next_number_key_offset) { // Autoincrement at key-start err = index_last(table->record[1]); - if (!err) + if (!err && !table->next_number_field->is_null(TS(table)->rec_buff_length)) { /* {PRE-INC} */ nr = (xtWord8) table->next_number_field->val_int_offset(TS(table)->rec_buff_length); + } } else { /* Do an index scan to find the largest value! */ @@ -2180,8 +2481,10 @@ void ha_pbxt::init_auto_increment(xtWord table->next_number_field = tmp_fie; table->in_use = tmp_thd; - if (xn_started) + if (xn_started) { + XT_PRINT0(self, "xt_xn_commit in init_auto_increment\n"); xt_xn_commit(self); + } } xt_spinlock_unlock(&tab->tab_ainc_lock); } @@ -2228,13 +2531,13 @@ void ha_pbxt::get_auto_increment(MX_ULON * insert into t1 values (-1); * insert into t1 values (NULL); */ -void ha_pbxt::set_auto_increment(Field *nr) +xtPublic void ha_set_auto_increment(XTOpenTablePtr ot, Field *nr) { register XTTableHPtr tab; MX_ULONGLONG_T nr_int_val; nr_int_val = nr->val_int(); - tab = pb_open_tab->ot_table; + tab = ot->ot_table; if (nr->cmp((const unsigned char *)&tab->tab_auto_inc) > 0) { xt_spinlock_lock(&tab->tab_ainc_lock); @@ -2260,9 +2563,9 @@ void ha_pbxt::set_auto_increment(Field * #else tab->tab_dic.dic_min_auto_inc = nr_int_val + 100; #endif - pb_open_tab->ot_thread = xt_get_self(); - if (!xt_tab_write_min_auto_inc(pb_open_tab)) - xt_log_and_clear_exception(pb_open_tab->ot_thread); + ot->ot_thread = xt_get_self(); + if (!xt_tab_write_min_auto_inc(ot)) + xt_log_and_clear_exception(ot->ot_thread); } } } @@ -2305,7 +2608,7 @@ int ha_pbxt::write_row(byte *buf) ASSERT_NS(pb_ex_in_use); - XT_PRINT1(pb_open_tab->ot_thread, "ha_pbxt::write_row %s\n", pb_share->sh_table_path->ps_path); + XT_PRINT1(pb_open_tab->ot_thread, "write_row (%s)\n", pb_share->sh_table_path->ps_path); XT_DISABLED_TRACE(("INSERT tx=%d val=%d\n", (int) pb_open_tab->ot_thread->st_xact_data->xd_start_xn_id, (int) XT_GET_DISK_4(&buf[1]))); //statistic_increment(ha_write_count,&LOCK_status); #ifdef PBMS_ENABLED @@ -2350,7 +2653,7 @@ int ha_pbxt::write_row(byte *buf) err = update_err; goto done; } - set_auto_increment(table->next_number_field); + ha_set_auto_increment(pb_open_tab, table->next_number_field); } if (!xt_tab_new_record(pb_open_tab, (xtWord1 *) buf)) { @@ -2423,7 +2726,7 @@ int ha_pbxt::update_row(const byte * old ASSERT_NS(pb_ex_in_use); - XT_PRINT1(self, "ha_pbxt::update_row %s\n", pb_share->sh_table_path->ps_path); + XT_PRINT1(self, "update_row (%s)\n", pb_share->sh_table_path->ps_path); XT_DISABLED_TRACE(("UPDATE tx=%d val=%d\n", (int) self->st_xact_data->xd_start_xn_id, (int) XT_GET_DISK_4(&new_data[1]))); //statistic_increment(ha_update_count,&LOCK_status); @@ -2472,7 +2775,7 @@ int ha_pbxt::update_row(const byte * old old_map = mx_tmp_use_all_columns(table, table->read_set); nr = table->found_next_number_field->val_int(); - set_auto_increment(table->found_next_number_field); + ha_set_auto_increment(pb_open_tab, table->found_next_number_field); mx_tmp_restore_column_map(table, old_map); } @@ -2504,7 +2807,7 @@ int ha_pbxt::delete_row(const byte * buf ASSERT_NS(pb_ex_in_use); - XT_PRINT1(pb_open_tab->ot_thread, "ha_pbxt::delete_row %s\n", pb_share->sh_table_path->ps_path); + XT_PRINT1(pb_open_tab->ot_thread, "delete_row (%s)\n", pb_share->sh_table_path->ps_path); XT_DISABLED_TRACE(("DELETE tx=%d val=%d\n", (int) pb_open_tab->ot_thread->st_xact_data->xd_start_xn_id, (int) XT_GET_DISK_4(&buf[1]))); //statistic_increment(ha_delete_count,&LOCK_status); @@ -2829,7 +3132,8 @@ int ha_pbxt::xt_index_prev_read(XTOpenTa int ha_pbxt::index_init(uint idx, bool XT_UNUSED(sorted)) { - XTIndexPtr ind; + XTIndexPtr ind; + XTThreadPtr thread = pb_open_tab->ot_thread; /* select count(*) from smalltab_PBXT; * ignores the error below, and continues to @@ -2851,6 +3155,15 @@ int ha_pbxt::index_init(uint idx, bool X printf("index_init %s index %d cols req=%d/%d read_bits=%X write_bits=%X index_bits=%X\n", pb_open_tab->ot_table->tab_name->ps_path, (int) idx, pb_open_tab->ot_cols_req, pb_open_tab->ot_cols_req, (int) *table->read_set->bitmap, (int) *table->write_set->bitmap, (int) *ind->mi_col_map.bitmap); #endif + /* Start a statement based transaction as soon + * as a read is done for a modify type statement! + * Previously, this was done too late! + */ + if (!thread->st_stat_trans) { + trans_register_ha(pb_mysql_thd, FALSE, pbxt_hton); + XT_PRINT0(thread, "ha_pbxt::update_row trans_register_ha all=FALSE\n"); + thread->st_stat_trans = TRUE; + } } else { pb_open_tab->ot_cols_req = ha_get_max_bit(table->read_set); @@ -2901,7 +3214,7 @@ int ha_pbxt::index_init(uint idx, bool X #endif } - xt_xlog_check_long_writer(pb_open_tab->ot_thread); + xt_xlog_check_long_writer(thread); pb_open_tab->ot_thread->st_statistics.st_scan_index++; return 0; @@ -2911,7 +3224,7 @@ int ha_pbxt::index_end() { int err = 0; - XT_TRACE_CALL(); + XT_TRACE_METHOD(); XTThreadPtr thread = pb_open_tab->ot_thread; @@ -2991,7 +3304,7 @@ int ha_pbxt::index_read_xt(byte * buf, u ASSERT_NS(pb_ex_in_use); - XT_PRINT1(pb_open_tab->ot_thread, "ha_pbxt::index_read_xt %s\n", pb_share->sh_table_path->ps_path); + XT_PRINT1(pb_open_tab->ot_thread, "index_read_xt (%s)\n", pb_share->sh_table_path->ps_path); XT_DISABLED_TRACE(("search tx=%d val=%d update=%d\n", (int) pb_open_tab->ot_thread->st_xact_data->xd_start_xn_id, (int) XT_GET_DISK_4(key), pb_modified)); ind = (XTIndexPtr) pb_share->sh_dic_keys[idx]; @@ -3072,7 +3385,7 @@ int ha_pbxt::index_next(byte * buf) int err = 0; XTIndexPtr ind; - XT_TRACE_CALL(); + XT_TRACE_METHOD(); //statistic_increment(ha_read_next_count,&LOCK_status); ASSERT_NS(pb_ex_in_use); @@ -3114,7 +3427,7 @@ int ha_pbxt::index_next_same(byte * buf, XTIndexPtr ind; XTIdxSearchKeyRec search_key; - XT_TRACE_CALL(); + XT_TRACE_METHOD(); //statistic_increment(ha_read_next_count,&LOCK_status); ASSERT_NS(pb_ex_in_use); @@ -3154,7 +3467,7 @@ int ha_pbxt::index_prev(byte * buf) int err = 0; XTIndexPtr ind; - XT_TRACE_CALL(); + XT_TRACE_METHOD(); //statistic_increment(ha_read_prev_count,&LOCK_status); ASSERT_NS(pb_ex_in_use); @@ -3188,7 +3501,7 @@ int ha_pbxt::index_first(byte * buf) XTIndexPtr ind; XTIdxSearchKeyRec search_key; - XT_TRACE_CALL(); + XT_TRACE_METHOD(); //statistic_increment(ha_read_first_count,&LOCK_status); ASSERT_NS(pb_ex_in_use); @@ -3228,7 +3541,7 @@ int ha_pbxt::index_last(byte * buf) XTIndexPtr ind; XTIdxSearchKeyRec search_key; - XT_TRACE_CALL(); + XT_TRACE_METHOD(); //statistic_increment(ha_read_last_count,&LOCK_status); ASSERT_NS(pb_ex_in_use); @@ -3275,10 +3588,11 @@ int ha_pbxt::index_last(byte * buf) */ int ha_pbxt::rnd_init(bool scan) { - int err = 0; + int err = 0; + XTThreadPtr thread = pb_open_tab->ot_thread; - XT_PRINT1(pb_open_tab->ot_thread, "ha_pbxt::rnd_init %s\n", pb_share->sh_table_path->ps_path); - XT_DISABLED_TRACE(("seq scan tx=%d\n", (int) pb_open_tab->ot_thread->st_xact_data->xd_start_xn_id)); + XT_PRINT1(thread, "rnd_init (%s)\n", pb_share->sh_table_path->ps_path); + XT_DISABLED_TRACE(("seq scan tx=%d\n", (int) thread->st_xact_data->xd_start_xn_id)); /* Call xt_tab_seq_exit() to make sure the resources used by the previous * scan are freed. In particular make sure cache page ref count is decremented. @@ -3296,8 +3610,18 @@ int ha_pbxt::rnd_init(bool scan) xt_tab_seq_exit(pb_open_tab); /* The number of columns required: */ - if (pb_open_tab->ot_is_modify) + if (pb_open_tab->ot_is_modify) { pb_open_tab->ot_cols_req = table->read_set->MX_BIT_SIZE(); + /* Start a statement based transaction as soon + * as a read is done for a modify type statement! + * Previously, this was done too late! + */ + if (!thread->st_stat_trans) { + trans_register_ha(pb_mysql_thd, FALSE, pbxt_hton); + XT_PRINT0(thread, "ha_pbxt::update_row trans_register_ha all=FALSE\n"); + thread->st_stat_trans = TRUE; + } + } else { pb_open_tab->ot_cols_req = ha_get_max_bit(table->read_set); @@ -3322,14 +3646,14 @@ int ha_pbxt::rnd_init(bool scan) else xt_tab_seq_reset(pb_open_tab); - xt_xlog_check_long_writer(pb_open_tab->ot_thread); + xt_xlog_check_long_writer(thread); return err; } int ha_pbxt::rnd_end() { - XT_TRACE_CALL(); + XT_TRACE_METHOD(); /* * make permanent the lock for the last scanned row @@ -3358,7 +3682,7 @@ int ha_pbxt::rnd_next(byte *buf) int err = 0; xtBool eof; - XT_TRACE_CALL(); + XT_TRACE_METHOD(); ASSERT_NS(pb_ex_in_use); //statistic_increment(ha_read_rnd_next_count, &LOCK_status); xt_xlog_check_long_writer(pb_open_tab->ot_thread); @@ -3393,7 +3717,7 @@ int ha_pbxt::rnd_next(byte *buf) */ void ha_pbxt::position(const byte *XT_UNUSED(record)) { - XT_TRACE_CALL(); + XT_TRACE_METHOD(); ASSERT_NS(pb_ex_in_use); /* * I changed this from using little endian to big endian. @@ -3429,10 +3753,10 @@ int ha_pbxt::rnd_pos(byte * buf, byte *p { int err = 0; - XT_TRACE_CALL(); + XT_TRACE_METHOD(); ASSERT_NS(pb_ex_in_use); //statistic_increment(ha_read_rnd_count, &LOCK_status); - XT_PRINT1(pb_open_tab->ot_thread, "ha_pbxt::rnd_pos %s\n", pb_share->sh_table_path->ps_path); + XT_PRINT1(pb_open_tab->ot_thread, "rnd_pos (%s)\n", pb_share->sh_table_path->ps_path); pb_open_tab->ot_curr_rec_id = mi_uint4korr((xtWord1 *) pos); switch (xt_tab_dirty_read_record(pb_open_tab, (xtWord1 *) buf)) { @@ -3509,7 +3833,7 @@ int ha_pbxt::info(uint flag) XTOpenTablePtr ot; int in_use; - XT_TRACE_CALL(); + XT_TRACE_METHOD(); if (!(in_use = pb_ex_in_use)) { pb_ex_in_use = 1; @@ -3535,7 +3859,7 @@ int ha_pbxt::info(uint flag) stats.index_file_length = xt_ind_node_to_offset(ot->ot_table, ot->ot_table->tab_ind_eof); stats.delete_length = ot->ot_table->tab_rec_fnum * ot->ot_rec_size; //check_time = info.check_time; - stats.mean_rec_length = ot->ot_rec_size; + stats.mean_rec_length = (ulong) ot->ot_rec_size; } if (flag & HA_STATUS_CONST) { @@ -3551,7 +3875,9 @@ int ha_pbxt::info(uint flag) stats.block_size = XT_INDEX_PAGE_SIZE; if (share->tmp_table == NO_TMP_TABLE) -#if MYSQL_VERSION_ID > 60005 +#ifdef DRIZZLED +#define WHICH_MUTEX mutex +#elif MYSQL_VERSION_ID >= 50404 #define WHICH_MUTEX LOCK_ha_data #else #define WHICH_MUTEX mutex @@ -3559,19 +3885,15 @@ int ha_pbxt::info(uint flag) #ifdef SAFE_MUTEX -#if MYSQL_VERSION_ID < 60000 +#if MYSQL_VERSION_ID < 50404 #if MYSQL_VERSION_ID < 50123 safe_mutex_lock(&share->mutex,__FILE__,__LINE__); #else safe_mutex_lock(&share->mutex,0,__FILE__,__LINE__); #endif #else -#if MYSQL_VERSION_ID < 60004 - safe_mutex_lock(&share->mutex,__FILE__,__LINE__); -#else safe_mutex_lock(&share->WHICH_MUTEX,0,__FILE__,__LINE__); #endif -#endif #else // SAFE_MUTEX @@ -3675,7 +3997,7 @@ int ha_pbxt::extra(enum ha_extra_functio { int err = 0; - XT_PRINT2(xt_get_self(), "ha_pbxt::extra %s operation=%d\n", pb_share->sh_table_path->ps_path, operation); + XT_PRINT2(xt_get_self(), "ha_pbxt::extra (%s) operation=%d\n", pb_share->sh_table_path->ps_path, operation); switch (operation) { case HA_EXTRA_RESET_STATE: @@ -3771,14 +4093,14 @@ int ha_pbxt::extra(enum ha_extra_functio */ int ha_pbxt::reset(void) { - XT_TRACE_CALL(); + XT_TRACE_METHOD(); extra(HA_EXTRA_RESET_STATE); XT_RETURN(0); } void ha_pbxt::unlock_row() { - XT_TRACE_CALL(); + XT_TRACE_METHOD(); if (pb_open_tab) pb_open_tab->ot_table->tab_locks.xt_remove_temp_lock(pb_open_tab, FALSE); } @@ -3802,7 +4124,7 @@ int ha_pbxt::delete_all_rows() XTDDTable *tab_def = NULL; char path[PATH_MAX]; - XT_TRACE_CALL(); + XT_TRACE_METHOD(); if (thd_sql_command(thd) != SQLCOM_TRUNCATE) { /* Just like InnoDB we only handle TRUNCATE TABLE @@ -3906,7 +4228,7 @@ int ha_pbxt::analyze(THD *thd, HA_CHECK_ xtXactID clean_xn_id = 0; uint cnt = 10; - XT_TRACE_CALL(); + XT_TRACE_METHOD(); if (!pb_open_tab) { if ((err = reopen())) @@ -4056,7 +4378,7 @@ xtPublic int ha_pbxt::external_lock(THD ASSERT_NS(pb_ex_in_use); */ - XT_PRINT1(self, "ha_pbxt::EXTERNAL_LOCK %s lock_type=UNLOCK\n", pb_share->sh_table_path->ps_path); + XT_PRINT1(self, "EXTERNAL_LOCK (%s) lock_type=UNLOCK\n", pb_share->sh_table_path->ps_path); /* Make any temporary locks on this table permanent. * @@ -4189,10 +4511,9 @@ xtPublic int ha_pbxt::external_lock(THD xt_broadcast_cond_ns((xt_cond_type *) pb_share->sh_ex_cond); } else { - XT_PRINT2(self, "ha_pbxt::EXTERNAL_LOCK %s lock_type=%d\n", pb_share->sh_table_path->ps_path, lock_type); + XT_PRINT2(self, "ha_pbxt::EXTERNAL_LOCK (%s) lock_type=%d\n", pb_share->sh_table_path->ps_path, lock_type); if (pb_lock_table) { - pb_ex_in_use = 1; try_(a) { if (!pb_table_locked) @@ -4243,14 +4564,18 @@ xtPublic int ha_pbxt::external_lock(THD if ((pb_open_tab->ot_for_update = (lock_type == F_WRLCK))) { switch ((int) thd_sql_command(thd)) { case SQLCOM_DELETE: +#ifndef DRIZZLED case SQLCOM_DELETE_MULTI: +#endif /* turn DELETE IGNORE into normal DELETE. The IGNORE option causes problems because * when a record is deleted we add an xlog record which we cannot "rollback" later * when we find that an FK-constraint has failed. */ thd->lex->ignore = false; case SQLCOM_UPDATE: +#ifndef DRIZZLED case SQLCOM_UPDATE_MULTI: +#endif case SQLCOM_REPLACE: case SQLCOM_REPLACE_SELECT: case SQLCOM_INSERT: @@ -4265,7 +4590,9 @@ xtPublic int ha_pbxt::external_lock(THD case SQLCOM_DROP_TABLE: case SQLCOM_DROP_INDEX: case SQLCOM_LOAD: +#ifndef DRIZZLED case SQLCOM_REPAIR: +#endif case SQLCOM_OPTIMIZE: self->st_stat_modify = TRUE; break; @@ -4316,11 +4643,16 @@ xtPublic int ha_pbxt::external_lock(THD self->st_xact_mode = thd_tx_isolation(thd) <= ISO_READ_COMMITTED ? XT_XACT_COMMITTED_READ : XT_XACT_REPEATABLE_READ; self->st_ignore_fkeys = (thd_test_options(thd,OPTION_NO_FOREIGN_KEY_CHECKS)) != 0; self->st_auto_commit = (thd_test_options(thd, (OPTION_NOT_AUTOCOMMIT | OPTION_BEGIN))) == 0; +#ifdef DRIZZLED + self->st_table_trans = FALSE; +#else self->st_table_trans = thd_sql_command(thd) == SQLCOM_LOCK_TABLES; +#endif self->st_abort_trans = FALSE; self->st_stat_ended = FALSE; self->st_stat_trans = FALSE; XT_PRINT0(self, "xt_xn_begin\n"); + xt_xres_wait_for_recovery(self, XT_RECOVER_SWEPT); if (!xt_xn_begin(self)) { err = xt_ha_pbxt_thread_error_for_mysql(thd, self, pb_ignore_dup_key); pb_ex_in_use = 0; @@ -4404,7 +4736,7 @@ int ha_pbxt::start_stmt(THD *thd, thr_lo if (!(self = ha_set_current_thread(thd, &err))) return xt_ha_pbxt_to_mysql_error(err); - XT_PRINT2(self, "ha_pbxt::start_stmt %s lock_type=%d\n", pb_share->sh_table_path->ps_path, (int) lock_type); + XT_PRINT2(self, "ha_pbxt::start_stmt (%s) lock_type=%d\n", pb_share->sh_table_path->ps_path, (int) lock_type); if (!pb_open_tab) { if ((err = reopen())) @@ -4430,12 +4762,12 @@ int ha_pbxt::start_stmt(THD *thd, thr_lo /* This section handles "auto-commit"... */ if (self->st_xact_data && self->st_auto_commit && self->st_table_trans) { if (self->st_abort_trans) { - XT_PRINT0(self, "xt_xn_rollback\n"); + XT_PRINT0(self, "xt_xn_rollback in start_stmt\n"); if (!xt_xn_rollback(self)) err = xt_ha_pbxt_thread_error_for_mysql(pb_mysql_thd, self, pb_ignore_dup_key); } else { - XT_PRINT0(self, "xt_xn_commit\n"); + XT_PRINT0(self, "xt_xn_commit in start_stmt\n"); if (!xt_xn_commit(self)) err = xt_ha_pbxt_thread_error_for_mysql(pb_mysql_thd, self, pb_ignore_dup_key); } @@ -4466,9 +4798,11 @@ int ha_pbxt::start_stmt(THD *thd, thr_lo if (pb_open_tab->ot_for_update) { switch ((int) thd_sql_command(thd)) { case SQLCOM_UPDATE: - case SQLCOM_UPDATE_MULTI: case SQLCOM_DELETE: +#ifndef DRIZZLED + case SQLCOM_UPDATE_MULTI: case SQLCOM_DELETE_MULTI: +#endif case SQLCOM_REPLACE: case SQLCOM_REPLACE_SELECT: case SQLCOM_INSERT: @@ -4483,14 +4817,15 @@ int ha_pbxt::start_stmt(THD *thd, thr_lo case SQLCOM_DROP_TABLE: case SQLCOM_DROP_INDEX: case SQLCOM_LOAD: +#ifndef DRIZZLED case SQLCOM_REPAIR: +#endif case SQLCOM_OPTIMIZE: self->st_stat_modify = TRUE; break; } } - /* (***) This is required at this level! * No matter how often it is called, it is still the start of a * statement. We need to make sure statements that are NOT mistaken @@ -4516,6 +4851,7 @@ int ha_pbxt::start_stmt(THD *thd, thr_lo self->st_stat_ended = FALSE; self->st_stat_trans = FALSE; XT_PRINT0(self, "xt_xn_begin\n"); + xt_xres_wait_for_recovery(self, XT_RECOVER_SWEPT); if (!xt_xn_begin(self)) { err = xt_ha_pbxt_thread_error_for_mysql(thd, self, pb_ignore_dup_key); goto complete; @@ -4673,7 +5009,9 @@ THR_LOCK_DATA **ha_pbxt::store_lock(THD * */ if ((lock_type >= TL_WRITE_CONCURRENT_INSERT && lock_type <= TL_WRITE) && +#ifndef DRIZZLED !(thd_in_lock_tables(thd) && thd_sql_command(thd) == SQLCOM_LOCK_TABLES) && +#endif !thd_tablespace_op(thd) && thd_sql_command(thd) != SQLCOM_TRUNCATE && thd_sql_command(thd) != SQLCOM_OPTIMIZE && @@ -4691,22 +5029,23 @@ THR_LOCK_DATA **ha_pbxt::store_lock(THD * Stewart: removed SQLCOM_CALL, not sure of implications. */ - if (lock_type == TL_READ_NO_INSERT && - (!thd_in_lock_tables(thd) + if (lock_type == TL_READ_NO_INSERT #ifndef DRIZZLED + && (!thd_in_lock_tables(thd) || thd_sql_command(thd) == SQLCOM_CALL + ) #endif - )) + ) { lock_type = TL_READ; } - XT_PRINT3(xt_get_self(), "ha_pbxt::store_lock %s %d->%d\n", pb_share->sh_table_path->ps_path, pb_lock.type, lock_type); + XT_PRINT3(xt_get_self(), "store_lock (%s) %d->%d\n", pb_share->sh_table_path->ps_path, pb_lock.type, lock_type); pb_lock.type = lock_type; } #ifdef PBXT_HANDLER_TRACE else { - XT_PRINT3(xt_get_self(), "ha_pbxt::store_lock %s %d->%d (ignore/unlock)\n", pb_share->sh_table_path->ps_path, lock_type, lock_type); + XT_PRINT3(xt_get_self(), "store_lock (%s) %d->%d (ignore/unlock)\n", pb_share->sh_table_path->ps_path, lock_type, lock_type); } #endif *to++= &pb_lock; @@ -4723,15 +5062,23 @@ THR_LOCK_DATA **ha_pbxt::store_lock(THD * during create if the table_flag HA_DROP_BEFORE_CREATE was specified for * the storage engine. */ +#ifdef DRIZZLED +int PBXTStorageEngine::doDropTable(Session &, std::string table_path_str) +#else int ha_pbxt::delete_table(const char *table_path) +#endif { THD *thd = current_thd; int err = 0; XTThreadPtr self = NULL; XTSharePtr share; +#ifdef DRIZZLED + const char *table_path = table_path_str.c_str(); +#endif + STAT_TRACE(self, *thd_query(thd)); - XT_PRINT1(self, "ha_pbxt::delete_table %s\n", table_path); + XT_PRINT1(self, "delete_table (%s)\n", table_path); if (XTSystemTableShare::isSystemTable(table_path)) return delete_system_table(table_path); @@ -4795,7 +5142,7 @@ int ha_pbxt::delete_table(const char *ta #endif } catch_(a) { - err = xt_ha_pbxt_thread_error_for_mysql(thd, self, pb_ignore_dup_key); + err = xt_ha_pbxt_thread_error_for_mysql(thd, self, FALSE); #ifdef DRIZZLED if (err == HA_ERR_NO_SUCH_TABLE) err = ENOENT; @@ -4819,7 +5166,11 @@ int ha_pbxt::delete_table(const char *ta return err; } +#ifdef DRIZZLED +int PBXTStorageEngine::delete_system_table(const char *table_path) +#else int ha_pbxt::delete_system_table(const char *table_path) +#endif { THD *thd = current_thd; XTExceptionRec e; @@ -4857,7 +5208,13 @@ int ha_pbxt::delete_system_table(const c * This function can be used to move a table from one database to * another. */ +#ifdef DRIZZLED +int PBXTStorageEngine::doRenameTable(Session *, + const char *from, + const char *to) +#else int ha_pbxt::rename_table(const char *from, const char *to) +#endif { THD *thd = current_thd; int err = 0; @@ -4865,15 +5222,13 @@ int ha_pbxt::rename_table(const char *fr XTSharePtr share; XTDatabaseHPtr to_db; - XT_TRACE_CALL(); - if (XTSystemTableShare::isSystemTable(from)) return rename_system_table(from, to); if (!(self = ha_set_current_thread(thd, &err))) return xt_ha_pbxt_to_mysql_error(err); - XT_PRINT2(self, "ha_pbxt::rename_table %s -> %s\n", from, to); + XT_PRINT2(self, "rename_table (%s -> %s)\n", from, to); #ifdef PBMS_ENABLED PBMSResultRec result; @@ -4929,7 +5284,7 @@ int ha_pbxt::rename_table(const char *fr #endif } catch_(a) { - err = xt_ha_pbxt_thread_error_for_mysql(thd, self, pb_ignore_dup_key); + err = xt_ha_pbxt_thread_error_for_mysql(thd, self, FALSE); } cont_(a); @@ -4940,7 +5295,11 @@ int ha_pbxt::rename_table(const char *fr XT_RETURN(err); } +#ifdef DRIZZLED +int PBXTStorageEngine::rename_system_table(const char *XT_UNUSED(from), const char *XT_UNUSED(to)) +#else int ha_pbxt::rename_system_table(const char *XT_UNUSED(from), const char *XT_UNUSED(to)) +#endif { return ER_NOT_SUPPORTED_YET; } @@ -5029,7 +5388,15 @@ ha_rows ha_pbxt::records_in_range(uint i * Called from handle.cc by ha_create_table(). */ +#ifdef DRIZZLED +int PBXTStorageEngine::doCreateTable(Session *, + const char *table_path, + Table &table_arg, + HA_CREATE_INFO &create_info, + drizzled::message::Table &XT_UNUSED(proto)) +#else int ha_pbxt::create(const char *table_path, TABLE *table_arg, HA_CREATE_INFO *create_info) +#endif { THD *thd = current_thd; int err = 0; @@ -5037,37 +5404,61 @@ int ha_pbxt::create(const char *table_pa XTDDTable *tab_def = NULL; XTDictionaryRec dic; - memset(&dic, 0, sizeof(dic)); + if ((strcmp(table_path, "./pbxt/location") == 0) || (strcmp(table_path, "./pbxt/statistics") == 0)) + return 0; - XT_TRACE_CALL(); + memset(&dic, 0, sizeof(dic)); if (!(self = ha_set_current_thread(thd, &err))) return xt_ha_pbxt_to_mysql_error(err); +#ifdef DRIZZLED + XT_PRINT2(self, "create (%s) %s\n", table_path, (create_info.options & HA_LEX_CREATE_TMP_TABLE) ? "temporary" : ""); +#else + XT_PRINT2(self, "create (%s) %s\n", table_path, (create_info->options & HA_LEX_CREATE_TMP_TABLE) ? "temporary" : ""); +#endif STAT_TRACE(self, *thd_query(thd)); - XT_PRINT1(self, "ha_pbxt::create %s\n", table_path); try_(a) { xt_ha_open_database_of_table(self, (XTPathStrPtr) table_path); +#ifdef DRIZZLED + for (uint i=0; i<TS(&table_arg)->keys; i++) { + if (table_arg.key_info[i].key_length > XT_INDEX_MAX_KEY_SIZE) + xt_throw_sulxterr(XT_CONTEXT, XT_ERR_KEY_TOO_LARGE, table_arg.key_info[i].name, (u_long) XT_INDEX_MAX_KEY_SIZE); + } +#else for (uint i=0; i<TS(table_arg)->keys; i++) { if (table_arg->key_info[i].key_length > XT_INDEX_MAX_KEY_SIZE) xt_throw_sulxterr(XT_CONTEXT, XT_ERR_KEY_TOO_LARGE, table_arg->key_info[i].name, (u_long) XT_INDEX_MAX_KEY_SIZE); } +#endif /* ($) auto_increment_value will be zero if * AUTO_INCREMENT is not used. Otherwise * Query was ALTER TABLE ... AUTO_INCREMENT = x; or * CREATE TABLE ... AUTO_INCREMENT = x; */ +#ifdef DRIZZLED + tab_def = xt_ri_create_table(self, true, (XTPathStrPtr) table_path, *thd_query(thd), myxt_create_table_from_table(self, &table_arg)); + tab_def->checkForeignKeys(self, create_info.options & HA_LEX_CREATE_TMP_TABLE); +#else tab_def = xt_ri_create_table(self, true, (XTPathStrPtr) table_path, *thd_query(thd), myxt_create_table_from_table(self, table_arg)); tab_def->checkForeignKeys(self, create_info->options & HA_LEX_CREATE_TMP_TABLE); +#endif dic.dic_table = tab_def; +#ifdef DRIZZLED + dic.dic_my_table = &table_arg; + dic.dic_tab_flags = (create_info.options & HA_LEX_CREATE_TMP_TABLE) ? XT_TAB_FLAGS_TEMP_TAB : 0; + dic.dic_min_auto_inc = (xtWord8) create_info.auto_increment_value; /* ($) */ + dic.dic_def_ave_row_size = table_arg.s->getAvgRowLength(); +#else dic.dic_my_table = table_arg; dic.dic_tab_flags = (create_info->options & HA_LEX_CREATE_TMP_TABLE) ? XT_TAB_FLAGS_TEMP_TAB : 0; dic.dic_min_auto_inc = (xtWord8) create_info->auto_increment_value; /* ($) */ dic.dic_def_ave_row_size = (xtWord8) table_arg->s->avg_row_length; +#endif myxt_setup_dictionary(self, &dic); /* @@ -5089,7 +5480,7 @@ int ha_pbxt::create(const char *table_pa if (tab_def) tab_def->finalize(self); dic.dic_table = NULL; - err = xt_ha_pbxt_thread_error_for_mysql(thd, self, pb_ignore_dup_key); + err = xt_ha_pbxt_thread_error_for_mysql(thd, self, FALSE); } cont_(a); @@ -5161,7 +5552,7 @@ bool ha_pbxt::get_error_message(int XT_U if (!self->t_exception.e_xt_err) return FALSE; - buf->copy(self->t_exception.e_err_msg, strlen(self->t_exception.e_err_msg), system_charset_info); + buf->copy(self->t_exception.e_err_msg, (uint32) strlen(self->t_exception.e_err_msg), system_charset_info); return TRUE; } @@ -5421,16 +5812,31 @@ static MYSQL_SYSVAR_INT(sweeper_priority #ifdef DRIZZLED static MYSQL_SYSVAR_INT(max_threads, pbxt_max_threads, - PLUGIN_VAR_OPCMDARG, + PLUGIN_VAR_OPCMDARG | PLUGIN_VAR_READONLY, "The maximum number of threads used by PBXT", NULL, NULL, 500, 20, 20000, 1); #else static MYSQL_SYSVAR_INT(max_threads, pbxt_max_threads, - PLUGIN_VAR_OPCMDARG, + PLUGIN_VAR_OPCMDARG | PLUGIN_VAR_READONLY, "The maximum number of threads used by PBXT, 0 = set according to MySQL max_connections.", NULL, NULL, 0, 0, 20000, 1); #endif +#ifndef DEBUG +static MYSQL_SYSVAR_BOOL(support_xa, pbxt_support_xa, + PLUGIN_VAR_OPCMDARG, + "Enable PBXT support for the XA two-phase commit, default is enabled", + NULL, NULL, TRUE); +#else +static MYSQL_SYSVAR_BOOL(support_xa, pbxt_support_xa, + PLUGIN_VAR_OPCMDARG, + "Enable PBXT support for the XA two-phase commit, default is disabled (due to assertion failure in MySQL)", + /* The problem is, in MySQL an assertion fails in debug mode: + * Assertion failed: (total_ha_2pc == (ulong) opt_bin_log+1), function ha_recover, file handler.cc, line 1557. + */ + NULL, NULL, FALSE); +#endif + static struct st_mysql_sys_var* pbxt_system_variables[] = { MYSQL_SYSVAR(index_cache_size), MYSQL_SYSVAR(record_cache_size), @@ -5448,6 +5854,7 @@ static struct st_mysql_sys_var* pbxt_sys MYSQL_SYSVAR(offline_log_function), MYSQL_SYSVAR(sweeper_priority), MYSQL_SYSVAR(max_threads), + MYSQL_SYSVAR(support_xa), NULL }; #endif @@ -5494,8 +5901,8 @@ mysql_declare_plugin(pbxt) "Paul McCullagh, PrimeBase Technologies GmbH", "PBXT internal system statitics", PLUGIN_LICENSE_GPL, - pbxt_init_statitics, /* plugin init */ - pbxt_exit_statitics, /* plugin deinit */ + pbxt_init_statistics, /* plugin init */ + pbxt_exit_statistics, /* plugin deinit */ #ifndef DRIZZLED 0x0005, #endif === modified file 'storage/pbxt/src/ha_pbxt.h' --- a/storage/pbxt/src/ha_pbxt.h 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/ha_pbxt.h 2009-11-24 10:55:06 +0000 @@ -27,9 +27,9 @@ #ifdef DRIZZLED #include <drizzled/common.h> -#include <drizzled/handler.h> -#include <drizzled/plugin/storage_engine.h> #include <mysys/thr_lock.h> +#include <drizzled/cursor.h> + #else #include "mysql_priv.h" #endif @@ -53,17 +53,31 @@ class ha_pbxt; #ifdef DRIZZLED -class PBXTStorageEngine : public StorageEngine { +class PBXTStorageEngine : public drizzled::plugin::StorageEngine +{ + + int delete_system_table(const char *table_path); + int rename_system_table(const char * from, const char * to); + public: PBXTStorageEngine(std::string name_arg) - : StorageEngine(name_arg, HTON_NO_FLAGS) {} + : drizzled::plugin::StorageEngine(name_arg, HTON_NO_FLAGS) {} + + void operator delete(void *) {} + void operator delete[] (void *) {} /* override */ int close_connection(Session *); /* override */ int commit(Session *, bool); /* override */ int rollback(Session *, bool); - /* override */ handler *create(TABLE_SHARE *, MEM_ROOT *); + /* override */ Cursor *create(TABLE_SHARE *, MEM_ROOT *); /* override */ void drop_database(char *); /* override */ bool show_status(Session *, stat_print_fn *, enum ha_stat_type); + /* override */ const char **bas_ext() const; + /* override */ int doCreateTable(Session *session, const char *table_name, + Table &table_arg, HA_CREATE_INFO + &create_info, drizzled::message::Table &proto); + /* override */ int doRenameTable(Session *, const char *from, const char *to); + /* override */ int doDropTable(Session &session, std::string table_path); }; typedef PBXTStorageEngine handlerton; @@ -139,9 +153,9 @@ class ha_pbxt: public handler * don't implement this method unless you really have indexes. */ const char *index_type(uint inx) { (void) inx; return "BTREE"; } - +#ifndef DRIZZLED const char **bas_ext() const; - +#endif MX_UINT8_T table_cache_type(); /* @@ -241,11 +255,13 @@ class ha_pbxt: public handler int optimize(THD* thd, HA_CHECK_OPT* check_opt); int check(THD* thd, HA_CHECK_OPT* check_opt); ha_rows records_in_range(uint inx, key_range *min_key, key_range *max_key); - int delete_table(const char *from); +#ifndef DRIZZLED int delete_system_table(const char *table_path); - int rename_table(const char * from, const char * to); + int delete_table(const char *from); int rename_system_table(const char * from, const char * to); + int rename_table(const char * from, const char * to); int create(const char *name, TABLE *form, HA_CREATE_INFO *create_info); //required +#endif void update_create_info(HA_CREATE_INFO *create_info); THR_LOCK_DATA **store_lock(THD *thd, THR_LOCK_DATA **to, enum thr_lock_type lock_type); //required @@ -277,6 +293,7 @@ struct XTThread *xt_ha_thd_to_self(THD* int xt_ha_pbxt_to_mysql_error(int xt_err); int xt_ha_pbxt_thread_error_for_mysql(THD *thd, const XTThreadPtr self, int ignore_dup_key); void xt_ha_all_threads_close_database(XTThreadPtr self, XTDatabase *db); +void ha_set_auto_increment(XTOpenTablePtr ot, Field *nr); /* * These hooks are suppossed to only be used by InnoDB: === modified file 'storage/pbxt/src/ha_xtsys.h' --- a/storage/pbxt/src/ha_xtsys.h 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/ha_xtsys.h 2009-11-24 10:55:06 +0000 @@ -30,8 +30,9 @@ #ifdef DRIZZLED #include <drizzled/common.h> -#include <drizzled/handler.h> +#include <drizzled/handler_structs.h> #include <drizzled/current_session.h> +#include <drizzled/cursor.h> #else #include "mysql_priv.h" #endif === modified file 'storage/pbxt/src/heap_xt.cc' --- a/storage/pbxt/src/heap_xt.cc 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/heap_xt.cc 2009-11-24 10:55:06 +0000 @@ -73,7 +73,7 @@ xtPublic void xt_check_heap(XTThreadPtr } #ifdef DEBUG_MEMORY -xtPublic void xt_mm_heap_reference(XTThreadPtr self, XTHeapPtr hp, u_int line, c_char *file) +xtPublic void xt_mm_heap_reference(XTThreadPtr XT_UNUSED(self), XTHeapPtr hp, u_int line, c_char *file) #else xtPublic void xt_heap_reference(XTThreadPtr, XTHeapPtr hp) #endif === modified file 'storage/pbxt/src/index_xt.cc' --- a/storage/pbxt/src/index_xt.cc 2009-08-18 07:46:53 +0000 +++ b/storage/pbxt/src/index_xt.cc 2009-11-24 10:55:06 +0000 @@ -829,14 +829,25 @@ static void idx_next_branch_item(XTTable result->sr_item.i_item_offset += result->sr_item.i_item_size + result->sr_item.i_node_ref_size; bitem = branch->tb_data + result->sr_item.i_item_offset; - if (ind->mi_fix_key) - ilen = result->sr_item.i_item_size; + if (result->sr_item.i_item_offset < result->sr_item.i_total_size) { + if (ind->mi_fix_key) + ilen = result->sr_item.i_item_size; + else { + ilen = myxt_get_key_length(ind, bitem) + XT_RECORD_REF_SIZE; + result->sr_item.i_item_size = ilen; + } + xt_get_res_record_ref(bitem + ilen - XT_RECORD_REF_SIZE, result); /* (Only valid if i_item_offset < i_total_size) */ + } else { - ilen = myxt_get_key_length(ind, bitem) + XT_RECORD_REF_SIZE; - result->sr_item.i_item_size = ilen; + result->sr_item.i_item_size = 0; + result->sr_rec_id = 0; + result->sr_row_id = 0; } - xt_get_res_record_ref(bitem + ilen - XT_RECORD_REF_SIZE, result); /* (Only valid if i_item_offset < i_total_size) */ - result->sr_branch = IDX_GET_NODE_REF(tab, bitem, result->sr_item.i_node_ref_size); + if (result->sr_item.i_node_ref_size) + /* IDX_GET_NODE_REF() loads the branch reference to the LEFT of the item. */ + result->sr_branch = IDX_GET_NODE_REF(tab, bitem, result->sr_item.i_node_ref_size); + else + result->sr_branch = 0; } xtPublic void xt_prev_branch_item_fix(XTTableHPtr XT_UNUSED(tab), XTIndexPtr XT_UNUSED(ind), XTIdxBranchDPtr branch, register XTIdxResultRec *result) @@ -3987,7 +3998,7 @@ xtPublic xtBool xt_flush_indices(XTOpenT * here. */ if (!(tab->tab_dic.dic_tab_flags & XT_TAB_FLAGS_TEMP_TAB)) { - if (!xt_xlog_flush_log(ot->ot_thread)) + if (!xt_xlog_flush_log(tab->tab_db, ot->ot_thread)) goto failed_2; if (!il->il_flush(ot)) goto failed_2; === modified file 'storage/pbxt/src/lock_xt.cc' --- a/storage/pbxt/src/lock_xt.cc 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/lock_xt.cc 2009-11-24 10:55:06 +0000 @@ -1246,7 +1246,7 @@ xtPublic void xt_spinlock_init(XTThreadP (void) self; spl->spl_lock = 0; #ifdef XT_NO_ATOMICS - xt_init_mutex(self, &spl->spl_mutex); + xt_init_mutex_with_autoname(self, &spl->spl_mutex); #endif #ifdef DEBUG spl->spl_locker = 0; === modified file 'storage/pbxt/src/locklist_xt.h' --- a/storage/pbxt/src/locklist_xt.h 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/locklist_xt.h 2009-11-27 15:37:02 +0000 @@ -24,11 +24,16 @@ #ifndef __xt_locklist_h__ #define __xt_locklist_h__ +/* + * XT_THREAD_LOCK_INFO and DEBUG_LOCKING code must be updated to avoid calls to xt_get_self() as it can be called before hton->slot is + * assigned by MySQL which is used by xt_get_self() + */ + #ifdef DEBUG -#define XT_THREAD_LOCK_INFO +//#define XT_THREAD_LOCK_INFO #ifndef XT_WIN /* We need DEBUG_LOCKING in order to enable pthread function wrappers */ -#define DEBUG_LOCKING +//#define DEBUG_LOCKING #endif #endif === modified file 'storage/pbxt/src/memory_xt.cc' --- a/storage/pbxt/src/memory_xt.cc 2009-09-03 06:15:03 +0000 +++ b/storage/pbxt/src/memory_xt.cc 2009-11-25 15:06:47 +0000 @@ -34,7 +34,7 @@ #include "trace_xt.h" #ifdef DEBUG -//#define RECORD_MM +#define RECORD_MM #endif #ifdef DEBUG @@ -367,9 +367,8 @@ static long mm_find_pointer(void *ptr) return(-1); } -static long mm_add_pointer(void *ptr, u_int id) +static long mm_add_pointer(void *ptr, u_int XT_UNUSED(id)) { -#pragma unused(id) register int i, n, guess; if (mm_nr_in_use == mm_total_allocated) { === modified file 'storage/pbxt/src/myxt_xt.cc' --- a/storage/pbxt/src/myxt_xt.cc 2009-09-03 06:15:03 +0000 +++ b/storage/pbxt/src/myxt_xt.cc 2009-12-01 09:50:46 +0000 @@ -36,7 +36,7 @@ #include <drizzled/current_session.h> #include <drizzled/sql_lex.h> #include <drizzled/session.h> -extern "C" struct charset_info_st *session_charset(Session *session); +//extern "C" struct charset_info_st *session_charset(Session *session); extern pthread_key_t THR_Session; #else #include "mysql_priv.h" @@ -171,7 +171,9 @@ xtPublic u_int myxt_create_key_from_key( for (u_int i=0; i<ind->mi_seg_count && (int) k_length > 0; i++, old += keyseg->length, keyseg++) { +#ifndef DRIZZLED enum ha_base_keytype type = (enum ha_base_keytype) keyseg->type; +#endif u_int length = keyseg->length < k_length ? keyseg->length : k_length; u_int char_length; xtWord1 *pos; @@ -192,14 +194,18 @@ xtPublic u_int myxt_create_key_from_key( pos = old; if (keyseg->flag & HA_SPACE_PACK) { uchar *end = pos + length; +#ifndef DRIZZLED if (type != HA_KEYTYPE_NUM) { +#endif while (end > pos && end[-1] == ' ') end--; +#ifndef DRIZZLED } else { while (pos < end && pos[0] == ' ') pos++; } +#endif k_length -= length; length = (u_int) (end-pos); FIX_LENGTH(cs, pos, length, char_length); @@ -276,6 +282,7 @@ xtPublic u_int myxt_create_key_from_row( char_length= ((cs && cs->mbmaxlen > 1) ? length/cs->mbmaxlen : length); pos = record + keyseg->start; +#ifndef DRIZZLED if (type == HA_KEYTYPE_BIT) { if (keyseg->bit_length) @@ -289,17 +296,22 @@ xtPublic u_int myxt_create_key_from_row( key+= length; continue; } +#endif if (keyseg->flag & HA_SPACE_PACK) { end = pos + length; +#ifndef DRIZZLED if (type != HA_KEYTYPE_NUM) { +#endif while (end > pos && end[-1] == ' ') end--; +#ifndef DRIZZLED } else { while (pos < end && pos[0] == ' ') pos++; } +#endif length = (u_int) (end-pos); FIX_LENGTH(cs, pos, length, char_length); store_key_length_inc(key,char_length); @@ -333,6 +345,7 @@ xtPublic u_int myxt_create_key_from_row( if (keyseg->flag & HA_SWAP_KEY) { /* Numerical column */ #ifdef HAVE_ISNAN +#ifndef DRIZZLED if (type == HA_KEYTYPE_FLOAT) { float nr; @@ -345,7 +358,9 @@ xtPublic u_int myxt_create_key_from_row( continue; } } - else if (type == HA_KEYTYPE_DOUBLE) { + else +#endif + if (type == HA_KEYTYPE_DOUBLE) { double nr; float8get(nr,pos); @@ -414,6 +429,7 @@ xtPublic u_int myxt_create_foreign_key_f char_length= ((cs && cs->mbmaxlen > 1) ? length/cs->mbmaxlen : length); pos = record + keyseg->start; +#ifndef DRIZZLED if (type == HA_KEYTYPE_BIT) { if (keyseg->bit_length) @@ -427,17 +443,22 @@ xtPublic u_int myxt_create_foreign_key_f key+= length; continue; } +#endif if (keyseg->flag & HA_SPACE_PACK) { end = pos + length; +#ifndef DRIZZLED if (type != HA_KEYTYPE_NUM) { +#endif while (end > pos && end[-1] == ' ') end--; +#ifndef DRIZZLED } else { while (pos < end && pos[0] == ' ') pos++; } +#endif length = (u_int) (end-pos); FIX_LENGTH(cs, pos, length, char_length); store_key_length_inc(key,char_length); @@ -471,6 +492,7 @@ xtPublic u_int myxt_create_foreign_key_f if (keyseg->flag & HA_SWAP_KEY) { /* Numerical column */ #ifdef HAVE_ISNAN +#ifndef DRIZZLED if (type == HA_KEYTYPE_FLOAT) { float nr; @@ -483,7 +505,9 @@ xtPublic u_int myxt_create_foreign_key_f continue; } } - else if (type == HA_KEYTYPE_DOUBLE) { + else +#endif + if (type == HA_KEYTYPE_DOUBLE) { double nr; float8get(nr,pos); @@ -622,7 +646,6 @@ static char *mx_get_length_and_data(Fiel case MYSQL_TYPE_SET: case MYSQL_TYPE_GEOMETRY: #else - case DRIZZLE_TYPE_TINY: case DRIZZLE_TYPE_LONG: case DRIZZLE_TYPE_DOUBLE: case DRIZZLE_TYPE_NULL: @@ -740,7 +763,6 @@ static void mx_set_length_and_data(Field case MYSQL_TYPE_SET: case MYSQL_TYPE_GEOMETRY: #else - case DRIZZLE_TYPE_TINY: case DRIZZLE_TYPE_LONG: case DRIZZLE_TYPE_DOUBLE: case DRIZZLE_TYPE_NULL: @@ -825,6 +847,7 @@ xtPublic xtBool myxt_create_row_from_key } record[keyseg->null_pos] &= ~keyseg->null_bit; } +#ifndef DRIZZLED if (keyseg->type == HA_KEYTYPE_BIT) { uint length = keyseg->length; @@ -845,6 +868,7 @@ xtPublic xtBool myxt_create_row_from_key key+= length; continue; } +#endif if (keyseg->flag & HA_SPACE_PACK) { uint length; @@ -854,16 +878,20 @@ xtPublic xtBool myxt_create_row_from_key goto err; #endif pos = record+keyseg->start; +#ifndef DRIZZLED if (keyseg->type != (int) HA_KEYTYPE_NUM) { +#endif memcpy(pos,key,(size_t) length); bfill(pos+length,keyseg->length-length,' '); +#ifndef DRIZZLED } else { bfill(pos,keyseg->length-length,' '); memcpy(pos+keyseg->length-length,key,(size_t) length); } +#endif key+=length; continue; } @@ -945,7 +973,7 @@ xtPublic xtBool myxt_create_row_from_key static int my_compare_bin(uchar *a, uint a_length, uchar *b, uint b_length, my_bool part_key, my_bool skip_end_space) { - uint length= min(a_length,b_length); + uint length= a_length < b_length ? a_length : b_length; uchar *end= a+ length; int flag; @@ -1023,6 +1051,7 @@ xtPublic u_int myxt_get_key_length(XTInd get_key_pack_length(seg_len, pack_len, key_data); key_data += seg_len; break; +#ifndef DRIZZLED case HA_KEYTYPE_NUM: { /* Numeric key */ if (keyseg->flag & HA_SPACE_PACK) @@ -1035,15 +1064,16 @@ xtPublic u_int myxt_get_key_length(XTInd case HA_KEYTYPE_INT8: case HA_KEYTYPE_SHORT_INT: case HA_KEYTYPE_USHORT_INT: + case HA_KEYTYPE_INT24: + case HA_KEYTYPE_FLOAT: + case HA_KEYTYPE_BIT: +#endif case HA_KEYTYPE_LONG_INT: case HA_KEYTYPE_ULONG_INT: - case HA_KEYTYPE_INT24: case HA_KEYTYPE_UINT24: - case HA_KEYTYPE_FLOAT: case HA_KEYTYPE_DOUBLE: case HA_KEYTYPE_LONGLONG: case HA_KEYTYPE_ULONGLONG: - case HA_KEYTYPE_BIT: key_data += keyseg->length; break; case HA_KEYTYPE_END: @@ -1190,6 +1220,7 @@ xtPublic int myxt_compare_key(XTIndexPtr b += b_length; break; } +#ifndef DRIZZLED case HA_KEYTYPE_INT8: { int i_1 = (int) *((signed char *) a); @@ -1218,6 +1249,7 @@ xtPublic int myxt_compare_key(XTIndexPtr b += keyseg->length; break; } +#endif case HA_KEYTYPE_LONG_INT: { int32 l_1 = sint4korr(a); int32 l_2 = sint4korr(b); @@ -1236,6 +1268,7 @@ xtPublic int myxt_compare_key(XTIndexPtr b += keyseg->length; break; } +#ifndef DRIZZLED case HA_KEYTYPE_INT24: { int32 l_1 = sint3korr(a); int32 l_2 = sint3korr(b); @@ -1245,6 +1278,7 @@ xtPublic int myxt_compare_key(XTIndexPtr b += keyseg->length; break; } +#endif case HA_KEYTYPE_UINT24: { int32 l_1 = uint3korr(a); int32 l_2 = uint3korr(b); @@ -1254,6 +1288,7 @@ xtPublic int myxt_compare_key(XTIndexPtr b += keyseg->length; break; } +#ifndef DRIZZLED case HA_KEYTYPE_FLOAT: { float f_1, f_2; @@ -1270,6 +1305,7 @@ xtPublic int myxt_compare_key(XTIndexPtr b += keyseg->length; break; } +#endif case HA_KEYTYPE_DOUBLE: { double d_1, d_2; @@ -1286,6 +1322,7 @@ xtPublic int myxt_compare_key(XTIndexPtr b += keyseg->length; break; } +#ifndef DRIZZLED case HA_KEYTYPE_NUM: { /* Numeric key */ if (keyseg->flag & HA_SPACE_PACK) { @@ -1339,6 +1376,7 @@ xtPublic int myxt_compare_key(XTIndexPtr b += b_length; break; } +#endif #ifdef HAVE_LONG_LONG case HA_KEYTYPE_LONGLONG: { longlong ll_a = sint8korr(a); @@ -1359,9 +1397,11 @@ xtPublic int myxt_compare_key(XTIndexPtr break; } #endif +#ifndef DRIZZLED case HA_KEYTYPE_BIT: /* TODO: What here? */ break; +#endif case HA_KEYTYPE_END: /* Ready */ goto end; } @@ -1410,16 +1450,19 @@ xtPublic u_int myxt_key_seg_length(XTInd key_length = has_null + a_length + pack_len; break; } +#ifndef DRIZZLED case HA_KEYTYPE_INT8: case HA_KEYTYPE_SHORT_INT: case HA_KEYTYPE_USHORT_INT: + case HA_KEYTYPE_INT24: + case HA_KEYTYPE_FLOAT: +#endif case HA_KEYTYPE_LONG_INT: case HA_KEYTYPE_ULONG_INT: - case HA_KEYTYPE_INT24: case HA_KEYTYPE_UINT24: - case HA_KEYTYPE_FLOAT: case HA_KEYTYPE_DOUBLE: break; +#ifndef DRIZZLED case HA_KEYTYPE_NUM: { /* Numeric key */ if (keyseg->flag & HA_SPACE_PACK) { @@ -1428,14 +1471,17 @@ xtPublic u_int myxt_key_seg_length(XTInd } break; } +#endif #ifdef HAVE_LONG_LONG case HA_KEYTYPE_LONGLONG: case HA_KEYTYPE_ULONGLONG: break; #endif +#ifndef DRIZZLED case HA_KEYTYPE_BIT: /* TODO: What here? */ break; +#endif case HA_KEYTYPE_END: /* Ready */ break; } @@ -1486,7 +1532,7 @@ xtPublic xtWord4 myxt_store_row_length(X return row_size; } -static xtWord4 mx_store_row(XTOpenTablePtr ot, xtWord4 row_size, char *rec_buff) +xtPublic xtWord4 myxt_store_row_data(XTOpenTablePtr ot, xtWord4 row_size, char *rec_buff) { TABLE *table = ot->ot_table->tab_dic.dic_my_table; char *sdata; @@ -1614,8 +1660,9 @@ xtPublic size_t myxt_load_row_length(XTO } /* Unload from PBXT variable length format to the MySQL row format. */ -xtPublic xtBool myxt_load_row(XTOpenTablePtr ot, xtWord1 *source_buf, xtWord1 *dest_buff, u_int col_cnt) +xtPublic xtWord4 myxt_load_row_data(XTOpenTablePtr ot, xtWord1 *source_buf, xtWord1 *dest_buff, u_int col_cnt) { + xtWord1 *input_buf = source_buf; TABLE *table; xtWord4 len; Field *curr_field; @@ -1624,7 +1671,7 @@ xtPublic xtBool myxt_load_row(XTOpenTabl if (!(table = ot->ot_table->tab_dic.dic_my_table)) { xt_register_taberr(XT_REG_CONTEXT, XT_ERR_NO_DICTIONARY, ot->ot_table->tab_name); - return FAILED; + return 0; } /* According to the InnoDB implementation: @@ -1657,7 +1704,7 @@ xtPublic xtBool myxt_load_row(XTOpenTabl default: // Length byte if (*source_buf > 240) { xt_register_xterr(XT_REG_CONTEXT, XT_ERR_BAD_RECORD_FORMAT); - return FAILED; + return 0; } len = *source_buf; source_buf++; @@ -1671,7 +1718,12 @@ xtPublic xtBool myxt_load_row(XTOpenTabl source_buf += len; } - return OK; + return (xtWord4) (source_buf - input_buf); +} + +xtPublic xtBool myxt_load_row(XTOpenTablePtr ot, xtWord1 *source_buf, xtWord1 *dest_buff, u_int col_cnt) +{ + return myxt_load_row_data(ot, source_buf, dest_buff, col_cnt) != 0; } xtPublic xtBool myxt_find_column(XTOpenTablePtr ot, u_int *col_idx, const char *col_name) @@ -1784,7 +1836,7 @@ xtPublic xtBool myxt_store_row(XTOpenTab else { xtWord4 row_size; - if (!(row_size = mx_store_row(ot, XT_REC_EXT_HEADER_SIZE, rec_buff))) + if (!(row_size = myxt_store_row_data(ot, XT_REC_EXT_HEADER_SIZE, rec_buff))) return FAILED; if (row_size - XT_REC_FIX_EXT_HEADER_DIFF <= ot->ot_rec_size) { rec_info->ri_fix_rec_buf = (XTTabRecFixDPtr) &ot->ot_row_wbuffer[XT_REC_FIX_EXT_HEADER_DIFF]; @@ -1951,7 +2003,7 @@ static TABLE *my_open_table(XTThreadPtr #ifdef DRIZZLED share->init(db_name, 0, name, path); - if ((error = open_table_def(thd, share)) || + if ((error = open_table_def(*thd, share)) || (error = open_table_from_share(thd, share, "", 0, (uint32_t) READ_ALL, 0, table, OTM_OPEN))) { xt_free(self, table); @@ -1995,7 +2047,7 @@ static TABLE *my_open_table(XTThreadPtr return NULL; } -#if MYSQL_VERSION_ID >= 60003 +#if MYSQL_VERSION_ID >= 50404 if ((error = open_table_from_share(thd, share, "", 0, (uint) READ_ALL, 0, table, OTM_OPEN))) #else if ((error = open_table_from_share(thd, share, "", 0, (uint) READ_ALL, 0, table, FALSE))) @@ -2145,7 +2197,10 @@ static XTIndexPtr my_create_index(XTThre if (options & HA_OPTION_PACK_KEYS || (index->flags & (HA_PACK_KEY | HA_BINARY_PACK_KEY | HA_SPACE_PACK_USED))) { - if (key_part->length > 8 && (type == HA_KEYTYPE_TEXT || type == HA_KEYTYPE_NUM || + if (key_part->length > 8 && (type == HA_KEYTYPE_TEXT || +#ifndef DRIZZLED + type == HA_KEYTYPE_NUM || +#endif (type == HA_KEYTYPE_BINARY && !field->zero_pack()))) { /* No blobs here */ @@ -2213,8 +2268,12 @@ static XTIndexPtr my_create_index(XTThre else if (field->type() == MYSQL_TYPE_ENUM) { switch (seg->length) { case 2: +#ifdef DRIZZLED + ASSERT_NS(FALSE); +#else seg->type = HA_KEYTYPE_USHORT_INT; break; +#endif case 3: seg->type = HA_KEYTYPE_UINT24; break; @@ -2675,7 +2734,11 @@ xtPublic xtBool myxt_load_dictionary(XTT if (!(my_tab = my_open_table(self, db, tab_path))) return FAILED; dic->dic_my_table = my_tab; +#ifdef DRIZZLED + dic->dic_def_ave_row_size = (xtWord8) my_tab->s->getAvgRowLength(); +#else dic->dic_def_ave_row_size = (xtWord8) my_tab->s->avg_row_length; +#endif myxt_setup_dictionary(self, dic); dic->dic_keys = (XTIndexPtr *) xt_calloc(self, sizeof(XTIndexPtr) * TS(my_tab)->keys); for (uint i=0; i<TS(my_tab)->keys; i++) @@ -2805,8 +2868,10 @@ static void ha_create_dd_index(XTThreadP static char *my_type_to_string(XTThreadPtr self, Field *field, TABLE *XT_UNUSED(my_tab)) { - char buffer[MAX_FIELD_WIDTH + 400], *ptr; + char buffer[MAX_FIELD_WIDTH + 400]; + const char *ptr; String type((char *) buffer, sizeof(buffer), system_charset_info); + xtWord4 len; /* GOTCHA: * - Above sets the string length to the same as the buffer, @@ -2817,10 +2882,17 @@ static char *my_type_to_string(XTThreadP */ type.length(0); field->sql_type(type); - ptr = type.c_ptr(); + ptr = type.ptr(); + len = type.length(); + + if (len >= sizeof(buffer)) + len = sizeof(buffer)-1; + if (ptr != buffer) - xt_strcpy(sizeof(buffer), buffer, ptr); + xt_strcpy(sizeof(buffer), buffer, ptr); + buffer[len] = 0; + if (field->has_charset()) { /* Always include the charset so that we can compare types * for FK/PK releations. @@ -2877,6 +2949,10 @@ xtPublic XTDDTable *myxt_create_table_fr xtPublic void myxt_static_convert_identifier(XTThreadPtr XT_UNUSED(self), MX_CHARSET_INFO *cs, char *from, char *to, size_t to_len) { +#ifdef DRIZZLED + ((void *)cs); + xt_strcpy(to_len, to, from); +#else uint errors; /* @@ -2888,11 +2964,16 @@ xtPublic void myxt_static_convert_identi xt_strcpy(to_len, to, from); else strconvert(cs, from, &my_charset_utf8_general_ci, to, to_len, &errors); +#endif } // cs == current_thd->charset() xtPublic char *myxt_convert_identifier(XTThreadPtr self, MX_CHARSET_INFO *cs, char *from) { +#ifdef DRIZZLED + char *to = xt_dup_string(self, from); + ((void *)cs); +#else uint errors; u_int len; char *to; @@ -2904,6 +2985,7 @@ xtPublic char *myxt_convert_identifier(X to = (char *) xt_malloc(self, len); strconvert(cs, from, &my_charset_utf8_general_ci, to, len, &errors); } +#endif return to; } @@ -2954,11 +3036,19 @@ xtPublic MX_CHARSET_INFO *myxt_getcharse THD *thd = current_thd; if (thd) - return thd_charset(thd); + return (MX_CHARSET_INFO *)thd_charset(thd); } - return &my_charset_utf8_general_ci; + return (MX_CHARSET_INFO *)&my_charset_utf8_general_ci; } +#ifdef DBUG_OFF +//typedef struct st_plugin_int *plugin_ref; +#define REF_MYSQL_PLUGIN(x) (x) +#else +//typedef struct st_plugin_int **plugin_ref; +#define REF_MYSQL_PLUGIN(x) (*(x)) +#endif + xtPublic void *myxt_create_thread() { #ifdef DRIZZLED @@ -3011,12 +3101,22 @@ xtPublic void *myxt_create_thread() return NULL; } - if (!(new_thd = new THD())) { + if (!(new_thd = new THD)) { my_thread_end(); xt_register_error(XT_REG_CONTEXT, XT_ERR_MYSQL_ERROR, 0, "Unable to create MySQL thread (THD)"); return NULL; } + /* + * If PBXT is the default storage engine, then creating any THD objects will add extra + * references to the PBXT plugin object and will effectively deadlock the plugin so + * that server will have to force plugin shutdown. To avoid deadlocking and forced shutdown + * we must dereference the plugin after creating THD objects. + */ + LEX_STRING& plugin_name = REF_MYSQL_PLUGIN(new_thd->variables.table_plugin)->name; + if ((plugin_name.length == 4) && (strncmp(plugin_name.str, "PBXT", plugin_name.length) == 0)) { + REF_MYSQL_PLUGIN(new_thd->variables.table_plugin)->ref_count--; + } new_thd->thread_stack = (char *) &new_thd; new_thd->store_globals(); lex_start(new_thd); @@ -3052,6 +3152,17 @@ xtPublic void myxt_destroy_thread(void * close_thread_tables(thd); #endif + /* + * In myxt_create_thread we decremented plugin ref-count to avoid dead-locking. + * Here we need to increment ref-count to avoid assertion failures. + */ + if (thd->variables.table_plugin) { + LEX_STRING& plugin_name = REF_MYSQL_PLUGIN(thd->variables.table_plugin)->name; + if ((plugin_name.length == 4) && (strncmp(plugin_name.str, "PBXT", plugin_name.length) == 0)) { + REF_MYSQL_PLUGIN(thd->variables.table_plugin)->ref_count++; + } + } + delete thd; /* Remember that we don't have a THD */ @@ -3128,11 +3239,12 @@ xtPublic int myxt_statistics_fill_table( const char *stat_name; u_llong stat_value; XTStatisticsRec statistics; + XTDatabaseHPtr db = self->st_database; xt_gather_statistics(&statistics); for (u_int rec_id=0; !err && rec_id<XT_STAT_CURRENT_MAX; rec_id++) { stat_name = xt_get_stat_meta_data(rec_id)->sm_name; - stat_value = xt_get_statistic(&statistics, self->st_database, rec_id); + stat_value = xt_get_statistic(&statistics, db, rec_id); col=0; mx_put_u_llong(table, col++, rec_id+1); @@ -3213,19 +3325,31 @@ static void myxt_bitmap_init(XTThreadPtr my_bitmap_map *buf; uint size_in_bytes = (((n_bits) + 31) / 32) * 4; - buf = (my_bitmap_map *) xt_malloc(self, size_in_bytes); + buf = (my_bitmap_map *) xt_malloc(self, size_in_bytes); + +#ifdef DRIZZLED + map->init(buf, n_bits); +#else map->bitmap= buf; map->n_bits= n_bits; create_last_word_mask(map); bitmap_clear_all(map); +#endif } static void myxt_bitmap_free(XTThreadPtr self, MX_BITMAP *map) { +#ifdef DRIZZLED + my_bitmap_map *buf = map->getBitmap(); + if (buf) + xt_free(self, buf); + map->setBitmap(NULL); +#else if (map->bitmap) { xt_free(self, map->bitmap); map->bitmap = NULL; } +#endif } /* @@ -3269,3 +3393,29 @@ XTDDColumn *XTDDColumnFactory::createFro return col; } +/* + * ----------------------------------------------------------------------- + * utilities + */ + +/* + * MySQL (not sure about Drizzle) first calls hton->init and then assigns the plugin a thread slot + * which is used by xt_get_self(). This is a problem as pbxt_init() starts a number of daemon threads + * which could try to use the slot before it is assigned. This code waits till slot is inited. + * We cannot directly check hton->slot as in some versions of MySQL it can be 0 before init which is a + * valid value. + */ +extern ulong total_ha; + +xtPublic void myxt_wait_pbxt_plugin_slot_assigned(XTThread *self) +{ +#ifdef DRIZZLED + static LEX_STRING plugin_name = { C_STRING_WITH_LEN("PBXT") }; + + while (!self->t_quit && !Registry::singleton().find(&plugin_name)) + xt_sleep_milli_second(1); +#else + while(!self->t_quit && (pbxt_hton->slot >= total_ha)) + xt_sleep_milli_second(1); +#endif +} === modified file 'storage/pbxt/src/myxt_xt.h' --- a/storage/pbxt/src/myxt_xt.h 2009-09-03 06:15:03 +0000 +++ b/storage/pbxt/src/myxt_xt.h 2009-11-27 15:37:02 +0000 @@ -52,8 +52,10 @@ void myxt_set_default_row_from_key(XTOp void myxt_print_key(XTIndexPtr ind, xtWord1 *key_value); xtWord4 myxt_store_row_length(XTOpenTablePtr ot, char *rec_buff); +xtWord4 myxt_store_row_data(XTOpenTablePtr ot, xtWord4 row_size, char *rec_buff); xtBool myxt_store_row(XTOpenTablePtr ot, XTTabRecInfoPtr rec_info, char *rec_buff); size_t myxt_load_row_length(XTOpenTablePtr ot, size_t buffer_size, xtWord1 *source_buf, u_int *ret_col_cnt); +xtWord4 myxt_load_row_data(XTOpenTablePtr ot, xtWord1 *source_buf, xtWord1 *dest_buff, u_int col_cnt); xtBool myxt_load_row(XTOpenTablePtr ot, xtWord1 *source_buf, xtWord1 *dest_buff, u_int col_cnt); xtBool myxt_find_column(XTOpenTablePtr ot, u_int *col_idx, const char *col_name); void myxt_get_column_name(XTOpenTablePtr ot, u_int col_idx, u_int len, char *col_name); @@ -93,4 +95,6 @@ public: static XTDDColumn *createFromMySQLField(XTThread *self, STRUCT_TABLE *, Field *); }; +void myxt_wait_pbxt_plugin_slot_assigned(XTThread *self); + #endif === modified file 'storage/pbxt/src/pbms.h' --- a/storage/pbxt/src/pbms.h 2009-08-18 07:46:53 +0000 +++ b/storage/pbxt/src/pbms.h 2009-11-24 10:55:06 +0000 @@ -344,16 +344,16 @@ public: int couldBeURL(char *blob_url, int size) { if (blob_url && (size < PBMS_BLOB_URL_SIZE)) { - char buffer[PBMS_BLOB_URL_SIZE+1]; - u_int32_t db_id = 0; - u_int32_t tab_id = 0; - u_int64_t blob_id = 0; - u_int64_t blob_ref_id = 0; - u_int64_t blob_size = 0; - u_int32_t auth_code = 0; - u_int32_t server_id = 0; - char type, junk[5]; - int scanned; + char buffer[PBMS_BLOB_URL_SIZE+1]; + unsigned long db_id = 0; + unsigned long tab_id = 0; + unsigned long long blob_id = 0; + unsigned long long blob_ref_id = 0; + unsigned long long blob_size = 0; + unsigned long auth_code = 0; + unsigned long server_id = 0; + char type, junk[5]; + int scanned; junk[0] = 0; if (blob_url[size]) { // There is no guarantee that the URL will be null terminated. @@ -364,12 +364,12 @@ public: scanned = sscanf(blob_url, URL_FMT"%4s", &db_id, &type, &tab_id, &blob_id, &auth_code, &server_id, &blob_ref_id, &blob_size, junk); if (scanned != 8) {// If junk is found at the end this will also result in an invalid URL. - printf("Bad URL \"%s\": scanned = %d, junk: %d, %d, %d, %d\n", blob_url, scanned, junk[0], junk[1], junk[2], junk[3]); + printf("Bad URL \"%s\": scanned = %d, junk: %d, %d, %d, %d\n", blob_url, scanned, junk[0], junk[1], junk[2], junk[3]); return 0; } if (junk[0] || (type != '~' && type != '_')) { - printf("Bad URL \"%s\": scanned = %d, junk: %d, %d, %d, %d\n", blob_url, scanned, junk[0], junk[1], junk[2], junk[3]); + printf("Bad URL \"%s\": scanned = %d, junk: %d, %d, %d, %d\n", blob_url, scanned, junk[0], junk[1], junk[2], junk[3]); return 0; } === modified file 'storage/pbxt/src/pbms_enabled.cc' --- a/storage/pbxt/src/pbms_enabled.cc 2009-10-06 15:16:01 +0000 +++ b/storage/pbxt/src/pbms_enabled.cc 2009-11-24 10:55:06 +0000 @@ -29,15 +29,10 @@ * */ -/* - The following two lines backported by psergey. Remove them when we merge from PBXT again. -*/ #include "xt_config.h" -#ifdef PBMS_ENABLED -#define PBMS_API pbms_enabled_api +#ifdef PBMS_ENABLED -#include "pbms_enabled.h" #ifdef DRIZZLED #include <sys/stat.h> #include <drizzled/common_includes.h> @@ -47,11 +42,15 @@ #include <mysql/plugin.h> #define session_alloc(sess, size) thd_alloc(sess, size); #define current_session current_thd -#endif +#endif -#define GET_BLOB_FIELD(t, i) (Field_blob *)(t->field[t->s->blob_field[i]]) -#define DB_NAME(f) (f->table->s->db.str) -#define TAB_NAME(f) (*(f->table_name)) +#define GET_BLOB_FIELD(t, i) (Field_blob *)(t->field[t->s->blob_field[i]]) +#define DB_NAME(f) (f->table->s->db.str) +#define TAB_NAME(f) (*(f->table_name)) + +#define PBMS_API pbms_enabled_api + +#include "pbms_enabled.h" static PBMS_API pbms_api; @@ -242,4 +241,4 @@ void pbms_completed(TABLE *table, bool o return ; } -#endif \ No newline at end of file +#endif // PBMS_ENABLED === modified file 'storage/pbxt/src/pbms_enabled.h' --- a/storage/pbxt/src/pbms_enabled.h 2009-08-18 07:46:53 +0000 +++ b/storage/pbxt/src/pbms_enabled.h 2009-11-24 10:55:06 +0000 @@ -35,13 +35,6 @@ #include "pbms.h" -#ifdef DRIZZLED -#include <drizzled/server_includes.h> -#define TABLE Table -#else -#include <mysql_priv.h> -#endif - /* * pbms_initialize() should be called from the engines plugIn's 'init()' function. * The engine_name is the name of your engine, "PBXT" or "InnoDB" for example. === modified file 'storage/pbxt/src/pthread_xt.cc' --- a/storage/pbxt/src/pthread_xt.cc 2009-08-18 07:46:53 +0000 +++ b/storage/pbxt/src/pthread_xt.cc 2009-11-24 10:55:06 +0000 @@ -578,8 +578,8 @@ xtPublic int xt_p_mutex_unlock(xt_mutex_ xtPublic int xt_p_mutex_destroy(xt_mutex_type *mutex) { - ASSERT_NS(mutex->mu_init == 12345); - mutex->mu_init = 89898; + //ASSERT_NS(mutex->mu_init == 12345); + mutex->mu_init = 11111; #ifdef XT_THREAD_LOCK_INFO xt_thread_lock_info_free(&mutex->mu_lock_info); #endif === modified file 'storage/pbxt/src/restart_xt.cc' --- a/storage/pbxt/src/restart_xt.cc 2009-09-03 06:15:03 +0000 +++ b/storage/pbxt/src/restart_xt.cc 2009-11-27 15:37:02 +0000 @@ -34,12 +34,16 @@ #include "ha_pbxt.h" +#ifdef DRIZZLED +#include <drizzled/data_home.h> +using drizzled::plugin::Registry; +#endif + #include "xactlog_xt.h" #include "database_xt.h" #include "util_xt.h" #include "strutil_xt.h" #include "filesys_xt.h" -#include "restart_xt.h" #include "myxt_xt.h" #include "trace_xt.h" @@ -57,13 +61,27 @@ //#define PRINTF xt_ftracef //#define PRINTF xt_trace -void xt_print_bytes(xtWord1 *buf, u_int len) +/* + * ----------------------------------------------------------------------- + * GLOBALS + */ + +xtPublic int pbxt_recovery_state; + +/* + * ----------------------------------------------------------------------- + * UTILITIES + */ + +#ifdef TRACE_RECORD_DATA +static void xt_print_bytes(xtWord1 *buf, u_int len) { for (u_int i=0; i<len; i++) { PRINTF("%02x ", (u_int) *buf); buf++; } } +#endif void xt_print_log_record(xtLogID log, xtLogOffset offset, XTXactLogBufferDPtr record) { @@ -252,7 +270,7 @@ void xt_print_log_record(xtLogID log, xt rec_type = "DELETE"; break; case XT_LOG_ENT_DELETE_FL: - rec_type = "DELETE-FL-BG"; + rec_type = "DELETE-FL"; break; case XT_LOG_ENT_UPDATE_BG: rec_type = "UPDATE-BG"; @@ -320,6 +338,11 @@ void xt_print_log_record(xtLogID log, xt case XT_LOG_ENT_END_OF_LOG: rec_type = "END OF LOG"; break; + case XT_LOG_ENT_PREPARE: + rec_type = "PREPARE"; + xn_id = XT_GET_DISK_4(record->xp.xp_xact_id_4); + xn_set = TRUE; + break; } if (log) @@ -327,6 +350,8 @@ void xt_print_log_record(xtLogID log, xt PRINTF("%s ", rec_type); if (type) PRINTF("op=%lu tab=%lu %s=%lu ", (u_long) op_no, (u_long) tab_id, type, (u_long) rec_id); + else if (tab_id) + PRINTF("tab=%lu ", (u_long) tab_id); if (row_id) PRINTF("row=%lu ", (u_long) row_id); if (log_id) @@ -459,6 +484,7 @@ static xtBool xres_open_table(XTThreadPt } return OK; } + ws->ws_tab_gone = tab_id; return FAILED; } @@ -629,6 +655,7 @@ static void xres_apply_change(XTThreadPt xtWord1 *rec_data = NULL; XTTabRecFreeDPtr free_data; + ASSERT(ot->ot_thread == self); if (tab->tab_dic.dic_key_count == 0) check_index = FALSE; @@ -1300,7 +1327,7 @@ static void xres_apply_operations(XTThre * These operations are applied even though operations * in sequence are missing. */ -xtBool xres_sync_operations(XTThreadPtr self, XTDatabaseHPtr db, XTWriterStatePtr ws) +static xtBool xres_sync_operations(XTThreadPtr self, XTDatabaseHPtr db, XTWriterStatePtr ws) { u_int edx; XTTableEntryPtr te_ptr; @@ -1881,15 +1908,6 @@ xtBool XTXactRestart::xres_check_checksu void XTXactRestart::xres_recover_progress(XTThreadPtr self, XTOpenFilePtr *of, int perc) { #ifdef XT_USE_GLOBAL_DB - if (!perc) { - char file_path[PATH_MAX]; - - xt_strcpy(PATH_MAX, file_path, xres_db->db_main_path); - xt_add_pbxt_file(PATH_MAX, file_path, "recovery-progress"); - *of = xt_open_file(self, file_path, XT_FS_CREATE | XT_FS_MAKE_PATH); - xt_set_eof_file(self, *of, 0); - } - if (perc > 100) { char file_path[PATH_MAX]; @@ -1905,6 +1923,15 @@ void XTXactRestart::xres_recover_progres else { char number[40]; + if (!*of) { + char file_path[PATH_MAX]; + + xt_strcpy(PATH_MAX, file_path, xres_db->db_main_path); + xt_add_pbxt_file(PATH_MAX, file_path, "recovery-progress"); + *of = xt_open_file(self, file_path, XT_FS_CREATE | XT_FS_MAKE_PATH); + xt_set_eof_file(self, *of, 0); + } + sprintf(number, "%d", perc); if (!xt_pwrite_file(*of, 0, strlen(number), number, &self->st_statistics.st_x, self)) xt_throw(self); @@ -1927,10 +1954,11 @@ xtBool XTXactRestart::xres_restart(XTThr off_t bytes_to_read; volatile xtBool print_progress = FALSE; volatile off_t perc_size = 0, next_goal = 0; - int perc_complete = 1; + int perc_complete = 1, perc_to_write = 1; XTOpenFilePtr progress_file = NULL; xtBool min_ram_xn_id_set = FALSE; u_int log_count; + time_t start_time; memset(&ws, 0, sizeof(ws)); @@ -1955,12 +1983,11 @@ xtBool XTXactRestart::xres_restart(XTThr /* Don't print anything about recovering an empty database: */ if (bytes_to_read != 0) xt_logf(XT_NT_INFO, "PBXT: Recovering from %lu-%llu, bytes to read: %llu\n", (u_long) xres_cp_log_id, (u_llong) xres_cp_log_offset, (u_llong) bytes_to_read); - if (bytes_to_read >= 10*1024*1024) { - print_progress = TRUE; - perc_size = bytes_to_read / 100; - next_goal = perc_size; - xres_recover_progress(self, &progress_file, 0); - } + + print_progress = FALSE; + start_time = time(NULL); + perc_size = bytes_to_read / 100; + next_goal = perc_size; if (!db->db_xlog.xlog_seq_start(&ws.ws_seqread, xres_cp_log_id, xres_cp_log_offset, FALSE)) { ok = FALSE; @@ -1983,17 +2010,28 @@ xtBool XTXactRestart::xres_restart(XTThr #ifdef PRINT_LOG_ON_RECOVERY xt_print_log_record(ws.ws_seqread.xseq_rec_log_id, ws.ws_seqread.xseq_rec_log_offset, record); #endif - if (print_progress && bytes_read > next_goal) { - if (((perc_complete - 1) % 25) == 0) - xt_logf(XT_NT_INFO, "PBXT: "); - if ((perc_complete % 25) == 0) - xt_logf(XT_NT_INFO, "%2d\n", (int) perc_complete); - else - xt_logf(XT_NT_INFO, "%2d ", (int) perc_complete); - xt_log_flush(self); - xres_recover_progress(self, &progress_file, perc_complete); - next_goal += perc_size; - perc_complete++; + if (bytes_read >= next_goal) { + while (bytes_read >= next_goal) { + next_goal += perc_size; + perc_complete++; + } + if (!print_progress) { + if (time(NULL) - start_time > 2) + print_progress = TRUE; + } + if (print_progress) { + while (perc_to_write < perc_complete) { + if (((perc_to_write - 1) % 25) == 0) + xt_logf(XT_NT_INFO, "PBXT: "); + if ((perc_to_write % 25) == 0) + xt_logf(XT_NT_INFO, "%2d\n", (int) perc_to_write); + else + xt_logf(XT_NT_INFO, "%2d ", (int) perc_to_write); + xt_log_flush(self); + xres_recover_progress(self, &progress_file, perc_to_write); + perc_to_write++; + } + } } switch (record->xl.xl_status_1) { case XT_LOG_ENT_HEADER: @@ -2053,8 +2091,11 @@ xtBool XTXactRestart::xres_restart(XTThr xact->xd_end_xn_id = xn_id; xact->xd_flags |= XT_XN_XAC_ENDED | XT_XN_XAC_SWEEP; xact->xd_flags &= ~XT_XN_XAC_RECOVERED; // We can expect an end record on cleanup! + xact->xd_flags &= ~XT_XN_XAC_PREPARED; // Prepared transactions cannot be swept! if (record->xl.xl_status_1 == XT_LOG_ENT_COMMIT) xact->xd_flags |= XT_XN_XAC_COMMITTED; + if (xt_sl_get_size(db->db_xn_xa_list) > 0) + xt_xn_delete_xa_data_by_xact(db, xn_id, self); } break; case XT_LOG_ENT_CLEANUP: @@ -2071,6 +2112,14 @@ xtBool XTXactRestart::xres_restart(XTThr rec_log_id = XT_GET_DISK_4(record->xl.xl_log_id_4); xt_dl_set_to_delete(self, db, rec_log_id); break; + case XT_LOG_ENT_PREPARE: + xn_id = XT_GET_DISK_4(record->xp.xp_xact_id_4); + if ((xact = xt_xn_get_xact(db, xn_id, self))) { + xact->xd_flags |= XT_XN_XAC_PREPARED; + if (!xt_xn_store_xa_data(db, xn_id, record->xp.xp_xa_len_1, record->xp.xp_xa_data, self)) + xt_throw(self); + } + break; default: xt_xres_apply_in_order(self, &ws, ws.ws_seqread.xseq_rec_log_id, ws.ws_seqread.xseq_rec_log_offset, record); break; @@ -2534,7 +2583,8 @@ static void xres_cp_main(XTThreadPtr sel /* This condition means we could checkpoint: */ if (!(xt_sl_get_size(db->db_datalogs.dlc_to_delete) == 0 && xt_sl_get_size(db->db_datalogs.dlc_deleted) == 0 && - xt_comp_log_pos(log_id, log_offset, db->db_restart.xres_cp_log_id, db->db_restart.xres_cp_log_offset) <= 0)) + xt_comp_log_pos(log_id, log_offset, db->db_restart.xres_cp_log_id, db->db_restart.xres_cp_log_offset) <= 0) && + xt_sl_get_size(db->db_xn_xa_list) == 0) break; xres_cp_wait_for_log_writer(self, db, 400); @@ -2654,7 +2704,7 @@ xtPublic xtBool xt_begin_checkpoint(XTDa * until they are flushed. */ /* This is an alternative to the above. - if (!xt_xlog_flush_log(self)) + if (!xt_xlog_flush_log(db, self)) xt_throw(self); */ xt_lock_mutex_ns(&db->db_wr_lock); @@ -2776,6 +2826,14 @@ xtPublic xtBool xt_end_checkpoint(XTData size_t chk_size = 0; u_int no_of_logs = 0; + /* As long as we have outstanding XA transactions, we may not checkpoint! */ + if (xt_sl_get_size(db->db_xn_xa_list) > 0) { +#ifdef DEBUG + printf("Checkpoint must wait\n"); +#endif + return OK; + } + #ifdef NEVER_CHECKPOINT return OK; #endif @@ -3183,7 +3241,7 @@ xtPublic void xt_dump_xlogs(XTDatabaseHP * D A T A B A S E R E C O V E R Y T H R E A D */ -extern XTDatabaseHPtr pbxt_database; + static XTThreadPtr xres_recovery_thread; static void *xn_xres_run_recovery_thread(XTThreadPtr self) @@ -3193,18 +3251,18 @@ static void *xn_xres_run_recovery_thread if (!(mysql_thread = (THD *) myxt_create_thread())) xt_throw(self); - while (!xres_recovery_thread->t_quit && !ha_resolve_by_legacy_type(mysql_thread, DB_TYPE_PBXT)) - xt_sleep_milli_second(1); + myxt_wait_pbxt_plugin_slot_assigned(self); if (!xres_recovery_thread->t_quit) { - /* {GLOBAL-DB} - * It can happen that something will just get in before this - * thread and open/recover the database! - */ - if (!pbxt_database) { - try_(a) { + try_(a) { + /* {GLOBAL-DB} + * It can happen that something will just get in before this + * thread and open/recover the database! + */ + if (!pbxt_database) { xt_open_database(self, mysql_real_data_home, TRUE); - /* This can be done at the same time by a foreground thread, + /* {GLOBAL-DB} + * This can be done at the same time as the recovery thread, * strictly speaking I need a lock. */ if (!pbxt_database) { @@ -3212,11 +3270,22 @@ static void *xn_xres_run_recovery_thread xt_heap_reference(self, pbxt_database); } } - catch_(a) { - xt_log_and_clear_exception(self); - } - cont_(a); + else + xt_use_database(self, pbxt_database, XT_FOR_USER); + + pbxt_recovery_state = XT_RECOVER_DONE; + + /* {WAIT-FOR-SW-AFTER-RECOV} + * Moved to here... + */ + xt_wait_for_sweeper(self, self->st_database, 0); + + pbxt_recovery_state = XT_RECOVER_SWEPT; } + catch_(a) { + xt_log_and_clear_exception(self); + } + cont_(a); } /* @@ -3261,11 +3330,12 @@ xtPublic void xt_xres_start_database_rec sprintf(name, "DB-RECOVERY-%s", xt_last_directory_of_path(mysql_real_data_home)); xt_remove_dir_char(name); + pbxt_recovery_state = XT_RECOVER_PENDING; xres_recovery_thread = xt_create_daemon(self, name); xt_run_thread(self, xres_recovery_thread, xn_xres_run_recovery_thread); } -xtPublic void xt_xres_wait_for_recovery(XTThreadPtr self) +xtPublic void xt_xres_terminate_recovery(XTThreadPtr self) { XTThreadPtr thr_rec; === modified file 'storage/pbxt/src/restart_xt.h' --- a/storage/pbxt/src/restart_xt.h 2009-09-03 06:15:03 +0000 +++ b/storage/pbxt/src/restart_xt.h 2009-11-24 10:55:06 +0000 @@ -37,6 +37,8 @@ struct XTOpenTable; struct XTDatabase; struct XTTable; +extern int pbxt_recovery_state; + typedef struct XTWriterState { struct XTDatabase *ws_db; xtBool ws_in_recover; @@ -132,6 +134,16 @@ void xt_print_log_record(xtLogID log, of void xt_dump_xlogs(struct XTDatabase *db, xtLogID start_log); void xt_xres_start_database_recovery(XTThreadPtr self); -void xt_xres_wait_for_recovery(XTThreadPtr self); +void xt_xres_terminate_recovery(XTThreadPtr self); + +#define XT_RECOVER_PENDING 0 +#define XT_RECOVER_DONE 1 +#define XT_RECOVER_SWEPT 2 + +inline void xt_xres_wait_for_recovery(XTThreadPtr XT_UNUSED(self), int state) +{ + while (pbxt_recovery_state < state) + xt_sleep_milli_second(100); +} #endif === modified file 'storage/pbxt/src/strutil_xt.cc' --- a/storage/pbxt/src/strutil_xt.cc 2009-10-26 11:35:42 +0000 +++ b/storage/pbxt/src/strutil_xt.cc 2009-11-24 10:55:06 +0000 @@ -21,8 +21,10 @@ * H&G2JCtL */ -#include "mysql_priv.h" #include "xt_config.h" + +#include <stdio.h> +#include <string.h> #include <ctype.h> #include "strutil_xt.h" @@ -107,13 +109,17 @@ xtPublic void xt_2nd_last_name_of_path(s *dest = 0; return; } - /* If temporary file */ - if (!is_prefix(path, mysql_data_home) && + + /* {INVALID-OLD-TABLE-FIX} + * I have changed the implementation of + * this bug fix (see {INVALID-OLD-TABLE-FIX}). + if (!is_prefix(path, mysql_data_home) && !is_prefix(path, mysql_real_data_home)) { *dest= 0; return; } + */ ptr = path + len - 1; while (ptr != path && !XT_IS_DIR_CHAR(*ptr)) @@ -374,7 +380,7 @@ xtPublic void xt_int8_to_byte_size(xtInt /* Version number must also be set in configure.in! */ xtPublic c_char *xt_get_version(void) { - return "1.0.08d RC"; + return "1.0.09f RC"; } /* Copy and URL decode! */ === modified file 'storage/pbxt/src/systab_xt.cc' --- a/storage/pbxt/src/systab_xt.cc 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/systab_xt.cc 2009-11-24 10:55:06 +0000 @@ -130,7 +130,7 @@ static int pbms_discover_handler(handler * MYSQL UTILITIES */ -void xt_my_set_notnull_in_record(Field *field, char *record) +static void xt_my_set_notnull_in_record(Field *field, char *record) { if (field->null_ptr) record[(uint) (field->null_ptr - (uchar *) field->table->record[0])] &= (uchar) ~field->null_bit; @@ -518,7 +518,7 @@ bool XTStatisticsTable::seqScanRead(xtWo * SYSTEM TABLE SHARES */ -void st_path_to_table_name(size_t size, char *buffer, const char *path) +static void st_path_to_table_name(size_t size, char *buffer, const char *path) { char *str; === modified file 'storage/pbxt/src/tabcache_xt.cc' --- a/storage/pbxt/src/tabcache_xt.cc 2009-09-03 06:15:03 +0000 +++ b/storage/pbxt/src/tabcache_xt.cc 2009-11-27 15:37:02 +0000 @@ -590,7 +590,7 @@ xtBool XTTabCache::tc_fetch(XT_ROW_REC_F * So there could be a deadlock if I don't flush the log! */ if ((self = xt_get_self())) { - if (!xt_xlog_flush_log(self)) + if (!xt_xlog_flush_log(tci_table->tab_db, self)) goto failed; } @@ -1150,6 +1150,8 @@ static void *tabc_fr_run_thread(XTThread int count; void *mysql_thread; + myxt_wait_pbxt_plugin_slot_assigned(self); + mysql_thread = myxt_create_thread(); while (!self->t_quit) { === modified file 'storage/pbxt/src/tabcache_xt.h' --- a/storage/pbxt/src/tabcache_xt.h 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/tabcache_xt.h 2009-11-24 10:55:06 +0000 @@ -168,7 +168,7 @@ typedef struct XTTableSeq { #define TAB_CAC_UNLOCK(i, o) xt_xsmutex_unlock(i, o) #elif defined(TAB_CAC_USE_PTHREAD_RW) #define TAB_CAC_LOCK_TYPE xt_rwlock_type -#define TAB_CAC_INIT_LOCK(s, i) xt_init_rwlock(s, i) +#define TAB_CAC_INIT_LOCK(s, i) xt_init_rwlock_with_autoname(s, i) #define TAB_CAC_FREE_LOCK(s, i) xt_free_rwlock(i) #define TAB_CAC_READ_LOCK(i, o) xt_slock_rwlock_ns(i) #define TAB_CAC_WRITE_LOCK(i, o) xt_xlock_rwlock_ns(i) === modified file 'storage/pbxt/src/table_xt.cc' --- a/storage/pbxt/src/table_xt.cc 2009-08-18 07:46:53 +0000 +++ b/storage/pbxt/src/table_xt.cc 2009-11-25 15:40:51 +0000 @@ -35,7 +35,6 @@ #include <drizzled/common.h> #include <mysys/thr_lock.h> #include <drizzled/dtcollation.h> -#include <drizzled/plugin/storage_engine.h> #else #include "mysql_priv.h" #endif @@ -48,7 +47,6 @@ #include "cache_xt.h" #include "trace_xt.h" #include "index_xt.h" -#include "restart_xt.h" #include "systab_xt.h" #ifdef DEBUG @@ -2347,7 +2345,7 @@ xtPublic void xt_flush_table(XTThreadPtr } -xtPublic XTOpenTablePtr tab_open_table(XTTableHPtr tab) +static XTOpenTablePtr tab_open_table(XTTableHPtr tab) { volatile XTOpenTablePtr ot; XTThreadPtr self; @@ -2588,7 +2586,7 @@ xtPublic xtBool xt_tab_put_log_op_rec_da return FAILED; } - return xt_xlog_modify_table(ot, status, op_seq, free_rec_id, rec_id, size, buffer); + return xt_xlog_modify_table(tab->tab_id, status, op_seq, free_rec_id, rec_id, size, buffer, ot->ot_thread); } xtPublic xtBool xt_tab_put_log_rec_data(XTOpenTablePtr ot, u_int status, xtRecordID free_rec_id, xtRecordID rec_id, size_t size, xtWord1 *buffer, xtOpSeqNo *op_seq) @@ -2606,7 +2604,7 @@ xtPublic xtBool xt_tab_put_log_rec_data( return FAILED; } - return xt_xlog_modify_table(ot, status, *op_seq, free_rec_id, rec_id, size, buffer); + return xt_xlog_modify_table(tab->tab_id, status, *op_seq, free_rec_id, rec_id, size, buffer, ot->ot_thread); } xtPublic xtBool xt_tab_get_rec_data(XTOpenTablePtr ot, xtRecordID rec_id, size_t size, xtWord1 *buffer) @@ -3541,7 +3539,7 @@ xtPublic xtBool xt_tab_free_row(XTOpenTa tab->tab_row_fnum++; xt_unlock_mutex_ns(&tab->tab_row_lock); - if (!xt_xlog_modify_table(ot, XT_LOG_ENT_ROW_FREED, op_seq, 0, row_id, sizeof(XTTabRowRefDRec), (xtWord1 *) &free_row)) + if (!xt_xlog_modify_table(tab->tab_id, XT_LOG_ENT_ROW_FREED, op_seq, 0, row_id, sizeof(XTTabRowRefDRec), (xtWord1 *) &free_row, ot->ot_thread)) return FAILED; return OK; @@ -3791,7 +3789,7 @@ xtPublic int xt_tab_remove_record(XTOpen xt_unlock_mutex_ns(&tab->tab_rec_lock); free_rec->rf_rec_type_1 = old_rec_type; - return xt_xlog_modify_table(ot, XT_LOG_ENT_REC_REMOVED_BI, op_seq, (xtRecordID) new_rec_type, rec_id, rec_size, ot->ot_row_rbuffer); + return xt_xlog_modify_table(tab->tab_id, XT_LOG_ENT_REC_REMOVED_BI, op_seq, (xtRecordID) new_rec_type, rec_id, rec_size, ot->ot_row_rbuffer, ot->ot_thread); } static xtRowID tab_new_row(XTOpenTablePtr ot, XTTableHPtr tab) @@ -3837,7 +3835,7 @@ static xtRowID tab_new_row(XTOpenTablePt op_seq = tab->tab_seq.ts_get_op_seq(); xt_unlock_mutex_ns(&tab->tab_row_lock); - if (!xt_xlog_modify_table(ot, status, op_seq, next_row_id, row_id, 0, NULL)) + if (!xt_xlog_modify_table(tab->tab_id, status, op_seq, next_row_id, row_id, 0, NULL, ot->ot_thread)) return 0; XT_DISABLED_TRACE(("new row tx=%d row=%d\n", (int) ot->ot_thread->st_xact_data->xd_start_xn_id, (int) row_id)); @@ -3868,7 +3866,7 @@ xtPublic xtBool xt_tab_set_row(XTOpenTab if (!tab->tab_rows.xt_tc_write(ot->ot_row_file, row_id, 0, sizeof(XTTabRowRefDRec), (xtWord1 *) &row_buf, &op_seq, TRUE, ot->ot_thread)) return FAILED; - return xt_xlog_modify_table(ot, status, op_seq, 0, row_id, sizeof(XTTabRowRefDRec), (xtWord1 *) &row_buf); + return xt_xlog_modify_table(tab->tab_id, status, op_seq, 0, row_id, sizeof(XTTabRowRefDRec), (xtWord1 *) &row_buf, ot->ot_thread); } xtPublic xtBool xt_tab_free_record(XTOpenTablePtr ot, u_int status, xtRecordID rec_id, xtBool clean_delete) @@ -3937,7 +3935,7 @@ xtPublic xtBool xt_tab_free_record(XTOpe tab->tab_rec_fnum++; xt_unlock_mutex_ns(&tab->tab_rec_lock); - if (!xt_xlog_modify_table(ot, status, op_seq, rec_id, rec_id, sizeof(XTactFreeRecEntryDRec) - offsetof(XTactFreeRecEntryDRec, fr_stat_id_1), &free_rec.fr_stat_id_1)) + if (!xt_xlog_modify_table(tab->tab_id, status, op_seq, rec_id, rec_id, sizeof(XTactFreeRecEntryDRec) - offsetof(XTactFreeRecEntryDRec, fr_stat_id_1), &free_rec.fr_stat_id_1, ot->ot_thread)) return FAILED; } return OK; @@ -4016,7 +4014,7 @@ static xtBool tab_add_record(XTOpenTable } xt_unlock_mutex_ns(&tab->tab_rec_lock); - if (!xt_xlog_modify_table(ot, status, op_seq, next_rec_id, rec_id, rec_info->ri_rec_buf_size, (xtWord1 *) rec_info->ri_fix_rec_buf)) + if (!xt_xlog_modify_table(tab->tab_id, status, op_seq, next_rec_id, rec_id, rec_info->ri_rec_buf_size, (xtWord1 *) rec_info->ri_fix_rec_buf, ot->ot_thread)) return FAILED; if (rec_info->ri_ext_rec) { @@ -4932,6 +4930,12 @@ xtPublic void xt_tab_seq_exit(XTOpenTabl #endif #endif +xtPublic void xt_tab_seq_repeat(XTOpenTablePtr ot) +{ + ot->ot_seq_rec_id--; + ot->ot_seq_offset -= ot->ot_table->tab_dic.dic_rec_size; +} + xtPublic xtBool xt_tab_seq_next(XTOpenTablePtr ot, xtWord1 *buffer, xtBool *eof) { register XTTableHPtr tab = ot->ot_table; @@ -5094,7 +5098,7 @@ static xtBool tab_exec_repair_pending(XT return FALSE; } else { - if (!xt_open_file_ns(&of, file_path, XT_FS_DEFAULT)) + if (!xt_open_file_ns(&of, file_path, XT_FS_DEFAULT | XT_FS_MISSING_OK)) return FALSE; } if (!of) @@ -5190,15 +5194,76 @@ static xtBool tab_exec_repair_pending(XT return FALSE; } -xtPublic void tab_make_table_name(XTTableHPtr tab, char *table_name, size_t size) +static void tab_make_table_name(XTTableHPtr tab, char *table_name, size_t size) { - char name_buf[XT_IDENTIFIER_NAME_SIZE*3+3]; + char *nptr; - xt_2nd_last_name_of_path(sizeof(name_buf), name_buf, tab->tab_name->ps_path); - myxt_static_convert_file_name(name_buf, table_name, size); - xt_strcat(size, table_name, "."); - myxt_static_convert_file_name(xt_last_name_of_path(tab->tab_name->ps_path), name_buf, sizeof(name_buf)); - xt_strcat(size, table_name, name_buf); + nptr = xt_last_name_of_path(tab->tab_name->ps_path); + if (xt_starts_with(nptr, "#sql")) { + /* {INVALID-OLD-TABLE-FIX} + * Temporary files can have strange paths, for example + * ..../var/tmp/mysqld.1/#sqldaec_1_6 + * This occurs, for example, occurs when the temp_table.test is + * run using the PBXT suite in MariaDB: + * ./mtr --suite=pbxt --do-test=temp_table + * + * Calling myxt_static_convert_file_name, with a '.', in the name + * causes the error: + * [ERROR] Invalid (old?) table or database name 'mysqld.1' + * To prevent this, we do not convert the temporary + * table names using the mysql functions. + * + * Note, this bug was found by Monty, and fixed by modifying + * xt_2nd_last_name_of_path(), see {INVALID-OLD-TABLE-FIX}. + * + */ + xt_2nd_last_name_of_path(size, table_name, tab->tab_name->ps_path); + xt_strcat(size, table_name, "."); + xt_strcat(size, table_name, nptr); + } + else { + char name_buf[XT_TABLE_NAME_SIZE*3+3]; + char *part_ptr; + size_t len; + + xt_2nd_last_name_of_path(sizeof(name_buf), name_buf, tab->tab_name->ps_path); + myxt_static_convert_file_name(name_buf, table_name, size); + xt_strcat(size, table_name, "."); + + /* Handle partition extensions to table names: */ + if ((part_ptr = strstr(nptr, "#P#"))) + xt_strncpy(sizeof(name_buf), name_buf, nptr, part_ptr - nptr); + else + xt_strcpy(sizeof(name_buf), name_buf, nptr); + + len = strlen(table_name); + myxt_static_convert_file_name(name_buf, table_name + len, size - len); + + if (part_ptr) { + /* Add the partition extension (which is relevant to the engine). */ + char *sub_part_ptr; + + part_ptr += 3; + if ((sub_part_ptr = strstr(part_ptr, "#SP#"))) + xt_strncpy(sizeof(name_buf), name_buf, part_ptr, sub_part_ptr - part_ptr); + else + xt_strcpy(sizeof(name_buf), name_buf, part_ptr); + + xt_strcat(size, table_name, " ("); + len = strlen(table_name); + myxt_static_convert_file_name(name_buf, table_name + len, size - len); + + if (sub_part_ptr) { + + sub_part_ptr += 4; + xt_strcat(size, table_name, " - "); + len = strlen(table_name); + myxt_static_convert_file_name(sub_part_ptr, table_name + len, size - len); + } + + xt_strcat(size, table_name, ")"); + } + } } xtPublic xtBool xt_tab_is_table_repair_pending(XTTableHPtr tab) === modified file 'storage/pbxt/src/table_xt.h' --- a/storage/pbxt/src/table_xt.h 2009-08-18 07:46:53 +0000 +++ b/storage/pbxt/src/table_xt.h 2009-11-24 10:55:06 +0000 @@ -127,7 +127,7 @@ struct XTTablePath; #define XT_TAB_ROW_UNLOCK(i, s) xt_xsmutex_unlock(i, (s)->t_id) #elif defined(XT_TAB_ROW_USE_PTHREAD_RW) #define XT_TAB_ROW_LOCK_TYPE xt_rwlock_type -#define XT_TAB_ROW_INIT_LOCK(s, i) xt_init_rwlock(s, i) +#define XT_TAB_ROW_INIT_LOCK(s, i) xt_init_rwlock_with_autoname(s, i) #define XT_TAB_ROW_FREE_LOCK(s, i) xt_free_rwlock(i) #define XT_TAB_ROW_READ_LOCK(i, s) xt_slock_rwlock_ns(i) #define XT_TAB_ROW_WRITE_LOCK(i, s) xt_xlock_rwlock_ns(i) @@ -528,13 +528,14 @@ xtBool xt_table_exists(struct XTDatab void xt_enum_tables_init(u_int *edx); XTTableEntryPtr xt_enum_tables_next(struct XTThread *self, struct XTDatabase *db, u_int *edx); -void xt_enum_files_of_tables_init(struct XTDatabase *db, char *tab_name, xtTableID tab_id, XTFilesOfTablePtr ft); +void xt_enum_files_of_tables_init(XTPathStrPtr tab_name, xtTableID tab_id, XTFilesOfTablePtr ft); xtBool xt_enum_files_of_tables_next(XTFilesOfTablePtr ft); xtBool xt_tab_seq_init(XTOpenTablePtr ot); void xt_tab_seq_reset(XTOpenTablePtr ot); void xt_tab_seq_exit(XTOpenTablePtr ot); xtBool xt_tab_seq_next(XTOpenTablePtr ot, xtWord1 *buffer, xtBool *eof); +void xt_tab_seq_repeat(XTOpenTablePtr ot); xtBool xt_tab_new_record(XTOpenTablePtr ot, xtWord1 *buffer); xtBool xt_tab_delete_record(XTOpenTablePtr ot, xtWord1 *buffer); === modified file 'storage/pbxt/src/thread_xt.cc' --- a/storage/pbxt/src/thread_xt.cc 2009-10-06 15:16:01 +0000 +++ b/storage/pbxt/src/thread_xt.cc 2009-11-24 10:55:06 +0000 @@ -75,6 +75,13 @@ static xt_mutex_type thr_array_lock; /* Global accumulated statistics: */ static XTStatisticsRec thr_statistics; +#ifdef DEBUG +static void break_in_assertion(c_char *expr, c_char *func, c_char *file, u_int line) +{ + printf("%s(%s:%d) %s\n", func, file, (int) line, expr); +} +#endif + /* * ----------------------------------------------------------------------- * Error logging @@ -658,6 +665,9 @@ static c_char *thr_get_err_string(int xt case XT_ERR_FK_REF_TEMP_TABLE: str = "Foreign key may not reference temporary table"; break; case XT_ERR_MYSQL_SHUTDOWN: str = "Cannot open table, MySQL has shutdown"; break; case XT_ERR_MYSQL_NO_THREAD: str = "Cannot create thread, MySQL has shutdown"; break; + case XT_ERR_BUFFER_TOO_SMALL: str = "System backup buffer too small"; break; + case XT_ERR_BAD_BACKUP_FORMAT: str = "Unknown or corrupt backup format, restore aborted"; break; + case XT_ERR_PBXT_NOT_INSTALLED: str = "PBXT plugin is not installed"; break; default: str = "Unknown XT error"; break; } return str; @@ -862,6 +872,11 @@ xtPublic xtBool xt_exception_errno(XTExc return FAILED; } +xtPublic void xt_exception_xterr(XTExceptionPtr e, XTThreadPtr self, c_char *func, c_char *file, u_int line, int xt_err) +{ + xt_exception_error(e, self, func, file, line, xt_err, 0, thr_get_err_string(xt_err)); +} + /* * ----------------------------------------------------------------------- * LOG ERRORS @@ -887,7 +902,7 @@ xtPublic xtBool xt_assert(XTThreadPtr se #ifdef DEBUG //xt_set_fflush(TRUE); //xt_dump_trace(); - printf("%s(%s:%d) %s\n", func, file, (int) line, expr); + break_in_assertion(expr, func, file, line); #ifdef CRASH_ON_ASSERT abort(); #endif === modified file 'storage/pbxt/src/thread_xt.h' --- a/storage/pbxt/src/thread_xt.h 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/thread_xt.h 2009-11-24 10:55:06 +0000 @@ -536,6 +536,8 @@ extern struct XTThread **xt_thr_array; * Function prototypes */ +extern "C" void *thr_main(void *data); + void xt_get_now(char *buffer, size_t len); xtBool xt_init_logging(void); void xt_exit_logging(void); @@ -583,6 +585,7 @@ void xt_register_xterr(c_char *func, c void xt_exceptionf(XTExceptionPtr e, XTThreadPtr self, c_char *func, c_char *file, u_int line, int xt_err, int sys_err, c_char *fmt, ...); void xt_exception_error(XTExceptionPtr e, XTThreadPtr self, c_char *func, c_char *file, u_int line, int xt_err, int sys_err, c_char *msg); xtBool xt_exception_errno(XTExceptionPtr e, XTThreadPtr self, c_char *func, c_char *file, u_int line, int err); +void xt_exception_xterr(XTExceptionPtr e, XTThreadPtr self, c_char *func, c_char *file, u_int line, int xt_err); void xt_log_errno(XTThreadPtr self, c_char *func, c_char *file, u_int line, int err); @@ -610,7 +613,7 @@ void xt_critical_wait(void); void xt_yield(void); void xt_sleep_milli_second(u_int t); xtBool xt_suspend(XTThreadPtr self); -xtBool xt_unsuspend(XTThreadPtr self, XTThreadPtr target); +xtBool xt_unsuspend(XTThreadPtr target); void xt_lock_thread(XTThreadPtr thread); void xt_unlock_thread(XTThreadPtr thread); xtBool xt_wait_thread(XTThreadPtr thread); === modified file 'storage/pbxt/src/util_xt.cc' --- a/storage/pbxt/src/util_xt.cc 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/util_xt.cc 2009-11-24 10:55:06 +0000 @@ -150,6 +150,23 @@ xtPublic xtWord1 xt_get_checksum1(xtWord return (xtWord1) (sum ^ (sum >> 24) ^ (sum >> 16) ^ (sum >> 8)); } +xtPublic xtWord4 xt_get_checksum4(xtWord1 *data, size_t len) +{ + register xtWord4 sum = 0, g; + xtWord1 *chk; + + chk = data + len - 1; + while (chk > data) { + sum = (sum << 4) + *chk; + if ((g = sum & 0xF0000000)) { + sum = sum ^ (g >> 24); + sum = sum ^ g; + } + chk--; + } + return sum; +} + /* * --------------- Data Buffer ------------------ */ === modified file 'storage/pbxt/src/util_xt.h' --- a/storage/pbxt/src/util_xt.h 2009-03-26 12:18:01 +0000 +++ b/storage/pbxt/src/util_xt.h 2009-11-24 10:55:06 +0000 @@ -39,6 +39,7 @@ xtWord4 xt_file_name_to_id(char *file_na xtBool xt_time_difference(register xtWord4 now, register xtWord4 then); xtWord2 xt_get_checksum(xtWord1 *data, size_t len, u_int interval); xtWord1 xt_get_checksum1(xtWord1 *data, size_t len); +xtWord4 xt_get_checksum4(xtWord1 *data, size_t len); typedef struct XTDataBuffer { size_t db_size; === modified file 'storage/pbxt/src/xaction_xt.cc' --- a/storage/pbxt/src/xaction_xt.cc 2009-09-03 06:15:03 +0000 +++ b/storage/pbxt/src/xaction_xt.cc 2009-11-24 10:55:06 +0000 @@ -1075,6 +1075,7 @@ xtPublic void xt_xn_init_db(XTThreadPtr #endif xt_spinlock_init_with_autoname(self, &db->db_xn_id_lock); xt_spinlock_init_with_autoname(self, &db->db_xn_wait_spinlock); + xt_init_mutex_with_autoname(self, &db->db_xn_xa_lock); //xt_init_mutex_with_autoname(self, &db->db_xn_wait_lock); //xt_init_cond(self, &db->db_xn_wait_cond); xt_init_mutex_with_autoname(self, &db->db_sw_lock); @@ -1096,6 +1097,9 @@ xtPublic void xt_xn_init_db(XTThreadPtr } } + /* Create a sorted list for XA transactions recovered: */ + db->db_xn_xa_list = xt_new_sortedlist(self, sizeof(XTXactXARec), 100, 50, xt_xn_xa_compare, db, NULL, FALSE, FALSE); + /* Initialize the data logs: */ db->db_datalogs.dlc_init(self, db); @@ -1146,6 +1150,7 @@ xtPublic void xt_xn_exit_db(XTThreadPtr printf("=========> MAX TXs NOT CLEAN: %lu\n", not_clean_max); printf("=========> MAX TXs IN RAM: %lu\n", in_ram_max); #endif + XTXactPreparePtr xap, xap_next; xt_stop_sweeper(self, db); // Should be done already! xt_stop_writer(self, db); // Should be done already! @@ -1187,6 +1192,19 @@ xtPublic void xt_xn_exit_db(XTThreadPtr xt_free_mutex(&db->db_sw_lock); //xt_free_cond(&db->db_xn_wait_cond); //xt_free_mutex(&db->db_xn_wait_lock); + xt_free_mutex(&db->db_xn_xa_lock); + for (u_int i=0; i<XT_XA_HASH_TAB_SIZE; i++) { + xap = db->db_xn_xa_table[i]; + while (xap) { + xap_next = xap->xp_next; + xt_free(self, xap); + xap = xap_next; + } + } + if (db->db_xn_xa_list) { + xt_free_sortedlist(self, db->db_xn_xa_list); + db->db_xn_xa_list = NULL; + } xt_spinlock_free(self, &db->db_xn_wait_spinlock); xt_spinlock_free(self, &db->db_xn_id_lock); #ifdef DEBUG_RAM_LIST @@ -1428,7 +1446,7 @@ static xtBool xn_end_xact(XTThreadPtr th wait_xn_id = thread->st_prev_xact[thread->st_last_xact]; thread->st_prev_xact[thread->st_last_xact] = xn_id; /* This works because XT_MAX_XACT_BEHIND == 2! */ - ASSERT_NS((thread->st_last_xact + 1) % XT_MAX_XACT_BEHIND == thread->st_last_xact ^ 1); + ASSERT_NS((thread->st_last_xact + 1) % XT_MAX_XACT_BEHIND == (thread->st_last_xact ^ 1)); thread->st_last_xact ^= 1; while (xt_xn_is_before(db->db_xn_to_clean_id, wait_xn_id) && (db->db_sw_faster & XT_SW_TOO_FAR_BEHIND)) { xt_critical_wait(); @@ -1548,6 +1566,149 @@ xtPublic int xt_xn_status(XTOpenTablePtr return XT_XN_ABORTED; } +/* ---------------------------------------------------------------------- + * XA Functionality + */ + +xtPublic int xt_xn_xa_compare(XTThreadPtr XT_UNUSED(self), register const void *XT_UNUSED(thunk), register const void *a, register const void *b) +{ + xtXactID *x = (xtXactID *) a; + XTXactXAPtr y = (XTXactXAPtr) b; + + if (*x == y->xx_xact_id) + return 0; + if (xt_xn_is_before(*x, y->xx_xact_id)) + return -1; + return 1; +} + +xtPublic xtBool xt_xn_prepare(int len, xtWord1 *xa_data, XTThreadPtr thread) +{ + XTXactDataPtr xact; + + ASSERT_NS(thread->st_xact_data); + if ((xact = thread->st_xact_data)) { + xtXactID xn_id = xact->xd_start_xn_id; + + /* Only makes sense if the transaction has already been logged: */ + if ((thread->st_xact_data->xd_flags & XT_XN_XAC_LOGGED)) { + if (!xt_xlog_modify_table(0, XT_LOG_ENT_PREPARE, xn_id, 0, 0, len, xa_data, thread)) + return FAILED; + } + } + return OK; +} + +xtPublic xtBool xt_xn_store_xa_data(XTDatabaseHPtr db, xtXactID xact_id, int len, xtWord1 *xa_data, XTThreadPtr XT_UNUSED(thread)) +{ + XTXactPreparePtr xap; + u_int idx; + XTXactXARec xx; + + if (!(xap = (XTXactPreparePtr) xt_malloc_ns(offsetof(XTXactPrepareRec, xp_xa_data) + len))) + return FAILED; + xap->xp_xact_id = xact_id; + xap->xp_hash = xt_get_checksum4(xa_data, len); + xap->xp_data_len = len; + memcpy(xap->xp_xa_data, xa_data, len); + xx.xx_xact_id = xact_id; + xx.xx_xa_ptr = xap; + + idx = xap->xp_hash % XT_XA_HASH_TAB_SIZE; + xt_lock_mutex_ns(&db->db_xn_xa_lock); + if (!xt_sl_insert(NULL, db->db_xn_xa_list, &xact_id, &xx)) { + xt_unlock_mutex_ns(&db->db_xn_xa_lock); + xt_free_ns(xap); + } + xap->xp_next = db->db_xn_xa_table[idx]; + db->db_xn_xa_table[idx] = xap; + xt_unlock_mutex_ns(&db->db_xn_xa_lock); + return OK; +} + +xtPublic void xt_xn_delete_xa_data_by_xact(XTDatabaseHPtr db, xtXactID xact_id, XTThreadPtr thread) +{ + XTXactXAPtr xx; + + xt_lock_mutex_ns(&db->db_xn_xa_lock); + if (!(xx = (XTXactXAPtr) xt_sl_find(NULL, db->db_xn_xa_list, &xact_id))) + return; + xt_xn_delete_xa_data(db, xx->xx_xa_ptr, TRUE, thread); +} + +xtPublic void xt_xn_delete_xa_data(XTDatabaseHPtr db, XTXactPreparePtr xap, xtBool unlock, XTThreadPtr XT_UNUSED(thread)) +{ + u_int idx; + XTXactPreparePtr xap_ptr, xap_pptr = NULL; + + xt_sl_delete(NULL, db->db_xn_xa_list, &xap->xp_xact_id); + idx = xap->xp_hash % XT_XA_HASH_TAB_SIZE; + xap_ptr = db->db_xn_xa_table[idx]; + while (xap_ptr) { + if (xap_ptr == xap) + break; + xap_pptr = xap_ptr; + xap_ptr = xap_ptr->xp_next; + } + if (xap_ptr) { + if (xap_pptr) + xap_pptr->xp_next = xap_ptr->xp_next; + else + db->db_xn_xa_table[idx] = xap_ptr->xp_next; + xt_free_ns(xap); + } + if (unlock) + xt_unlock_mutex_ns(&db->db_xn_xa_lock); +} + +xtPublic XTXactPreparePtr xt_xn_find_xa_data(XTDatabaseHPtr db, int len, xtWord1 *xa_data, xtBool lock, XTThreadPtr XT_UNUSED(thread)) +{ + xtWord4 hash; + XTXactPreparePtr xap; + u_int idx; + + if (lock) + xt_lock_mutex_ns(&db->db_xn_xa_lock); + hash = xt_get_checksum4(xa_data, len); + idx = hash % XT_XA_HASH_TAB_SIZE; + xap = db->db_xn_xa_table[idx]; + while (xap) { + if (xap->xp_hash == hash && + xap->xp_data_len == len && + memcmp(xap->xp_xa_data, xa_data, len) == 0) { + break; + } + xap = xap->xp_next; + } + + return xap; +} + +xtPublic XTXactPreparePtr xt_xn_enum_xa_data(XTDatabaseHPtr db, XTXactEnumXAPtr exa) +{ + XTXactXAPtr xx; + + if (!exa->exa_locked) { + xt_lock_mutex_ns(&db->db_xn_xa_lock); + exa->exa_locked = TRUE; + } + + if ((xx = (XTXactXAPtr) xt_sl_item_at(db->db_xn_xa_list, exa->exa_index))) { + exa->exa_index++; + return xx->xx_xa_ptr; + } + + if (exa->exa_locked) { + exa->exa_locked = FALSE; + xt_unlock_mutex_ns(&db->db_xn_xa_lock); + } + return NULL; +} + +/* ---------------------------------------------------------------------- + * S W E E P E R F U N C T I O N S + */ + xtPublic xtWord8 xt_xn_bytes_to_sweep(XTDatabaseHPtr db, XTThreadPtr thread) { xtXactID xn_id; @@ -2047,7 +2208,7 @@ static xtBool xn_sw_cleanup_variation(XT if(!tab->tab_recs.xt_tc_write_cond(self, ot->ot_rec_file, rec_id, rec_head.tr_rec_type_1, &op_seq, xn_id, row_id, stat_id, rec_type)) /* this means record was not updated by xt_tc_write_bor and doesn't need to */ break; - if (!xt_xlog_modify_table(ot, XT_LOG_ENT_REC_CLEANED_1, op_seq, 0, rec_id, 1, &rec_head.tr_rec_type_1)) + if (!xt_xlog_modify_table(tab->tab_id, XT_LOG_ENT_REC_CLEANED_1, op_seq, 0, rec_id, 1, &rec_head.tr_rec_type_1, self)) throw_(); xn_sw_clean_indices(self, ot, rec_id, row_id, rec_buf, ss->ss_databuf.db_data); break; @@ -2397,8 +2558,10 @@ static void xn_sw_main(XTThreadPtr self) if ((xact = xt_xn_get_xact(db, db->db_xn_to_clean_id, self))) { xtXactID xn_id; - if (!(xact->xd_flags & XT_XN_XAC_SWEEP)) - /* Transaction has not yet ending, and ready to sweep. */ + /* The sweep flag is set when the transaction is ready for sweeping. + * Prepared transactions may not be swept! + */ + if (!(xact->xd_flags & XT_XN_XAC_SWEEP) || (xact->xd_flags & XT_XN_XAC_PREPARED)) goto sleep; /* Check if we can cleanup the transaction. @@ -2493,7 +2656,7 @@ static void xn_sw_main(XTThreadPtr self) * we flush the log. */ if (now >= idle_start + 2) { - if (!xt_xlog_flush_log(self)) + if (!xt_xlog_flush_log(db, self)) xt_throw(self); ss->ss_flush_pending = FALSE; } @@ -2516,7 +2679,7 @@ static void xn_sw_main(XTThreadPtr self) } if (ss->ss_flush_pending) { - xt_xlog_flush_log(self); + xt_xlog_flush_log(db, self); ss->ss_flush_pending = FALSE; } === modified file 'storage/pbxt/src/xaction_xt.h' --- a/storage/pbxt/src/xaction_xt.h 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/xaction_xt.h 2009-11-24 10:55:06 +0000 @@ -87,6 +87,7 @@ struct XTOpenTable; #define XT_XN_XAC_CLEANED 8 /* The transaction has been cleaned. */ #define XT_XN_XAC_RECOVERED 16 /* This transaction was detected on recovery. */ #define XT_XN_XAC_SWEEP 32 /* End ID has been set, OK to sweep. */ +#define XT_XN_XAC_PREPARED 64 /* The transaction was prepared (used only by recovery). */ #define XT_XN_VISIBLE 0 /* The transaction is committed, and the record is visible. */ #define XT_XN_NOT_VISIBLE 1 /* The transaction is committed, but not visible. */ @@ -95,6 +96,24 @@ struct XTOpenTable; #define XT_XN_OTHER_UPDATE 4 /* The record was updated by someone else. */ #define XT_XN_REREAD 5 /* The transaction is not longer in RAM, status is unkown, retry. */ +typedef struct XTXactPrepare { + xtXactID xp_xact_id; + xtWord4 xp_hash; + struct XTXactPrepare *xp_next; /* Next item in hash table. */ + int xp_data_len; + xtWord1 xp_xa_data[XT_MAX_XA_DATA_SIZE]; +} XTXactPrepareRec, *XTXactPreparePtr; + +typedef struct XTXactXA { + xtXactID xx_xact_id; + XTXactPreparePtr xx_xa_ptr; +} XTXactXARec, *XTXactXAPtr; + +typedef struct XTXactEnumXA { + u_int exa_index; + xtBool exa_locked; +} XTXactEnumXARec, *XTXactEnumXAPtr; + typedef struct XTXactData { xtXactID xd_start_xn_id; /* Note: may be zero!. */ xtXactID xd_end_xn_id; /* Note: may be zero!. */ @@ -105,6 +124,7 @@ typedef struct XTXactData { int xd_flags; xtWord4 xd_end_time; xtThreadID xd_thread_id; + xtWord4 xd_xa_hash; /* 0 if no XA transaction. */ /* A transaction may be indexed twice in the hash table. * Once on the start sequence number, and once on the @@ -123,7 +143,7 @@ typedef struct XTXactData { #if defined(XT_XACT_USE_PTHREAD_RW) #define XT_XACT_LOCK_TYPE xt_rwlock_type -#define XT_XACT_INIT_LOCK(s, i) xt_init_rwlock(s, i) +#define XT_XACT_INIT_LOCK(s, i) xt_init_rwlock_with_autoname(s, i) #define XT_XACT_FREE_LOCK(s, i) xt_free_rwlock(i) #define XT_XACT_READ_LOCK(i, s) xt_slock_rwlock_ns(i) #define XT_XACT_WRITE_LOCK(i, s) xt_xlock_rwlock_ns(i) @@ -183,6 +203,14 @@ void xt_xn_wakeup_thread(xtThreadID th xtXactID xt_xn_get_curr_id(struct XTDatabase *db); xtWord8 xt_xn_bytes_to_sweep(struct XTDatabase *db, struct XTThread *thread); +int xt_xn_xa_compare(struct XTThread *self, register const void *thunk, register const void *a, register const void *b); +xtBool xt_xn_prepare(int len, xtWord1 *xa_data, struct XTThread *thread); +xtBool xt_xn_store_xa_data(struct XTDatabase *db, xtXactID xn_id, int len, xtWord1 *xa_data, struct XTThread *thread); +void xt_xn_delete_xa_data_by_xact(struct XTDatabase *db, xtXactID xact_id, struct XTThread *thread); +void xt_xn_delete_xa_data(struct XTDatabase *db, XTXactPreparePtr xap, xtBool unlock, struct XTThread *thread); +XTXactPreparePtr xt_xn_find_xa_data(struct XTDatabase *db, int len, xtWord1 *xa_data, xtBool lock, struct XTThread *thread); +XTXactPreparePtr xt_xn_enum_xa_data(struct XTDatabase *db, XTXactEnumXAPtr exa); + XTXactDataPtr xt_xn_add_old_xact(struct XTDatabase *db, xtXactID xn_id, struct XTThread *thread); XTXactDataPtr xt_xn_get_xact(struct XTDatabase *db, xtXactID xn_id, struct XTThread *thread); xtBool xt_xn_delete_xact(struct XTDatabase *db, xtXactID xn_id, struct XTThread *thread); === modified file 'storage/pbxt/src/xactlog_xt.cc' --- a/storage/pbxt/src/xactlog_xt.cc 2009-09-03 06:15:03 +0000 +++ b/storage/pbxt/src/xactlog_xt.cc 2009-11-24 10:55:06 +0000 @@ -1108,6 +1108,9 @@ xtBool XTDatabaseLog::xlog_append(XTThre if ((part_size = xl_write_buf_pos % 512)) { part_size = 512 - part_size; xl_write_buffer[xl_write_buf_pos] = XT_LOG_ENT_END_OF_LOG; +#ifdef HAVE_valgrind + memset(xl_write_buffer + xl_write_buf_pos + 1, 0x66, part_size); +#endif if (!xt_pwrite_file(xl_log_file, xl_write_log_offset, xl_write_buf_pos+part_size, xl_write_buffer, &thread->st_statistics.st_xlog, thread)) goto write_failed; } @@ -1477,9 +1480,9 @@ void XTDatabaseLog::xlog_name(size_t siz * T H R E A D T R A N S A C T I O N B U F F E R */ -xtPublic xtBool xt_xlog_flush_log(XTThreadPtr thread) +xtPublic xtBool xt_xlog_flush_log(struct XTDatabase *db, XTThreadPtr thread) { - return thread->st_database->db_xlog.xlog_flush(thread); + return db->db_xlog.xlog_flush(thread); } xtPublic xtBool xt_xlog_log_data(XTThreadPtr thread, size_t size, XTXactLogBufferDPtr log_entry, xtBool commit) @@ -1488,15 +1491,14 @@ xtPublic xtBool xt_xlog_log_data(XTThrea } /* Allocate a record from the free list. */ -xtPublic xtBool xt_xlog_modify_table(struct XTOpenTable *ot, u_int status, xtOpSeqNo op_seq, xtRecordID free_rec_id, xtRecordID rec_id, size_t size, xtWord1 *data) +xtPublic xtBool xt_xlog_modify_table(xtTableID tab_id, u_int status, xtOpSeqNo op_seq, xtRecordID free_rec_id, xtRecordID rec_id, size_t size, xtWord1 *data, XTThreadPtr thread) { XTXactLogBufferDRec log_entry; - XTThreadPtr thread = ot->ot_thread; - XTTableHPtr tab = ot->ot_table; size_t len; xtWord4 sum = 0; int check_size = 1; XTXactDataPtr xact = NULL; + xtBool commit = FALSE; switch (status) { case XT_LOG_ENT_REC_MODIFIED: @@ -1505,7 +1507,7 @@ xtPublic xtBool xt_xlog_modify_table(str case XT_LOG_ENT_DELETE: check_size = 2; XT_SET_DISK_4(log_entry.xu.xu_op_seq_4, op_seq); - XT_SET_DISK_4(log_entry.xu.xu_tab_id_4, tab->tab_id); + XT_SET_DISK_4(log_entry.xu.xu_tab_id_4, tab_id); XT_SET_DISK_4(log_entry.xu.xu_rec_id_4, rec_id); XT_SET_DISK_2(log_entry.xu.xu_size_2, size); len = offsetof(XTactUpdateEntryDRec, xu_rec_type_1); @@ -1521,7 +1523,7 @@ xtPublic xtBool xt_xlog_modify_table(str case XT_LOG_ENT_DELETE_FL: check_size = 2; XT_SET_DISK_4(log_entry.xf.xf_op_seq_4, op_seq); - XT_SET_DISK_4(log_entry.xf.xf_tab_id_4, tab->tab_id); + XT_SET_DISK_4(log_entry.xf.xf_tab_id_4, tab_id); XT_SET_DISK_4(log_entry.xf.xf_rec_id_4, rec_id); XT_SET_DISK_2(log_entry.xf.xf_size_2, size); XT_SET_DISK_4(log_entry.xf.xf_free_rec_id_4, free_rec_id); @@ -1539,14 +1541,14 @@ xtPublic xtBool xt_xlog_modify_table(str case XT_LOG_ENT_REC_REMOVED_EXT: ASSERT_NS(size == 1 + XT_XACT_ID_SIZE + sizeof(XTTabRecFreeDRec)); XT_SET_DISK_4(log_entry.fr.fr_op_seq_4, op_seq); - XT_SET_DISK_4(log_entry.fr.fr_tab_id_4, tab->tab_id); + XT_SET_DISK_4(log_entry.fr.fr_tab_id_4, tab_id); XT_SET_DISK_4(log_entry.fr.fr_rec_id_4, rec_id); len = offsetof(XTactFreeRecEntryDRec, fr_stat_id_1); break; case XT_LOG_ENT_REC_REMOVED_BI: check_size = 2; XT_SET_DISK_4(log_entry.rb.rb_op_seq_4, op_seq); - XT_SET_DISK_4(log_entry.rb.rb_tab_id_4, tab->tab_id); + XT_SET_DISK_4(log_entry.rb.rb_tab_id_4, tab_id); XT_SET_DISK_4(log_entry.rb.rb_rec_id_4, rec_id); XT_SET_DISK_2(log_entry.rb.rb_size_2, size); log_entry.rb.rb_new_rec_type_1 = (xtWord1) free_rec_id; @@ -1556,42 +1558,42 @@ xtPublic xtBool xt_xlog_modify_table(str case XT_LOG_ENT_REC_MOVED: ASSERT_NS(size == 8); XT_SET_DISK_4(log_entry.xw.xw_op_seq_4, op_seq); - XT_SET_DISK_4(log_entry.xw.xw_tab_id_4, tab->tab_id); + XT_SET_DISK_4(log_entry.xw.xw_tab_id_4, tab_id); XT_SET_DISK_4(log_entry.xw.xw_rec_id_4, rec_id); len = offsetof(XTactWriteRecEntryDRec, xw_rec_type_1); break; case XT_LOG_ENT_REC_CLEANED: ASSERT_NS(size == offsetof(XTTabRecHeadDRec, tr_prev_rec_id_4) + XT_RECORD_ID_SIZE); XT_SET_DISK_4(log_entry.xw.xw_op_seq_4, op_seq); - XT_SET_DISK_4(log_entry.xw.xw_tab_id_4, tab->tab_id); + XT_SET_DISK_4(log_entry.xw.xw_tab_id_4, tab_id); XT_SET_DISK_4(log_entry.xw.xw_rec_id_4, rec_id); len = offsetof(XTactWriteRecEntryDRec, xw_rec_type_1); break; case XT_LOG_ENT_REC_CLEANED_1: ASSERT_NS(size == 1); XT_SET_DISK_4(log_entry.xw.xw_op_seq_4, op_seq); - XT_SET_DISK_4(log_entry.xw.xw_tab_id_4, tab->tab_id); + XT_SET_DISK_4(log_entry.xw.xw_tab_id_4, tab_id); XT_SET_DISK_4(log_entry.xw.xw_rec_id_4, rec_id); len = offsetof(XTactWriteRecEntryDRec, xw_rec_type_1); break; case XT_LOG_ENT_REC_UNLINKED: ASSERT_NS(size == offsetof(XTTabRecHeadDRec, tr_prev_rec_id_4) + XT_RECORD_ID_SIZE); XT_SET_DISK_4(log_entry.xw.xw_op_seq_4, op_seq); - XT_SET_DISK_4(log_entry.xw.xw_tab_id_4, tab->tab_id); + XT_SET_DISK_4(log_entry.xw.xw_tab_id_4, tab_id); XT_SET_DISK_4(log_entry.xw.xw_rec_id_4, rec_id); len = offsetof(XTactWriteRecEntryDRec, xw_rec_type_1); break; case XT_LOG_ENT_ROW_NEW: ASSERT_NS(size == 0); XT_SET_DISK_4(log_entry.xa.xa_op_seq_4, op_seq); - XT_SET_DISK_4(log_entry.xa.xa_tab_id_4, tab->tab_id); + XT_SET_DISK_4(log_entry.xa.xa_tab_id_4, tab_id); XT_SET_DISK_4(log_entry.xa.xa_row_id_4, rec_id); len = offsetof(XTactRowAddedEntryDRec, xa_row_id_4) + XT_ROW_ID_SIZE; break; case XT_LOG_ENT_ROW_NEW_FL: ASSERT_NS(size == 0); XT_SET_DISK_4(log_entry.xa.xa_op_seq_4, op_seq); - XT_SET_DISK_4(log_entry.xa.xa_tab_id_4, tab->tab_id); + XT_SET_DISK_4(log_entry.xa.xa_tab_id_4, tab_id); XT_SET_DISK_4(log_entry.xa.xa_row_id_4, rec_id); XT_SET_DISK_4(log_entry.xa.xa_free_list_4, free_rec_id); sum ^= XT_CHECKSUM4_REC(free_rec_id); @@ -1602,10 +1604,17 @@ xtPublic xtBool xt_xlog_modify_table(str case XT_LOG_ENT_ROW_FREED: ASSERT_NS(size == sizeof(XTTabRowRefDRec)); XT_SET_DISK_4(log_entry.wr.wr_op_seq_4, op_seq); - XT_SET_DISK_4(log_entry.wr.wr_tab_id_4, tab->tab_id); + XT_SET_DISK_4(log_entry.wr.wr_tab_id_4, tab_id); XT_SET_DISK_4(log_entry.wr.wr_row_id_4, rec_id); len = offsetof(XTactWriteRowEntryDRec, wr_ref_id_4); break; + case XT_LOG_ENT_PREPARE: + check_size = 2; + XT_SET_DISK_4(log_entry.xp.xp_xact_id_4, op_seq); + log_entry.xp.xp_xa_len_1 = (xtWord1) size; + len = offsetof(XTXactPrepareEntryDRec, xp_xa_data); + commit = TRUE; + break; default: ASSERT_NS(FALSE); len = 0; @@ -1615,7 +1624,7 @@ xtPublic xtBool xt_xlog_modify_table(str xtWord1 *dptr = data; xtWord4 g; - sum ^= op_seq ^ (tab->tab_id << 8) ^ XT_CHECKSUM4_REC(rec_id); + sum ^= op_seq ^ (tab_id << 8) ^ XT_CHECKSUM4_REC(rec_id); if ((g = sum & 0xF0000000)) { sum = sum ^ (g >> 24); sum = sum ^ g; @@ -1643,9 +1652,9 @@ xtPublic xtBool xt_xlog_modify_table(str xt_print_log_record(0, 0, &log_entry); #endif if (xact) - return thread->st_database->db_xlog.xlog_append(thread, len, (xtWord1 *) &log_entry, size, data, FALSE, &xact->xd_begin_log, &xact->xd_begin_offset); + return thread->st_database->db_xlog.xlog_append(thread, len, (xtWord1 *) &log_entry, size, data, commit, &xact->xd_begin_log, &xact->xd_begin_offset); - return thread->st_database->db_xlog.xlog_append(thread, len, (xtWord1 *) &log_entry, size, data, FALSE, NULL, NULL); + return thread->st_database->db_xlog.xlog_append(thread, len, (xtWord1 *) &log_entry, size, data, commit, NULL, NULL); } /* @@ -1905,6 +1914,7 @@ xtBool XTDatabaseLog::xlog_verify(XTXact xtRecordID rec_id, free_rec_id; int check_size = 1; xtWord1 *dptr; + xtWord4 g; switch (record->xh.xh_status_1) { case XT_LOG_ENT_HEADER: @@ -2019,13 +2029,19 @@ xtBool XTDatabaseLog::xlog_verify(XTXact return record->xe.xe_checksum_1 == (XT_CHECKSUM_1(sum) ^ XT_CHECKSUM_1(log_id)); case XT_LOG_ENT_END_OF_LOG: return FALSE; + case XT_LOG_ENT_PREPARE: + check_size = 2; + op_seq = XT_GET_DISK_4(record->xp.xp_xact_id_4); + tab_id = 0; + rec_id = 0; + dptr = record->xp.xp_xa_data; + rec_size -= offsetof(XTXactPrepareEntryDRec, xp_xa_data); + break; default: ASSERT_NS(FALSE); return FALSE; } - xtWord4 g; - sum ^= (xtWord4) op_seq ^ ((xtWord4) tab_id << 8) ^ XT_CHECKSUM4_REC(rec_id); if ((g = sum & 0xF0000000)) { @@ -2193,6 +2209,14 @@ xtBool XTDatabaseLog::xlog_seq_next(XTXa } goto return_empty; } + case XT_LOG_ENT_PREPARE: + check_size = 2; + len = offsetof(XTXactPrepareEntryDRec, xp_xa_data); + if (len > max_rec_len) + /* The size is not in the buffer: */ + goto read_more; + len += (size_t) record->xp.xp_xa_len_1; + break; default: /* It is possible to land here after a crash, if the * log was not completely written. @@ -2231,7 +2255,7 @@ xtBool XTDatabaseLog::xlog_seq_next(XTXa goto return_empty; } - /* The record is not completely in the buffer: */ + /* The record is now completely in the buffer: */ seq->xseq_record_len = len; *ret_entry = (XTXactLogBufferDPtr) seq->xseq_buffer; return OK; @@ -2428,7 +2452,7 @@ static void xlog_wr_wait_for_log_flush(X if (reason == XT_LOG_CACHE_FULL || reason == XT_TIME_TO_WRITE || reason == XT_CHECKPOINT_REQ) { /* Make sure that we have something to write: */ if (db->db_xlog.xlog_bytes_to_write() < 2 * 1204 * 1024) - xt_xlog_flush_log(self); + xt_xlog_flush_log(db, self); } #ifdef TRACE_WRITER_ACTIVITY @@ -2529,6 +2553,7 @@ static void xlog_wr_main(XTThreadPtr sel case XT_LOG_ENT_ABORT: case XT_LOG_ENT_CLEANUP: case XT_LOG_ENT_OP_SYNC: + case XT_LOG_ENT_PREPARE: break; case XT_LOG_ENT_DEL_LOG: xtLogID log_id; === modified file 'storage/pbxt/src/xactlog_xt.h' --- a/storage/pbxt/src/xactlog_xt.h 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/xactlog_xt.h 2009-11-24 10:55:06 +0000 @@ -160,6 +160,7 @@ typedef struct XTXLogCache { #define XT_LOG_ENT_END_OF_LOG 37 /* This is a record that indicates the end of the log, and * fills to the end of a 512 byte block. */ +#define XT_LOG_ENT_PREPARE 39 /* XA prepare log entry. */ #define XT_LOG_FILE_MAGIC 0xAE88FE12 #define XT_LOG_VERSION_NO 1 @@ -201,6 +202,14 @@ typedef struct XTXactEndEntry { XTDiskValue4 xe_not_used_4; /* Was the end sequence number (no longer used - v1.0.04+), set to zero). */ } XTXactEndEntryDRec, *XTXactEndEntryDPtr; +typedef struct XTXactPrepareEntry { + xtWord1 xp_status_1; /* XT_LOG_ENT_PREPARE */ + XTDiskValue2 xp_checksum_2; + XTDiskValue4 xp_xact_id_4; /* The transaction. */ + xtWord1 xp_xa_len_1; /* The length of the XA data. */ + xtWord1 xp_xa_data[XT_MAX_XA_DATA_SIZE]; +} XTXactPrepareEntryDRec, *XTXactPrepareEntryDPtr; + typedef struct XTXactCleanupEntry { xtWord1 xc_status_1; /* XT_LOG_ENT_CLEANUP */ xtWord1 xc_checksum_1; @@ -344,6 +353,7 @@ typedef union XTXactLogBuffer { XTactOpSyncEntryDRec os; XTactExtRecEntryDRec er; XTactNoOpEntryDRec no; + XTXactPrepareEntryDRec xp; } XTXactLogBufferDRec, *XTXactLogBufferDPtr; /* ---------------------------------------- */ @@ -453,9 +463,9 @@ private: xtBool xlog_open_log(xtLogID log_id, off_t curr_eof, struct XTThread *thread); } XTDatabaseLogRec, *XTDatabaseLogPtr; -xtBool xt_xlog_flush_log(struct XTThread *thread); +xtBool xt_xlog_flush_log(struct XTDatabase *db, struct XTThread *thread); xtBool xt_xlog_log_data(struct XTThread *thread, size_t len, XTXactLogBufferDPtr log_entry, xtBool commit); -xtBool xt_xlog_modify_table(struct XTOpenTable *ot, u_int status, xtOpSeqNo op_seq, xtRecordID free_list, xtRecordID address, size_t size, xtWord1 *data); +xtBool xt_xlog_modify_table(xtTableID tab_id, u_int status, xtOpSeqNo op_seq, xtRecordID free_list, xtRecordID address, size_t size, xtWord1 *data, struct XTThread *thread); void xt_xlog_init(struct XTThread *self, size_t cache_size); void xt_xlog_exit(struct XTThread *self); === modified file 'storage/pbxt/src/xt_config.h' --- a/storage/pbxt/src/xt_config.h 2009-08-31 11:07:44 +0000 +++ b/storage/pbxt/src/xt_config.h 2009-11-24 10:55:06 +0000 @@ -50,7 +50,9 @@ const int max_connections = 500; /* * Make sure we use the thread safe version of the library. */ +#ifndef _THREAD_SAFE // Seems to be defined by some Drizzle header #define _THREAD_SAFE +#endif /* * This causes things to be defined like stuff in inttypes.h @@ -72,12 +74,12 @@ const int max_connections = 500; #define XT_MAC #endif -#if defined(MSDOS) || defined(__WIN__) +#if defined(MSDOS) || defined(__WIN__) || defined(_WIN64) #define XT_WIN #endif #ifdef XT_WIN -#ifdef _DEBUG +#if defined(_DEBUG) && !defined(DEBUG) #define DEBUG #endif // _DEBUG #else @@ -101,8 +103,13 @@ const int max_connections = 500; * Definition of which atomic operations to use: */ #ifdef XT_WIN +#ifdef _WIN64 +/* 64-bit Windows atomic ops are not yet supported: */ +#define XT_NO_ATOMICS +#else /* MS Studio style embedded assembler for x86 */ #define XT_ATOMIC_WIN32_X86 +#endif #elif defined(__GNUC__) && (defined(__x86_64__) || defined(__i386__)) /* Use GNU style embedded assembler for x86 */ #define XT_ATOMIC_GNUC_X86 @@ -115,4 +122,10 @@ const int max_connections = 500; #define XT_NO_ATOMICS #endif +#ifndef DRIZZLED +#if MYSQL_VERSION_ID >= 50404 +#define MYSQL_SUPPORTS_BACKUP +#endif +#endif + #endif === modified file 'storage/pbxt/src/xt_defs.h' --- a/storage/pbxt/src/xt_defs.h 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/xt_defs.h 2009-11-24 10:55:06 +0000 @@ -379,6 +379,19 @@ typedef struct XTPathStr { */ //#define XT_IMPLEMENT_NO_ACTION +/* Define this value if online-backup should be supported. + * Note that, online backup is currently only supported + * by MySQL 6.0.9 or later + */ +#define XT_ENABLE_ONLINE_BACKUP + +/* Define this switch if you don't want to use atomic + * synchronisation. + */ +#ifndef XT_NO_ATOMICS +//#define XT_NO_ATOMICS +#endif + /* ---------------------------------------------------------------------- * GLOBAL CONSTANTS */ @@ -406,6 +419,8 @@ typedef struct XTPathStr { #define XT_ADD_PTR(p, l) ((void *) ((char *) (p) + (l))) +#define XT_MAX_XA_DATA_SIZE (3*4 + 128) /* Corresponds to the maximum size of struct xid_t in handler.h. */ + /* ---------------------------------------------------------------------- * DEFINES DEPENDENT ON CONSTANTS */ @@ -744,6 +759,7 @@ extern xtBool pbxt_crash_debug; #define MYSQL_PLUGIN_VAR_HEADER DRIZZLE_PLUGIN_VAR_HEADER #define MYSQL_SYSVAR_STR DRIZZLE_SYSVAR_STR #define MYSQL_SYSVAR_INT DRIZZLE_SYSVAR_INT +#define MYSQL_SYSVAR_BOOL DRIZZLE_SYSVAR_BOOL #define MYSQL_SYSVAR DRIZZLE_SYSVAR #define MYSQL_STORAGE_ENGINE_PLUGIN DRIZZLE_STORAGE_ENGINE_PLUGIN #define MYSQL_INFORMATION_SCHEMA_PLUGIN DRIZZLE_INFORMATION_SCHEMA_PLUGIN @@ -752,9 +768,8 @@ extern xtBool pbxt_crash_debug; #define mx_tmp_use_all_columns(x, y) (x)->use_all_columns(y) #define mx_tmp_restore_column_map(x, y) (x)->restore_column_map(y) -#define MX_BIT_FAST_TEST_AND_SET(x, y) bitmap_test_and_set(x, y) -#define MX_TABLE_TYPES_T handler::Table_flags +#define MX_TABLE_TYPES_T Cursor::Table_flags #define MX_UINT8_T uint8_t #define MX_ULONG_T uint32_t #define MX_ULONGLONG_T uint64_t @@ -762,6 +777,10 @@ extern xtBool pbxt_crash_debug; #define MX_CHARSET_INFO struct charset_info_st #define MX_CONST_CHARSET_INFO const struct charset_info_st #define MX_CONST const +#define MX_BITMAP MyBitmap +#define MX_BIT_SIZE() numOfBitsInMap() +#define MX_BIT_SET(x, y) (x)->setBit(y) +#define MX_BIT_FAST_TEST_AND_SET(x, y) (x)->testAndSet(y) #define my_bool bool #define int16 int16_t @@ -771,6 +790,7 @@ extern xtBool pbxt_crash_debug; #define uchar unsigned char #define longlong int64_t #define ulonglong uint64_t +#define handler Cursor #define HAVE_LONG_LONG @@ -823,10 +843,13 @@ extern xtBool pbxt_crash_debug; class PBXTStorageEngine; typedef PBXTStorageEngine handlerton; +class Session; + +extern "C" void session_mark_transaction_to_rollback(Session *session, bool all); #else // DRIZZLED /* The MySQL case: */ -#if MYSQL_VERSION_ID >= 60008 +#if MYSQL_VERSION_ID >= 50404 #define STRUCT_TABLE struct TABLE #else #define STRUCT_TABLE struct st_table @@ -844,13 +867,13 @@ typedef PBXTStorageEngine handlerton; #define MX_CHARSET_INFO CHARSET_INFO #define MX_CONST_CHARSET_INFO struct charset_info_st #define MX_CONST +#define MX_BITMAP MY_BITMAP +#define MX_BIT_SIZE() n_bits +#define MX_BIT_SET(x, y) bitmap_set_bit(x, y) #endif // DRIZZLED -#define MX_BITMAP MY_BITMAP -#define MX_BIT_SIZE() n_bits #define MX_BIT_IS_SUBSET(x, y) bitmap_is_subset(x, y) -#define MX_BIT_SET(x, y) bitmap_set_bit(x, y) #ifndef XT_SCAN_CORE_DEFINED #define XT_SCAN_CORE_DEFINED === modified file 'storage/pbxt/src/xt_errno.h' --- a/storage/pbxt/src/xt_errno.h 2009-09-03 06:15:03 +0000 +++ b/storage/pbxt/src/xt_errno.h 2009-11-24 10:55:06 +0000 @@ -119,6 +119,9 @@ #define XT_ERR_FK_REF_TEMP_TABLE -95 #define XT_ERR_MYSQL_SHUTDOWN -98 #define XT_ERR_MYSQL_NO_THREAD -99 +#define XT_ERR_BUFFER_TOO_SMALL -100 +#define XT_ERR_BAD_BACKUP_FORMAT -101 +#define XT_ERR_PBXT_NOT_INSTALLED -102 #ifdef XT_WIN #define XT_ENOMEM ERROR_NOT_ENOUGH_MEMORY

1 0

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (knielsen:2774)
by knielsen＠knielsen-hq.org 02 Dec '09

02 Dec '09

#At lp:maria 2774 knielsen(a)knielsen-hq.org 2009-12-02 [merge] Merge PBXT 1.0.09f RC3 into MariaDB. added: storage/pbxt/src/backup_xt.cc storage/pbxt/src/backup_xt.h modified: sql/handler.cc storage/pbxt/ChangeLog storage/pbxt/src/Makefile.am storage/pbxt/src/cache_xt.cc storage/pbxt/src/cache_xt.h storage/pbxt/src/database_xt.cc storage/pbxt/src/database_xt.h storage/pbxt/src/datadic_xt.cc storage/pbxt/src/datalog_xt.cc storage/pbxt/src/discover_xt.cc storage/pbxt/src/filesys_xt.cc storage/pbxt/src/filesys_xt.h storage/pbxt/src/ha_pbxt.cc storage/pbxt/src/ha_pbxt.h storage/pbxt/src/ha_xtsys.h storage/pbxt/src/heap_xt.cc storage/pbxt/src/index_xt.cc storage/pbxt/src/lock_xt.cc storage/pbxt/src/locklist_xt.h storage/pbxt/src/memory_xt.cc storage/pbxt/src/myxt_xt.cc storage/pbxt/src/myxt_xt.h storage/pbxt/src/pbms.h storage/pbxt/src/pbms_enabled.cc storage/pbxt/src/pbms_enabled.h storage/pbxt/src/pthread_xt.cc storage/pbxt/src/restart_xt.cc storage/pbxt/src/restart_xt.h storage/pbxt/src/strutil_xt.cc storage/pbxt/src/systab_xt.cc storage/pbxt/src/tabcache_xt.cc storage/pbxt/src/tabcache_xt.h storage/pbxt/src/table_xt.cc storage/pbxt/src/table_xt.h storage/pbxt/src/thread_xt.cc storage/pbxt/src/thread_xt.h storage/pbxt/src/util_xt.cc storage/pbxt/src/util_xt.h storage/pbxt/src/xaction_xt.cc storage/pbxt/src/xaction_xt.h storage/pbxt/src/xactlog_xt.cc storage/pbxt/src/xactlog_xt.h storage/pbxt/src/xt_config.h storage/pbxt/src/xt_defs.h storage/pbxt/src/xt_errno.h === modified file 'sql/handler.cc' --- a/sql/handler.cc 2009-11-16 20:49:51 +0000 +++ b/sql/handler.cc 2009-12-02 11:50:40 +0000 @@ -1584,16 +1584,6 @@ int ha_recover(HASH *commit_list) if (info.commit_list) sql_print_information("Starting crash recovery..."); -#ifndef WILL_BE_DELETED_LATER - /* - for now, only InnoDB supports 2pc. It means we can always safely - rollback all pending transactions, without risking inconsistent data - */ - DBUG_ASSERT(total_ha_2pc == (ulong) opt_bin_log+1); // only InnoDB and binlog - tc_heuristic_recover= TC_HEURISTIC_RECOVER_ROLLBACK; // forcing ROLLBACK - info.dry_run=FALSE; -#endif - for (info.len= MAX_XID_LIST_SIZE ; info.list==0 && info.len > MIN_XID_LIST_SIZE; info.len/=2) { === modified file 'storage/pbxt/ChangeLog' --- a/storage/pbxt/ChangeLog 2009-09-03 06:15:03 +0000 +++ b/storage/pbxt/ChangeLog 2009-12-01 09:50:46 +0000 @@ -1,6 +1,58 @@ PBXT Release Notes ================== +------- 1.0.09f RC3 - 2009-11-30 + +RN291: Fixed bug #489088: On shutdown MySQL reports: [Warning] Plugin 'PBXT' will be forced to shutdown. + +RN290: Fixed bug #345524: pbxt does not compile on 64 bit windows. Currently atomic operations are not supported on this platform. + +RN286: Fixed a bug introduced in RN281, which could cause an index scan to hang. The original change was to prevent a warning in Valgrind. + +RN285: Merged changes required to compile with Drizzle. + +RN284: Fixed bug that cause the error "[ERROR] Invalid (old?) table or database name 'mysqld.1'", when running temp_table.test under MariaDB (thanks to Monty for his initial bug fix). Added a fix for partition table names as well. + +RN283: Added win_inttypes.h to the distribution. This file is only required for the Windows build. + +RN282: Fixed bug #451101: jump or move depends on uninitialised value in myxt_get_key_length + +RN281: Fixed bug #451080: Uninitialised memory write in XTDatabaseLog::xlog_append + +RN280: Fixed bug #451085: jump or move depends on uninitialised value in my_type_to_string + +RN279: Fixed bug #441000: xtstat crashes with segmentation fault on startup if max_pbxt_threads exceeded. + +------- 1.0.09e RC3 - 2009-11-20 + +RN278: Fixed compile error with MySQL 5.1.41. + +------- 1.0.09d RC3 - 2009-09-30 + +RN277: Added r/o flag to pbxt_max_threads server variable (this fix is related to bug #430637) + +RN276: Added test case for replication on tables w/o PKs (see bug #430716) + +RN275: Fixed bug #430600: 'Failed to read auto-increment value from storage engine' error. + +RN274: Fixed bug #431240: This report is public edit xtstat fails if no PBXT table has been created. xtstat now accepts --database=information_schema or --database=pbxt. Depending on this setting PBXT will either use the information_schema.pbxt_statistics or the pbxt.statistics table. If information_schema is used, then the statistics are displayed even when no PBXT table exists. Recovery activity is also displayed, unless pbxt_support_xa=1, in which case MySQL will wait for PBXT recovery to complete before allowing connections. + +RN273: Fixed bug #430633: XA_RBDEADLOCK is not returned on XA END after the transacting ended with a deadlock. + +RN272: Fixed bug #430596: Backup/restore does not work well even on a basic PBXT table with auto-increment. + +------- 1.0.09c RC3 - 2009-09-16 + +RN271: Windows build update: now you can simply put the pbxt directory under <mysql-root>/storage and build the PBXT engine as a part of the source tree. The engine will be linked statically. Be sure to specify the WITH_PBXT_STORAGE_ENGINE option when running win\configure.js + +RN270: Correctly disabled PBMS so that this version now compiles under Windows. If PBMS_ENABLED is defined, PBXT will not compile under Windows becaause of a getpid() call in pbms.h. + +------- 1.0.09 RC3 - 2009-09-09 + +RN269: Implemented online backup. A native online backup driver now performs BACKUP and RESTORE DATABASE operations for PBXT. NOTE: This feature is only supported by MySQL 6.0.9 or later. + +RN268: Implemented XA support. PBXT now supports all XA related MySQL statements. The variable pbxt_support_xa determines if XA support is enabled. Note: due to MySQL bug #47134, enabling XA support could lead to a crash. + ------- 1.0.08d RC2 - 2009-09-02 RN267: Fixed a bug that caused MySQL to crash on shutdown, after an incorrect command line parameter was given. The crash occurred because the background recovery task was not cleaned up before the PBXT engine was de-initialized. === modified file 'storage/pbxt/src/Makefile.am' --- a/storage/pbxt/src/Makefile.am 2009-10-07 07:40:56 +0000 +++ b/storage/pbxt/src/Makefile.am 2009-11-24 10:55:06 +0000 @@ -22,7 +22,7 @@ noinst_HEADERS = bsearch_xt.h cache_xt. pbms_enabled.h sortedlist_xt.h strutil_xt.h \ tabcache_xt.h table_xt.h trace_xt.h thread_xt.h \ util_xt.h xaction_xt.h xactlog_xt.h lock_xt.h \ - systab_xt.h ha_xtsys.h discover_xt.h \ + systab_xt.h ha_xtsys.h discover_xt.h backup_xt.h \ pbms.h xt_config.h xt_defs.h xt_errno.h locklist_xt.h EXTRA_LTLIBRARIES = libpbxt.la @@ -32,7 +32,7 @@ libpbxt_la_SOURCES = bsearch_xt.cc cache memory_xt.cc myxt_xt.cc pthread_xt.cc restart_xt.cc \ sortedlist_xt.cc strutil_xt.cc \ tabcache_xt.cc table_xt.cc trace_xt.cc thread_xt.cc \ - systab_xt.cc ha_xtsys.cc discover_xt.cc \ + systab_xt.cc ha_xtsys.cc discover_xt.cc backup_xt.cc \ util_xt.cc xaction_xt.cc xactlog_xt.cc lock_xt.cc locklist_xt.cc libpbxt_la_LDFLAGS = -module === added file 'storage/pbxt/src/backup_xt.cc' --- a/storage/pbxt/src/backup_xt.cc 1970-01-01 00:00:00 +0000 +++ b/storage/pbxt/src/backup_xt.cc 2009-11-24 10:55:06 +0000 @@ -0,0 +1,802 @@ +/* Copyright (c) 2009 PrimeBase Technologies GmbH + * + * PrimeBase XT + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * + * 2009-09-07 Paul McCullagh + * + * H&G2JCtL + */ + +#include "xt_config.h" + +#ifdef MYSQL_SUPPORTS_BACKUP + +#include <string.h> +#include <stdio.h> +#include <stdlib.h> +#include <time.h> +#include <ctype.h> + +#include "mysql_priv.h" +#include <backup/api_types.h> +#include <backup/backup_engine.h> +#include <backup/backup_aux.h> // for build_table_list() +#include <hash.h> + +#include "ha_pbxt.h" + +#include "backup_xt.h" +#include "pthread_xt.h" +#include "filesys_xt.h" +#include "database_xt.h" +#include "strutil_xt.h" +#include "memory_xt.h" +#include "trace_xt.h" +#include "myxt_xt.h" + +#ifdef OK +#undef OK +#endif + +#ifdef byte +#undef byte +#endif + +#ifdef DEBUG +//#define TRACE_BACKUP_CALLS +//#define TEST_SMALL_BLOCK 100000 +#endif + +using backup::byte; +using backup::result_t; +using backup::version_t; +using backup::Table_list; +using backup::Table_ref; +using backup::Buffer; + +#ifdef TRACE_BACKUP_CALLS +#define XT_TRACE_CALL() ha_trace_function(__FUNC__, NULL) +#else +#define XT_TRACE_CALL() +#endif + +#define XT_RESTORE_BATCH_SIZE 10000 + +#define BUP_STATE_BEFORE_LOCK 0 +#define BUP_STATE_AFTER_LOCK 1 + +#define BUP_STANDARD_VAR_RECORD 1 +#define BUP_RECORD_BLOCK_4_START 2 // Part of a record, with a 4 byte total length, and 4 byte data length +#define BUP_RECORD_BLOCK_4 3 // Part of a record, with a 4 byte length +#define BUP_RECORD_BLOCK_4_END 4 // Last part of a record with a 4 byte length + +/* + * ----------------------------------------------------------------------- + * UTILITIES + */ + +#ifdef TRACE_BACKUP_CALLS +static void ha_trace_function(const char *function, char *table) +{ + char func_buf[50], *ptr; + XTThreadPtr thread = xt_get_self(); + + if ((ptr = strchr(function, '('))) { + ptr--; + while (ptr > function) { + if (!(isalnum(*ptr) || *ptr == '_')) + break; + ptr--; + } + ptr++; + xt_strcpy(50, func_buf, ptr); + if ((ptr = strchr(func_buf, '('))) + *ptr = 0; + } + else + xt_strcpy(50, func_buf, function); + if (table) + printf("%s %s (%s)\n", thread ? thread->t_name : "-unknown-", func_buf, table); + else + printf("%s %s\n", thread ? thread->t_name : "-unknown-", func_buf); +} +#endif + +/* + * ----------------------------------------------------------------------- + * BACKUP DRIVER + */ + +class PBXTBackupDriver: public Backup_driver +{ + public: + PBXTBackupDriver(const Table_list &); + virtual ~PBXTBackupDriver(); + + virtual size_t size(); + virtual size_t init_size(); + virtual result_t begin(const size_t); + virtual result_t end(); + virtual result_t get_data(Buffer &); + virtual result_t prelock(); + virtual result_t lock(); + virtual result_t unlock(); + virtual result_t cancel(); + virtual void free(); + void lock_tables_TL_READ_NO_INSERT(); + + private: + XTThreadPtr bd_thread; + int bd_state; + u_int bd_table_no; + XTOpenTablePtr bd_ot; + xtWord1 *bd_row_buf; + + /* Non-zero if we last returned only part of + * a row. + */ + xtWord1 *db_write_block(xtWord1 *buffer, xtWord1 bup_type, size_t *size, xtWord4 row_len); + xtWord1 *db_write_block(xtWord1 *buffer, xtWord1 bup_type, size_t *size, xtWord4 total_len, xtWord4 row_len); + + xtWord4 bd_row_offset; + xtWord4 bd_row_size; +}; + + +PBXTBackupDriver::PBXTBackupDriver(const Table_list &tables): +Backup_driver(tables), +bd_state(BUP_STATE_BEFORE_LOCK), +bd_table_no(0), +bd_ot(NULL), +bd_row_buf(NULL), +bd_row_offset(0), +bd_row_size(0) +{ +} + +PBXTBackupDriver::~PBXTBackupDriver() +{ +} + +/** Estimates total size of backup. @todo improve it */ +size_t PBXTBackupDriver::size() +{ + XT_TRACE_CALL(); + return UNKNOWN_SIZE; +} + +/** Estimates size of backup before lock. @todo improve it */ +size_t PBXTBackupDriver::init_size() +{ + XT_TRACE_CALL(); + return 0; +} + +result_t PBXTBackupDriver::begin(const size_t) +{ + THD *thd = current_thd; + XTExceptionRec e; + + XT_TRACE_CALL(); + + if (!(bd_thread = xt_ha_set_current_thread(thd, &e))) { + xt_log_exception(NULL, &e, XT_LOG_DEFAULT); + return backup::ERROR; + } + + return backup::OK; +} + +result_t PBXTBackupDriver::end() +{ + XT_TRACE_CALL(); + if (bd_ot) { + xt_tab_seq_exit(bd_ot); + xt_db_return_table_to_pool_ns(bd_ot); + bd_ot = NULL; + } + if (bd_thread->st_xact_data) { + if (!xt_xn_commit(bd_thread)) + return backup::ERROR; + } + return backup::OK; +} + +xtWord1 *PBXTBackupDriver::db_write_block(xtWord1 *buffer, xtWord1 bup_type, size_t *ret_size, xtWord4 row_len) +{ + register size_t size = *ret_size; + + *buffer = bup_type; // Record type identifier. + buffer++; + size--; + memcpy(buffer, bd_ot->ot_row_wbuffer, row_len); + buffer += row_len; + size -= row_len; + *ret_size = size; + return buffer; +} + +xtWord1 *PBXTBackupDriver::db_write_block(xtWord1 *buffer, xtWord1 bup_type, size_t *ret_size, xtWord4 total_len, xtWord4 row_len) +{ + register size_t size = *ret_size; + + *buffer = bup_type; // Record type identifier. + buffer++; + size--; + if (bup_type == BUP_RECORD_BLOCK_4_START) { + XT_SET_DISK_4(buffer, total_len); + buffer += 4; + size -= 4; + } + XT_SET_DISK_4(buffer, row_len); + buffer += 4; + size -= 4; + memcpy(buffer, bd_ot->ot_row_wbuffer+bd_row_offset, row_len); + buffer += row_len; + size -= row_len; + bd_row_size -= row_len; + bd_row_offset += row_len; + *ret_size = size; + return buffer; +} + +result_t PBXTBackupDriver::get_data(Buffer &buf) +{ + xtBool eof = FALSE; + size_t size; + xtWord4 row_len; + xtWord1 *buffer; + + XT_TRACE_CALL(); + + if (bd_state == BUP_STATE_BEFORE_LOCK) { + buf.table_num = 0; + buf.size = 0; + buf.last = FALSE; + return backup::READY; + } + + /* Open the backup table: */ + if (!bd_ot) { + XTThreadPtr self = bd_thread; + XTTableHPtr tab; + char path[PATH_MAX]; + + if (bd_table_no == m_tables.count()) { + buf.size = 0; + buf.table_num = 0; + buf.last = TRUE; + return backup::DONE; + } + + m_tables[bd_table_no].internal_name(path, sizeof(path)); + bd_table_no++; + try_(a) { + xt_ha_open_database_of_table(self, (XTPathStrPtr) path); + tab = xt_use_table(self, (XTPathStrPtr) path, FALSE, FALSE, NULL); + pushr_(xt_heap_release, tab); + if (!(bd_ot = xt_db_open_table_using_tab(tab, bd_thread))) + xt_throw(self); + freer_(); // xt_heap_release(tab) + + /* Prepare the seqential scan: */ + xt_tab_seq_exit(bd_ot); + if (!xt_tab_seq_init(bd_ot)) + xt_throw(self); + + if (bd_row_buf) { + xt_free(self, bd_row_buf); + bd_row_buf = NULL; + } + bd_row_buf = (xtWord1 *) xt_malloc(self, bd_ot->ot_table->tab_dic.dic_mysql_buf_size); + bd_ot->ot_cols_req = bd_ot->ot_table->tab_dic.dic_no_of_cols; + } + catch_(a) { + ; + } + cont_(a); + + if (!bd_ot) + goto failed; + } + + buf.table_num = bd_table_no; +#ifdef TEST_SMALL_BLOCK + buf.size = TEST_SMALL_BLOCK; +#endif + size = buf.size; + buffer = (xtWord1 *) buf.data; + ASSERT_NS(size > 9); + + /* First check of a record was partically written + * last time. + */ + write_row: + if (bd_row_size > 0) { + row_len = bd_row_size; + if (bd_row_offset == 0) { + if (row_len+1 > size) { + ASSERT_NS(size > 9); + row_len = size - 9; + buffer = db_write_block(buffer, BUP_RECORD_BLOCK_4_START, &size, bd_row_size, row_len); + goto done; + } + buffer = db_write_block(buffer, BUP_STANDARD_VAR_RECORD, &size, row_len); + bd_row_size = 0; + } + else { + if (row_len+5 > size) { + row_len = size - 5; + buffer = db_write_block(buffer, BUP_RECORD_BLOCK_4, &size, 0, row_len); + goto done; + } + buffer = db_write_block(buffer, BUP_RECORD_BLOCK_4_END, &size, 0, row_len); + } + } + + /* Now continue with the sequential scan. */ + while (size > 1) { + if (!xt_tab_seq_next(bd_ot, bd_row_buf, &eof)) + goto failed; + if (eof) { + /* We will go the next table, on the next call. */ + xt_tab_seq_exit(bd_ot); + xt_db_return_table_to_pool_ns(bd_ot); + bd_ot = NULL; + break; + } + if (!(row_len = myxt_store_row_data(bd_ot, 0, (char *) bd_row_buf))) + goto failed; + if (row_len+1 > size) { + /* Does not fit: */ + bd_row_offset = 0; + bd_row_size = row_len; + /* Only add part of the row, if there is still + * quite a bit of space left: + */ + if (size >= (32 * 1024)) + goto write_row; + break; + } + buffer = db_write_block(buffer, BUP_STANDARD_VAR_RECORD, &size, row_len); + } + + done: + buf.size = buf.size - size; + /* This indicates wnd of data for a table! */ + buf.last = eof; + + return backup::OK; + + failed: + xt_log_and_clear_exception(bd_thread); + return backup::ERROR; +} + +result_t PBXTBackupDriver::prelock() +{ + XT_TRACE_CALL(); + return backup::READY; +} + +result_t PBXTBackupDriver::lock() +{ + XT_TRACE_CALL(); + bd_thread->st_xact_mode = XT_XACT_COMMITTED_READ; + bd_thread->st_ignore_fkeys = FALSE; + bd_thread->st_auto_commit = FALSE; + bd_thread->st_table_trans = FALSE; + bd_thread->st_abort_trans = FALSE; + bd_thread->st_stat_ended = FALSE; + bd_thread->st_stat_trans = FALSE; + bd_thread->st_is_update = FALSE; + if (!xt_xn_begin(bd_thread)) + return backup::ERROR; + bd_state = BUP_STATE_AFTER_LOCK; + return backup::OK; +} + +result_t PBXTBackupDriver::unlock() +{ + XT_TRACE_CALL(); + return backup::OK; +} + +result_t PBXTBackupDriver::cancel() +{ + XT_TRACE_CALL(); + return backup::OK; // free() will be called and suffice +} + +void PBXTBackupDriver::free() +{ + XT_TRACE_CALL(); + if (bd_ot) { + xt_tab_seq_exit(bd_ot); + xt_db_return_table_to_pool_ns(bd_ot); + bd_ot = NULL; + } + if (bd_row_buf) { + xt_free_ns(bd_row_buf); + bd_row_buf = NULL; + } + if (bd_thread->st_xact_data) + xt_xn_rollback(bd_thread); + delete this; +} + +void PBXTBackupDriver::lock_tables_TL_READ_NO_INSERT() +{ + XT_TRACE_CALL(); +} + +/* + * ----------------------------------------------------------------------- + * BACKUP DRIVER + */ + +class PBXTRestoreDriver: public Restore_driver +{ + public: + PBXTRestoreDriver(const Table_list &tables); + virtual ~PBXTRestoreDriver(); + + virtual result_t begin(const size_t); + virtual result_t end(); + virtual result_t send_data(Buffer &buf); + virtual result_t cancel(); + virtual void free(); + + private: + XTThreadPtr rd_thread; + u_int rd_table_no; + XTOpenTablePtr rd_ot; + STRUCT_TABLE *rd_my_table; + xtWord1 *rb_row_buf; + u_int rb_col_cnt; + u_int rb_insert_count; + + /* Long rows are accumulated here: */ + xtWord4 rb_row_len; + xtWord4 rb_data_size; + xtWord1 *rb_row_data; +}; + +PBXTRestoreDriver::PBXTRestoreDriver(const Table_list &tables): +Restore_driver(tables), +rd_thread(NULL), +rd_table_no(0), +rd_ot(NULL), +rb_row_buf(NULL), +rb_row_len(0), +rb_data_size(0), +rb_row_data(NULL) +{ +} + +PBXTRestoreDriver::~PBXTRestoreDriver() +{ +} + +result_t PBXTRestoreDriver::begin(const size_t) +{ + THD *thd = current_thd; + XTExceptionRec e; + + XT_TRACE_CALL(); + + if (!(rd_thread = xt_ha_set_current_thread(thd, &e))) { + xt_log_exception(NULL, &e, XT_LOG_DEFAULT); + return backup::ERROR; + } + + return backup::OK; +} + +result_t PBXTRestoreDriver::end() +{ + XT_TRACE_CALL(); + if (rd_ot) { + xt_db_return_table_to_pool_ns(rd_ot); + rd_ot = NULL; + } + //if (rb_row_buf) { + // xt_free_ns(rb_row_buf); + // rb_row_buf = NULL; + //} + if (rb_row_data) { + xt_free_ns(rb_row_data); + rb_row_data = NULL; + } + if (rd_thread->st_xact_data) { + if (!xt_xn_commit(rd_thread)) + return backup::ERROR; + } + return backup::OK; +} + + +result_t PBXTRestoreDriver::send_data(Buffer &buf) +{ + size_t size; + xtWord1 type; + xtWord1 *buffer; + xtWord4 row_len; + xtWord1 *rec_data; + + XT_TRACE_CALL(); + + if (buf.table_num != rd_table_no) { + XTThreadPtr self = rd_thread; + XTTableHPtr tab; + char path[PATH_MAX]; + + if (rd_ot) { + xt_db_return_table_to_pool_ns(rd_ot); + rd_ot = NULL; + } + + if (rd_thread->st_xact_data) { + if (!xt_xn_commit(rd_thread)) + goto failed; + } + if (!xt_xn_begin(rd_thread)) + goto failed; + rb_insert_count = 0; + + rd_table_no = buf.table_num; + m_tables[rd_table_no-1].internal_name(path, sizeof(path)); + try_(a) { + xt_ha_open_database_of_table(self, (XTPathStrPtr) path); + tab = xt_use_table(self, (XTPathStrPtr) path, FALSE, FALSE, NULL); + pushr_(xt_heap_release, tab); + if (!(rd_ot = xt_db_open_table_using_tab(tab, rd_thread))) + xt_throw(self); + freer_(); // xt_heap_release(tab) + + rd_my_table = rd_ot->ot_table->tab_dic.dic_my_table; + if (rd_my_table->found_next_number_field) { + rd_my_table->in_use = current_thd; + rd_my_table->next_number_field = rd_my_table->found_next_number_field; + rd_my_table->mark_columns_used_by_index_no_reset(rd_my_table->s->next_number_index, rd_my_table->read_set); + } + + /* This is safe because only one thread can restore a table at + * a time! + */ + rb_row_buf = (xtWord1 *) rd_my_table->record[0]; + //if (rb_row_buf) { + // xt_free(self, rb_row_buf); + // rb_row_buf = NULL; + //} + //rb_row_buf = (xtWord1 *) xt_malloc(self, rd_ot->ot_table->tab_dic.dic_mysql_buf_size); + + rb_col_cnt = rd_ot->ot_table->tab_dic.dic_no_of_cols; + + } + catch_(a) { + ; + } + cont_(a); + + if (!rd_ot) + goto failed; + } + + buffer = (xtWord1 *) buf.data; + size = buf.size; + + while (size > 0) { + type = *buffer; + switch (type) { + case BUP_STANDARD_VAR_RECORD: + rec_data = buffer + 1; + break; + case BUP_RECORD_BLOCK_4_START: + buffer++; + row_len = XT_GET_DISK_4(buffer); + buffer += 4; + if (rb_data_size < row_len) { + if (!xt_realloc_ns((void **) &rb_row_data, row_len)) + goto failed; + rb_data_size = row_len; + } + row_len = XT_GET_DISK_4(buffer); + buffer += 4; + ASSERT_NS(row_len <= rb_data_size); + if (row_len > rb_data_size) { + xt_register_xterr(XT_REG_CONTEXT, XT_ERR_BAD_BACKUP_FORMAT); + goto failed; + } + memcpy(rb_row_data, buffer, row_len); + rb_row_len = row_len; + buffer += row_len; + if (row_len + 9 > size) { + xt_register_xterr(XT_REG_CONTEXT, XT_ERR_BAD_BACKUP_FORMAT); + goto failed; + } + size -= row_len + 9; + continue; + case BUP_RECORD_BLOCK_4: + buffer++; + row_len = XT_GET_DISK_4(buffer); + buffer += 4; + ASSERT_NS(rb_row_len + row_len <= rb_data_size); + if (rb_row_len + row_len > rb_data_size) { + xt_register_xterr(XT_REG_CONTEXT, XT_ERR_BAD_BACKUP_FORMAT); + goto failed; + } + memcpy(rb_row_data + rb_row_len, buffer, row_len); + rb_row_len += row_len; + buffer += row_len; + if (row_len + 5 > size) { + xt_register_xterr(XT_REG_CONTEXT, XT_ERR_BAD_BACKUP_FORMAT); + goto failed; + } + size -= row_len + 5; + continue; + case BUP_RECORD_BLOCK_4_END: + buffer++; + row_len = XT_GET_DISK_4(buffer); + buffer += 4; + ASSERT_NS(rb_row_len + row_len <= rb_data_size); + if (rb_row_len + row_len > rb_data_size) { + xt_register_xterr(XT_REG_CONTEXT, XT_ERR_BAD_BACKUP_FORMAT); + goto failed; + } + memcpy(rb_row_data + rb_row_len, buffer, row_len); + buffer += row_len; + if (row_len + 5 > size) { + xt_register_xterr(XT_REG_CONTEXT, XT_ERR_BAD_BACKUP_FORMAT); + goto failed; + } + size -= row_len + 5; + rec_data = rb_row_data; + break; + default: + xt_register_xterr(XT_REG_CONTEXT, XT_ERR_BAD_BACKUP_FORMAT); + goto failed; + } + + if (!(row_len = myxt_load_row_data(rd_ot, rec_data, rb_row_buf, rb_col_cnt))) + goto failed; + + if (rd_ot->ot_table->tab_dic.dic_my_table->found_next_number_field) + ha_set_auto_increment(rd_ot, rd_ot->ot_table->tab_dic.dic_my_table->found_next_number_field); + + if (!xt_tab_new_record(rd_ot, rb_row_buf)) + goto failed; + + if (type == BUP_STANDARD_VAR_RECORD) { + buffer += row_len+1; + if (row_len + 1 > size) { + xt_register_xterr(XT_REG_CONTEXT, XT_ERR_BAD_BACKUP_FORMAT); + goto failed; + } + size -= row_len + 1; + } + + rb_insert_count++; + if (rb_insert_count == XT_RESTORE_BATCH_SIZE) { + if (!xt_xn_commit(rd_thread)) + goto failed; + if (!xt_xn_begin(rd_thread)) + goto failed; + rb_insert_count = 0; + } + } + + return backup::OK; + + failed: + xt_log_and_clear_exception(rd_thread); + return backup::ERROR; +} + + +result_t PBXTRestoreDriver::cancel() +{ + XT_TRACE_CALL(); + /* Nothing to do in cancel(); free() will suffice */ + return backup::OK; +} + +void PBXTRestoreDriver::free() +{ + XT_TRACE_CALL(); + if (rd_ot) { + xt_db_return_table_to_pool_ns(rd_ot); + rd_ot = NULL; + } + //if (rb_row_buf) { + // xt_free_ns(rb_row_buf); + // rb_row_buf = NULL; + //} + if (rb_row_data) { + xt_free_ns(rb_row_data); + rb_row_data = NULL; + } + if (rd_thread->st_xact_data) + xt_xn_rollback(rd_thread); + delete this; +} + +/* + * ----------------------------------------------------------------------- + * BACKUP ENGINE FACTORY + */ + +#define PBXT_BACKUP_VERSION 1 + + +class PBXTBackupEngine: public Backup_engine +{ + public: + PBXTBackupEngine() { }; + + virtual version_t version() const { + return PBXT_BACKUP_VERSION; + }; + + virtual result_t get_backup(const uint32, const Table_list &, Backup_driver* &); + + virtual result_t get_restore(const version_t, const uint32, const Table_list &,Restore_driver* &); + + virtual void free() + { + delete this; + } +}; + +result_t PBXTBackupEngine::get_backup(const u_int count, const Table_list &tables, Backup_driver* &drv) +{ + PBXTBackupDriver *ptr = new PBXTBackupDriver(tables); + + if (!ptr) + return backup::ERROR; + drv = ptr; + return backup::OK; +} + +result_t PBXTBackupEngine::get_restore(const version_t ver, const uint32, + const Table_list &tables, Restore_driver* &drv) +{ + if (ver > PBXT_BACKUP_VERSION) + { + return backup::ERROR; + } + + PBXTRestoreDriver *ptr = new PBXTRestoreDriver(tables); + + if (!ptr) + return backup::ERROR; + drv = (Restore_driver *) ptr; + return backup::OK; +} + + +Backup_result_t pbxt_backup_engine(handlerton *self, Backup_engine* &be) +{ + be = new PBXTBackupEngine(); + + if (!be) + return backup::ERROR; + + return backup::OK; +} + +#endif === added file 'storage/pbxt/src/backup_xt.h' --- a/storage/pbxt/src/backup_xt.h 1970-01-01 00:00:00 +0000 +++ b/storage/pbxt/src/backup_xt.h 2009-11-24 10:55:06 +0000 @@ -0,0 +1,34 @@ +/* Copyright (c) 2009 PrimeBase Technologies GmbH + * + * PrimeBase XT + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * + * 2009-09-07 Paul McCullagh + * + * H&G2JCtL + */ + +#ifndef __backup_xt_h__ +#define __backup_xt_h__ + +#include "xt_defs.h" + +#ifdef MYSQL_SUPPORTS_BACKUP + +Backup_result_t pbxt_backup_engine(handlerton *self, Backup_engine* &be); + +#endif +#endif === modified file 'storage/pbxt/src/cache_xt.cc' --- a/storage/pbxt/src/cache_xt.cc 2009-10-30 18:50:56 +0000 +++ b/storage/pbxt/src/cache_xt.cc 2009-11-24 10:55:06 +0000 @@ -73,7 +73,7 @@ #define IDX_CAC_UNLOCK(i, o) xt_xsmutex_unlock(&(i)->cs_lock, (o)->t_id) #elif defined(IDX_CAC_USE_PTHREAD_RW) #define IDX_CAC_LOCK_TYPE xt_rwlock_type -#define IDX_CAC_INIT_LOCK(s, i) xt_init_rwlock(s, &(i)->cs_lock) +#define IDX_CAC_INIT_LOCK(s, i) xt_init_rwlock_with_autoname(s, &(i)->cs_lock) #define IDX_CAC_FREE_LOCK(s, i) xt_free_rwlock(&(i)->cs_lock) #define IDX_CAC_READ_LOCK(i, o) xt_slock_rwlock_ns(&(i)->cs_lock) #define IDX_CAC_WRITE_LOCK(i, o) xt_xlock_rwlock_ns(&(i)->cs_lock) @@ -94,8 +94,12 @@ #define IDX_CAC_UNLOCK(i, s) xt_spinxslock_unlock(&(i)->cs_lock, (s)->t_id) #endif +#ifdef XT_NO_ATOMICS +#define ID_HANDLE_USE_PTHREAD_RW +#else #define ID_HANDLE_USE_SPINLOCK //#define ID_HANDLE_USE_PTHREAD_RW +#endif #if defined(ID_HANDLE_USE_PTHREAD_RW) #define ID_HANDLE_LOCK_TYPE xt_mutex_type @@ -374,7 +378,7 @@ xtPublic void xt_ind_release_handle(XTIn { DcHandleSlotPtr hs; XTIndBlockPtr block = NULL; - u_int hash_idx = 0; + u_int hash_idx = 0; DcSegmentPtr seg = NULL; XTIndBlockPtr xblock; @@ -1379,7 +1383,7 @@ xtPublic xtBool xt_ind_fetch(XTOpenTable ASSERT_NS(iref->ir_xlock == 2); #endif if (!(block = ind_cac_fetch(ot, ind, address, &seg, TRUE))) - return 0; + return FAILED; branch_size = XT_GET_DISK_2(((XTIdxBranchDPtr) block->cb_data)->tb_size_2); if (XT_GET_INDEX_BLOCK_LEN(branch_size) < 2 || XT_GET_INDEX_BLOCK_LEN(branch_size) > XT_INDEX_PAGE_SIZE) { === modified file 'storage/pbxt/src/cache_xt.h' --- a/storage/pbxt/src/cache_xt.h 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/cache_xt.h 2009-11-24 10:55:06 +0000 @@ -62,7 +62,7 @@ struct XTIdxReadBuffer; #define XT_IPAGE_UNLOCK(i, x) xt_atomicrwlock_unlock(i, x) #elif defined(XT_IPAGE_USE_PTHREAD_RW) #define XT_IPAGE_LOCK_TYPE xt_rwlock_type -#define XT_IPAGE_INIT_LOCK(s, i) xt_init_rwlock(s, i) +#define XT_IPAGE_INIT_LOCK(s, i) xt_init_rwlock_with_autoname(s, i) #define XT_IPAGE_FREE_LOCK(s, i) xt_free_rwlock(i) #define XT_IPAGE_READ_LOCK(i) xt_slock_rwlock_ns(i) #define XT_IPAGE_WRITE_LOCK(i, s) xt_xlock_rwlock_ns(i) === modified file 'storage/pbxt/src/database_xt.cc' --- a/storage/pbxt/src/database_xt.cc 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/database_xt.cc 2009-11-24 10:55:06 +0000 @@ -54,6 +54,8 @@ * GLOBALS */ +xtPublic XTDatabaseHPtr pbxt_database = NULL; // The global open database + xtPublic xtLogOffset xt_db_log_file_threshold; xtPublic size_t xt_db_log_buffer_size; xtPublic size_t xt_db_transaction_buffer_size; @@ -505,6 +507,15 @@ xtPublic XTDatabaseHPtr xt_get_database( * all index entries that are not visible have * been removed. * + * REASON WHY WE SET ROWID ON RECOVERY: + * The row ID is set on recovery because the + * change to the index may be lost after a crash. + * The change to the index is done by the sweeper, and + * there is no record of this change in the log. + * The sweeper will not "re-sweep" all transations + * that are recovered. As a result, this upadte + * of the index by the sweeper may be lost. + * * {OPEN-DB-SWEEPER-WAIT} * This has been moved to after the release of the open * database lock because: @@ -518,9 +529,12 @@ xtPublic XTDatabaseHPtr xt_get_database( * - To open the database it needs the open database * lock. */ + /* + * This has been moved, see: {WAIT-FOR-SW-AFTER-RECOV} pushr_(xt_heap_release, db); xt_wait_for_sweeper(self, db, 0); popr_(); + */ return db; } === modified file 'storage/pbxt/src/database_xt.h' --- a/storage/pbxt/src/database_xt.h 2009-04-02 10:03:14 +0000 +++ b/storage/pbxt/src/database_xt.h 2009-11-24 10:55:06 +0000 @@ -105,6 +105,8 @@ typedef struct XTTablePath { #define XT_THREAD_IDLE 1 #define XT_THREAD_INERR 2 +#define XT_XA_HASH_TAB_SIZE 223 + typedef struct XTDatabase : public XTHeap { char *db_name; /* The name of the database, last component of the path! */ char *db_main_path; @@ -131,6 +133,9 @@ typedef struct XTDatabase : public XTHea u_int db_stat_sweep_waits; /* STATISTICS: count the sweeper waits. */ XTDatabaseLogRec db_xlog; /* The transaction log for this database. */ XTXactRestartRec db_restart; /* Database recovery stuff. */ + xt_mutex_type db_xn_xa_lock; + XTXactPreparePtr db_xn_xa_table[XT_XA_HASH_TAB_SIZE]; + XTSortedListPtr db_xn_xa_list; /* The "wait-for" list, of transactions waiting for other transactions. */ XTSortedListPtr db_xn_wait_for; /* The "wait-for" list, of transactions waiting for other transactions. */ u_int db_xn_call_start; /* Start of the post wait calls. */ @@ -198,6 +203,7 @@ void xt_check_database(XTThreadPtr se void xt_add_pbxt_file(size_t size, char *path, const char *file); void xt_add_location_file(size_t size, char *path); +void xt_add_pbxt_dir(size_t size, char *path); void xt_add_system_dir(size_t size, char *path); void xt_add_data_dir(size_t size, char *path); @@ -244,4 +250,6 @@ inline void xt_xlog_check_long_writer(XT } } +extern XTDatabaseHPtr pbxt_database; // The global open database + #endif === modified file 'storage/pbxt/src/datadic_xt.cc' --- a/storage/pbxt/src/datadic_xt.cc 2009-08-18 07:46:53 +0000 +++ b/storage/pbxt/src/datadic_xt.cc 2009-11-24 10:55:06 +0000 @@ -35,7 +35,7 @@ #ifdef DEBUG #ifdef DRIZZLED -#include <drizzled/common_includes.h> +//#include <drizzled/common_includes.h> #else #include "mysql_priv.h" #endif @@ -437,11 +437,6 @@ class XTTokenizer { XTToken *nextToken(XTThreadPtr self, c_char *keyword, XTToken *tk); }; -void ri_free_token(XTThreadPtr XT_UNUSED(self), XTToken *tk) -{ - delete tk; -} - XTToken *XTTokenizer::newToken(XTThreadPtr self, u_int type, char *start, char *end) { if (!tkn_current) { === modified file 'storage/pbxt/src/datalog_xt.cc' --- a/storage/pbxt/src/datalog_xt.cc 2009-09-04 07:29:34 +0000 +++ b/storage/pbxt/src/datalog_xt.cc 2009-11-24 10:55:06 +0000 @@ -410,7 +410,7 @@ static void dl_recover_log(XTThreadPtr s ASSERT_NS(seq_read.sl_log_eof == seq_read.sl_rec_log_offset); data_log->dlf_log_eof = seq_read.sl_rec_log_offset; - if ((size_t) data_log->dlf_log_eof < sizeof(XTXactLogHeaderDRec)) { + if (data_log->dlf_log_eof < (off_t) sizeof(XTXactLogHeaderDRec)) { data_log->dlf_log_eof = sizeof(XTXactLogHeaderDRec); if (!dl_create_log_header(data_log, seq_read.sl_log_file, self)) xt_throw(self); @@ -1162,7 +1162,7 @@ xtBool XTDataLogBuffer::dlb_close_log(XT /* When I use 'thread' instead of 'self', this means * that I will not throw an error. */ -xtBool XTDataLogBuffer::dlb_get_log_offset(xtLogID *log_id, xtLogOffset *out_offset, size_t req_size, struct XTThread *thread) +xtBool XTDataLogBuffer::dlb_get_log_offset(xtLogID *log_id, xtLogOffset *out_offset, size_t XT_UNUSED(req_size), struct XTThread *thread) { /* Note, I am allowing a log to grow beyond the threshold. * The amount depends on the maximum extended record size. @@ -1757,7 +1757,7 @@ static xtBool dl_collect_garbage(XTThrea freer_(); // xt_unlock_mutex(&db->db_co_dlog_lock) /* Flush the transaction log. */ - if (!xt_xlog_flush_log(self)) + if (!xt_xlog_flush_log(db, self)) xt_throw(self); xt_lock_mutex_ns(&db->db_datalogs.dlc_head_lock); @@ -1891,7 +1891,7 @@ static xtBool dl_collect_garbage(XTThrea freer_(); // xt_unlock_mutex(&db->db_co_dlog_lock) /* Flush the transaction log. */ - if (!xt_xlog_flush_log(self)) + if (!xt_xlog_flush_log(db, self)) xt_throw(self); /* Save state in source log header. */ === modified file 'storage/pbxt/src/discover_xt.cc' --- a/storage/pbxt/src/discover_xt.cc 2009-11-16 20:49:51 +0000 +++ b/storage/pbxt/src/discover_xt.cc 2009-12-02 11:50:40 +0000 @@ -31,6 +31,9 @@ #include <drizzled/session.h> #include <drizzled/server_includes.h> #include <drizzled/sql_base.h> +#include <drizzled/statement/alter_table.h> +#include <algorithm> +#include <sstream> #endif #include "strutil_xt.h" @@ -39,18 +42,273 @@ #include "ha_xtsys.h" #ifndef DRIZZLED -#if MYSQL_VERSION_ID > 60005 +#if MYSQL_VERSION_ID >= 50404 #define DOT_STR(x) x.str #else #define DOT_STR(x) x #endif #endif -#ifndef DRIZZLED +//#ifndef DRIZZLED #define LOCK_OPEN_HACK_REQUIRED -#endif // DRIZZLED +//#endif // DRIZZLED #ifdef LOCK_OPEN_HACK_REQUIRED +#ifdef DRIZZLED + +using namespace drizzled; +using namespace std; + +#define mysql_create_table_no_lock hacked_mysql_create_table_no_lock + +namespace drizzled { + +int rea_create_table(Session *session, const char *path, + const char *db, const char *table_name, + message::Table *table_proto, + HA_CREATE_INFO *create_info, + List<CreateField> &create_field, + uint32_t key_count,KEY *key_info); +} + +static uint32_t build_tmptable_filename(Session* session, + char *buff, size_t bufflen) +{ + uint32_t length; + ostringstream path_str, post_tmpdir_str; + string tmp; + + path_str << drizzle_tmpdir; + post_tmpdir_str << "/" << TMP_FILE_PREFIX << current_pid; + post_tmpdir_str << session->thread_id << session->tmp_table++; + tmp= post_tmpdir_str.str(); + + transform(tmp.begin(), tmp.end(), tmp.begin(), ::tolower); + + path_str << tmp; + + if (bufflen < path_str.str().length()) + length= 0; + else + length= unpack_filename(buff, path_str.str().c_str()); + + return length; +} + +static bool mysql_create_table_no_lock(Session *session, + const char *db, const char *table_name, + HA_CREATE_INFO *create_info, + message::Table *table_proto, + AlterInfo *alter_info, + bool internal_tmp_table, + uint32_t select_field_count) +{ + char path[FN_REFLEN]; + uint32_t path_length; + uint db_options, key_count; + KEY *key_info_buffer; + Cursor *file; + bool error= true; + /* Check for duplicate fields and check type of table to create */ + if (!alter_info->create_list.elements) + { + my_message(ER_TABLE_MUST_HAVE_COLUMNS, ER(ER_TABLE_MUST_HAVE_COLUMNS), + MYF(0)); + return true; + } + assert(strcmp(table_name,table_proto->name().c_str())==0); + if (check_engine(session, table_name, create_info)) + return true; + db_options= create_info->table_options; + if (create_info->row_type == ROW_TYPE_DYNAMIC) + db_options|=HA_OPTION_PACK_RECORD; + + /*if (!(file= create_info->db_type->getCursor((TableShare*) 0, session->mem_root))) + { + my_error(ER_OUTOFMEMORY, MYF(0), sizeof(Cursor)); + return true; + }*/ + + /* PMC - Done to avoid getting the partition handler by mistake! */ + if (!(file= new (session->mem_root) ha_xtsys(pbxt_hton, NULL))) + { + my_error(ER_OUTOFMEMORY, MYF(0), sizeof(Cursor)); + return true; + } + + set_table_default_charset(create_info, (char*) db); + + if (mysql_prepare_create_table(session, + create_info, + table_proto, + alter_info, + internal_tmp_table, + &db_options, file, + &key_info_buffer, &key_count, + select_field_count)) + goto err; + + /* Check if table exists */ + if (create_info->options & HA_LEX_CREATE_TMP_TABLE) + { + path_length= build_tmptable_filename(session, path, sizeof(path)); + } + else + { + #ifdef FN_DEVCHAR + /* check if the table name contains FN_DEVCHAR when defined */ + if (strchr(table_name, FN_DEVCHAR)) + { + my_error(ER_WRONG_TABLE_NAME, MYF(0), table_name); + return true; + } +#endif + path_length= build_table_filename(path, sizeof(path), db, table_name, internal_tmp_table); + } + + /* Check if table already exists */ + if ((create_info->options & HA_LEX_CREATE_TMP_TABLE) && + session->find_temporary_table(db, table_name)) + { + if (create_info->options & HA_LEX_CREATE_IF_NOT_EXISTS) + { + create_info->table_existed= 1; // Mark that table existed + push_warning_printf(session, DRIZZLE_ERROR::WARN_LEVEL_NOTE, + ER_TABLE_EXISTS_ERROR, ER(ER_TABLE_EXISTS_ERROR), + table_name); + error= 0; + goto err; + } + my_error(ER_TABLE_EXISTS_ERROR, MYF(0), table_name); + goto err; + } + + //pthread_mutex_lock(&LOCK_open); /* CREATE TABLE (some confussion on naming, double check) */ + if (!internal_tmp_table && !(create_info->options & HA_LEX_CREATE_TMP_TABLE)) + { + if (plugin::StorageEngine::getTableDefinition(*session, + path, + db, + table_name, + internal_tmp_table) == EEXIST) + { + if (create_info->options & HA_LEX_CREATE_IF_NOT_EXISTS) + { + error= false; + push_warning_printf(session, DRIZZLE_ERROR::WARN_LEVEL_NOTE, + ER_TABLE_EXISTS_ERROR, ER(ER_TABLE_EXISTS_ERROR), + table_name); + create_info->table_existed= 1; // Mark that table existed + } + else + my_error(ER_TABLE_EXISTS_ERROR,MYF(0),table_name); + + goto unlock_and_end; + } + /* + * We don't assert here, but check the result, because the table could be + * in the table definition cache and in the same time the .frm could be + * missing from the disk, in case of manual intervention which deletes + * the .frm file. The user has to use FLUSH TABLES; to clear the cache. + * Then she could create the table. This case is pretty obscure and + * therefore we don't introduce a new error message only for it. + * */ + if (TableShare::getShare(db, table_name)) + { + my_error(ER_TABLE_EXISTS_ERROR, MYF(0), table_name); + goto unlock_and_end; + } + } + /* + * Check that table with given name does not already + * exist in any storage engine. In such a case it should + * be discovered and the error ER_TABLE_EXISTS_ERROR be returned + * unless user specified CREATE TABLE IF EXISTS + * The LOCK_open mutex has been locked to make sure no + * one else is attempting to discover the table. Since + * it's not on disk as a frm file, no one could be using it! + * */ + if (!(create_info->options & HA_LEX_CREATE_TMP_TABLE)) + { + bool create_if_not_exists = + create_info->options & HA_LEX_CREATE_IF_NOT_EXISTS; + + char table_path[FN_REFLEN]; + uint32_t table_path_length; + + table_path_length= build_table_filename(table_path, sizeof(table_path), + db, table_name, false); + + int retcode= plugin::StorageEngine::getTableDefinition(*session, + table_path, + db, + table_name, + false); + switch (retcode) + { + case ENOENT: + /* Normal case, no table exists. we can go and create it */ + break; + case EEXIST: + if (create_if_not_exists) + { + error= false; + push_warning_printf(session, DRIZZLE_ERROR::WARN_LEVEL_NOTE, + ER_TABLE_EXISTS_ERROR, ER(ER_TABLE_EXISTS_ERROR), + table_name); + create_info->table_existed= 1; // Mark that table existed + goto unlock_and_end; + } + my_error(ER_TABLE_EXISTS_ERROR,MYF(0),table_name); + goto unlock_and_end; + default: + my_error(retcode, MYF(0),table_name); + goto unlock_and_end; + } + } + + session->set_proc_info("creating table"); + create_info->table_existed= 0; // Mark that table is created + + create_info->table_options=db_options; + + if (rea_create_table(session, path, db, table_name, + table_proto, + create_info, alter_info->create_list, + key_count, key_info_buffer)) + goto unlock_and_end; + + if (create_info->options & HA_LEX_CREATE_TMP_TABLE) + { + /* Open table and put in temporary table list */ + if (!(session->open_temporary_table(path, db, table_name, 1, OTM_OPEN))) + { + (void) session->rm_temporary_table(create_info->db_type, path); + goto unlock_and_end; + } + } + + /* + * Don't write statement if: + * - It is an internal temporary table, + * - Row-based logging is used and it we are creating a temporary table, or + * - The binary log is not open. + * Otherwise, the statement shall be binlogged. + * */ + if (!internal_tmp_table && + ((!(create_info->options & HA_LEX_CREATE_TMP_TABLE)))) + write_bin_log(session, session->query, session->query_length); + error= false; +unlock_and_end: + //pthread_mutex_unlock(&LOCK_open); + +err: + session->set_proc_info("After create"); + delete file; + return(error); +} + +#else // MySQL case /////////////////////////////// /* * Unfortunately I cannot use the standard mysql_create_table_no_lock() because it will lock "LOCK_open" @@ -1229,13 +1487,13 @@ static bool mysql_create_table_no_lock(T if (create_info->options & HA_LEX_CREATE_TMP_TABLE) { /* Open table and put in temporary table list */ -#if MYSQL_VERSION_ID > 60005 +#if MYSQL_VERSION_ID >= 50404 if (!(open_temporary_table(thd, path, db, table_name, 1, OTM_OPEN))) #else if (!(open_temporary_table(thd, path, db, table_name, 1))) #endif { -#if MYSQL_VERSION_ID > 60005 +#if MYSQL_VERSION_ID >= 50404 (void) rm_temporary_table(create_info->db_type, path, false); #else (void) rm_temporary_table(create_info->db_type, path); @@ -1252,11 +1510,21 @@ static bool mysql_create_table_no_lock(T - The binary log is not open. Otherwise, the statement shall be binlogged. */ + /* PBXT 1.0.09e + * Firstly we had a compile problem with MySQL 5.1.42 and + * the write_bin_log() call below: + * discover_xt.cc:1259: error: argument of type 'char* (Statement::)()' does not match 'const char*' + * + * And secondly, we should no write the BINLOG anyway because this is + * an internal PBXT system table. + * + * So I am just commenting out the code altogether. if (!internal_tmp_table && (!thd->current_stmt_binlog_row_based || (thd->current_stmt_binlog_row_based && !(create_info->options & HA_LEX_CREATE_TMP_TABLE)))) - write_bin_log(thd, TRUE, thd->query(), thd->query_length()); + write_bin_log(thd, TRUE, thd->query, thd->query_length); + */ error= FALSE; unlock_and_end: pthread_mutex_unlock(&LOCK_open); @@ -1279,37 +1547,51 @@ warn: ////// END OF CUT AND PASTES FROM sql_table.cc //////// //////////////////////////////////////////////////////// +#endif // DRIZZLED #endif // LOCK_OPEN_HACK_REQUIRED //------------------------------ int xt_create_table_frm(handlerton *hton, THD* thd, const char *db, const char *name, DT_FIELD_INFO *info, DT_KEY_INFO *XT_UNUSED(keys), xtBool skip_existing) { #ifdef DRIZZLED - drizzled::message::Table table_proto; +#define MYLEX_CREATE_INFO create_info +#else +#define MYLEX_CREATE_INFO mylex.create_info +#endif + +#ifdef DRIZZLED + drizzled::statement::AlterTable *stmt = new drizzled::statement::AlterTable(thd); + HA_CREATE_INFO create_info; + //AlterInfo alter_info; + drizzled::message::Table table_proto; static const char *ext = ".dfe"; static const int ext_len = 4; + + table_proto.mutable_engine()->mutable_name()->assign("PBXT"); #else static const char *ext = ".frm"; static const int ext_len = 4; #endif int err = 1; - //HA_CREATE_INFO create_info = {0}; - //Alter_info alter_info; char field_length_buffer[12], *field_length_ptr; LEX *save_lex= thd->lex, mylex; - - memset(&mylex.create_info, 0, sizeof(HA_CREATE_INFO)); + + memset(&MYLEX_CREATE_INFO, 0, sizeof(HA_CREATE_INFO)); thd->lex = &mylex; - lex_start(thd); + lex_start(thd); +#ifdef DRIZZLED + mylex.statement = stmt; +#endif /* setup the create info */ - mylex.create_info.db_type = hton; + MYLEX_CREATE_INFO.db_type = hton; + #ifndef DRIZZLED mylex.create_info.frm_only = 1; #endif - mylex.create_info.default_table_charset = system_charset_info; + MYLEX_CREATE_INFO.default_table_charset = system_charset_info; /* setup the column info. */ while (info->field_name) { @@ -1335,7 +1617,7 @@ int xt_create_table_frm(handlerton *hton #else if (add_field_to_list(thd, &field_name, info->field_type, field_length_ptr, info->field_decimal_length, info->field_flags, -#if MYSQL_VERSION_ID > 60005 +#if MYSQL_VERSION_ID >= 50404 HA_SM_DISK, COLUMN_FORMAT_TYPE_FIXED, #endif @@ -1369,7 +1651,7 @@ int xt_create_table_frm(handlerton *hton table_proto.set_name(name); table_proto.set_type(drizzled::message::Table::STANDARD); - if (mysql_create_table_no_lock(thd, db, name, &mylex.create_info, &table_proto, &mylex.alter_info, 1, 0, false)) + if (mysql_create_table_no_lock(thd, db, name, &create_info, &table_proto, &stmt->alter_info, 1, 0)) goto error; #else if (mysql_create_table_no_lock(thd, db, name, &mylex.create_info, &mylex.alter_info, 1, 0)) === modified file 'storage/pbxt/src/filesys_xt.cc' --- a/storage/pbxt/src/filesys_xt.cc 2009-09-03 06:15:03 +0000 +++ b/storage/pbxt/src/filesys_xt.cc 2009-11-24 10:55:06 +0000 @@ -56,6 +56,8 @@ //#define DEBUG_TRACE_FILES //#define INJECT_WRITE_REMAP_ERROR /* This is required to make testing on the Mac faster: */ +/* It turns of full file sync. */ +#define DEBUG_FAST_MAC #endif #ifdef DEBUG_TRACE_FILES @@ -63,10 +65,6 @@ #define PRINTF xt_trace #endif -#if defined(XT_MAC) && defined(F_FULLFSYNC) -#undef F_FULLFSYNC -#endif - #ifdef INJECT_WRITE_REMAP_ERROR #define INJECT_REMAP_FILE_SIZE 1000000 #define INJECT_REMAP_FILE_TYPE "xtd" @@ -883,7 +881,7 @@ xtPublic xtBool xt_flush_file(XTOpenFile * fsync didn't really flush index pages to disk. fcntl(F_FULLFSYNC) is considered more effective * in such case. */ -#ifdef F_FULLFSYNC +#if defined(F_FULLFSYNC) && !defined(DEBUG_FAST_MAC) if (fcntl(of->of_filedes, F_FULLFSYNC, 0) == -1) { xt_register_ferrno(XT_REG_CONTEXT, errno, xt_file_path(of)); goto failed; === modified file 'storage/pbxt/src/filesys_xt.h' --- a/storage/pbxt/src/filesys_xt.h 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/filesys_xt.h 2009-11-24 10:55:06 +0000 @@ -102,7 +102,7 @@ xtBool xt_fs_rename(struct XTThread *s #define FILE_MAP_UNLOCK(i, o) xt_xsmutex_unlock(i, o) #elif defined(FILE_MAP_USE_PTHREAD_RW) #define FILE_MAP_LOCK_TYPE xt_rwlock_type -#define FILE_MAP_INIT_LOCK(s, i) xt_init_rwlock(s, i) +#define FILE_MAP_INIT_LOCK(s, i) xt_init_rwlock_with_autoname(s, i) #define FILE_MAP_FREE_LOCK(s, i) xt_free_rwlock(i) #define FILE_MAP_READ_LOCK(i, o) xt_slock_rwlock_ns(i) #define FILE_MAP_WRITE_LOCK(i, o) xt_xlock_rwlock_ns(i) === modified file 'storage/pbxt/src/ha_pbxt.cc' --- a/storage/pbxt/src/ha_pbxt.cc 2009-09-03 06:15:03 +0000 +++ b/storage/pbxt/src/ha_pbxt.cc 2009-11-27 15:37:02 +0000 @@ -35,6 +35,7 @@ #include <stdlib.h> #include <time.h> +#include <ctype.h> #ifdef DRIZZLED #include <drizzled/common.h> @@ -42,14 +43,21 @@ #include <mysys/my_alloc.h> #include <mysys/hash.h> #include <drizzled/field.h> -#include <drizzled/current_session.h> +#include <drizzled/session.h> #include <drizzled/data_home.h> #include <drizzled/error.h> #include <drizzled/table.h> #include <drizzled/field/timestamp.h> #include <drizzled/server_includes.h> +#include <drizzled/plugin/info_schema_table.h> extern "C" char **session_query(Session *session); #define my_strdup(a,b) strdup(a) + +using drizzled::plugin::Registry; +using drizzled::plugin::ColumnInfo; +using drizzled::plugin::InfoSchemaTable; +using drizzled::plugin::InfoSchemaMethods; + #else #include "mysql_priv.h" #include <mysql/plugin.h> @@ -71,7 +79,7 @@ extern "C" char **session_query(Session #include "tabcache_xt.h" #include "systab_xt.h" #include "xaction_xt.h" -#include "restart_xt.h" +#include "backup_xt.h" #ifdef DEBUG //#define XT_USE_SYS_PAR_DEBUG_SIZES @@ -101,6 +109,10 @@ static void pbxt_drop_database(handlert static int pbxt_close_connection(handlerton *hton, THD* thd); static int pbxt_commit(handlerton *hton, THD *thd, bool all); static int pbxt_rollback(handlerton *hton, THD *thd, bool all); +static int pbxt_prepare(handlerton *hton, THD *thd, bool all); +static int pbxt_recover(handlerton *hton, XID *xid_list, uint len); +static int pbxt_commit_by_xid(handlerton *hton, XID *xid); +static int pbxt_rollback_by_xid(handlerton *hton, XID *xid); #endif static void ha_aquire_exclusive_use(XTThreadPtr self, XTSharePtr share, ha_pbxt *mine); static void ha_release_exclusive_use(XTThreadPtr self, XTSharePtr share); @@ -123,7 +135,9 @@ static void ha_close_open_tables(XTThre #ifdef PBXT_HANDLER_TRACE #define PBXT_ALLOW_PRINTING -#define XT_TRACE_CALL() do { XTThreadPtr s = xt_get_self(); printf("%s %s\n", s ? s->t_name : "-unknown-", __FUNC__); } while (0) +#define XT_TRACE_CALL() ha_trace_function(__FUNC__, NULL) +#define XT_TRACE_METHOD() ha_trace_function(__FUNC__, pb_share->sh_table_path->ps_path) + #ifdef PBXT_TRACE_RETURN #define XT_RETURN(x) do { printf("%d\n", (int) (x)); return (x); } while (0) #define XT_RETURN_VOID do { printf("out\n"); return; } while (0) @@ -135,6 +149,7 @@ static void ha_close_open_tables(XTThre #else #define XT_TRACE_CALL() +#define XT_TRACE_METHOD() #define XT_RETURN(x) return (x) #define XT_RETURN_VOID return @@ -165,10 +180,10 @@ xtBool pbxt_crash_debug = TRUE; xtBool pbxt_crash_debug = FALSE; #endif + /* Variables for pbxt share methods */ static xt_mutex_type pbxt_database_mutex; // Prevent a database from being opened while it is being dropped static XTHashTabPtr pbxt_share_tables; // Hash used to track open tables -XTDatabaseHPtr pbxt_database = NULL; // The global open database static char *pbxt_index_cache_size; static char *pbxt_record_cache_size; static char *pbxt_log_cache_size; @@ -180,6 +195,12 @@ static char *pbxt_data_log_threshold; static char *pbxt_data_file_grow_size; static char *pbxt_row_file_grow_size; static int pbxt_max_threads; +static my_bool pbxt_support_xa; + +#ifndef DRIZZLED +// drizzle complains it's not used +static XTXactEnumXARec pbxt_xa_enum; +#endif #ifdef DEBUG #define XT_SHARE_LOCK_WAIT 5000 @@ -259,6 +280,33 @@ static HAVarParamsRec vp_row_file_grow_s //#define XT_AUTO_INCREMENT_DEF 1 #endif +#ifdef PBXT_HANDLER_TRACE +static void ha_trace_function(const char *function, char *table) +{ + char func_buf[50], *ptr; + XTThreadPtr thread = xt_get_self(); + + if ((ptr = strchr(function, '('))) { + ptr--; + while (ptr > function) { + if (!(isalnum(*ptr) || *ptr == '_')) + break; + ptr--; + } + ptr++; + xt_strcpy(50, func_buf, ptr); + if ((ptr = strchr(func_buf, '('))) + *ptr = 0; + } + else + xt_strcpy(50, func_buf, function); + if (table) + printf("%s %s (%s)\n", thread ? thread->t_name : "-unknown-", func_buf, table); + else + printf("%s %s\n", thread ? thread->t_name : "-unknown-", func_buf); +} +#endif + /* * ----------------------------------------------------------------------- * SHARED TABLE DATA @@ -584,6 +632,9 @@ xtPublic XTThreadPtr xt_ha_thd_to_self(T /* The first bit is 1. */ static u_int ha_get_max_bit(MX_BITMAP *map) { +#ifdef DRIZZLED + return map->getFirstSet(); +#else my_bitmap_map *data_ptr = map->bitmap; my_bitmap_map *end_ptr = map->last_word_ptr; my_bitmap_map b; @@ -612,6 +663,7 @@ static u_int ha_get_max_bit(MX_BITMAP *m cnt -= 32; } return 0; +#endif } /* @@ -684,9 +736,10 @@ xtPublic int xt_ha_pbxt_to_mysql_error(i return(-1); // Unknown error } -xtPublic int xt_ha_pbxt_thread_error_for_mysql(THD *XT_UNUSED(thd), const XTThreadPtr self, int ignore_dup_key) +xtPublic int xt_ha_pbxt_thread_error_for_mysql(THD *thd, const XTThreadPtr self, int ignore_dup_key) { - int xt_err = self->t_exception.e_xt_err; + int xt_err = self->t_exception.e_xt_err; + xtBool dup_key = FALSE; XT_PRINT2(self, "xt_ha_pbxt_thread_error_for_mysql xt_err=%d auto commit=%d\n", (int) xt_err, (int) self->st_auto_commit); switch (xt_err) { @@ -725,6 +778,7 @@ xtPublic int xt_ha_pbxt_thread_error_for /* If we are in auto-commit mode (and we are not ignoring * duplicate keys) then rollback the transaction automatically. */ + dup_key = TRUE; if (!ignore_dup_key && self->st_auto_commit) goto abort_transaction; break; @@ -790,26 +844,20 @@ xtPublic int xt_ha_pbxt_thread_error_for /* Locks are held on tables. * Only rollback after locks are released. */ - self->st_auto_commit = TRUE; + /* I do not think this is required, because + * I tell mysql to rollback below, + * besides it is a hack! + self->st_auto_commit = TRUE; + */ self->st_abort_trans = TRUE; } -#ifdef xxxx -/* DBUG_ASSERT(thd->transaction.stmt.ha_list == NULL || - trans == &thd->transaction.stmt); in handler.cc now - * fails, and I don't know if this function can be called anymore! */ - /* Cause any other DBs to do a rollback as well... */ - if (thd) { - /* - * GOTCHA: - * This is a BUG in MySQL. I cannot rollback a transaction if - * pb_mysql_thd->in_sub_stmt! But I must....?! - */ -#ifdef MYSQL_SERVER - if (!thd->in_sub_stmt) - ha_rollback(thd); -#endif + /* Only tell MySQL to rollback if we automatically rollback. + * Note: calling this with (thd, FALSE), cause sp.test to fail. + */ + if (!dup_key) { + if (thd) + thd_mark_transaction_to_rollback(thd, TRUE); } -#endif } break; } @@ -908,7 +956,11 @@ static void pbxt_call_init(XTThreadPtr s xt_db_data_file_grow_size = (size_t) data_file_grow_size; xt_db_row_file_grow_size = (size_t) row_file_grow_size; +#ifdef DRIZZLED + pbxt_ignore_case = TRUE; +#else pbxt_ignore_case = lower_case_table_names != 0; +#endif if (pbxt_ignore_case) pbxt_share_tables = xt_new_hashtable(self, ha_hash_comp_ci, ha_hash_ci, ha_hash_free, TRUE, FALSE); else @@ -968,7 +1020,7 @@ static void pbxt_call_exit(XTThreadPtr s */ static void ha_exit(XTThreadPtr self) { - xt_xres_wait_for_recovery(self); + xt_xres_terminate_recovery(self); /* Wrap things up... */ xt_unuse_database(self, self); /* Just in case the main thread has a database in use (for testing)? */ @@ -1024,7 +1076,7 @@ static bool pbxt_show_status(handlerton cont_(a); if (!not_ok) { - if (stat_print(thd, "PBXT", 4, "", 0, strbuf.sb_cstring, strbuf.sb_len)) + if (stat_print(thd, "PBXT", 4, "", 0, strbuf.sb_cstring, (uint) strbuf.sb_len)) not_ok = TRUE; } xt_sb_set_size(self, &strbuf, 0); @@ -1038,14 +1090,14 @@ static bool pbxt_show_status(handlerton * return 1 on error, else 0. */ #ifdef DRIZZLED -static int pbxt_init(PluginRegistry &registry) +static int pbxt_init(Registry &registry) #else static int pbxt_init(void *p) #endif { int init_err = 0; - XT_TRACE_CALL(); + XT_PRINT0(NULL, "pbxt_init\n"); if (sizeof(xtWordPS) != sizeof(void *)) { printf("PBXT: This won't work, I require that sizeof(xtWordPS) == sizeof(void *)!\n"); @@ -1076,11 +1128,27 @@ static int pbxt_init(void *p) pbxt_hton->close_connection = pbxt_close_connection; /* close_connection, cleanup thread related data. */ pbxt_hton->commit = pbxt_commit; /* commit */ pbxt_hton->rollback = pbxt_rollback; /* rollback */ + if (pbxt_support_xa) { + pbxt_hton->prepare = pbxt_prepare; + pbxt_hton->recover = pbxt_recover; + pbxt_hton->commit_by_xid = pbxt_commit_by_xid; + pbxt_hton->rollback_by_xid = pbxt_rollback_by_xid; + } + else { + pbxt_hton->prepare = NULL; + pbxt_hton->recover = NULL; + pbxt_hton->commit_by_xid = NULL; + pbxt_hton->rollback_by_xid = NULL; + } pbxt_hton->create = pbxt_create_handler; /* Create a new handler */ pbxt_hton->drop_database = pbxt_drop_database; /* Drop a database */ pbxt_hton->panic = pbxt_panic; /* Panic call */ pbxt_hton->show_status = pbxt_show_status; pbxt_hton->flags = HTON_NO_FLAGS; /* HTON_CAN_RECREATE - Without this flags TRUNCATE uses delete_all_rows() */ + pbxt_hton->slot = (uint)-1; /* assign invald value, so we know when it's inited later */ +#if defined(MYSQL_SUPPORTS_BACKUP) && defined(XT_ENABLE_ONLINE_BACKUP) + pbxt_hton->get_backup_engine = pbxt_backup_engine; +#endif #endif if (!xt_init_logging()) /* Initialize logging */ goto error_1; @@ -1160,8 +1228,10 @@ static int pbxt_init(void *p) * Only real problem, 2 threads try to load the same * plugin at the same time. */ +#if MYSQL_VERSION_ID < 60014 myxt_mutex_unlock(&LOCK_plugin); #endif +#endif /* Can't do this here yet, because I need a THD! */ try_(b) { @@ -1195,8 +1265,10 @@ static int pbxt_init(void *p) if (thd) myxt_destroy_thread(thd, FALSE); #ifndef DRIZZLED +#if MYSQL_VERSION_ID < 60014 myxt_mutex_lock(&LOCK_plugin); #endif +#endif } #endif } @@ -1262,7 +1334,7 @@ static int pbxt_init(void *p) } #ifdef DRIZZLED -static int pbxt_end(PluginRegistry &registry) +static int pbxt_end(Registry &registry) #else static int pbxt_end(void *) #endif @@ -1378,7 +1450,7 @@ static int pbxt_commit(handlerton *hton, * transaction (!all && !self->st_auto_commit). */ if (all || self->st_auto_commit) { - XT_PRINT0(self, "xt_xn_commit\n"); + XT_PRINT0(self, "xt_xn_commit in pbxt_commit\n"); if (!xt_xn_commit(self)) err = xt_ha_pbxt_thread_error_for_mysql(thd, self, FALSE); @@ -1402,7 +1474,7 @@ static int pbxt_rollback(handlerton *hto XTThreadPtr self; if ((self = (XTThreadPtr) *thd_ha_data(thd, hton))) { - XT_PRINT1(self, "pbxt_rollback all=%d\n", all); + XT_PRINT1(self, "pbxt_rollback all=%d in pbxt_commit\n", all); if (self->st_xact_data) { /* There are no table locks, rollback immediately in all cases @@ -1431,7 +1503,7 @@ static int pbxt_rollback(handlerton *hto } #ifdef DRIZZLED -handler *PBXTStorageEngine::create(TABLE_SHARE *table, MEM_ROOT *mem_root) +Cursor *PBXTStorageEngine::create(TABLE_SHARE *table, MEM_ROOT *mem_root) { PBXTStorageEngine * const hton = this; #else @@ -1446,6 +1518,182 @@ static handler *pbxt_create_handler(hand /* * ----------------------------------------------------------------------- + * 2-PHASE COMMIT + * + */ + +#ifndef DRIZZLED + +static int pbxt_prepare(handlerton *hton, THD *thd, bool all) +{ + int err = 0; + XTThreadPtr self; + + XT_TRACE_CALL(); + if ((self = (XTThreadPtr) *thd_ha_data(thd, hton))) { + XT_PRINT1(self, "pbxt_commit all=%d\n", all); + + if (self->st_xact_data) { + /* There are no table locks, commit immediately in all cases + * except when this is a statement commit with an explicit + * transaction (!all && !self->st_auto_commit). + */ + if (all) { + XID xid; + + XT_PRINT0(self, "xt_xn_prepare in pbxt_prepare\n"); + thd_get_xid(thd, (MYSQL_XID*) &xid); + + if (!xt_xn_prepare(xid.length(), (xtWord1 *) &xid, self)) + err = xt_ha_pbxt_thread_error_for_mysql(thd, self, FALSE); + } + } + } + return err; +} + +static XTThreadPtr ha_temp_open_global_database(handlerton *hton, THD **ret_thd, int *temp_thread, char *thread_name, int *err) +{ + THD *thd; + XTThreadPtr self = NULL; + + *temp_thread = 0; + if ((thd = current_thd)) + self = (XTThreadPtr) *thd_ha_data(thd, hton); + else { + //thd = (THD *) myxt_create_thread(); + //*temp_thread |= 2; + } + + if (!self) { + XTExceptionRec e; + + if (!(self = xt_create_thread(thread_name, FALSE, TRUE, &e))) { + *err = xt_ha_pbxt_to_mysql_error(e.e_xt_err); + xt_log_exception(NULL, &e, XT_LOG_DEFAULT); + return NULL; + } + *temp_thread |= 1; + } + + xt_xres_wait_for_recovery(self, XT_RECOVER_DONE); + + try_(a) { + xt_open_database(self, mysql_real_data_home, TRUE); + } + catch_(a) { + *err = xt_ha_pbxt_thread_error_for_mysql(thd, self, FALSE); + if ((*temp_thread & 1)) + xt_free_thread(self); + if (*temp_thread & 2) + myxt_destroy_thread(thd, FALSE); + self = NULL; + } + cont_(a); + + *ret_thd = thd; + return self; +} + +static void ha_temp_close_database(XTThreadPtr self, THD *thd, int temp_thread) +{ + xt_unuse_database(self, self); + if (temp_thread & 1) + xt_free_thread(self); + if (temp_thread & 2) + myxt_destroy_thread(thd, TRUE); +} + +/* Return all prepared transactions, found during recovery. + * This function returns a count. If len is returned, the + * function will be called again. + */ +static int pbxt_recover(handlerton *hton, XID *xid_list, uint len) +{ + xtBool temp_thread; + XTThreadPtr self; + XTDatabaseHPtr db; + uint count = 0; + XTXactPreparePtr xap; + int err; + THD *thd; + + if (!(self = ha_temp_open_global_database(hton, &thd, &temp_thread, "TempForRecover", &err))) + return 0; + + db = self->st_database; + + for (count=0; count<len; count++) { + xap = xt_xn_enum_xa_data(db, &pbxt_xa_enum); + if (!xap) + break; + memcpy(&xid_list[count], xap->xp_xa_data, xap->xp_data_len); + } + + ha_temp_close_database(self, thd, temp_thread); + return (int) count; +} + +static int pbxt_commit_by_xid(handlerton *hton, XID *xid) +{ + xtBool temp_thread; + XTThreadPtr self; + XTDatabaseHPtr db; + int err = 0; + XTXactPreparePtr xap; + THD *thd; + + XT_TRACE_CALL(); + + if (!(self = ha_temp_open_global_database(hton, &thd, &temp_thread, "TempForCommitXA", &err))) + return err; + db = self->st_database; + + if ((xap = xt_xn_find_xa_data(db, xid->length(), (xtWord1 *) xid, TRUE, self))) { + if ((self->st_xact_data = xt_xn_get_xact(db, xap->xp_xact_id, self))) { + self->st_xact_data->xd_flags &= ~XT_XN_XAC_PREPARED; // Prepared transactions cannot be swept! + if (!xt_xn_commit(self)) + err = xt_ha_pbxt_thread_error_for_mysql(thd, self, FALSE); + } + xt_xn_delete_xa_data(db, xap, TRUE, self); + } + + ha_temp_close_database(self, thd, temp_thread); + return 0; +} + +static int pbxt_rollback_by_xid(handlerton *hton, XID *xid) +{ + int temp_thread; + XTThreadPtr self; + XTDatabaseHPtr db; + int err = 0; + XTXactPreparePtr xap; + THD *thd; + + XT_TRACE_CALL(); + + if (!(self = ha_temp_open_global_database(hton, &thd, &temp_thread, "TempForRollbackXA", &err))) + return err; + db = self->st_database; + + if ((xap = xt_xn_find_xa_data(db, xid->length(), (xtWord1 *) xid, TRUE, self))) { + if ((self->st_xact_data = xt_xn_get_xact(db, xap->xp_xact_id, self))) { + self->st_xact_data->xd_flags &= ~XT_XN_XAC_PREPARED; // Prepared transactions cannot be swept! + if (!xt_xn_rollback(self)) + err = xt_ha_pbxt_thread_error_for_mysql(thd, self, FALSE); + } + xt_xn_delete_xa_data(db, xap, TRUE, self); + } + + ha_temp_close_database(self, thd, temp_thread); + return 0; +} + +#endif + +/* + * ----------------------------------------------------------------------- * HANDLER LOCKING FUNCTIONS * * These functions are used get a lock on all handles of a particular table. @@ -1497,7 +1745,7 @@ static void ha_aquire_exclusive_use(XTTh ha_pbxt *handler; time_t end_time = time(NULL) + XT_SHARE_LOCK_TIMEOUT / 1000; - XT_PRINT1(self, "ha_aquire_exclusive_use %s PBXT X lock\n", share->sh_table_path->ps_path); + XT_PRINT1(self, "ha_aquire_exclusive_use (%s) PBXT X lock\n", share->sh_table_path->ps_path); /* GOTCHA: It is possible to hang here, if you hold * onto the sh_ex_mutex lock, before we really * have the exclusive lock (i.e. before all @@ -1578,7 +1826,7 @@ static void ha_release_exclusive_use(XTT static void ha_release_exclusive_use(XTThreadPtr XT_UNUSED(self), XTSharePtr share) #endif { - XT_PRINT1(self, "ha_release_exclusive_use %s PBXT X UNLOCK\n", share->sh_table_path->ps_path); + XT_PRINT1(self, "ha_release_exclusive_use (%s) PBXT X UNLOCK\n", share->sh_table_path->ps_path); xt_lock_mutex_ns((xt_mutex_type *) share->sh_ex_mutex); share->sh_table_lock = FALSE; xt_broadcast_cond_ns((xt_cond_type *) share->sh_ex_cond); @@ -1589,7 +1837,7 @@ static xtBool ha_wait_for_shared_use(ha_ { time_t end_time = time(NULL) + XT_SHARE_LOCK_TIMEOUT / 1000; - XT_PRINT1(xt_get_self(), "ha_wait_for_shared_use %s share lock wait...\n", share->sh_table_path->ps_path); + XT_PRINT1(xt_get_self(), "ha_wait_for_shared_use (%s) share lock wait...\n", share->sh_table_path->ps_path); mine->pb_ex_in_use = 0; xt_lock_mutex_ns((xt_mutex_type *) share->sh_ex_mutex); while (share->sh_table_lock) { @@ -1667,14 +1915,38 @@ xtPublic int ha_pbxt::reopen() * */ -int pbxt_statistics_fill_table(THD *thd, TABLE_LIST *tables, COND *cond) +static int pbxt_statistics_fill_table(THD *thd, TABLE_LIST *tables, COND *cond) { - XTThreadPtr self; + XTThreadPtr self = NULL; int err = 0; + if (!pbxt_hton) { + /* Can't do if PBXT is not loaded! */ + XTExceptionRec e; + + xt_exception_xterr(&e, XT_CONTEXT, XT_ERR_PBXT_NOT_INSTALLED); + xt_log_exception(NULL, &e, XT_LOG_DEFAULT); + /* Just return an empty set: */ + return 0; + } + if (!(self = ha_set_current_thread(thd, &err))) return xt_ha_pbxt_to_mysql_error(err); + + try_(a) { + /* If the thread has no open database, and the global + * database is already open, then open + * the database. Otherwise the statement will be + * executed without an open database, which means + * that the related statistics will be missing. + * + * This includes all background threads. + */ + if (!self->st_database && pbxt_database) { + xt_ha_open_database_of_table(self, (XTPathStrPtr) NULL); + } + err = myxt_statistics_fill_table(self, thd, tables, cond, system_charset_info); } catch_(a) { @@ -1684,6 +1956,24 @@ int pbxt_statistics_fill_table(THD *thd, return err; } +#ifdef DRIZZLED +ColumnInfo pbxt_statistics_fields_info[]= +{ + ColumnInfo("ID", 4, MYSQL_TYPE_LONG, 0, 0, "The ID of the statistic", SKIP_OPEN_TABLE), + ColumnInfo("Name", 40, MYSQL_TYPE_STRING, 0, 0, "The name of the statistic", SKIP_OPEN_TABLE), + ColumnInfo("Value", 8, MYSQL_TYPE_LONGLONG, 0, 0, "The accumulated value", SKIP_OPEN_TABLE), + ColumnInfo() +}; + +class PBXTStatisticsMethods : public InfoSchemaMethods +{ +public: + int fillTable(Session *session, TableList *tables, COND *cond) + { + return pbxt_statistics_fill_table(session, tables, cond); + } +}; +#else ST_FIELD_INFO pbxt_statistics_fields_info[]= { { "ID", 4, MYSQL_TYPE_LONG, 0, 0, "The ID of the statistic", SKIP_OPEN_TABLE}, @@ -1691,24 +1981,28 @@ ST_FIELD_INFO pbxt_statistics_fields_inf { "Value", 8, MYSQL_TYPE_LONGLONG, 0, 0, "The accumulated value", SKIP_OPEN_TABLE}, { 0, 0, MYSQL_TYPE_STRING, 0, 0, 0, SKIP_OPEN_TABLE} }; +#endif #ifdef DRIZZLED static InfoSchemaTable *pbxt_statistics_table; - -int pbxt_init_statitics(PluginRegistry &registry) +static PBXTStatisticsMethods pbxt_statistics_methods; +static int pbxt_init_statistics(Registry &registry) #else -int pbxt_init_statitics(void *p) +static int pbxt_init_statistics(void *p) #endif { #ifdef DRIZZLED - pbxt_statistics_table = (InfoSchemaTable *)xt_calloc_ns(sizeof(InfoSchemaTable)); - pbxt_statistics_table->table_name= "PBXT_STATISTICS"; + //pbxt_statistics_table = (InfoSchemaTable *)xt_calloc_ns(sizeof(InfoSchemaTable)); + //pbxt_statistics_table->table_name= "PBXT_STATISTICS"; + pbxt_statistics_table = new InfoSchemaTable("PBXT_STATISTICS"); + pbxt_statistics_table->setColumnInfo(pbxt_statistics_fields_info); + pbxt_statistics_table->setInfoSchemaMethods(&pbxt_statistics_methods); registry.add(pbxt_statistics_table); #else ST_SCHEMA_TABLE *pbxt_statistics_table = (ST_SCHEMA_TABLE *) p; -#endif pbxt_statistics_table->fields_info = pbxt_statistics_fields_info; pbxt_statistics_table->fill_table = pbxt_statistics_fill_table; +#endif #if defined(XT_WIN) && defined(XT_COREDUMP) void register_crash_filter(); @@ -1721,14 +2015,14 @@ int pbxt_init_statitics(void *p) } #ifdef DRIZZLED -int pbxt_exit_statitics(PluginRegistry &registry) +static int pbxt_exit_statistics(Registry &registry) #else -int pbxt_exit_statitics(void *XT_UNUSED(p)) +static int pbxt_exit_statistics(void *XT_UNUSED(p)) #endif { #ifdef DRIZZLED registry.remove(pbxt_statistics_table); - xt_free_ns(pbxt_statistics_table); + delete pbxt_statistics_table; #endif return(0); } @@ -1758,7 +2052,11 @@ ha_pbxt::ha_pbxt(handlerton *hton, TABLE * exist for the storage engine. This is also used by the default rename_table and * delete_table method in handler.cc. */ +#ifdef DRIZZLED +const char **PBXTStorageEngine::bas_ext() const +#else const char **ha_pbxt::bas_ext() const +#endif { return pbxt_extensions; } @@ -1800,11 +2098,13 @@ MX_TABLE_TYPES_T ha_pbxt::table_flags() * purposes! HA_NOT_EXACT_COUNT | */ +#ifndef DRIZZLED /* * This basically means we have a file with the name of * database table (which we do). */ HA_FILE_BASED | +#endif /* * Not sure what this does (but MyISAM and InnoDB have it)?! * Could it mean that we support the handler functions. @@ -1971,7 +2271,7 @@ int ha_pbxt::open(const char *table_path if (!(self = ha_set_current_thread(thd, &err))) return xt_ha_pbxt_to_mysql_error(err); - XT_PRINT1(self, "ha_pbxt::open %s\n", table_path); + XT_PRINT1(self, "open (%s)\n", table_path); pb_ex_in_use = 1; try_(a) { @@ -2049,7 +2349,7 @@ int ha_pbxt::close(void) } } - XT_PRINT1(self, "ha_pbxt::close %s\n", pb_share && pb_share->sh_table_path->ps_path ? pb_share->sh_table_path->ps_path : "unknown"); + XT_PRINT1(self, "close (%s)\n", pb_share && pb_share->sh_table_path->ps_path ? pb_share->sh_table_path->ps_path : "unknown"); if (self) { try_(a) { @@ -2125,9 +2425,10 @@ void ha_pbxt::init_auto_increment(xtWord if (!TS(table)->next_number_key_offset) { // Autoincrement at key-start err = index_last(table->record[1]); - if (!err) + if (!err && !table->next_number_field->is_null(TS(table)->rec_buff_length)) { /* {PRE-INC} */ nr = (xtWord8) table->next_number_field->val_int_offset(TS(table)->rec_buff_length); + } } else { /* Do an index scan to find the largest value! */ @@ -2180,8 +2481,10 @@ void ha_pbxt::init_auto_increment(xtWord table->next_number_field = tmp_fie; table->in_use = tmp_thd; - if (xn_started) + if (xn_started) { + XT_PRINT0(self, "xt_xn_commit in init_auto_increment\n"); xt_xn_commit(self); + } } xt_spinlock_unlock(&tab->tab_ainc_lock); } @@ -2228,13 +2531,13 @@ void ha_pbxt::get_auto_increment(MX_ULON * insert into t1 values (-1); * insert into t1 values (NULL); */ -void ha_pbxt::set_auto_increment(Field *nr) +xtPublic void ha_set_auto_increment(XTOpenTablePtr ot, Field *nr) { register XTTableHPtr tab; MX_ULONGLONG_T nr_int_val; nr_int_val = nr->val_int(); - tab = pb_open_tab->ot_table; + tab = ot->ot_table; if (nr->cmp((const unsigned char *)&tab->tab_auto_inc) > 0) { xt_spinlock_lock(&tab->tab_ainc_lock); @@ -2260,9 +2563,9 @@ void ha_pbxt::set_auto_increment(Field * #else tab->tab_dic.dic_min_auto_inc = nr_int_val + 100; #endif - pb_open_tab->ot_thread = xt_get_self(); - if (!xt_tab_write_min_auto_inc(pb_open_tab)) - xt_log_and_clear_exception(pb_open_tab->ot_thread); + ot->ot_thread = xt_get_self(); + if (!xt_tab_write_min_auto_inc(ot)) + xt_log_and_clear_exception(ot->ot_thread); } } } @@ -2305,7 +2608,7 @@ int ha_pbxt::write_row(byte *buf) ASSERT_NS(pb_ex_in_use); - XT_PRINT1(pb_open_tab->ot_thread, "ha_pbxt::write_row %s\n", pb_share->sh_table_path->ps_path); + XT_PRINT1(pb_open_tab->ot_thread, "write_row (%s)\n", pb_share->sh_table_path->ps_path); XT_DISABLED_TRACE(("INSERT tx=%d val=%d\n", (int) pb_open_tab->ot_thread->st_xact_data->xd_start_xn_id, (int) XT_GET_DISK_4(&buf[1]))); //statistic_increment(ha_write_count,&LOCK_status); #ifdef PBMS_ENABLED @@ -2350,7 +2653,7 @@ int ha_pbxt::write_row(byte *buf) err = update_err; goto done; } - set_auto_increment(table->next_number_field); + ha_set_auto_increment(pb_open_tab, table->next_number_field); } if (!xt_tab_new_record(pb_open_tab, (xtWord1 *) buf)) { @@ -2423,7 +2726,7 @@ int ha_pbxt::update_row(const byte * old ASSERT_NS(pb_ex_in_use); - XT_PRINT1(self, "ha_pbxt::update_row %s\n", pb_share->sh_table_path->ps_path); + XT_PRINT1(self, "update_row (%s)\n", pb_share->sh_table_path->ps_path); XT_DISABLED_TRACE(("UPDATE tx=%d val=%d\n", (int) self->st_xact_data->xd_start_xn_id, (int) XT_GET_DISK_4(&new_data[1]))); //statistic_increment(ha_update_count,&LOCK_status); @@ -2472,7 +2775,7 @@ int ha_pbxt::update_row(const byte * old old_map = mx_tmp_use_all_columns(table, table->read_set); nr = table->found_next_number_field->val_int(); - set_auto_increment(table->found_next_number_field); + ha_set_auto_increment(pb_open_tab, table->found_next_number_field); mx_tmp_restore_column_map(table, old_map); } @@ -2504,7 +2807,7 @@ int ha_pbxt::delete_row(const byte * buf ASSERT_NS(pb_ex_in_use); - XT_PRINT1(pb_open_tab->ot_thread, "ha_pbxt::delete_row %s\n", pb_share->sh_table_path->ps_path); + XT_PRINT1(pb_open_tab->ot_thread, "delete_row (%s)\n", pb_share->sh_table_path->ps_path); XT_DISABLED_TRACE(("DELETE tx=%d val=%d\n", (int) pb_open_tab->ot_thread->st_xact_data->xd_start_xn_id, (int) XT_GET_DISK_4(&buf[1]))); //statistic_increment(ha_delete_count,&LOCK_status); @@ -2829,7 +3132,8 @@ int ha_pbxt::xt_index_prev_read(XTOpenTa int ha_pbxt::index_init(uint idx, bool XT_UNUSED(sorted)) { - XTIndexPtr ind; + XTIndexPtr ind; + XTThreadPtr thread = pb_open_tab->ot_thread; /* select count(*) from smalltab_PBXT; * ignores the error below, and continues to @@ -2851,6 +3155,15 @@ int ha_pbxt::index_init(uint idx, bool X printf("index_init %s index %d cols req=%d/%d read_bits=%X write_bits=%X index_bits=%X\n", pb_open_tab->ot_table->tab_name->ps_path, (int) idx, pb_open_tab->ot_cols_req, pb_open_tab->ot_cols_req, (int) *table->read_set->bitmap, (int) *table->write_set->bitmap, (int) *ind->mi_col_map.bitmap); #endif + /* Start a statement based transaction as soon + * as a read is done for a modify type statement! + * Previously, this was done too late! + */ + if (!thread->st_stat_trans) { + trans_register_ha(pb_mysql_thd, FALSE, pbxt_hton); + XT_PRINT0(thread, "ha_pbxt::update_row trans_register_ha all=FALSE\n"); + thread->st_stat_trans = TRUE; + } } else { pb_open_tab->ot_cols_req = ha_get_max_bit(table->read_set); @@ -2901,7 +3214,7 @@ int ha_pbxt::index_init(uint idx, bool X #endif } - xt_xlog_check_long_writer(pb_open_tab->ot_thread); + xt_xlog_check_long_writer(thread); pb_open_tab->ot_thread->st_statistics.st_scan_index++; return 0; @@ -2911,7 +3224,7 @@ int ha_pbxt::index_end() { int err = 0; - XT_TRACE_CALL(); + XT_TRACE_METHOD(); XTThreadPtr thread = pb_open_tab->ot_thread; @@ -2991,7 +3304,7 @@ int ha_pbxt::index_read_xt(byte * buf, u ASSERT_NS(pb_ex_in_use); - XT_PRINT1(pb_open_tab->ot_thread, "ha_pbxt::index_read_xt %s\n", pb_share->sh_table_path->ps_path); + XT_PRINT1(pb_open_tab->ot_thread, "index_read_xt (%s)\n", pb_share->sh_table_path->ps_path); XT_DISABLED_TRACE(("search tx=%d val=%d update=%d\n", (int) pb_open_tab->ot_thread->st_xact_data->xd_start_xn_id, (int) XT_GET_DISK_4(key), pb_modified)); ind = (XTIndexPtr) pb_share->sh_dic_keys[idx]; @@ -3072,7 +3385,7 @@ int ha_pbxt::index_next(byte * buf) int err = 0; XTIndexPtr ind; - XT_TRACE_CALL(); + XT_TRACE_METHOD(); //statistic_increment(ha_read_next_count,&LOCK_status); ASSERT_NS(pb_ex_in_use); @@ -3114,7 +3427,7 @@ int ha_pbxt::index_next_same(byte * buf, XTIndexPtr ind; XTIdxSearchKeyRec search_key; - XT_TRACE_CALL(); + XT_TRACE_METHOD(); //statistic_increment(ha_read_next_count,&LOCK_status); ASSERT_NS(pb_ex_in_use); @@ -3154,7 +3467,7 @@ int ha_pbxt::index_prev(byte * buf) int err = 0; XTIndexPtr ind; - XT_TRACE_CALL(); + XT_TRACE_METHOD(); //statistic_increment(ha_read_prev_count,&LOCK_status); ASSERT_NS(pb_ex_in_use); @@ -3188,7 +3501,7 @@ int ha_pbxt::index_first(byte * buf) XTIndexPtr ind; XTIdxSearchKeyRec search_key; - XT_TRACE_CALL(); + XT_TRACE_METHOD(); //statistic_increment(ha_read_first_count,&LOCK_status); ASSERT_NS(pb_ex_in_use); @@ -3228,7 +3541,7 @@ int ha_pbxt::index_last(byte * buf) XTIndexPtr ind; XTIdxSearchKeyRec search_key; - XT_TRACE_CALL(); + XT_TRACE_METHOD(); //statistic_increment(ha_read_last_count,&LOCK_status); ASSERT_NS(pb_ex_in_use); @@ -3275,10 +3588,11 @@ int ha_pbxt::index_last(byte * buf) */ int ha_pbxt::rnd_init(bool scan) { - int err = 0; + int err = 0; + XTThreadPtr thread = pb_open_tab->ot_thread; - XT_PRINT1(pb_open_tab->ot_thread, "ha_pbxt::rnd_init %s\n", pb_share->sh_table_path->ps_path); - XT_DISABLED_TRACE(("seq scan tx=%d\n", (int) pb_open_tab->ot_thread->st_xact_data->xd_start_xn_id)); + XT_PRINT1(thread, "rnd_init (%s)\n", pb_share->sh_table_path->ps_path); + XT_DISABLED_TRACE(("seq scan tx=%d\n", (int) thread->st_xact_data->xd_start_xn_id)); /* Call xt_tab_seq_exit() to make sure the resources used by the previous * scan are freed. In particular make sure cache page ref count is decremented. @@ -3296,8 +3610,18 @@ int ha_pbxt::rnd_init(bool scan) xt_tab_seq_exit(pb_open_tab); /* The number of columns required: */ - if (pb_open_tab->ot_is_modify) + if (pb_open_tab->ot_is_modify) { pb_open_tab->ot_cols_req = table->read_set->MX_BIT_SIZE(); + /* Start a statement based transaction as soon + * as a read is done for a modify type statement! + * Previously, this was done too late! + */ + if (!thread->st_stat_trans) { + trans_register_ha(pb_mysql_thd, FALSE, pbxt_hton); + XT_PRINT0(thread, "ha_pbxt::update_row trans_register_ha all=FALSE\n"); + thread->st_stat_trans = TRUE; + } + } else { pb_open_tab->ot_cols_req = ha_get_max_bit(table->read_set); @@ -3322,14 +3646,14 @@ int ha_pbxt::rnd_init(bool scan) else xt_tab_seq_reset(pb_open_tab); - xt_xlog_check_long_writer(pb_open_tab->ot_thread); + xt_xlog_check_long_writer(thread); return err; } int ha_pbxt::rnd_end() { - XT_TRACE_CALL(); + XT_TRACE_METHOD(); /* * make permanent the lock for the last scanned row @@ -3358,7 +3682,7 @@ int ha_pbxt::rnd_next(byte *buf) int err = 0; xtBool eof; - XT_TRACE_CALL(); + XT_TRACE_METHOD(); ASSERT_NS(pb_ex_in_use); //statistic_increment(ha_read_rnd_next_count, &LOCK_status); xt_xlog_check_long_writer(pb_open_tab->ot_thread); @@ -3393,7 +3717,7 @@ int ha_pbxt::rnd_next(byte *buf) */ void ha_pbxt::position(const byte *XT_UNUSED(record)) { - XT_TRACE_CALL(); + XT_TRACE_METHOD(); ASSERT_NS(pb_ex_in_use); /* * I changed this from using little endian to big endian. @@ -3429,10 +3753,10 @@ int ha_pbxt::rnd_pos(byte * buf, byte *p { int err = 0; - XT_TRACE_CALL(); + XT_TRACE_METHOD(); ASSERT_NS(pb_ex_in_use); //statistic_increment(ha_read_rnd_count, &LOCK_status); - XT_PRINT1(pb_open_tab->ot_thread, "ha_pbxt::rnd_pos %s\n", pb_share->sh_table_path->ps_path); + XT_PRINT1(pb_open_tab->ot_thread, "rnd_pos (%s)\n", pb_share->sh_table_path->ps_path); pb_open_tab->ot_curr_rec_id = mi_uint4korr((xtWord1 *) pos); switch (xt_tab_dirty_read_record(pb_open_tab, (xtWord1 *) buf)) { @@ -3509,7 +3833,7 @@ int ha_pbxt::info(uint flag) XTOpenTablePtr ot; int in_use; - XT_TRACE_CALL(); + XT_TRACE_METHOD(); if (!(in_use = pb_ex_in_use)) { pb_ex_in_use = 1; @@ -3535,7 +3859,7 @@ int ha_pbxt::info(uint flag) stats.index_file_length = xt_ind_node_to_offset(ot->ot_table, ot->ot_table->tab_ind_eof); stats.delete_length = ot->ot_table->tab_rec_fnum * ot->ot_rec_size; //check_time = info.check_time; - stats.mean_rec_length = ot->ot_rec_size; + stats.mean_rec_length = (ulong) ot->ot_rec_size; } if (flag & HA_STATUS_CONST) { @@ -3551,7 +3875,9 @@ int ha_pbxt::info(uint flag) stats.block_size = XT_INDEX_PAGE_SIZE; if (share->tmp_table == NO_TMP_TABLE) -#if MYSQL_VERSION_ID > 60005 +#ifdef DRIZZLED +#define WHICH_MUTEX mutex +#elif MYSQL_VERSION_ID >= 50404 #define WHICH_MUTEX LOCK_ha_data #else #define WHICH_MUTEX mutex @@ -3559,19 +3885,15 @@ int ha_pbxt::info(uint flag) #ifdef SAFE_MUTEX -#if MYSQL_VERSION_ID < 60000 +#if MYSQL_VERSION_ID < 50404 #if MYSQL_VERSION_ID < 50123 safe_mutex_lock(&share->mutex,__FILE__,__LINE__); #else safe_mutex_lock(&share->mutex,0,__FILE__,__LINE__); #endif #else -#if MYSQL_VERSION_ID < 60004 - safe_mutex_lock(&share->mutex,__FILE__,__LINE__); -#else safe_mutex_lock(&share->WHICH_MUTEX,0,__FILE__,__LINE__); #endif -#endif #else // SAFE_MUTEX @@ -3675,7 +3997,7 @@ int ha_pbxt::extra(enum ha_extra_functio { int err = 0; - XT_PRINT2(xt_get_self(), "ha_pbxt::extra %s operation=%d\n", pb_share->sh_table_path->ps_path, operation); + XT_PRINT2(xt_get_self(), "ha_pbxt::extra (%s) operation=%d\n", pb_share->sh_table_path->ps_path, operation); switch (operation) { case HA_EXTRA_RESET_STATE: @@ -3771,14 +4093,14 @@ int ha_pbxt::extra(enum ha_extra_functio */ int ha_pbxt::reset(void) { - XT_TRACE_CALL(); + XT_TRACE_METHOD(); extra(HA_EXTRA_RESET_STATE); XT_RETURN(0); } void ha_pbxt::unlock_row() { - XT_TRACE_CALL(); + XT_TRACE_METHOD(); if (pb_open_tab) pb_open_tab->ot_table->tab_locks.xt_remove_temp_lock(pb_open_tab, FALSE); } @@ -3802,7 +4124,7 @@ int ha_pbxt::delete_all_rows() XTDDTable *tab_def = NULL; char path[PATH_MAX]; - XT_TRACE_CALL(); + XT_TRACE_METHOD(); if (thd_sql_command(thd) != SQLCOM_TRUNCATE) { /* Just like InnoDB we only handle TRUNCATE TABLE @@ -3906,7 +4228,7 @@ int ha_pbxt::analyze(THD *thd, HA_CHECK_ xtXactID clean_xn_id = 0; uint cnt = 10; - XT_TRACE_CALL(); + XT_TRACE_METHOD(); if (!pb_open_tab) { if ((err = reopen())) @@ -4056,7 +4378,7 @@ xtPublic int ha_pbxt::external_lock(THD ASSERT_NS(pb_ex_in_use); */ - XT_PRINT1(self, "ha_pbxt::EXTERNAL_LOCK %s lock_type=UNLOCK\n", pb_share->sh_table_path->ps_path); + XT_PRINT1(self, "EXTERNAL_LOCK (%s) lock_type=UNLOCK\n", pb_share->sh_table_path->ps_path); /* Make any temporary locks on this table permanent. * @@ -4189,10 +4511,9 @@ xtPublic int ha_pbxt::external_lock(THD xt_broadcast_cond_ns((xt_cond_type *) pb_share->sh_ex_cond); } else { - XT_PRINT2(self, "ha_pbxt::EXTERNAL_LOCK %s lock_type=%d\n", pb_share->sh_table_path->ps_path, lock_type); + XT_PRINT2(self, "ha_pbxt::EXTERNAL_LOCK (%s) lock_type=%d\n", pb_share->sh_table_path->ps_path, lock_type); if (pb_lock_table) { - pb_ex_in_use = 1; try_(a) { if (!pb_table_locked) @@ -4243,14 +4564,18 @@ xtPublic int ha_pbxt::external_lock(THD if ((pb_open_tab->ot_for_update = (lock_type == F_WRLCK))) { switch ((int) thd_sql_command(thd)) { case SQLCOM_DELETE: +#ifndef DRIZZLED case SQLCOM_DELETE_MULTI: +#endif /* turn DELETE IGNORE into normal DELETE. The IGNORE option causes problems because * when a record is deleted we add an xlog record which we cannot "rollback" later * when we find that an FK-constraint has failed. */ thd->lex->ignore = false; case SQLCOM_UPDATE: +#ifndef DRIZZLED case SQLCOM_UPDATE_MULTI: +#endif case SQLCOM_REPLACE: case SQLCOM_REPLACE_SELECT: case SQLCOM_INSERT: @@ -4265,7 +4590,9 @@ xtPublic int ha_pbxt::external_lock(THD case SQLCOM_DROP_TABLE: case SQLCOM_DROP_INDEX: case SQLCOM_LOAD: +#ifndef DRIZZLED case SQLCOM_REPAIR: +#endif case SQLCOM_OPTIMIZE: self->st_stat_modify = TRUE; break; @@ -4316,11 +4643,16 @@ xtPublic int ha_pbxt::external_lock(THD self->st_xact_mode = thd_tx_isolation(thd) <= ISO_READ_COMMITTED ? XT_XACT_COMMITTED_READ : XT_XACT_REPEATABLE_READ; self->st_ignore_fkeys = (thd_test_options(thd,OPTION_NO_FOREIGN_KEY_CHECKS)) != 0; self->st_auto_commit = (thd_test_options(thd, (OPTION_NOT_AUTOCOMMIT | OPTION_BEGIN))) == 0; +#ifdef DRIZZLED + self->st_table_trans = FALSE; +#else self->st_table_trans = thd_sql_command(thd) == SQLCOM_LOCK_TABLES; +#endif self->st_abort_trans = FALSE; self->st_stat_ended = FALSE; self->st_stat_trans = FALSE; XT_PRINT0(self, "xt_xn_begin\n"); + xt_xres_wait_for_recovery(self, XT_RECOVER_SWEPT); if (!xt_xn_begin(self)) { err = xt_ha_pbxt_thread_error_for_mysql(thd, self, pb_ignore_dup_key); pb_ex_in_use = 0; @@ -4404,7 +4736,7 @@ int ha_pbxt::start_stmt(THD *thd, thr_lo if (!(self = ha_set_current_thread(thd, &err))) return xt_ha_pbxt_to_mysql_error(err); - XT_PRINT2(self, "ha_pbxt::start_stmt %s lock_type=%d\n", pb_share->sh_table_path->ps_path, (int) lock_type); + XT_PRINT2(self, "ha_pbxt::start_stmt (%s) lock_type=%d\n", pb_share->sh_table_path->ps_path, (int) lock_type); if (!pb_open_tab) { if ((err = reopen())) @@ -4430,12 +4762,12 @@ int ha_pbxt::start_stmt(THD *thd, thr_lo /* This section handles "auto-commit"... */ if (self->st_xact_data && self->st_auto_commit && self->st_table_trans) { if (self->st_abort_trans) { - XT_PRINT0(self, "xt_xn_rollback\n"); + XT_PRINT0(self, "xt_xn_rollback in start_stmt\n"); if (!xt_xn_rollback(self)) err = xt_ha_pbxt_thread_error_for_mysql(pb_mysql_thd, self, pb_ignore_dup_key); } else { - XT_PRINT0(self, "xt_xn_commit\n"); + XT_PRINT0(self, "xt_xn_commit in start_stmt\n"); if (!xt_xn_commit(self)) err = xt_ha_pbxt_thread_error_for_mysql(pb_mysql_thd, self, pb_ignore_dup_key); } @@ -4466,9 +4798,11 @@ int ha_pbxt::start_stmt(THD *thd, thr_lo if (pb_open_tab->ot_for_update) { switch ((int) thd_sql_command(thd)) { case SQLCOM_UPDATE: - case SQLCOM_UPDATE_MULTI: case SQLCOM_DELETE: +#ifndef DRIZZLED + case SQLCOM_UPDATE_MULTI: case SQLCOM_DELETE_MULTI: +#endif case SQLCOM_REPLACE: case SQLCOM_REPLACE_SELECT: case SQLCOM_INSERT: @@ -4483,14 +4817,15 @@ int ha_pbxt::start_stmt(THD *thd, thr_lo case SQLCOM_DROP_TABLE: case SQLCOM_DROP_INDEX: case SQLCOM_LOAD: +#ifndef DRIZZLED case SQLCOM_REPAIR: +#endif case SQLCOM_OPTIMIZE: self->st_stat_modify = TRUE; break; } } - /* (***) This is required at this level! * No matter how often it is called, it is still the start of a * statement. We need to make sure statements that are NOT mistaken @@ -4516,6 +4851,7 @@ int ha_pbxt::start_stmt(THD *thd, thr_lo self->st_stat_ended = FALSE; self->st_stat_trans = FALSE; XT_PRINT0(self, "xt_xn_begin\n"); + xt_xres_wait_for_recovery(self, XT_RECOVER_SWEPT); if (!xt_xn_begin(self)) { err = xt_ha_pbxt_thread_error_for_mysql(thd, self, pb_ignore_dup_key); goto complete; @@ -4673,7 +5009,9 @@ THR_LOCK_DATA **ha_pbxt::store_lock(THD * */ if ((lock_type >= TL_WRITE_CONCURRENT_INSERT && lock_type <= TL_WRITE) && +#ifndef DRIZZLED !(thd_in_lock_tables(thd) && thd_sql_command(thd) == SQLCOM_LOCK_TABLES) && +#endif !thd_tablespace_op(thd) && thd_sql_command(thd) != SQLCOM_TRUNCATE && thd_sql_command(thd) != SQLCOM_OPTIMIZE && @@ -4691,22 +5029,23 @@ THR_LOCK_DATA **ha_pbxt::store_lock(THD * Stewart: removed SQLCOM_CALL, not sure of implications. */ - if (lock_type == TL_READ_NO_INSERT && - (!thd_in_lock_tables(thd) + if (lock_type == TL_READ_NO_INSERT #ifndef DRIZZLED + && (!thd_in_lock_tables(thd) || thd_sql_command(thd) == SQLCOM_CALL + ) #endif - )) + ) { lock_type = TL_READ; } - XT_PRINT3(xt_get_self(), "ha_pbxt::store_lock %s %d->%d\n", pb_share->sh_table_path->ps_path, pb_lock.type, lock_type); + XT_PRINT3(xt_get_self(), "store_lock (%s) %d->%d\n", pb_share->sh_table_path->ps_path, pb_lock.type, lock_type); pb_lock.type = lock_type; } #ifdef PBXT_HANDLER_TRACE else { - XT_PRINT3(xt_get_self(), "ha_pbxt::store_lock %s %d->%d (ignore/unlock)\n", pb_share->sh_table_path->ps_path, lock_type, lock_type); + XT_PRINT3(xt_get_self(), "store_lock (%s) %d->%d (ignore/unlock)\n", pb_share->sh_table_path->ps_path, lock_type, lock_type); } #endif *to++= &pb_lock; @@ -4723,15 +5062,23 @@ THR_LOCK_DATA **ha_pbxt::store_lock(THD * during create if the table_flag HA_DROP_BEFORE_CREATE was specified for * the storage engine. */ +#ifdef DRIZZLED +int PBXTStorageEngine::doDropTable(Session &, std::string table_path_str) +#else int ha_pbxt::delete_table(const char *table_path) +#endif { THD *thd = current_thd; int err = 0; XTThreadPtr self = NULL; XTSharePtr share; +#ifdef DRIZZLED + const char *table_path = table_path_str.c_str(); +#endif + STAT_TRACE(self, *thd_query(thd)); - XT_PRINT1(self, "ha_pbxt::delete_table %s\n", table_path); + XT_PRINT1(self, "delete_table (%s)\n", table_path); if (XTSystemTableShare::isSystemTable(table_path)) return delete_system_table(table_path); @@ -4795,7 +5142,7 @@ int ha_pbxt::delete_table(const char *ta #endif } catch_(a) { - err = xt_ha_pbxt_thread_error_for_mysql(thd, self, pb_ignore_dup_key); + err = xt_ha_pbxt_thread_error_for_mysql(thd, self, FALSE); #ifdef DRIZZLED if (err == HA_ERR_NO_SUCH_TABLE) err = ENOENT; @@ -4819,7 +5166,11 @@ int ha_pbxt::delete_table(const char *ta return err; } +#ifdef DRIZZLED +int PBXTStorageEngine::delete_system_table(const char *table_path) +#else int ha_pbxt::delete_system_table(const char *table_path) +#endif { THD *thd = current_thd; XTExceptionRec e; @@ -4857,7 +5208,13 @@ int ha_pbxt::delete_system_table(const c * This function can be used to move a table from one database to * another. */ +#ifdef DRIZZLED +int PBXTStorageEngine::doRenameTable(Session *, + const char *from, + const char *to) +#else int ha_pbxt::rename_table(const char *from, const char *to) +#endif { THD *thd = current_thd; int err = 0; @@ -4865,15 +5222,13 @@ int ha_pbxt::rename_table(const char *fr XTSharePtr share; XTDatabaseHPtr to_db; - XT_TRACE_CALL(); - if (XTSystemTableShare::isSystemTable(from)) return rename_system_table(from, to); if (!(self = ha_set_current_thread(thd, &err))) return xt_ha_pbxt_to_mysql_error(err); - XT_PRINT2(self, "ha_pbxt::rename_table %s -> %s\n", from, to); + XT_PRINT2(self, "rename_table (%s -> %s)\n", from, to); #ifdef PBMS_ENABLED PBMSResultRec result; @@ -4929,7 +5284,7 @@ int ha_pbxt::rename_table(const char *fr #endif } catch_(a) { - err = xt_ha_pbxt_thread_error_for_mysql(thd, self, pb_ignore_dup_key); + err = xt_ha_pbxt_thread_error_for_mysql(thd, self, FALSE); } cont_(a); @@ -4940,7 +5295,11 @@ int ha_pbxt::rename_table(const char *fr XT_RETURN(err); } +#ifdef DRIZZLED +int PBXTStorageEngine::rename_system_table(const char *XT_UNUSED(from), const char *XT_UNUSED(to)) +#else int ha_pbxt::rename_system_table(const char *XT_UNUSED(from), const char *XT_UNUSED(to)) +#endif { return ER_NOT_SUPPORTED_YET; } @@ -5029,7 +5388,15 @@ ha_rows ha_pbxt::records_in_range(uint i * Called from handle.cc by ha_create_table(). */ +#ifdef DRIZZLED +int PBXTStorageEngine::doCreateTable(Session *, + const char *table_path, + Table &table_arg, + HA_CREATE_INFO &create_info, + drizzled::message::Table &XT_UNUSED(proto)) +#else int ha_pbxt::create(const char *table_path, TABLE *table_arg, HA_CREATE_INFO *create_info) +#endif { THD *thd = current_thd; int err = 0; @@ -5037,37 +5404,61 @@ int ha_pbxt::create(const char *table_pa XTDDTable *tab_def = NULL; XTDictionaryRec dic; - memset(&dic, 0, sizeof(dic)); + if ((strcmp(table_path, "./pbxt/location") == 0) || (strcmp(table_path, "./pbxt/statistics") == 0)) + return 0; - XT_TRACE_CALL(); + memset(&dic, 0, sizeof(dic)); if (!(self = ha_set_current_thread(thd, &err))) return xt_ha_pbxt_to_mysql_error(err); +#ifdef DRIZZLED + XT_PRINT2(self, "create (%s) %s\n", table_path, (create_info.options & HA_LEX_CREATE_TMP_TABLE) ? "temporary" : ""); +#else + XT_PRINT2(self, "create (%s) %s\n", table_path, (create_info->options & HA_LEX_CREATE_TMP_TABLE) ? "temporary" : ""); +#endif STAT_TRACE(self, *thd_query(thd)); - XT_PRINT1(self, "ha_pbxt::create %s\n", table_path); try_(a) { xt_ha_open_database_of_table(self, (XTPathStrPtr) table_path); +#ifdef DRIZZLED + for (uint i=0; i<TS(&table_arg)->keys; i++) { + if (table_arg.key_info[i].key_length > XT_INDEX_MAX_KEY_SIZE) + xt_throw_sulxterr(XT_CONTEXT, XT_ERR_KEY_TOO_LARGE, table_arg.key_info[i].name, (u_long) XT_INDEX_MAX_KEY_SIZE); + } +#else for (uint i=0; i<TS(table_arg)->keys; i++) { if (table_arg->key_info[i].key_length > XT_INDEX_MAX_KEY_SIZE) xt_throw_sulxterr(XT_CONTEXT, XT_ERR_KEY_TOO_LARGE, table_arg->key_info[i].name, (u_long) XT_INDEX_MAX_KEY_SIZE); } +#endif /* ($) auto_increment_value will be zero if * AUTO_INCREMENT is not used. Otherwise * Query was ALTER TABLE ... AUTO_INCREMENT = x; or * CREATE TABLE ... AUTO_INCREMENT = x; */ +#ifdef DRIZZLED + tab_def = xt_ri_create_table(self, true, (XTPathStrPtr) table_path, *thd_query(thd), myxt_create_table_from_table(self, &table_arg)); + tab_def->checkForeignKeys(self, create_info.options & HA_LEX_CREATE_TMP_TABLE); +#else tab_def = xt_ri_create_table(self, true, (XTPathStrPtr) table_path, *thd_query(thd), myxt_create_table_from_table(self, table_arg)); tab_def->checkForeignKeys(self, create_info->options & HA_LEX_CREATE_TMP_TABLE); +#endif dic.dic_table = tab_def; +#ifdef DRIZZLED + dic.dic_my_table = &table_arg; + dic.dic_tab_flags = (create_info.options & HA_LEX_CREATE_TMP_TABLE) ? XT_TAB_FLAGS_TEMP_TAB : 0; + dic.dic_min_auto_inc = (xtWord8) create_info.auto_increment_value; /* ($) */ + dic.dic_def_ave_row_size = table_arg.s->getAvgRowLength(); +#else dic.dic_my_table = table_arg; dic.dic_tab_flags = (create_info->options & HA_LEX_CREATE_TMP_TABLE) ? XT_TAB_FLAGS_TEMP_TAB : 0; dic.dic_min_auto_inc = (xtWord8) create_info->auto_increment_value; /* ($) */ dic.dic_def_ave_row_size = (xtWord8) table_arg->s->avg_row_length; +#endif myxt_setup_dictionary(self, &dic); /* @@ -5089,7 +5480,7 @@ int ha_pbxt::create(const char *table_pa if (tab_def) tab_def->finalize(self); dic.dic_table = NULL; - err = xt_ha_pbxt_thread_error_for_mysql(thd, self, pb_ignore_dup_key); + err = xt_ha_pbxt_thread_error_for_mysql(thd, self, FALSE); } cont_(a); @@ -5161,7 +5552,7 @@ bool ha_pbxt::get_error_message(int XT_U if (!self->t_exception.e_xt_err) return FALSE; - buf->copy(self->t_exception.e_err_msg, strlen(self->t_exception.e_err_msg), system_charset_info); + buf->copy(self->t_exception.e_err_msg, (uint32) strlen(self->t_exception.e_err_msg), system_charset_info); return TRUE; } @@ -5421,16 +5812,31 @@ static MYSQL_SYSVAR_INT(sweeper_priority #ifdef DRIZZLED static MYSQL_SYSVAR_INT(max_threads, pbxt_max_threads, - PLUGIN_VAR_OPCMDARG, + PLUGIN_VAR_OPCMDARG | PLUGIN_VAR_READONLY, "The maximum number of threads used by PBXT", NULL, NULL, 500, 20, 20000, 1); #else static MYSQL_SYSVAR_INT(max_threads, pbxt_max_threads, - PLUGIN_VAR_OPCMDARG, + PLUGIN_VAR_OPCMDARG | PLUGIN_VAR_READONLY, "The maximum number of threads used by PBXT, 0 = set according to MySQL max_connections.", NULL, NULL, 0, 0, 20000, 1); #endif +#ifndef DEBUG +static MYSQL_SYSVAR_BOOL(support_xa, pbxt_support_xa, + PLUGIN_VAR_OPCMDARG, + "Enable PBXT support for the XA two-phase commit, default is enabled", + NULL, NULL, TRUE); +#else +static MYSQL_SYSVAR_BOOL(support_xa, pbxt_support_xa, + PLUGIN_VAR_OPCMDARG, + "Enable PBXT support for the XA two-phase commit, default is disabled (due to assertion failure in MySQL)", + /* The problem is, in MySQL an assertion fails in debug mode: + * Assertion failed: (total_ha_2pc == (ulong) opt_bin_log+1), function ha_recover, file handler.cc, line 1557. + */ + NULL, NULL, FALSE); +#endif + static struct st_mysql_sys_var* pbxt_system_variables[] = { MYSQL_SYSVAR(index_cache_size), MYSQL_SYSVAR(record_cache_size), @@ -5448,6 +5854,7 @@ static struct st_mysql_sys_var* pbxt_sys MYSQL_SYSVAR(offline_log_function), MYSQL_SYSVAR(sweeper_priority), MYSQL_SYSVAR(max_threads), + MYSQL_SYSVAR(support_xa), NULL }; #endif @@ -5494,8 +5901,8 @@ mysql_declare_plugin(pbxt) "Paul McCullagh, PrimeBase Technologies GmbH", "PBXT internal system statitics", PLUGIN_LICENSE_GPL, - pbxt_init_statitics, /* plugin init */ - pbxt_exit_statitics, /* plugin deinit */ + pbxt_init_statistics, /* plugin init */ + pbxt_exit_statistics, /* plugin deinit */ #ifndef DRIZZLED 0x0005, #endif === modified file 'storage/pbxt/src/ha_pbxt.h' --- a/storage/pbxt/src/ha_pbxt.h 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/ha_pbxt.h 2009-11-24 10:55:06 +0000 @@ -27,9 +27,9 @@ #ifdef DRIZZLED #include <drizzled/common.h> -#include <drizzled/handler.h> -#include <drizzled/plugin/storage_engine.h> #include <mysys/thr_lock.h> +#include <drizzled/cursor.h> + #else #include "mysql_priv.h" #endif @@ -53,17 +53,31 @@ class ha_pbxt; #ifdef DRIZZLED -class PBXTStorageEngine : public StorageEngine { +class PBXTStorageEngine : public drizzled::plugin::StorageEngine +{ + + int delete_system_table(const char *table_path); + int rename_system_table(const char * from, const char * to); + public: PBXTStorageEngine(std::string name_arg) - : StorageEngine(name_arg, HTON_NO_FLAGS) {} + : drizzled::plugin::StorageEngine(name_arg, HTON_NO_FLAGS) {} + + void operator delete(void *) {} + void operator delete[] (void *) {} /* override */ int close_connection(Session *); /* override */ int commit(Session *, bool); /* override */ int rollback(Session *, bool); - /* override */ handler *create(TABLE_SHARE *, MEM_ROOT *); + /* override */ Cursor *create(TABLE_SHARE *, MEM_ROOT *); /* override */ void drop_database(char *); /* override */ bool show_status(Session *, stat_print_fn *, enum ha_stat_type); + /* override */ const char **bas_ext() const; + /* override */ int doCreateTable(Session *session, const char *table_name, + Table &table_arg, HA_CREATE_INFO + &create_info, drizzled::message::Table &proto); + /* override */ int doRenameTable(Session *, const char *from, const char *to); + /* override */ int doDropTable(Session &session, std::string table_path); }; typedef PBXTStorageEngine handlerton; @@ -139,9 +153,9 @@ class ha_pbxt: public handler * don't implement this method unless you really have indexes. */ const char *index_type(uint inx) { (void) inx; return "BTREE"; } - +#ifndef DRIZZLED const char **bas_ext() const; - +#endif MX_UINT8_T table_cache_type(); /* @@ -241,11 +255,13 @@ class ha_pbxt: public handler int optimize(THD* thd, HA_CHECK_OPT* check_opt); int check(THD* thd, HA_CHECK_OPT* check_opt); ha_rows records_in_range(uint inx, key_range *min_key, key_range *max_key); - int delete_table(const char *from); +#ifndef DRIZZLED int delete_system_table(const char *table_path); - int rename_table(const char * from, const char * to); + int delete_table(const char *from); int rename_system_table(const char * from, const char * to); + int rename_table(const char * from, const char * to); int create(const char *name, TABLE *form, HA_CREATE_INFO *create_info); //required +#endif void update_create_info(HA_CREATE_INFO *create_info); THR_LOCK_DATA **store_lock(THD *thd, THR_LOCK_DATA **to, enum thr_lock_type lock_type); //required @@ -277,6 +293,7 @@ struct XTThread *xt_ha_thd_to_self(THD* int xt_ha_pbxt_to_mysql_error(int xt_err); int xt_ha_pbxt_thread_error_for_mysql(THD *thd, const XTThreadPtr self, int ignore_dup_key); void xt_ha_all_threads_close_database(XTThreadPtr self, XTDatabase *db); +void ha_set_auto_increment(XTOpenTablePtr ot, Field *nr); /* * These hooks are suppossed to only be used by InnoDB: === modified file 'storage/pbxt/src/ha_xtsys.h' --- a/storage/pbxt/src/ha_xtsys.h 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/ha_xtsys.h 2009-11-24 10:55:06 +0000 @@ -30,8 +30,9 @@ #ifdef DRIZZLED #include <drizzled/common.h> -#include <drizzled/handler.h> +#include <drizzled/handler_structs.h> #include <drizzled/current_session.h> +#include <drizzled/cursor.h> #else #include "mysql_priv.h" #endif === modified file 'storage/pbxt/src/heap_xt.cc' --- a/storage/pbxt/src/heap_xt.cc 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/heap_xt.cc 2009-11-24 10:55:06 +0000 @@ -73,7 +73,7 @@ xtPublic void xt_check_heap(XTThreadPtr } #ifdef DEBUG_MEMORY -xtPublic void xt_mm_heap_reference(XTThreadPtr self, XTHeapPtr hp, u_int line, c_char *file) +xtPublic void xt_mm_heap_reference(XTThreadPtr XT_UNUSED(self), XTHeapPtr hp, u_int line, c_char *file) #else xtPublic void xt_heap_reference(XTThreadPtr, XTHeapPtr hp) #endif === modified file 'storage/pbxt/src/index_xt.cc' --- a/storage/pbxt/src/index_xt.cc 2009-08-18 07:46:53 +0000 +++ b/storage/pbxt/src/index_xt.cc 2009-11-24 10:55:06 +0000 @@ -829,14 +829,25 @@ static void idx_next_branch_item(XTTable result->sr_item.i_item_offset += result->sr_item.i_item_size + result->sr_item.i_node_ref_size; bitem = branch->tb_data + result->sr_item.i_item_offset; - if (ind->mi_fix_key) - ilen = result->sr_item.i_item_size; + if (result->sr_item.i_item_offset < result->sr_item.i_total_size) { + if (ind->mi_fix_key) + ilen = result->sr_item.i_item_size; + else { + ilen = myxt_get_key_length(ind, bitem) + XT_RECORD_REF_SIZE; + result->sr_item.i_item_size = ilen; + } + xt_get_res_record_ref(bitem + ilen - XT_RECORD_REF_SIZE, result); /* (Only valid if i_item_offset < i_total_size) */ + } else { - ilen = myxt_get_key_length(ind, bitem) + XT_RECORD_REF_SIZE; - result->sr_item.i_item_size = ilen; + result->sr_item.i_item_size = 0; + result->sr_rec_id = 0; + result->sr_row_id = 0; } - xt_get_res_record_ref(bitem + ilen - XT_RECORD_REF_SIZE, result); /* (Only valid if i_item_offset < i_total_size) */ - result->sr_branch = IDX_GET_NODE_REF(tab, bitem, result->sr_item.i_node_ref_size); + if (result->sr_item.i_node_ref_size) + /* IDX_GET_NODE_REF() loads the branch reference to the LEFT of the item. */ + result->sr_branch = IDX_GET_NODE_REF(tab, bitem, result->sr_item.i_node_ref_size); + else + result->sr_branch = 0; } xtPublic void xt_prev_branch_item_fix(XTTableHPtr XT_UNUSED(tab), XTIndexPtr XT_UNUSED(ind), XTIdxBranchDPtr branch, register XTIdxResultRec *result) @@ -3987,7 +3998,7 @@ xtPublic xtBool xt_flush_indices(XTOpenT * here. */ if (!(tab->tab_dic.dic_tab_flags & XT_TAB_FLAGS_TEMP_TAB)) { - if (!xt_xlog_flush_log(ot->ot_thread)) + if (!xt_xlog_flush_log(tab->tab_db, ot->ot_thread)) goto failed_2; if (!il->il_flush(ot)) goto failed_2; === modified file 'storage/pbxt/src/lock_xt.cc' --- a/storage/pbxt/src/lock_xt.cc 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/lock_xt.cc 2009-11-24 10:55:06 +0000 @@ -1246,7 +1246,7 @@ xtPublic void xt_spinlock_init(XTThreadP (void) self; spl->spl_lock = 0; #ifdef XT_NO_ATOMICS - xt_init_mutex(self, &spl->spl_mutex); + xt_init_mutex_with_autoname(self, &spl->spl_mutex); #endif #ifdef DEBUG spl->spl_locker = 0; === modified file 'storage/pbxt/src/locklist_xt.h' --- a/storage/pbxt/src/locklist_xt.h 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/locklist_xt.h 2009-11-27 15:37:02 +0000 @@ -24,11 +24,16 @@ #ifndef __xt_locklist_h__ #define __xt_locklist_h__ +/* + * XT_THREAD_LOCK_INFO and DEBUG_LOCKING code must be updated to avoid calls to xt_get_self() as it can be called before hton->slot is + * assigned by MySQL which is used by xt_get_self() + */ + #ifdef DEBUG -#define XT_THREAD_LOCK_INFO +//#define XT_THREAD_LOCK_INFO #ifndef XT_WIN /* We need DEBUG_LOCKING in order to enable pthread function wrappers */ -#define DEBUG_LOCKING +//#define DEBUG_LOCKING #endif #endif === modified file 'storage/pbxt/src/memory_xt.cc' --- a/storage/pbxt/src/memory_xt.cc 2009-09-03 06:15:03 +0000 +++ b/storage/pbxt/src/memory_xt.cc 2009-11-25 15:06:47 +0000 @@ -34,7 +34,7 @@ #include "trace_xt.h" #ifdef DEBUG -//#define RECORD_MM +#define RECORD_MM #endif #ifdef DEBUG @@ -367,9 +367,8 @@ static long mm_find_pointer(void *ptr) return(-1); } -static long mm_add_pointer(void *ptr, u_int id) +static long mm_add_pointer(void *ptr, u_int XT_UNUSED(id)) { -#pragma unused(id) register int i, n, guess; if (mm_nr_in_use == mm_total_allocated) { === modified file 'storage/pbxt/src/myxt_xt.cc' --- a/storage/pbxt/src/myxt_xt.cc 2009-09-03 06:15:03 +0000 +++ b/storage/pbxt/src/myxt_xt.cc 2009-12-01 09:50:46 +0000 @@ -36,7 +36,7 @@ #include <drizzled/current_session.h> #include <drizzled/sql_lex.h> #include <drizzled/session.h> -extern "C" struct charset_info_st *session_charset(Session *session); +//extern "C" struct charset_info_st *session_charset(Session *session); extern pthread_key_t THR_Session; #else #include "mysql_priv.h" @@ -171,7 +171,9 @@ xtPublic u_int myxt_create_key_from_key( for (u_int i=0; i<ind->mi_seg_count && (int) k_length > 0; i++, old += keyseg->length, keyseg++) { +#ifndef DRIZZLED enum ha_base_keytype type = (enum ha_base_keytype) keyseg->type; +#endif u_int length = keyseg->length < k_length ? keyseg->length : k_length; u_int char_length; xtWord1 *pos; @@ -192,14 +194,18 @@ xtPublic u_int myxt_create_key_from_key( pos = old; if (keyseg->flag & HA_SPACE_PACK) { uchar *end = pos + length; +#ifndef DRIZZLED if (type != HA_KEYTYPE_NUM) { +#endif while (end > pos && end[-1] == ' ') end--; +#ifndef DRIZZLED } else { while (pos < end && pos[0] == ' ') pos++; } +#endif k_length -= length; length = (u_int) (end-pos); FIX_LENGTH(cs, pos, length, char_length); @@ -276,6 +282,7 @@ xtPublic u_int myxt_create_key_from_row( char_length= ((cs && cs->mbmaxlen > 1) ? length/cs->mbmaxlen : length); pos = record + keyseg->start; +#ifndef DRIZZLED if (type == HA_KEYTYPE_BIT) { if (keyseg->bit_length) @@ -289,17 +296,22 @@ xtPublic u_int myxt_create_key_from_row( key+= length; continue; } +#endif if (keyseg->flag & HA_SPACE_PACK) { end = pos + length; +#ifndef DRIZZLED if (type != HA_KEYTYPE_NUM) { +#endif while (end > pos && end[-1] == ' ') end--; +#ifndef DRIZZLED } else { while (pos < end && pos[0] == ' ') pos++; } +#endif length = (u_int) (end-pos); FIX_LENGTH(cs, pos, length, char_length); store_key_length_inc(key,char_length); @@ -333,6 +345,7 @@ xtPublic u_int myxt_create_key_from_row( if (keyseg->flag & HA_SWAP_KEY) { /* Numerical column */ #ifdef HAVE_ISNAN +#ifndef DRIZZLED if (type == HA_KEYTYPE_FLOAT) { float nr; @@ -345,7 +358,9 @@ xtPublic u_int myxt_create_key_from_row( continue; } } - else if (type == HA_KEYTYPE_DOUBLE) { + else +#endif + if (type == HA_KEYTYPE_DOUBLE) { double nr; float8get(nr,pos); @@ -414,6 +429,7 @@ xtPublic u_int myxt_create_foreign_key_f char_length= ((cs && cs->mbmaxlen > 1) ? length/cs->mbmaxlen : length); pos = record + keyseg->start; +#ifndef DRIZZLED if (type == HA_KEYTYPE_BIT) { if (keyseg->bit_length) @@ -427,17 +443,22 @@ xtPublic u_int myxt_create_foreign_key_f key+= length; continue; } +#endif if (keyseg->flag & HA_SPACE_PACK) { end = pos + length; +#ifndef DRIZZLED if (type != HA_KEYTYPE_NUM) { +#endif while (end > pos && end[-1] == ' ') end--; +#ifndef DRIZZLED } else { while (pos < end && pos[0] == ' ') pos++; } +#endif length = (u_int) (end-pos); FIX_LENGTH(cs, pos, length, char_length); store_key_length_inc(key,char_length); @@ -471,6 +492,7 @@ xtPublic u_int myxt_create_foreign_key_f if (keyseg->flag & HA_SWAP_KEY) { /* Numerical column */ #ifdef HAVE_ISNAN +#ifndef DRIZZLED if (type == HA_KEYTYPE_FLOAT) { float nr; @@ -483,7 +505,9 @@ xtPublic u_int myxt_create_foreign_key_f continue; } } - else if (type == HA_KEYTYPE_DOUBLE) { + else +#endif + if (type == HA_KEYTYPE_DOUBLE) { double nr; float8get(nr,pos); @@ -622,7 +646,6 @@ static char *mx_get_length_and_data(Fiel case MYSQL_TYPE_SET: case MYSQL_TYPE_GEOMETRY: #else - case DRIZZLE_TYPE_TINY: case DRIZZLE_TYPE_LONG: case DRIZZLE_TYPE_DOUBLE: case DRIZZLE_TYPE_NULL: @@ -740,7 +763,6 @@ static void mx_set_length_and_data(Field case MYSQL_TYPE_SET: case MYSQL_TYPE_GEOMETRY: #else - case DRIZZLE_TYPE_TINY: case DRIZZLE_TYPE_LONG: case DRIZZLE_TYPE_DOUBLE: case DRIZZLE_TYPE_NULL: @@ -825,6 +847,7 @@ xtPublic xtBool myxt_create_row_from_key } record[keyseg->null_pos] &= ~keyseg->null_bit; } +#ifndef DRIZZLED if (keyseg->type == HA_KEYTYPE_BIT) { uint length = keyseg->length; @@ -845,6 +868,7 @@ xtPublic xtBool myxt_create_row_from_key key+= length; continue; } +#endif if (keyseg->flag & HA_SPACE_PACK) { uint length; @@ -854,16 +878,20 @@ xtPublic xtBool myxt_create_row_from_key goto err; #endif pos = record+keyseg->start; +#ifndef DRIZZLED if (keyseg->type != (int) HA_KEYTYPE_NUM) { +#endif memcpy(pos,key,(size_t) length); bfill(pos+length,keyseg->length-length,' '); +#ifndef DRIZZLED } else { bfill(pos,keyseg->length-length,' '); memcpy(pos+keyseg->length-length,key,(size_t) length); } +#endif key+=length; continue; } @@ -945,7 +973,7 @@ xtPublic xtBool myxt_create_row_from_key static int my_compare_bin(uchar *a, uint a_length, uchar *b, uint b_length, my_bool part_key, my_bool skip_end_space) { - uint length= min(a_length,b_length); + uint length= a_length < b_length ? a_length : b_length; uchar *end= a+ length; int flag; @@ -1023,6 +1051,7 @@ xtPublic u_int myxt_get_key_length(XTInd get_key_pack_length(seg_len, pack_len, key_data); key_data += seg_len; break; +#ifndef DRIZZLED case HA_KEYTYPE_NUM: { /* Numeric key */ if (keyseg->flag & HA_SPACE_PACK) @@ -1035,15 +1064,16 @@ xtPublic u_int myxt_get_key_length(XTInd case HA_KEYTYPE_INT8: case HA_KEYTYPE_SHORT_INT: case HA_KEYTYPE_USHORT_INT: + case HA_KEYTYPE_INT24: + case HA_KEYTYPE_FLOAT: + case HA_KEYTYPE_BIT: +#endif case HA_KEYTYPE_LONG_INT: case HA_KEYTYPE_ULONG_INT: - case HA_KEYTYPE_INT24: case HA_KEYTYPE_UINT24: - case HA_KEYTYPE_FLOAT: case HA_KEYTYPE_DOUBLE: case HA_KEYTYPE_LONGLONG: case HA_KEYTYPE_ULONGLONG: - case HA_KEYTYPE_BIT: key_data += keyseg->length; break; case HA_KEYTYPE_END: @@ -1190,6 +1220,7 @@ xtPublic int myxt_compare_key(XTIndexPtr b += b_length; break; } +#ifndef DRIZZLED case HA_KEYTYPE_INT8: { int i_1 = (int) *((signed char *) a); @@ -1218,6 +1249,7 @@ xtPublic int myxt_compare_key(XTIndexPtr b += keyseg->length; break; } +#endif case HA_KEYTYPE_LONG_INT: { int32 l_1 = sint4korr(a); int32 l_2 = sint4korr(b); @@ -1236,6 +1268,7 @@ xtPublic int myxt_compare_key(XTIndexPtr b += keyseg->length; break; } +#ifndef DRIZZLED case HA_KEYTYPE_INT24: { int32 l_1 = sint3korr(a); int32 l_2 = sint3korr(b); @@ -1245,6 +1278,7 @@ xtPublic int myxt_compare_key(XTIndexPtr b += keyseg->length; break; } +#endif case HA_KEYTYPE_UINT24: { int32 l_1 = uint3korr(a); int32 l_2 = uint3korr(b); @@ -1254,6 +1288,7 @@ xtPublic int myxt_compare_key(XTIndexPtr b += keyseg->length; break; } +#ifndef DRIZZLED case HA_KEYTYPE_FLOAT: { float f_1, f_2; @@ -1270,6 +1305,7 @@ xtPublic int myxt_compare_key(XTIndexPtr b += keyseg->length; break; } +#endif case HA_KEYTYPE_DOUBLE: { double d_1, d_2; @@ -1286,6 +1322,7 @@ xtPublic int myxt_compare_key(XTIndexPtr b += keyseg->length; break; } +#ifndef DRIZZLED case HA_KEYTYPE_NUM: { /* Numeric key */ if (keyseg->flag & HA_SPACE_PACK) { @@ -1339,6 +1376,7 @@ xtPublic int myxt_compare_key(XTIndexPtr b += b_length; break; } +#endif #ifdef HAVE_LONG_LONG case HA_KEYTYPE_LONGLONG: { longlong ll_a = sint8korr(a); @@ -1359,9 +1397,11 @@ xtPublic int myxt_compare_key(XTIndexPtr break; } #endif +#ifndef DRIZZLED case HA_KEYTYPE_BIT: /* TODO: What here? */ break; +#endif case HA_KEYTYPE_END: /* Ready */ goto end; } @@ -1410,16 +1450,19 @@ xtPublic u_int myxt_key_seg_length(XTInd key_length = has_null + a_length + pack_len; break; } +#ifndef DRIZZLED case HA_KEYTYPE_INT8: case HA_KEYTYPE_SHORT_INT: case HA_KEYTYPE_USHORT_INT: + case HA_KEYTYPE_INT24: + case HA_KEYTYPE_FLOAT: +#endif case HA_KEYTYPE_LONG_INT: case HA_KEYTYPE_ULONG_INT: - case HA_KEYTYPE_INT24: case HA_KEYTYPE_UINT24: - case HA_KEYTYPE_FLOAT: case HA_KEYTYPE_DOUBLE: break; +#ifndef DRIZZLED case HA_KEYTYPE_NUM: { /* Numeric key */ if (keyseg->flag & HA_SPACE_PACK) { @@ -1428,14 +1471,17 @@ xtPublic u_int myxt_key_seg_length(XTInd } break; } +#endif #ifdef HAVE_LONG_LONG case HA_KEYTYPE_LONGLONG: case HA_KEYTYPE_ULONGLONG: break; #endif +#ifndef DRIZZLED case HA_KEYTYPE_BIT: /* TODO: What here? */ break; +#endif case HA_KEYTYPE_END: /* Ready */ break; } @@ -1486,7 +1532,7 @@ xtPublic xtWord4 myxt_store_row_length(X return row_size; } -static xtWord4 mx_store_row(XTOpenTablePtr ot, xtWord4 row_size, char *rec_buff) +xtPublic xtWord4 myxt_store_row_data(XTOpenTablePtr ot, xtWord4 row_size, char *rec_buff) { TABLE *table = ot->ot_table->tab_dic.dic_my_table; char *sdata; @@ -1614,8 +1660,9 @@ xtPublic size_t myxt_load_row_length(XTO } /* Unload from PBXT variable length format to the MySQL row format. */ -xtPublic xtBool myxt_load_row(XTOpenTablePtr ot, xtWord1 *source_buf, xtWord1 *dest_buff, u_int col_cnt) +xtPublic xtWord4 myxt_load_row_data(XTOpenTablePtr ot, xtWord1 *source_buf, xtWord1 *dest_buff, u_int col_cnt) { + xtWord1 *input_buf = source_buf; TABLE *table; xtWord4 len; Field *curr_field; @@ -1624,7 +1671,7 @@ xtPublic xtBool myxt_load_row(XTOpenTabl if (!(table = ot->ot_table->tab_dic.dic_my_table)) { xt_register_taberr(XT_REG_CONTEXT, XT_ERR_NO_DICTIONARY, ot->ot_table->tab_name); - return FAILED; + return 0; } /* According to the InnoDB implementation: @@ -1657,7 +1704,7 @@ xtPublic xtBool myxt_load_row(XTOpenTabl default: // Length byte if (*source_buf > 240) { xt_register_xterr(XT_REG_CONTEXT, XT_ERR_BAD_RECORD_FORMAT); - return FAILED; + return 0; } len = *source_buf; source_buf++; @@ -1671,7 +1718,12 @@ xtPublic xtBool myxt_load_row(XTOpenTabl source_buf += len; } - return OK; + return (xtWord4) (source_buf - input_buf); +} + +xtPublic xtBool myxt_load_row(XTOpenTablePtr ot, xtWord1 *source_buf, xtWord1 *dest_buff, u_int col_cnt) +{ + return myxt_load_row_data(ot, source_buf, dest_buff, col_cnt) != 0; } xtPublic xtBool myxt_find_column(XTOpenTablePtr ot, u_int *col_idx, const char *col_name) @@ -1784,7 +1836,7 @@ xtPublic xtBool myxt_store_row(XTOpenTab else { xtWord4 row_size; - if (!(row_size = mx_store_row(ot, XT_REC_EXT_HEADER_SIZE, rec_buff))) + if (!(row_size = myxt_store_row_data(ot, XT_REC_EXT_HEADER_SIZE, rec_buff))) return FAILED; if (row_size - XT_REC_FIX_EXT_HEADER_DIFF <= ot->ot_rec_size) { rec_info->ri_fix_rec_buf = (XTTabRecFixDPtr) &ot->ot_row_wbuffer[XT_REC_FIX_EXT_HEADER_DIFF]; @@ -1951,7 +2003,7 @@ static TABLE *my_open_table(XTThreadPtr #ifdef DRIZZLED share->init(db_name, 0, name, path); - if ((error = open_table_def(thd, share)) || + if ((error = open_table_def(*thd, share)) || (error = open_table_from_share(thd, share, "", 0, (uint32_t) READ_ALL, 0, table, OTM_OPEN))) { xt_free(self, table); @@ -1995,7 +2047,7 @@ static TABLE *my_open_table(XTThreadPtr return NULL; } -#if MYSQL_VERSION_ID >= 60003 +#if MYSQL_VERSION_ID >= 50404 if ((error = open_table_from_share(thd, share, "", 0, (uint) READ_ALL, 0, table, OTM_OPEN))) #else if ((error = open_table_from_share(thd, share, "", 0, (uint) READ_ALL, 0, table, FALSE))) @@ -2145,7 +2197,10 @@ static XTIndexPtr my_create_index(XTThre if (options & HA_OPTION_PACK_KEYS || (index->flags & (HA_PACK_KEY | HA_BINARY_PACK_KEY | HA_SPACE_PACK_USED))) { - if (key_part->length > 8 && (type == HA_KEYTYPE_TEXT || type == HA_KEYTYPE_NUM || + if (key_part->length > 8 && (type == HA_KEYTYPE_TEXT || +#ifndef DRIZZLED + type == HA_KEYTYPE_NUM || +#endif (type == HA_KEYTYPE_BINARY && !field->zero_pack()))) { /* No blobs here */ @@ -2213,8 +2268,12 @@ static XTIndexPtr my_create_index(XTThre else if (field->type() == MYSQL_TYPE_ENUM) { switch (seg->length) { case 2: +#ifdef DRIZZLED + ASSERT_NS(FALSE); +#else seg->type = HA_KEYTYPE_USHORT_INT; break; +#endif case 3: seg->type = HA_KEYTYPE_UINT24; break; @@ -2675,7 +2734,11 @@ xtPublic xtBool myxt_load_dictionary(XTT if (!(my_tab = my_open_table(self, db, tab_path))) return FAILED; dic->dic_my_table = my_tab; +#ifdef DRIZZLED + dic->dic_def_ave_row_size = (xtWord8) my_tab->s->getAvgRowLength(); +#else dic->dic_def_ave_row_size = (xtWord8) my_tab->s->avg_row_length; +#endif myxt_setup_dictionary(self, dic); dic->dic_keys = (XTIndexPtr *) xt_calloc(self, sizeof(XTIndexPtr) * TS(my_tab)->keys); for (uint i=0; i<TS(my_tab)->keys; i++) @@ -2805,8 +2868,10 @@ static void ha_create_dd_index(XTThreadP static char *my_type_to_string(XTThreadPtr self, Field *field, TABLE *XT_UNUSED(my_tab)) { - char buffer[MAX_FIELD_WIDTH + 400], *ptr; + char buffer[MAX_FIELD_WIDTH + 400]; + const char *ptr; String type((char *) buffer, sizeof(buffer), system_charset_info); + xtWord4 len; /* GOTCHA: * - Above sets the string length to the same as the buffer, @@ -2817,10 +2882,17 @@ static char *my_type_to_string(XTThreadP */ type.length(0); field->sql_type(type); - ptr = type.c_ptr(); + ptr = type.ptr(); + len = type.length(); + + if (len >= sizeof(buffer)) + len = sizeof(buffer)-1; + if (ptr != buffer) - xt_strcpy(sizeof(buffer), buffer, ptr); + xt_strcpy(sizeof(buffer), buffer, ptr); + buffer[len] = 0; + if (field->has_charset()) { /* Always include the charset so that we can compare types * for FK/PK releations. @@ -2877,6 +2949,10 @@ xtPublic XTDDTable *myxt_create_table_fr xtPublic void myxt_static_convert_identifier(XTThreadPtr XT_UNUSED(self), MX_CHARSET_INFO *cs, char *from, char *to, size_t to_len) { +#ifdef DRIZZLED + ((void *)cs); + xt_strcpy(to_len, to, from); +#else uint errors; /* @@ -2888,11 +2964,16 @@ xtPublic void myxt_static_convert_identi xt_strcpy(to_len, to, from); else strconvert(cs, from, &my_charset_utf8_general_ci, to, to_len, &errors); +#endif } // cs == current_thd->charset() xtPublic char *myxt_convert_identifier(XTThreadPtr self, MX_CHARSET_INFO *cs, char *from) { +#ifdef DRIZZLED + char *to = xt_dup_string(self, from); + ((void *)cs); +#else uint errors; u_int len; char *to; @@ -2904,6 +2985,7 @@ xtPublic char *myxt_convert_identifier(X to = (char *) xt_malloc(self, len); strconvert(cs, from, &my_charset_utf8_general_ci, to, len, &errors); } +#endif return to; } @@ -2954,11 +3036,19 @@ xtPublic MX_CHARSET_INFO *myxt_getcharse THD *thd = current_thd; if (thd) - return thd_charset(thd); + return (MX_CHARSET_INFO *)thd_charset(thd); } - return &my_charset_utf8_general_ci; + return (MX_CHARSET_INFO *)&my_charset_utf8_general_ci; } +#ifdef DBUG_OFF +//typedef struct st_plugin_int *plugin_ref; +#define REF_MYSQL_PLUGIN(x) (x) +#else +//typedef struct st_plugin_int **plugin_ref; +#define REF_MYSQL_PLUGIN(x) (*(x)) +#endif + xtPublic void *myxt_create_thread() { #ifdef DRIZZLED @@ -3011,12 +3101,22 @@ xtPublic void *myxt_create_thread() return NULL; } - if (!(new_thd = new THD())) { + if (!(new_thd = new THD)) { my_thread_end(); xt_register_error(XT_REG_CONTEXT, XT_ERR_MYSQL_ERROR, 0, "Unable to create MySQL thread (THD)"); return NULL; } + /* + * If PBXT is the default storage engine, then creating any THD objects will add extra + * references to the PBXT plugin object and will effectively deadlock the plugin so + * that server will have to force plugin shutdown. To avoid deadlocking and forced shutdown + * we must dereference the plugin after creating THD objects. + */ + LEX_STRING& plugin_name = REF_MYSQL_PLUGIN(new_thd->variables.table_plugin)->name; + if ((plugin_name.length == 4) && (strncmp(plugin_name.str, "PBXT", plugin_name.length) == 0)) { + REF_MYSQL_PLUGIN(new_thd->variables.table_plugin)->ref_count--; + } new_thd->thread_stack = (char *) &new_thd; new_thd->store_globals(); lex_start(new_thd); @@ -3052,6 +3152,17 @@ xtPublic void myxt_destroy_thread(void * close_thread_tables(thd); #endif + /* + * In myxt_create_thread we decremented plugin ref-count to avoid dead-locking. + * Here we need to increment ref-count to avoid assertion failures. + */ + if (thd->variables.table_plugin) { + LEX_STRING& plugin_name = REF_MYSQL_PLUGIN(thd->variables.table_plugin)->name; + if ((plugin_name.length == 4) && (strncmp(plugin_name.str, "PBXT", plugin_name.length) == 0)) { + REF_MYSQL_PLUGIN(thd->variables.table_plugin)->ref_count++; + } + } + delete thd; /* Remember that we don't have a THD */ @@ -3128,11 +3239,12 @@ xtPublic int myxt_statistics_fill_table( const char *stat_name; u_llong stat_value; XTStatisticsRec statistics; + XTDatabaseHPtr db = self->st_database; xt_gather_statistics(&statistics); for (u_int rec_id=0; !err && rec_id<XT_STAT_CURRENT_MAX; rec_id++) { stat_name = xt_get_stat_meta_data(rec_id)->sm_name; - stat_value = xt_get_statistic(&statistics, self->st_database, rec_id); + stat_value = xt_get_statistic(&statistics, db, rec_id); col=0; mx_put_u_llong(table, col++, rec_id+1); @@ -3213,19 +3325,31 @@ static void myxt_bitmap_init(XTThreadPtr my_bitmap_map *buf; uint size_in_bytes = (((n_bits) + 31) / 32) * 4; - buf = (my_bitmap_map *) xt_malloc(self, size_in_bytes); + buf = (my_bitmap_map *) xt_malloc(self, size_in_bytes); + +#ifdef DRIZZLED + map->init(buf, n_bits); +#else map->bitmap= buf; map->n_bits= n_bits; create_last_word_mask(map); bitmap_clear_all(map); +#endif } static void myxt_bitmap_free(XTThreadPtr self, MX_BITMAP *map) { +#ifdef DRIZZLED + my_bitmap_map *buf = map->getBitmap(); + if (buf) + xt_free(self, buf); + map->setBitmap(NULL); +#else if (map->bitmap) { xt_free(self, map->bitmap); map->bitmap = NULL; } +#endif } /* @@ -3269,3 +3393,29 @@ XTDDColumn *XTDDColumnFactory::createFro return col; } +/* + * ----------------------------------------------------------------------- + * utilities + */ + +/* + * MySQL (not sure about Drizzle) first calls hton->init and then assigns the plugin a thread slot + * which is used by xt_get_self(). This is a problem as pbxt_init() starts a number of daemon threads + * which could try to use the slot before it is assigned. This code waits till slot is inited. + * We cannot directly check hton->slot as in some versions of MySQL it can be 0 before init which is a + * valid value. + */ +extern ulong total_ha; + +xtPublic void myxt_wait_pbxt_plugin_slot_assigned(XTThread *self) +{ +#ifdef DRIZZLED + static LEX_STRING plugin_name = { C_STRING_WITH_LEN("PBXT") }; + + while (!self->t_quit && !Registry::singleton().find(&plugin_name)) + xt_sleep_milli_second(1); +#else + while(!self->t_quit && (pbxt_hton->slot >= total_ha)) + xt_sleep_milli_second(1); +#endif +} === modified file 'storage/pbxt/src/myxt_xt.h' --- a/storage/pbxt/src/myxt_xt.h 2009-09-03 06:15:03 +0000 +++ b/storage/pbxt/src/myxt_xt.h 2009-11-27 15:37:02 +0000 @@ -52,8 +52,10 @@ void myxt_set_default_row_from_key(XTOp void myxt_print_key(XTIndexPtr ind, xtWord1 *key_value); xtWord4 myxt_store_row_length(XTOpenTablePtr ot, char *rec_buff); +xtWord4 myxt_store_row_data(XTOpenTablePtr ot, xtWord4 row_size, char *rec_buff); xtBool myxt_store_row(XTOpenTablePtr ot, XTTabRecInfoPtr rec_info, char *rec_buff); size_t myxt_load_row_length(XTOpenTablePtr ot, size_t buffer_size, xtWord1 *source_buf, u_int *ret_col_cnt); +xtWord4 myxt_load_row_data(XTOpenTablePtr ot, xtWord1 *source_buf, xtWord1 *dest_buff, u_int col_cnt); xtBool myxt_load_row(XTOpenTablePtr ot, xtWord1 *source_buf, xtWord1 *dest_buff, u_int col_cnt); xtBool myxt_find_column(XTOpenTablePtr ot, u_int *col_idx, const char *col_name); void myxt_get_column_name(XTOpenTablePtr ot, u_int col_idx, u_int len, char *col_name); @@ -93,4 +95,6 @@ public: static XTDDColumn *createFromMySQLField(XTThread *self, STRUCT_TABLE *, Field *); }; +void myxt_wait_pbxt_plugin_slot_assigned(XTThread *self); + #endif === modified file 'storage/pbxt/src/pbms.h' --- a/storage/pbxt/src/pbms.h 2009-08-18 07:46:53 +0000 +++ b/storage/pbxt/src/pbms.h 2009-11-24 10:55:06 +0000 @@ -344,16 +344,16 @@ public: int couldBeURL(char *blob_url, int size) { if (blob_url && (size < PBMS_BLOB_URL_SIZE)) { - char buffer[PBMS_BLOB_URL_SIZE+1]; - u_int32_t db_id = 0; - u_int32_t tab_id = 0; - u_int64_t blob_id = 0; - u_int64_t blob_ref_id = 0; - u_int64_t blob_size = 0; - u_int32_t auth_code = 0; - u_int32_t server_id = 0; - char type, junk[5]; - int scanned; + char buffer[PBMS_BLOB_URL_SIZE+1]; + unsigned long db_id = 0; + unsigned long tab_id = 0; + unsigned long long blob_id = 0; + unsigned long long blob_ref_id = 0; + unsigned long long blob_size = 0; + unsigned long auth_code = 0; + unsigned long server_id = 0; + char type, junk[5]; + int scanned; junk[0] = 0; if (blob_url[size]) { // There is no guarantee that the URL will be null terminated. @@ -364,12 +364,12 @@ public: scanned = sscanf(blob_url, URL_FMT"%4s", &db_id, &type, &tab_id, &blob_id, &auth_code, &server_id, &blob_ref_id, &blob_size, junk); if (scanned != 8) {// If junk is found at the end this will also result in an invalid URL. - printf("Bad URL \"%s\": scanned = %d, junk: %d, %d, %d, %d\n", blob_url, scanned, junk[0], junk[1], junk[2], junk[3]); + printf("Bad URL \"%s\": scanned = %d, junk: %d, %d, %d, %d\n", blob_url, scanned, junk[0], junk[1], junk[2], junk[3]); return 0; } if (junk[0] || (type != '~' && type != '_')) { - printf("Bad URL \"%s\": scanned = %d, junk: %d, %d, %d, %d\n", blob_url, scanned, junk[0], junk[1], junk[2], junk[3]); + printf("Bad URL \"%s\": scanned = %d, junk: %d, %d, %d, %d\n", blob_url, scanned, junk[0], junk[1], junk[2], junk[3]); return 0; } === modified file 'storage/pbxt/src/pbms_enabled.cc' --- a/storage/pbxt/src/pbms_enabled.cc 2009-10-06 15:16:01 +0000 +++ b/storage/pbxt/src/pbms_enabled.cc 2009-11-24 10:55:06 +0000 @@ -29,15 +29,10 @@ * */ -/* - The following two lines backported by psergey. Remove them when we merge from PBXT again. -*/ #include "xt_config.h" -#ifdef PBMS_ENABLED -#define PBMS_API pbms_enabled_api +#ifdef PBMS_ENABLED -#include "pbms_enabled.h" #ifdef DRIZZLED #include <sys/stat.h> #include <drizzled/common_includes.h> @@ -47,11 +42,15 @@ #include <mysql/plugin.h> #define session_alloc(sess, size) thd_alloc(sess, size); #define current_session current_thd -#endif +#endif -#define GET_BLOB_FIELD(t, i) (Field_blob *)(t->field[t->s->blob_field[i]]) -#define DB_NAME(f) (f->table->s->db.str) -#define TAB_NAME(f) (*(f->table_name)) +#define GET_BLOB_FIELD(t, i) (Field_blob *)(t->field[t->s->blob_field[i]]) +#define DB_NAME(f) (f->table->s->db.str) +#define TAB_NAME(f) (*(f->table_name)) + +#define PBMS_API pbms_enabled_api + +#include "pbms_enabled.h" static PBMS_API pbms_api; @@ -242,4 +241,4 @@ void pbms_completed(TABLE *table, bool o return ; } -#endif \ No newline at end of file +#endif // PBMS_ENABLED === modified file 'storage/pbxt/src/pbms_enabled.h' --- a/storage/pbxt/src/pbms_enabled.h 2009-08-18 07:46:53 +0000 +++ b/storage/pbxt/src/pbms_enabled.h 2009-11-24 10:55:06 +0000 @@ -35,13 +35,6 @@ #include "pbms.h" -#ifdef DRIZZLED -#include <drizzled/server_includes.h> -#define TABLE Table -#else -#include <mysql_priv.h> -#endif - /* * pbms_initialize() should be called from the engines plugIn's 'init()' function. * The engine_name is the name of your engine, "PBXT" or "InnoDB" for example. === modified file 'storage/pbxt/src/pthread_xt.cc' --- a/storage/pbxt/src/pthread_xt.cc 2009-08-18 07:46:53 +0000 +++ b/storage/pbxt/src/pthread_xt.cc 2009-11-24 10:55:06 +0000 @@ -578,8 +578,8 @@ xtPublic int xt_p_mutex_unlock(xt_mutex_ xtPublic int xt_p_mutex_destroy(xt_mutex_type *mutex) { - ASSERT_NS(mutex->mu_init == 12345); - mutex->mu_init = 89898; + //ASSERT_NS(mutex->mu_init == 12345); + mutex->mu_init = 11111; #ifdef XT_THREAD_LOCK_INFO xt_thread_lock_info_free(&mutex->mu_lock_info); #endif === modified file 'storage/pbxt/src/restart_xt.cc' --- a/storage/pbxt/src/restart_xt.cc 2009-09-03 06:15:03 +0000 +++ b/storage/pbxt/src/restart_xt.cc 2009-11-27 15:37:02 +0000 @@ -34,12 +34,16 @@ #include "ha_pbxt.h" +#ifdef DRIZZLED +#include <drizzled/data_home.h> +using drizzled::plugin::Registry; +#endif + #include "xactlog_xt.h" #include "database_xt.h" #include "util_xt.h" #include "strutil_xt.h" #include "filesys_xt.h" -#include "restart_xt.h" #include "myxt_xt.h" #include "trace_xt.h" @@ -57,13 +61,27 @@ //#define PRINTF xt_ftracef //#define PRINTF xt_trace -void xt_print_bytes(xtWord1 *buf, u_int len) +/* + * ----------------------------------------------------------------------- + * GLOBALS + */ + +xtPublic int pbxt_recovery_state; + +/* + * ----------------------------------------------------------------------- + * UTILITIES + */ + +#ifdef TRACE_RECORD_DATA +static void xt_print_bytes(xtWord1 *buf, u_int len) { for (u_int i=0; i<len; i++) { PRINTF("%02x ", (u_int) *buf); buf++; } } +#endif void xt_print_log_record(xtLogID log, xtLogOffset offset, XTXactLogBufferDPtr record) { @@ -252,7 +270,7 @@ void xt_print_log_record(xtLogID log, xt rec_type = "DELETE"; break; case XT_LOG_ENT_DELETE_FL: - rec_type = "DELETE-FL-BG"; + rec_type = "DELETE-FL"; break; case XT_LOG_ENT_UPDATE_BG: rec_type = "UPDATE-BG"; @@ -320,6 +338,11 @@ void xt_print_log_record(xtLogID log, xt case XT_LOG_ENT_END_OF_LOG: rec_type = "END OF LOG"; break; + case XT_LOG_ENT_PREPARE: + rec_type = "PREPARE"; + xn_id = XT_GET_DISK_4(record->xp.xp_xact_id_4); + xn_set = TRUE; + break; } if (log) @@ -327,6 +350,8 @@ void xt_print_log_record(xtLogID log, xt PRINTF("%s ", rec_type); if (type) PRINTF("op=%lu tab=%lu %s=%lu ", (u_long) op_no, (u_long) tab_id, type, (u_long) rec_id); + else if (tab_id) + PRINTF("tab=%lu ", (u_long) tab_id); if (row_id) PRINTF("row=%lu ", (u_long) row_id); if (log_id) @@ -459,6 +484,7 @@ static xtBool xres_open_table(XTThreadPt } return OK; } + ws->ws_tab_gone = tab_id; return FAILED; } @@ -629,6 +655,7 @@ static void xres_apply_change(XTThreadPt xtWord1 *rec_data = NULL; XTTabRecFreeDPtr free_data; + ASSERT(ot->ot_thread == self); if (tab->tab_dic.dic_key_count == 0) check_index = FALSE; @@ -1300,7 +1327,7 @@ static void xres_apply_operations(XTThre * These operations are applied even though operations * in sequence are missing. */ -xtBool xres_sync_operations(XTThreadPtr self, XTDatabaseHPtr db, XTWriterStatePtr ws) +static xtBool xres_sync_operations(XTThreadPtr self, XTDatabaseHPtr db, XTWriterStatePtr ws) { u_int edx; XTTableEntryPtr te_ptr; @@ -1881,15 +1908,6 @@ xtBool XTXactRestart::xres_check_checksu void XTXactRestart::xres_recover_progress(XTThreadPtr self, XTOpenFilePtr *of, int perc) { #ifdef XT_USE_GLOBAL_DB - if (!perc) { - char file_path[PATH_MAX]; - - xt_strcpy(PATH_MAX, file_path, xres_db->db_main_path); - xt_add_pbxt_file(PATH_MAX, file_path, "recovery-progress"); - *of = xt_open_file(self, file_path, XT_FS_CREATE | XT_FS_MAKE_PATH); - xt_set_eof_file(self, *of, 0); - } - if (perc > 100) { char file_path[PATH_MAX]; @@ -1905,6 +1923,15 @@ void XTXactRestart::xres_recover_progres else { char number[40]; + if (!*of) { + char file_path[PATH_MAX]; + + xt_strcpy(PATH_MAX, file_path, xres_db->db_main_path); + xt_add_pbxt_file(PATH_MAX, file_path, "recovery-progress"); + *of = xt_open_file(self, file_path, XT_FS_CREATE | XT_FS_MAKE_PATH); + xt_set_eof_file(self, *of, 0); + } + sprintf(number, "%d", perc); if (!xt_pwrite_file(*of, 0, strlen(number), number, &self->st_statistics.st_x, self)) xt_throw(self); @@ -1927,10 +1954,11 @@ xtBool XTXactRestart::xres_restart(XTThr off_t bytes_to_read; volatile xtBool print_progress = FALSE; volatile off_t perc_size = 0, next_goal = 0; - int perc_complete = 1; + int perc_complete = 1, perc_to_write = 1; XTOpenFilePtr progress_file = NULL; xtBool min_ram_xn_id_set = FALSE; u_int log_count; + time_t start_time; memset(&ws, 0, sizeof(ws)); @@ -1955,12 +1983,11 @@ xtBool XTXactRestart::xres_restart(XTThr /* Don't print anything about recovering an empty database: */ if (bytes_to_read != 0) xt_logf(XT_NT_INFO, "PBXT: Recovering from %lu-%llu, bytes to read: %llu\n", (u_long) xres_cp_log_id, (u_llong) xres_cp_log_offset, (u_llong) bytes_to_read); - if (bytes_to_read >= 10*1024*1024) { - print_progress = TRUE; - perc_size = bytes_to_read / 100; - next_goal = perc_size; - xres_recover_progress(self, &progress_file, 0); - } + + print_progress = FALSE; + start_time = time(NULL); + perc_size = bytes_to_read / 100; + next_goal = perc_size; if (!db->db_xlog.xlog_seq_start(&ws.ws_seqread, xres_cp_log_id, xres_cp_log_offset, FALSE)) { ok = FALSE; @@ -1983,17 +2010,28 @@ xtBool XTXactRestart::xres_restart(XTThr #ifdef PRINT_LOG_ON_RECOVERY xt_print_log_record(ws.ws_seqread.xseq_rec_log_id, ws.ws_seqread.xseq_rec_log_offset, record); #endif - if (print_progress && bytes_read > next_goal) { - if (((perc_complete - 1) % 25) == 0) - xt_logf(XT_NT_INFO, "PBXT: "); - if ((perc_complete % 25) == 0) - xt_logf(XT_NT_INFO, "%2d\n", (int) perc_complete); - else - xt_logf(XT_NT_INFO, "%2d ", (int) perc_complete); - xt_log_flush(self); - xres_recover_progress(self, &progress_file, perc_complete); - next_goal += perc_size; - perc_complete++; + if (bytes_read >= next_goal) { + while (bytes_read >= next_goal) { + next_goal += perc_size; + perc_complete++; + } + if (!print_progress) { + if (time(NULL) - start_time > 2) + print_progress = TRUE; + } + if (print_progress) { + while (perc_to_write < perc_complete) { + if (((perc_to_write - 1) % 25) == 0) + xt_logf(XT_NT_INFO, "PBXT: "); + if ((perc_to_write % 25) == 0) + xt_logf(XT_NT_INFO, "%2d\n", (int) perc_to_write); + else + xt_logf(XT_NT_INFO, "%2d ", (int) perc_to_write); + xt_log_flush(self); + xres_recover_progress(self, &progress_file, perc_to_write); + perc_to_write++; + } + } } switch (record->xl.xl_status_1) { case XT_LOG_ENT_HEADER: @@ -2053,8 +2091,11 @@ xtBool XTXactRestart::xres_restart(XTThr xact->xd_end_xn_id = xn_id; xact->xd_flags |= XT_XN_XAC_ENDED | XT_XN_XAC_SWEEP; xact->xd_flags &= ~XT_XN_XAC_RECOVERED; // We can expect an end record on cleanup! + xact->xd_flags &= ~XT_XN_XAC_PREPARED; // Prepared transactions cannot be swept! if (record->xl.xl_status_1 == XT_LOG_ENT_COMMIT) xact->xd_flags |= XT_XN_XAC_COMMITTED; + if (xt_sl_get_size(db->db_xn_xa_list) > 0) + xt_xn_delete_xa_data_by_xact(db, xn_id, self); } break; case XT_LOG_ENT_CLEANUP: @@ -2071,6 +2112,14 @@ xtBool XTXactRestart::xres_restart(XTThr rec_log_id = XT_GET_DISK_4(record->xl.xl_log_id_4); xt_dl_set_to_delete(self, db, rec_log_id); break; + case XT_LOG_ENT_PREPARE: + xn_id = XT_GET_DISK_4(record->xp.xp_xact_id_4); + if ((xact = xt_xn_get_xact(db, xn_id, self))) { + xact->xd_flags |= XT_XN_XAC_PREPARED; + if (!xt_xn_store_xa_data(db, xn_id, record->xp.xp_xa_len_1, record->xp.xp_xa_data, self)) + xt_throw(self); + } + break; default: xt_xres_apply_in_order(self, &ws, ws.ws_seqread.xseq_rec_log_id, ws.ws_seqread.xseq_rec_log_offset, record); break; @@ -2534,7 +2583,8 @@ static void xres_cp_main(XTThreadPtr sel /* This condition means we could checkpoint: */ if (!(xt_sl_get_size(db->db_datalogs.dlc_to_delete) == 0 && xt_sl_get_size(db->db_datalogs.dlc_deleted) == 0 && - xt_comp_log_pos(log_id, log_offset, db->db_restart.xres_cp_log_id, db->db_restart.xres_cp_log_offset) <= 0)) + xt_comp_log_pos(log_id, log_offset, db->db_restart.xres_cp_log_id, db->db_restart.xres_cp_log_offset) <= 0) && + xt_sl_get_size(db->db_xn_xa_list) == 0) break; xres_cp_wait_for_log_writer(self, db, 400); @@ -2654,7 +2704,7 @@ xtPublic xtBool xt_begin_checkpoint(XTDa * until they are flushed. */ /* This is an alternative to the above. - if (!xt_xlog_flush_log(self)) + if (!xt_xlog_flush_log(db, self)) xt_throw(self); */ xt_lock_mutex_ns(&db->db_wr_lock); @@ -2776,6 +2826,14 @@ xtPublic xtBool xt_end_checkpoint(XTData size_t chk_size = 0; u_int no_of_logs = 0; + /* As long as we have outstanding XA transactions, we may not checkpoint! */ + if (xt_sl_get_size(db->db_xn_xa_list) > 0) { +#ifdef DEBUG + printf("Checkpoint must wait\n"); +#endif + return OK; + } + #ifdef NEVER_CHECKPOINT return OK; #endif @@ -3183,7 +3241,7 @@ xtPublic void xt_dump_xlogs(XTDatabaseHP * D A T A B A S E R E C O V E R Y T H R E A D */ -extern XTDatabaseHPtr pbxt_database; + static XTThreadPtr xres_recovery_thread; static void *xn_xres_run_recovery_thread(XTThreadPtr self) @@ -3193,18 +3251,18 @@ static void *xn_xres_run_recovery_thread if (!(mysql_thread = (THD *) myxt_create_thread())) xt_throw(self); - while (!xres_recovery_thread->t_quit && !ha_resolve_by_legacy_type(mysql_thread, DB_TYPE_PBXT)) - xt_sleep_milli_second(1); + myxt_wait_pbxt_plugin_slot_assigned(self); if (!xres_recovery_thread->t_quit) { - /* {GLOBAL-DB} - * It can happen that something will just get in before this - * thread and open/recover the database! - */ - if (!pbxt_database) { - try_(a) { + try_(a) { + /* {GLOBAL-DB} + * It can happen that something will just get in before this + * thread and open/recover the database! + */ + if (!pbxt_database) { xt_open_database(self, mysql_real_data_home, TRUE); - /* This can be done at the same time by a foreground thread, + /* {GLOBAL-DB} + * This can be done at the same time as the recovery thread, * strictly speaking I need a lock. */ if (!pbxt_database) { @@ -3212,11 +3270,22 @@ static void *xn_xres_run_recovery_thread xt_heap_reference(self, pbxt_database); } } - catch_(a) { - xt_log_and_clear_exception(self); - } - cont_(a); + else + xt_use_database(self, pbxt_database, XT_FOR_USER); + + pbxt_recovery_state = XT_RECOVER_DONE; + + /* {WAIT-FOR-SW-AFTER-RECOV} + * Moved to here... + */ + xt_wait_for_sweeper(self, self->st_database, 0); + + pbxt_recovery_state = XT_RECOVER_SWEPT; } + catch_(a) { + xt_log_and_clear_exception(self); + } + cont_(a); } /* @@ -3261,11 +3330,12 @@ xtPublic void xt_xres_start_database_rec sprintf(name, "DB-RECOVERY-%s", xt_last_directory_of_path(mysql_real_data_home)); xt_remove_dir_char(name); + pbxt_recovery_state = XT_RECOVER_PENDING; xres_recovery_thread = xt_create_daemon(self, name); xt_run_thread(self, xres_recovery_thread, xn_xres_run_recovery_thread); } -xtPublic void xt_xres_wait_for_recovery(XTThreadPtr self) +xtPublic void xt_xres_terminate_recovery(XTThreadPtr self) { XTThreadPtr thr_rec; === modified file 'storage/pbxt/src/restart_xt.h' --- a/storage/pbxt/src/restart_xt.h 2009-09-03 06:15:03 +0000 +++ b/storage/pbxt/src/restart_xt.h 2009-11-24 10:55:06 +0000 @@ -37,6 +37,8 @@ struct XTOpenTable; struct XTDatabase; struct XTTable; +extern int pbxt_recovery_state; + typedef struct XTWriterState { struct XTDatabase *ws_db; xtBool ws_in_recover; @@ -132,6 +134,16 @@ void xt_print_log_record(xtLogID log, of void xt_dump_xlogs(struct XTDatabase *db, xtLogID start_log); void xt_xres_start_database_recovery(XTThreadPtr self); -void xt_xres_wait_for_recovery(XTThreadPtr self); +void xt_xres_terminate_recovery(XTThreadPtr self); + +#define XT_RECOVER_PENDING 0 +#define XT_RECOVER_DONE 1 +#define XT_RECOVER_SWEPT 2 + +inline void xt_xres_wait_for_recovery(XTThreadPtr XT_UNUSED(self), int state) +{ + while (pbxt_recovery_state < state) + xt_sleep_milli_second(100); +} #endif === modified file 'storage/pbxt/src/strutil_xt.cc' --- a/storage/pbxt/src/strutil_xt.cc 2009-10-26 11:35:42 +0000 +++ b/storage/pbxt/src/strutil_xt.cc 2009-11-24 10:55:06 +0000 @@ -21,8 +21,10 @@ * H&G2JCtL */ -#include "mysql_priv.h" #include "xt_config.h" + +#include <stdio.h> +#include <string.h> #include <ctype.h> #include "strutil_xt.h" @@ -107,13 +109,17 @@ xtPublic void xt_2nd_last_name_of_path(s *dest = 0; return; } - /* If temporary file */ - if (!is_prefix(path, mysql_data_home) && + + /* {INVALID-OLD-TABLE-FIX} + * I have changed the implementation of + * this bug fix (see {INVALID-OLD-TABLE-FIX}). + if (!is_prefix(path, mysql_data_home) && !is_prefix(path, mysql_real_data_home)) { *dest= 0; return; } + */ ptr = path + len - 1; while (ptr != path && !XT_IS_DIR_CHAR(*ptr)) @@ -374,7 +380,7 @@ xtPublic void xt_int8_to_byte_size(xtInt /* Version number must also be set in configure.in! */ xtPublic c_char *xt_get_version(void) { - return "1.0.08d RC"; + return "1.0.09f RC"; } /* Copy and URL decode! */ === modified file 'storage/pbxt/src/systab_xt.cc' --- a/storage/pbxt/src/systab_xt.cc 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/systab_xt.cc 2009-11-24 10:55:06 +0000 @@ -130,7 +130,7 @@ static int pbms_discover_handler(handler * MYSQL UTILITIES */ -void xt_my_set_notnull_in_record(Field *field, char *record) +static void xt_my_set_notnull_in_record(Field *field, char *record) { if (field->null_ptr) record[(uint) (field->null_ptr - (uchar *) field->table->record[0])] &= (uchar) ~field->null_bit; @@ -518,7 +518,7 @@ bool XTStatisticsTable::seqScanRead(xtWo * SYSTEM TABLE SHARES */ -void st_path_to_table_name(size_t size, char *buffer, const char *path) +static void st_path_to_table_name(size_t size, char *buffer, const char *path) { char *str; === modified file 'storage/pbxt/src/tabcache_xt.cc' --- a/storage/pbxt/src/tabcache_xt.cc 2009-09-03 06:15:03 +0000 +++ b/storage/pbxt/src/tabcache_xt.cc 2009-11-27 15:37:02 +0000 @@ -590,7 +590,7 @@ xtBool XTTabCache::tc_fetch(XT_ROW_REC_F * So there could be a deadlock if I don't flush the log! */ if ((self = xt_get_self())) { - if (!xt_xlog_flush_log(self)) + if (!xt_xlog_flush_log(tci_table->tab_db, self)) goto failed; } @@ -1150,6 +1150,8 @@ static void *tabc_fr_run_thread(XTThread int count; void *mysql_thread; + myxt_wait_pbxt_plugin_slot_assigned(self); + mysql_thread = myxt_create_thread(); while (!self->t_quit) { === modified file 'storage/pbxt/src/tabcache_xt.h' --- a/storage/pbxt/src/tabcache_xt.h 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/tabcache_xt.h 2009-11-24 10:55:06 +0000 @@ -168,7 +168,7 @@ typedef struct XTTableSeq { #define TAB_CAC_UNLOCK(i, o) xt_xsmutex_unlock(i, o) #elif defined(TAB_CAC_USE_PTHREAD_RW) #define TAB_CAC_LOCK_TYPE xt_rwlock_type -#define TAB_CAC_INIT_LOCK(s, i) xt_init_rwlock(s, i) +#define TAB_CAC_INIT_LOCK(s, i) xt_init_rwlock_with_autoname(s, i) #define TAB_CAC_FREE_LOCK(s, i) xt_free_rwlock(i) #define TAB_CAC_READ_LOCK(i, o) xt_slock_rwlock_ns(i) #define TAB_CAC_WRITE_LOCK(i, o) xt_xlock_rwlock_ns(i) === modified file 'storage/pbxt/src/table_xt.cc' --- a/storage/pbxt/src/table_xt.cc 2009-08-18 07:46:53 +0000 +++ b/storage/pbxt/src/table_xt.cc 2009-11-25 15:40:51 +0000 @@ -35,7 +35,6 @@ #include <drizzled/common.h> #include <mysys/thr_lock.h> #include <drizzled/dtcollation.h> -#include <drizzled/plugin/storage_engine.h> #else #include "mysql_priv.h" #endif @@ -48,7 +47,6 @@ #include "cache_xt.h" #include "trace_xt.h" #include "index_xt.h" -#include "restart_xt.h" #include "systab_xt.h" #ifdef DEBUG @@ -2347,7 +2345,7 @@ xtPublic void xt_flush_table(XTThreadPtr } -xtPublic XTOpenTablePtr tab_open_table(XTTableHPtr tab) +static XTOpenTablePtr tab_open_table(XTTableHPtr tab) { volatile XTOpenTablePtr ot; XTThreadPtr self; @@ -2588,7 +2586,7 @@ xtPublic xtBool xt_tab_put_log_op_rec_da return FAILED; } - return xt_xlog_modify_table(ot, status, op_seq, free_rec_id, rec_id, size, buffer); + return xt_xlog_modify_table(tab->tab_id, status, op_seq, free_rec_id, rec_id, size, buffer, ot->ot_thread); } xtPublic xtBool xt_tab_put_log_rec_data(XTOpenTablePtr ot, u_int status, xtRecordID free_rec_id, xtRecordID rec_id, size_t size, xtWord1 *buffer, xtOpSeqNo *op_seq) @@ -2606,7 +2604,7 @@ xtPublic xtBool xt_tab_put_log_rec_data( return FAILED; } - return xt_xlog_modify_table(ot, status, *op_seq, free_rec_id, rec_id, size, buffer); + return xt_xlog_modify_table(tab->tab_id, status, *op_seq, free_rec_id, rec_id, size, buffer, ot->ot_thread); } xtPublic xtBool xt_tab_get_rec_data(XTOpenTablePtr ot, xtRecordID rec_id, size_t size, xtWord1 *buffer) @@ -3541,7 +3539,7 @@ xtPublic xtBool xt_tab_free_row(XTOpenTa tab->tab_row_fnum++; xt_unlock_mutex_ns(&tab->tab_row_lock); - if (!xt_xlog_modify_table(ot, XT_LOG_ENT_ROW_FREED, op_seq, 0, row_id, sizeof(XTTabRowRefDRec), (xtWord1 *) &free_row)) + if (!xt_xlog_modify_table(tab->tab_id, XT_LOG_ENT_ROW_FREED, op_seq, 0, row_id, sizeof(XTTabRowRefDRec), (xtWord1 *) &free_row, ot->ot_thread)) return FAILED; return OK; @@ -3791,7 +3789,7 @@ xtPublic int xt_tab_remove_record(XTOpen xt_unlock_mutex_ns(&tab->tab_rec_lock); free_rec->rf_rec_type_1 = old_rec_type; - return xt_xlog_modify_table(ot, XT_LOG_ENT_REC_REMOVED_BI, op_seq, (xtRecordID) new_rec_type, rec_id, rec_size, ot->ot_row_rbuffer); + return xt_xlog_modify_table(tab->tab_id, XT_LOG_ENT_REC_REMOVED_BI, op_seq, (xtRecordID) new_rec_type, rec_id, rec_size, ot->ot_row_rbuffer, ot->ot_thread); } static xtRowID tab_new_row(XTOpenTablePtr ot, XTTableHPtr tab) @@ -3837,7 +3835,7 @@ static xtRowID tab_new_row(XTOpenTablePt op_seq = tab->tab_seq.ts_get_op_seq(); xt_unlock_mutex_ns(&tab->tab_row_lock); - if (!xt_xlog_modify_table(ot, status, op_seq, next_row_id, row_id, 0, NULL)) + if (!xt_xlog_modify_table(tab->tab_id, status, op_seq, next_row_id, row_id, 0, NULL, ot->ot_thread)) return 0; XT_DISABLED_TRACE(("new row tx=%d row=%d\n", (int) ot->ot_thread->st_xact_data->xd_start_xn_id, (int) row_id)); @@ -3868,7 +3866,7 @@ xtPublic xtBool xt_tab_set_row(XTOpenTab if (!tab->tab_rows.xt_tc_write(ot->ot_row_file, row_id, 0, sizeof(XTTabRowRefDRec), (xtWord1 *) &row_buf, &op_seq, TRUE, ot->ot_thread)) return FAILED; - return xt_xlog_modify_table(ot, status, op_seq, 0, row_id, sizeof(XTTabRowRefDRec), (xtWord1 *) &row_buf); + return xt_xlog_modify_table(tab->tab_id, status, op_seq, 0, row_id, sizeof(XTTabRowRefDRec), (xtWord1 *) &row_buf, ot->ot_thread); } xtPublic xtBool xt_tab_free_record(XTOpenTablePtr ot, u_int status, xtRecordID rec_id, xtBool clean_delete) @@ -3937,7 +3935,7 @@ xtPublic xtBool xt_tab_free_record(XTOpe tab->tab_rec_fnum++; xt_unlock_mutex_ns(&tab->tab_rec_lock); - if (!xt_xlog_modify_table(ot, status, op_seq, rec_id, rec_id, sizeof(XTactFreeRecEntryDRec) - offsetof(XTactFreeRecEntryDRec, fr_stat_id_1), &free_rec.fr_stat_id_1)) + if (!xt_xlog_modify_table(tab->tab_id, status, op_seq, rec_id, rec_id, sizeof(XTactFreeRecEntryDRec) - offsetof(XTactFreeRecEntryDRec, fr_stat_id_1), &free_rec.fr_stat_id_1, ot->ot_thread)) return FAILED; } return OK; @@ -4016,7 +4014,7 @@ static xtBool tab_add_record(XTOpenTable } xt_unlock_mutex_ns(&tab->tab_rec_lock); - if (!xt_xlog_modify_table(ot, status, op_seq, next_rec_id, rec_id, rec_info->ri_rec_buf_size, (xtWord1 *) rec_info->ri_fix_rec_buf)) + if (!xt_xlog_modify_table(tab->tab_id, status, op_seq, next_rec_id, rec_id, rec_info->ri_rec_buf_size, (xtWord1 *) rec_info->ri_fix_rec_buf, ot->ot_thread)) return FAILED; if (rec_info->ri_ext_rec) { @@ -4932,6 +4930,12 @@ xtPublic void xt_tab_seq_exit(XTOpenTabl #endif #endif +xtPublic void xt_tab_seq_repeat(XTOpenTablePtr ot) +{ + ot->ot_seq_rec_id--; + ot->ot_seq_offset -= ot->ot_table->tab_dic.dic_rec_size; +} + xtPublic xtBool xt_tab_seq_next(XTOpenTablePtr ot, xtWord1 *buffer, xtBool *eof) { register XTTableHPtr tab = ot->ot_table; @@ -5094,7 +5098,7 @@ static xtBool tab_exec_repair_pending(XT return FALSE; } else { - if (!xt_open_file_ns(&of, file_path, XT_FS_DEFAULT)) + if (!xt_open_file_ns(&of, file_path, XT_FS_DEFAULT | XT_FS_MISSING_OK)) return FALSE; } if (!of) @@ -5190,15 +5194,76 @@ static xtBool tab_exec_repair_pending(XT return FALSE; } -xtPublic void tab_make_table_name(XTTableHPtr tab, char *table_name, size_t size) +static void tab_make_table_name(XTTableHPtr tab, char *table_name, size_t size) { - char name_buf[XT_IDENTIFIER_NAME_SIZE*3+3]; + char *nptr; - xt_2nd_last_name_of_path(sizeof(name_buf), name_buf, tab->tab_name->ps_path); - myxt_static_convert_file_name(name_buf, table_name, size); - xt_strcat(size, table_name, "."); - myxt_static_convert_file_name(xt_last_name_of_path(tab->tab_name->ps_path), name_buf, sizeof(name_buf)); - xt_strcat(size, table_name, name_buf); + nptr = xt_last_name_of_path(tab->tab_name->ps_path); + if (xt_starts_with(nptr, "#sql")) { + /* {INVALID-OLD-TABLE-FIX} + * Temporary files can have strange paths, for example + * ..../var/tmp/mysqld.1/#sqldaec_1_6 + * This occurs, for example, occurs when the temp_table.test is + * run using the PBXT suite in MariaDB: + * ./mtr --suite=pbxt --do-test=temp_table + * + * Calling myxt_static_convert_file_name, with a '.', in the name + * causes the error: + * [ERROR] Invalid (old?) table or database name 'mysqld.1' + * To prevent this, we do not convert the temporary + * table names using the mysql functions. + * + * Note, this bug was found by Monty, and fixed by modifying + * xt_2nd_last_name_of_path(), see {INVALID-OLD-TABLE-FIX}. + * + */ + xt_2nd_last_name_of_path(size, table_name, tab->tab_name->ps_path); + xt_strcat(size, table_name, "."); + xt_strcat(size, table_name, nptr); + } + else { + char name_buf[XT_TABLE_NAME_SIZE*3+3]; + char *part_ptr; + size_t len; + + xt_2nd_last_name_of_path(sizeof(name_buf), name_buf, tab->tab_name->ps_path); + myxt_static_convert_file_name(name_buf, table_name, size); + xt_strcat(size, table_name, "."); + + /* Handle partition extensions to table names: */ + if ((part_ptr = strstr(nptr, "#P#"))) + xt_strncpy(sizeof(name_buf), name_buf, nptr, part_ptr - nptr); + else + xt_strcpy(sizeof(name_buf), name_buf, nptr); + + len = strlen(table_name); + myxt_static_convert_file_name(name_buf, table_name + len, size - len); + + if (part_ptr) { + /* Add the partition extension (which is relevant to the engine). */ + char *sub_part_ptr; + + part_ptr += 3; + if ((sub_part_ptr = strstr(part_ptr, "#SP#"))) + xt_strncpy(sizeof(name_buf), name_buf, part_ptr, sub_part_ptr - part_ptr); + else + xt_strcpy(sizeof(name_buf), name_buf, part_ptr); + + xt_strcat(size, table_name, " ("); + len = strlen(table_name); + myxt_static_convert_file_name(name_buf, table_name + len, size - len); + + if (sub_part_ptr) { + + sub_part_ptr += 4; + xt_strcat(size, table_name, " - "); + len = strlen(table_name); + myxt_static_convert_file_name(sub_part_ptr, table_name + len, size - len); + } + + xt_strcat(size, table_name, ")"); + } + } } xtPublic xtBool xt_tab_is_table_repair_pending(XTTableHPtr tab) === modified file 'storage/pbxt/src/table_xt.h' --- a/storage/pbxt/src/table_xt.h 2009-08-18 07:46:53 +0000 +++ b/storage/pbxt/src/table_xt.h 2009-11-24 10:55:06 +0000 @@ -127,7 +127,7 @@ struct XTTablePath; #define XT_TAB_ROW_UNLOCK(i, s) xt_xsmutex_unlock(i, (s)->t_id) #elif defined(XT_TAB_ROW_USE_PTHREAD_RW) #define XT_TAB_ROW_LOCK_TYPE xt_rwlock_type -#define XT_TAB_ROW_INIT_LOCK(s, i) xt_init_rwlock(s, i) +#define XT_TAB_ROW_INIT_LOCK(s, i) xt_init_rwlock_with_autoname(s, i) #define XT_TAB_ROW_FREE_LOCK(s, i) xt_free_rwlock(i) #define XT_TAB_ROW_READ_LOCK(i, s) xt_slock_rwlock_ns(i) #define XT_TAB_ROW_WRITE_LOCK(i, s) xt_xlock_rwlock_ns(i) @@ -528,13 +528,14 @@ xtBool xt_table_exists(struct XTDatab void xt_enum_tables_init(u_int *edx); XTTableEntryPtr xt_enum_tables_next(struct XTThread *self, struct XTDatabase *db, u_int *edx); -void xt_enum_files_of_tables_init(struct XTDatabase *db, char *tab_name, xtTableID tab_id, XTFilesOfTablePtr ft); +void xt_enum_files_of_tables_init(XTPathStrPtr tab_name, xtTableID tab_id, XTFilesOfTablePtr ft); xtBool xt_enum_files_of_tables_next(XTFilesOfTablePtr ft); xtBool xt_tab_seq_init(XTOpenTablePtr ot); void xt_tab_seq_reset(XTOpenTablePtr ot); void xt_tab_seq_exit(XTOpenTablePtr ot); xtBool xt_tab_seq_next(XTOpenTablePtr ot, xtWord1 *buffer, xtBool *eof); +void xt_tab_seq_repeat(XTOpenTablePtr ot); xtBool xt_tab_new_record(XTOpenTablePtr ot, xtWord1 *buffer); xtBool xt_tab_delete_record(XTOpenTablePtr ot, xtWord1 *buffer); === modified file 'storage/pbxt/src/thread_xt.cc' --- a/storage/pbxt/src/thread_xt.cc 2009-10-06 15:16:01 +0000 +++ b/storage/pbxt/src/thread_xt.cc 2009-11-24 10:55:06 +0000 @@ -75,6 +75,13 @@ static xt_mutex_type thr_array_lock; /* Global accumulated statistics: */ static XTStatisticsRec thr_statistics; +#ifdef DEBUG +static void break_in_assertion(c_char *expr, c_char *func, c_char *file, u_int line) +{ + printf("%s(%s:%d) %s\n", func, file, (int) line, expr); +} +#endif + /* * ----------------------------------------------------------------------- * Error logging @@ -658,6 +665,9 @@ static c_char *thr_get_err_string(int xt case XT_ERR_FK_REF_TEMP_TABLE: str = "Foreign key may not reference temporary table"; break; case XT_ERR_MYSQL_SHUTDOWN: str = "Cannot open table, MySQL has shutdown"; break; case XT_ERR_MYSQL_NO_THREAD: str = "Cannot create thread, MySQL has shutdown"; break; + case XT_ERR_BUFFER_TOO_SMALL: str = "System backup buffer too small"; break; + case XT_ERR_BAD_BACKUP_FORMAT: str = "Unknown or corrupt backup format, restore aborted"; break; + case XT_ERR_PBXT_NOT_INSTALLED: str = "PBXT plugin is not installed"; break; default: str = "Unknown XT error"; break; } return str; @@ -862,6 +872,11 @@ xtPublic xtBool xt_exception_errno(XTExc return FAILED; } +xtPublic void xt_exception_xterr(XTExceptionPtr e, XTThreadPtr self, c_char *func, c_char *file, u_int line, int xt_err) +{ + xt_exception_error(e, self, func, file, line, xt_err, 0, thr_get_err_string(xt_err)); +} + /* * ----------------------------------------------------------------------- * LOG ERRORS @@ -887,7 +902,7 @@ xtPublic xtBool xt_assert(XTThreadPtr se #ifdef DEBUG //xt_set_fflush(TRUE); //xt_dump_trace(); - printf("%s(%s:%d) %s\n", func, file, (int) line, expr); + break_in_assertion(expr, func, file, line); #ifdef CRASH_ON_ASSERT abort(); #endif === modified file 'storage/pbxt/src/thread_xt.h' --- a/storage/pbxt/src/thread_xt.h 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/thread_xt.h 2009-11-24 10:55:06 +0000 @@ -536,6 +536,8 @@ extern struct XTThread **xt_thr_array; * Function prototypes */ +extern "C" void *thr_main(void *data); + void xt_get_now(char *buffer, size_t len); xtBool xt_init_logging(void); void xt_exit_logging(void); @@ -583,6 +585,7 @@ void xt_register_xterr(c_char *func, c void xt_exceptionf(XTExceptionPtr e, XTThreadPtr self, c_char *func, c_char *file, u_int line, int xt_err, int sys_err, c_char *fmt, ...); void xt_exception_error(XTExceptionPtr e, XTThreadPtr self, c_char *func, c_char *file, u_int line, int xt_err, int sys_err, c_char *msg); xtBool xt_exception_errno(XTExceptionPtr e, XTThreadPtr self, c_char *func, c_char *file, u_int line, int err); +void xt_exception_xterr(XTExceptionPtr e, XTThreadPtr self, c_char *func, c_char *file, u_int line, int xt_err); void xt_log_errno(XTThreadPtr self, c_char *func, c_char *file, u_int line, int err); @@ -610,7 +613,7 @@ void xt_critical_wait(void); void xt_yield(void); void xt_sleep_milli_second(u_int t); xtBool xt_suspend(XTThreadPtr self); -xtBool xt_unsuspend(XTThreadPtr self, XTThreadPtr target); +xtBool xt_unsuspend(XTThreadPtr target); void xt_lock_thread(XTThreadPtr thread); void xt_unlock_thread(XTThreadPtr thread); xtBool xt_wait_thread(XTThreadPtr thread); === modified file 'storage/pbxt/src/util_xt.cc' --- a/storage/pbxt/src/util_xt.cc 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/util_xt.cc 2009-11-24 10:55:06 +0000 @@ -150,6 +150,23 @@ xtPublic xtWord1 xt_get_checksum1(xtWord return (xtWord1) (sum ^ (sum >> 24) ^ (sum >> 16) ^ (sum >> 8)); } +xtPublic xtWord4 xt_get_checksum4(xtWord1 *data, size_t len) +{ + register xtWord4 sum = 0, g; + xtWord1 *chk; + + chk = data + len - 1; + while (chk > data) { + sum = (sum << 4) + *chk; + if ((g = sum & 0xF0000000)) { + sum = sum ^ (g >> 24); + sum = sum ^ g; + } + chk--; + } + return sum; +} + /* * --------------- Data Buffer ------------------ */ === modified file 'storage/pbxt/src/util_xt.h' --- a/storage/pbxt/src/util_xt.h 2009-03-26 12:18:01 +0000 +++ b/storage/pbxt/src/util_xt.h 2009-11-24 10:55:06 +0000 @@ -39,6 +39,7 @@ xtWord4 xt_file_name_to_id(char *file_na xtBool xt_time_difference(register xtWord4 now, register xtWord4 then); xtWord2 xt_get_checksum(xtWord1 *data, size_t len, u_int interval); xtWord1 xt_get_checksum1(xtWord1 *data, size_t len); +xtWord4 xt_get_checksum4(xtWord1 *data, size_t len); typedef struct XTDataBuffer { size_t db_size; === modified file 'storage/pbxt/src/xaction_xt.cc' --- a/storage/pbxt/src/xaction_xt.cc 2009-09-03 06:15:03 +0000 +++ b/storage/pbxt/src/xaction_xt.cc 2009-11-24 10:55:06 +0000 @@ -1075,6 +1075,7 @@ xtPublic void xt_xn_init_db(XTThreadPtr #endif xt_spinlock_init_with_autoname(self, &db->db_xn_id_lock); xt_spinlock_init_with_autoname(self, &db->db_xn_wait_spinlock); + xt_init_mutex_with_autoname(self, &db->db_xn_xa_lock); //xt_init_mutex_with_autoname(self, &db->db_xn_wait_lock); //xt_init_cond(self, &db->db_xn_wait_cond); xt_init_mutex_with_autoname(self, &db->db_sw_lock); @@ -1096,6 +1097,9 @@ xtPublic void xt_xn_init_db(XTThreadPtr } } + /* Create a sorted list for XA transactions recovered: */ + db->db_xn_xa_list = xt_new_sortedlist(self, sizeof(XTXactXARec), 100, 50, xt_xn_xa_compare, db, NULL, FALSE, FALSE); + /* Initialize the data logs: */ db->db_datalogs.dlc_init(self, db); @@ -1146,6 +1150,7 @@ xtPublic void xt_xn_exit_db(XTThreadPtr printf("=========> MAX TXs NOT CLEAN: %lu\n", not_clean_max); printf("=========> MAX TXs IN RAM: %lu\n", in_ram_max); #endif + XTXactPreparePtr xap, xap_next; xt_stop_sweeper(self, db); // Should be done already! xt_stop_writer(self, db); // Should be done already! @@ -1187,6 +1192,19 @@ xtPublic void xt_xn_exit_db(XTThreadPtr xt_free_mutex(&db->db_sw_lock); //xt_free_cond(&db->db_xn_wait_cond); //xt_free_mutex(&db->db_xn_wait_lock); + xt_free_mutex(&db->db_xn_xa_lock); + for (u_int i=0; i<XT_XA_HASH_TAB_SIZE; i++) { + xap = db->db_xn_xa_table[i]; + while (xap) { + xap_next = xap->xp_next; + xt_free(self, xap); + xap = xap_next; + } + } + if (db->db_xn_xa_list) { + xt_free_sortedlist(self, db->db_xn_xa_list); + db->db_xn_xa_list = NULL; + } xt_spinlock_free(self, &db->db_xn_wait_spinlock); xt_spinlock_free(self, &db->db_xn_id_lock); #ifdef DEBUG_RAM_LIST @@ -1428,7 +1446,7 @@ static xtBool xn_end_xact(XTThreadPtr th wait_xn_id = thread->st_prev_xact[thread->st_last_xact]; thread->st_prev_xact[thread->st_last_xact] = xn_id; /* This works because XT_MAX_XACT_BEHIND == 2! */ - ASSERT_NS((thread->st_last_xact + 1) % XT_MAX_XACT_BEHIND == thread->st_last_xact ^ 1); + ASSERT_NS((thread->st_last_xact + 1) % XT_MAX_XACT_BEHIND == (thread->st_last_xact ^ 1)); thread->st_last_xact ^= 1; while (xt_xn_is_before(db->db_xn_to_clean_id, wait_xn_id) && (db->db_sw_faster & XT_SW_TOO_FAR_BEHIND)) { xt_critical_wait(); @@ -1548,6 +1566,149 @@ xtPublic int xt_xn_status(XTOpenTablePtr return XT_XN_ABORTED; } +/* ---------------------------------------------------------------------- + * XA Functionality + */ + +xtPublic int xt_xn_xa_compare(XTThreadPtr XT_UNUSED(self), register const void *XT_UNUSED(thunk), register const void *a, register const void *b) +{ + xtXactID *x = (xtXactID *) a; + XTXactXAPtr y = (XTXactXAPtr) b; + + if (*x == y->xx_xact_id) + return 0; + if (xt_xn_is_before(*x, y->xx_xact_id)) + return -1; + return 1; +} + +xtPublic xtBool xt_xn_prepare(int len, xtWord1 *xa_data, XTThreadPtr thread) +{ + XTXactDataPtr xact; + + ASSERT_NS(thread->st_xact_data); + if ((xact = thread->st_xact_data)) { + xtXactID xn_id = xact->xd_start_xn_id; + + /* Only makes sense if the transaction has already been logged: */ + if ((thread->st_xact_data->xd_flags & XT_XN_XAC_LOGGED)) { + if (!xt_xlog_modify_table(0, XT_LOG_ENT_PREPARE, xn_id, 0, 0, len, xa_data, thread)) + return FAILED; + } + } + return OK; +} + +xtPublic xtBool xt_xn_store_xa_data(XTDatabaseHPtr db, xtXactID xact_id, int len, xtWord1 *xa_data, XTThreadPtr XT_UNUSED(thread)) +{ + XTXactPreparePtr xap; + u_int idx; + XTXactXARec xx; + + if (!(xap = (XTXactPreparePtr) xt_malloc_ns(offsetof(XTXactPrepareRec, xp_xa_data) + len))) + return FAILED; + xap->xp_xact_id = xact_id; + xap->xp_hash = xt_get_checksum4(xa_data, len); + xap->xp_data_len = len; + memcpy(xap->xp_xa_data, xa_data, len); + xx.xx_xact_id = xact_id; + xx.xx_xa_ptr = xap; + + idx = xap->xp_hash % XT_XA_HASH_TAB_SIZE; + xt_lock_mutex_ns(&db->db_xn_xa_lock); + if (!xt_sl_insert(NULL, db->db_xn_xa_list, &xact_id, &xx)) { + xt_unlock_mutex_ns(&db->db_xn_xa_lock); + xt_free_ns(xap); + } + xap->xp_next = db->db_xn_xa_table[idx]; + db->db_xn_xa_table[idx] = xap; + xt_unlock_mutex_ns(&db->db_xn_xa_lock); + return OK; +} + +xtPublic void xt_xn_delete_xa_data_by_xact(XTDatabaseHPtr db, xtXactID xact_id, XTThreadPtr thread) +{ + XTXactXAPtr xx; + + xt_lock_mutex_ns(&db->db_xn_xa_lock); + if (!(xx = (XTXactXAPtr) xt_sl_find(NULL, db->db_xn_xa_list, &xact_id))) + return; + xt_xn_delete_xa_data(db, xx->xx_xa_ptr, TRUE, thread); +} + +xtPublic void xt_xn_delete_xa_data(XTDatabaseHPtr db, XTXactPreparePtr xap, xtBool unlock, XTThreadPtr XT_UNUSED(thread)) +{ + u_int idx; + XTXactPreparePtr xap_ptr, xap_pptr = NULL; + + xt_sl_delete(NULL, db->db_xn_xa_list, &xap->xp_xact_id); + idx = xap->xp_hash % XT_XA_HASH_TAB_SIZE; + xap_ptr = db->db_xn_xa_table[idx]; + while (xap_ptr) { + if (xap_ptr == xap) + break; + xap_pptr = xap_ptr; + xap_ptr = xap_ptr->xp_next; + } + if (xap_ptr) { + if (xap_pptr) + xap_pptr->xp_next = xap_ptr->xp_next; + else + db->db_xn_xa_table[idx] = xap_ptr->xp_next; + xt_free_ns(xap); + } + if (unlock) + xt_unlock_mutex_ns(&db->db_xn_xa_lock); +} + +xtPublic XTXactPreparePtr xt_xn_find_xa_data(XTDatabaseHPtr db, int len, xtWord1 *xa_data, xtBool lock, XTThreadPtr XT_UNUSED(thread)) +{ + xtWord4 hash; + XTXactPreparePtr xap; + u_int idx; + + if (lock) + xt_lock_mutex_ns(&db->db_xn_xa_lock); + hash = xt_get_checksum4(xa_data, len); + idx = hash % XT_XA_HASH_TAB_SIZE; + xap = db->db_xn_xa_table[idx]; + while (xap) { + if (xap->xp_hash == hash && + xap->xp_data_len == len && + memcmp(xap->xp_xa_data, xa_data, len) == 0) { + break; + } + xap = xap->xp_next; + } + + return xap; +} + +xtPublic XTXactPreparePtr xt_xn_enum_xa_data(XTDatabaseHPtr db, XTXactEnumXAPtr exa) +{ + XTXactXAPtr xx; + + if (!exa->exa_locked) { + xt_lock_mutex_ns(&db->db_xn_xa_lock); + exa->exa_locked = TRUE; + } + + if ((xx = (XTXactXAPtr) xt_sl_item_at(db->db_xn_xa_list, exa->exa_index))) { + exa->exa_index++; + return xx->xx_xa_ptr; + } + + if (exa->exa_locked) { + exa->exa_locked = FALSE; + xt_unlock_mutex_ns(&db->db_xn_xa_lock); + } + return NULL; +} + +/* ---------------------------------------------------------------------- + * S W E E P E R F U N C T I O N S + */ + xtPublic xtWord8 xt_xn_bytes_to_sweep(XTDatabaseHPtr db, XTThreadPtr thread) { xtXactID xn_id; @@ -2047,7 +2208,7 @@ static xtBool xn_sw_cleanup_variation(XT if(!tab->tab_recs.xt_tc_write_cond(self, ot->ot_rec_file, rec_id, rec_head.tr_rec_type_1, &op_seq, xn_id, row_id, stat_id, rec_type)) /* this means record was not updated by xt_tc_write_bor and doesn't need to */ break; - if (!xt_xlog_modify_table(ot, XT_LOG_ENT_REC_CLEANED_1, op_seq, 0, rec_id, 1, &rec_head.tr_rec_type_1)) + if (!xt_xlog_modify_table(tab->tab_id, XT_LOG_ENT_REC_CLEANED_1, op_seq, 0, rec_id, 1, &rec_head.tr_rec_type_1, self)) throw_(); xn_sw_clean_indices(self, ot, rec_id, row_id, rec_buf, ss->ss_databuf.db_data); break; @@ -2397,8 +2558,10 @@ static void xn_sw_main(XTThreadPtr self) if ((xact = xt_xn_get_xact(db, db->db_xn_to_clean_id, self))) { xtXactID xn_id; - if (!(xact->xd_flags & XT_XN_XAC_SWEEP)) - /* Transaction has not yet ending, and ready to sweep. */ + /* The sweep flag is set when the transaction is ready for sweeping. + * Prepared transactions may not be swept! + */ + if (!(xact->xd_flags & XT_XN_XAC_SWEEP) || (xact->xd_flags & XT_XN_XAC_PREPARED)) goto sleep; /* Check if we can cleanup the transaction. @@ -2493,7 +2656,7 @@ static void xn_sw_main(XTThreadPtr self) * we flush the log. */ if (now >= idle_start + 2) { - if (!xt_xlog_flush_log(self)) + if (!xt_xlog_flush_log(db, self)) xt_throw(self); ss->ss_flush_pending = FALSE; } @@ -2516,7 +2679,7 @@ static void xn_sw_main(XTThreadPtr self) } if (ss->ss_flush_pending) { - xt_xlog_flush_log(self); + xt_xlog_flush_log(db, self); ss->ss_flush_pending = FALSE; } === modified file 'storage/pbxt/src/xaction_xt.h' --- a/storage/pbxt/src/xaction_xt.h 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/xaction_xt.h 2009-11-24 10:55:06 +0000 @@ -87,6 +87,7 @@ struct XTOpenTable; #define XT_XN_XAC_CLEANED 8 /* The transaction has been cleaned. */ #define XT_XN_XAC_RECOVERED 16 /* This transaction was detected on recovery. */ #define XT_XN_XAC_SWEEP 32 /* End ID has been set, OK to sweep. */ +#define XT_XN_XAC_PREPARED 64 /* The transaction was prepared (used only by recovery). */ #define XT_XN_VISIBLE 0 /* The transaction is committed, and the record is visible. */ #define XT_XN_NOT_VISIBLE 1 /* The transaction is committed, but not visible. */ @@ -95,6 +96,24 @@ struct XTOpenTable; #define XT_XN_OTHER_UPDATE 4 /* The record was updated by someone else. */ #define XT_XN_REREAD 5 /* The transaction is not longer in RAM, status is unkown, retry. */ +typedef struct XTXactPrepare { + xtXactID xp_xact_id; + xtWord4 xp_hash; + struct XTXactPrepare *xp_next; /* Next item in hash table. */ + int xp_data_len; + xtWord1 xp_xa_data[XT_MAX_XA_DATA_SIZE]; +} XTXactPrepareRec, *XTXactPreparePtr; + +typedef struct XTXactXA { + xtXactID xx_xact_id; + XTXactPreparePtr xx_xa_ptr; +} XTXactXARec, *XTXactXAPtr; + +typedef struct XTXactEnumXA { + u_int exa_index; + xtBool exa_locked; +} XTXactEnumXARec, *XTXactEnumXAPtr; + typedef struct XTXactData { xtXactID xd_start_xn_id; /* Note: may be zero!. */ xtXactID xd_end_xn_id; /* Note: may be zero!. */ @@ -105,6 +124,7 @@ typedef struct XTXactData { int xd_flags; xtWord4 xd_end_time; xtThreadID xd_thread_id; + xtWord4 xd_xa_hash; /* 0 if no XA transaction. */ /* A transaction may be indexed twice in the hash table. * Once on the start sequence number, and once on the @@ -123,7 +143,7 @@ typedef struct XTXactData { #if defined(XT_XACT_USE_PTHREAD_RW) #define XT_XACT_LOCK_TYPE xt_rwlock_type -#define XT_XACT_INIT_LOCK(s, i) xt_init_rwlock(s, i) +#define XT_XACT_INIT_LOCK(s, i) xt_init_rwlock_with_autoname(s, i) #define XT_XACT_FREE_LOCK(s, i) xt_free_rwlock(i) #define XT_XACT_READ_LOCK(i, s) xt_slock_rwlock_ns(i) #define XT_XACT_WRITE_LOCK(i, s) xt_xlock_rwlock_ns(i) @@ -183,6 +203,14 @@ void xt_xn_wakeup_thread(xtThreadID th xtXactID xt_xn_get_curr_id(struct XTDatabase *db); xtWord8 xt_xn_bytes_to_sweep(struct XTDatabase *db, struct XTThread *thread); +int xt_xn_xa_compare(struct XTThread *self, register const void *thunk, register const void *a, register const void *b); +xtBool xt_xn_prepare(int len, xtWord1 *xa_data, struct XTThread *thread); +xtBool xt_xn_store_xa_data(struct XTDatabase *db, xtXactID xn_id, int len, xtWord1 *xa_data, struct XTThread *thread); +void xt_xn_delete_xa_data_by_xact(struct XTDatabase *db, xtXactID xact_id, struct XTThread *thread); +void xt_xn_delete_xa_data(struct XTDatabase *db, XTXactPreparePtr xap, xtBool unlock, struct XTThread *thread); +XTXactPreparePtr xt_xn_find_xa_data(struct XTDatabase *db, int len, xtWord1 *xa_data, xtBool lock, struct XTThread *thread); +XTXactPreparePtr xt_xn_enum_xa_data(struct XTDatabase *db, XTXactEnumXAPtr exa); + XTXactDataPtr xt_xn_add_old_xact(struct XTDatabase *db, xtXactID xn_id, struct XTThread *thread); XTXactDataPtr xt_xn_get_xact(struct XTDatabase *db, xtXactID xn_id, struct XTThread *thread); xtBool xt_xn_delete_xact(struct XTDatabase *db, xtXactID xn_id, struct XTThread *thread); === modified file 'storage/pbxt/src/xactlog_xt.cc' --- a/storage/pbxt/src/xactlog_xt.cc 2009-09-03 06:15:03 +0000 +++ b/storage/pbxt/src/xactlog_xt.cc 2009-11-24 10:55:06 +0000 @@ -1108,6 +1108,9 @@ xtBool XTDatabaseLog::xlog_append(XTThre if ((part_size = xl_write_buf_pos % 512)) { part_size = 512 - part_size; xl_write_buffer[xl_write_buf_pos] = XT_LOG_ENT_END_OF_LOG; +#ifdef HAVE_valgrind + memset(xl_write_buffer + xl_write_buf_pos + 1, 0x66, part_size); +#endif if (!xt_pwrite_file(xl_log_file, xl_write_log_offset, xl_write_buf_pos+part_size, xl_write_buffer, &thread->st_statistics.st_xlog, thread)) goto write_failed; } @@ -1477,9 +1480,9 @@ void XTDatabaseLog::xlog_name(size_t siz * T H R E A D T R A N S A C T I O N B U F F E R */ -xtPublic xtBool xt_xlog_flush_log(XTThreadPtr thread) +xtPublic xtBool xt_xlog_flush_log(struct XTDatabase *db, XTThreadPtr thread) { - return thread->st_database->db_xlog.xlog_flush(thread); + return db->db_xlog.xlog_flush(thread); } xtPublic xtBool xt_xlog_log_data(XTThreadPtr thread, size_t size, XTXactLogBufferDPtr log_entry, xtBool commit) @@ -1488,15 +1491,14 @@ xtPublic xtBool xt_xlog_log_data(XTThrea } /* Allocate a record from the free list. */ -xtPublic xtBool xt_xlog_modify_table(struct XTOpenTable *ot, u_int status, xtOpSeqNo op_seq, xtRecordID free_rec_id, xtRecordID rec_id, size_t size, xtWord1 *data) +xtPublic xtBool xt_xlog_modify_table(xtTableID tab_id, u_int status, xtOpSeqNo op_seq, xtRecordID free_rec_id, xtRecordID rec_id, size_t size, xtWord1 *data, XTThreadPtr thread) { XTXactLogBufferDRec log_entry; - XTThreadPtr thread = ot->ot_thread; - XTTableHPtr tab = ot->ot_table; size_t len; xtWord4 sum = 0; int check_size = 1; XTXactDataPtr xact = NULL; + xtBool commit = FALSE; switch (status) { case XT_LOG_ENT_REC_MODIFIED: @@ -1505,7 +1507,7 @@ xtPublic xtBool xt_xlog_modify_table(str case XT_LOG_ENT_DELETE: check_size = 2; XT_SET_DISK_4(log_entry.xu.xu_op_seq_4, op_seq); - XT_SET_DISK_4(log_entry.xu.xu_tab_id_4, tab->tab_id); + XT_SET_DISK_4(log_entry.xu.xu_tab_id_4, tab_id); XT_SET_DISK_4(log_entry.xu.xu_rec_id_4, rec_id); XT_SET_DISK_2(log_entry.xu.xu_size_2, size); len = offsetof(XTactUpdateEntryDRec, xu_rec_type_1); @@ -1521,7 +1523,7 @@ xtPublic xtBool xt_xlog_modify_table(str case XT_LOG_ENT_DELETE_FL: check_size = 2; XT_SET_DISK_4(log_entry.xf.xf_op_seq_4, op_seq); - XT_SET_DISK_4(log_entry.xf.xf_tab_id_4, tab->tab_id); + XT_SET_DISK_4(log_entry.xf.xf_tab_id_4, tab_id); XT_SET_DISK_4(log_entry.xf.xf_rec_id_4, rec_id); XT_SET_DISK_2(log_entry.xf.xf_size_2, size); XT_SET_DISK_4(log_entry.xf.xf_free_rec_id_4, free_rec_id); @@ -1539,14 +1541,14 @@ xtPublic xtBool xt_xlog_modify_table(str case XT_LOG_ENT_REC_REMOVED_EXT: ASSERT_NS(size == 1 + XT_XACT_ID_SIZE + sizeof(XTTabRecFreeDRec)); XT_SET_DISK_4(log_entry.fr.fr_op_seq_4, op_seq); - XT_SET_DISK_4(log_entry.fr.fr_tab_id_4, tab->tab_id); + XT_SET_DISK_4(log_entry.fr.fr_tab_id_4, tab_id); XT_SET_DISK_4(log_entry.fr.fr_rec_id_4, rec_id); len = offsetof(XTactFreeRecEntryDRec, fr_stat_id_1); break; case XT_LOG_ENT_REC_REMOVED_BI: check_size = 2; XT_SET_DISK_4(log_entry.rb.rb_op_seq_4, op_seq); - XT_SET_DISK_4(log_entry.rb.rb_tab_id_4, tab->tab_id); + XT_SET_DISK_4(log_entry.rb.rb_tab_id_4, tab_id); XT_SET_DISK_4(log_entry.rb.rb_rec_id_4, rec_id); XT_SET_DISK_2(log_entry.rb.rb_size_2, size); log_entry.rb.rb_new_rec_type_1 = (xtWord1) free_rec_id; @@ -1556,42 +1558,42 @@ xtPublic xtBool xt_xlog_modify_table(str case XT_LOG_ENT_REC_MOVED: ASSERT_NS(size == 8); XT_SET_DISK_4(log_entry.xw.xw_op_seq_4, op_seq); - XT_SET_DISK_4(log_entry.xw.xw_tab_id_4, tab->tab_id); + XT_SET_DISK_4(log_entry.xw.xw_tab_id_4, tab_id); XT_SET_DISK_4(log_entry.xw.xw_rec_id_4, rec_id); len = offsetof(XTactWriteRecEntryDRec, xw_rec_type_1); break; case XT_LOG_ENT_REC_CLEANED: ASSERT_NS(size == offsetof(XTTabRecHeadDRec, tr_prev_rec_id_4) + XT_RECORD_ID_SIZE); XT_SET_DISK_4(log_entry.xw.xw_op_seq_4, op_seq); - XT_SET_DISK_4(log_entry.xw.xw_tab_id_4, tab->tab_id); + XT_SET_DISK_4(log_entry.xw.xw_tab_id_4, tab_id); XT_SET_DISK_4(log_entry.xw.xw_rec_id_4, rec_id); len = offsetof(XTactWriteRecEntryDRec, xw_rec_type_1); break; case XT_LOG_ENT_REC_CLEANED_1: ASSERT_NS(size == 1); XT_SET_DISK_4(log_entry.xw.xw_op_seq_4, op_seq); - XT_SET_DISK_4(log_entry.xw.xw_tab_id_4, tab->tab_id); + XT_SET_DISK_4(log_entry.xw.xw_tab_id_4, tab_id); XT_SET_DISK_4(log_entry.xw.xw_rec_id_4, rec_id); len = offsetof(XTactWriteRecEntryDRec, xw_rec_type_1); break; case XT_LOG_ENT_REC_UNLINKED: ASSERT_NS(size == offsetof(XTTabRecHeadDRec, tr_prev_rec_id_4) + XT_RECORD_ID_SIZE); XT_SET_DISK_4(log_entry.xw.xw_op_seq_4, op_seq); - XT_SET_DISK_4(log_entry.xw.xw_tab_id_4, tab->tab_id); + XT_SET_DISK_4(log_entry.xw.xw_tab_id_4, tab_id); XT_SET_DISK_4(log_entry.xw.xw_rec_id_4, rec_id); len = offsetof(XTactWriteRecEntryDRec, xw_rec_type_1); break; case XT_LOG_ENT_ROW_NEW: ASSERT_NS(size == 0); XT_SET_DISK_4(log_entry.xa.xa_op_seq_4, op_seq); - XT_SET_DISK_4(log_entry.xa.xa_tab_id_4, tab->tab_id); + XT_SET_DISK_4(log_entry.xa.xa_tab_id_4, tab_id); XT_SET_DISK_4(log_entry.xa.xa_row_id_4, rec_id); len = offsetof(XTactRowAddedEntryDRec, xa_row_id_4) + XT_ROW_ID_SIZE; break; case XT_LOG_ENT_ROW_NEW_FL: ASSERT_NS(size == 0); XT_SET_DISK_4(log_entry.xa.xa_op_seq_4, op_seq); - XT_SET_DISK_4(log_entry.xa.xa_tab_id_4, tab->tab_id); + XT_SET_DISK_4(log_entry.xa.xa_tab_id_4, tab_id); XT_SET_DISK_4(log_entry.xa.xa_row_id_4, rec_id); XT_SET_DISK_4(log_entry.xa.xa_free_list_4, free_rec_id); sum ^= XT_CHECKSUM4_REC(free_rec_id); @@ -1602,10 +1604,17 @@ xtPublic xtBool xt_xlog_modify_table(str case XT_LOG_ENT_ROW_FREED: ASSERT_NS(size == sizeof(XTTabRowRefDRec)); XT_SET_DISK_4(log_entry.wr.wr_op_seq_4, op_seq); - XT_SET_DISK_4(log_entry.wr.wr_tab_id_4, tab->tab_id); + XT_SET_DISK_4(log_entry.wr.wr_tab_id_4, tab_id); XT_SET_DISK_4(log_entry.wr.wr_row_id_4, rec_id); len = offsetof(XTactWriteRowEntryDRec, wr_ref_id_4); break; + case XT_LOG_ENT_PREPARE: + check_size = 2; + XT_SET_DISK_4(log_entry.xp.xp_xact_id_4, op_seq); + log_entry.xp.xp_xa_len_1 = (xtWord1) size; + len = offsetof(XTXactPrepareEntryDRec, xp_xa_data); + commit = TRUE; + break; default: ASSERT_NS(FALSE); len = 0; @@ -1615,7 +1624,7 @@ xtPublic xtBool xt_xlog_modify_table(str xtWord1 *dptr = data; xtWord4 g; - sum ^= op_seq ^ (tab->tab_id << 8) ^ XT_CHECKSUM4_REC(rec_id); + sum ^= op_seq ^ (tab_id << 8) ^ XT_CHECKSUM4_REC(rec_id); if ((g = sum & 0xF0000000)) { sum = sum ^ (g >> 24); sum = sum ^ g; @@ -1643,9 +1652,9 @@ xtPublic xtBool xt_xlog_modify_table(str xt_print_log_record(0, 0, &log_entry); #endif if (xact) - return thread->st_database->db_xlog.xlog_append(thread, len, (xtWord1 *) &log_entry, size, data, FALSE, &xact->xd_begin_log, &xact->xd_begin_offset); + return thread->st_database->db_xlog.xlog_append(thread, len, (xtWord1 *) &log_entry, size, data, commit, &xact->xd_begin_log, &xact->xd_begin_offset); - return thread->st_database->db_xlog.xlog_append(thread, len, (xtWord1 *) &log_entry, size, data, FALSE, NULL, NULL); + return thread->st_database->db_xlog.xlog_append(thread, len, (xtWord1 *) &log_entry, size, data, commit, NULL, NULL); } /* @@ -1905,6 +1914,7 @@ xtBool XTDatabaseLog::xlog_verify(XTXact xtRecordID rec_id, free_rec_id; int check_size = 1; xtWord1 *dptr; + xtWord4 g; switch (record->xh.xh_status_1) { case XT_LOG_ENT_HEADER: @@ -2019,13 +2029,19 @@ xtBool XTDatabaseLog::xlog_verify(XTXact return record->xe.xe_checksum_1 == (XT_CHECKSUM_1(sum) ^ XT_CHECKSUM_1(log_id)); case XT_LOG_ENT_END_OF_LOG: return FALSE; + case XT_LOG_ENT_PREPARE: + check_size = 2; + op_seq = XT_GET_DISK_4(record->xp.xp_xact_id_4); + tab_id = 0; + rec_id = 0; + dptr = record->xp.xp_xa_data; + rec_size -= offsetof(XTXactPrepareEntryDRec, xp_xa_data); + break; default: ASSERT_NS(FALSE); return FALSE; } - xtWord4 g; - sum ^= (xtWord4) op_seq ^ ((xtWord4) tab_id << 8) ^ XT_CHECKSUM4_REC(rec_id); if ((g = sum & 0xF0000000)) { @@ -2193,6 +2209,14 @@ xtBool XTDatabaseLog::xlog_seq_next(XTXa } goto return_empty; } + case XT_LOG_ENT_PREPARE: + check_size = 2; + len = offsetof(XTXactPrepareEntryDRec, xp_xa_data); + if (len > max_rec_len) + /* The size is not in the buffer: */ + goto read_more; + len += (size_t) record->xp.xp_xa_len_1; + break; default: /* It is possible to land here after a crash, if the * log was not completely written. @@ -2231,7 +2255,7 @@ xtBool XTDatabaseLog::xlog_seq_next(XTXa goto return_empty; } - /* The record is not completely in the buffer: */ + /* The record is now completely in the buffer: */ seq->xseq_record_len = len; *ret_entry = (XTXactLogBufferDPtr) seq->xseq_buffer; return OK; @@ -2428,7 +2452,7 @@ static void xlog_wr_wait_for_log_flush(X if (reason == XT_LOG_CACHE_FULL || reason == XT_TIME_TO_WRITE || reason == XT_CHECKPOINT_REQ) { /* Make sure that we have something to write: */ if (db->db_xlog.xlog_bytes_to_write() < 2 * 1204 * 1024) - xt_xlog_flush_log(self); + xt_xlog_flush_log(db, self); } #ifdef TRACE_WRITER_ACTIVITY @@ -2529,6 +2553,7 @@ static void xlog_wr_main(XTThreadPtr sel case XT_LOG_ENT_ABORT: case XT_LOG_ENT_CLEANUP: case XT_LOG_ENT_OP_SYNC: + case XT_LOG_ENT_PREPARE: break; case XT_LOG_ENT_DEL_LOG: xtLogID log_id; === modified file 'storage/pbxt/src/xactlog_xt.h' --- a/storage/pbxt/src/xactlog_xt.h 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/xactlog_xt.h 2009-11-24 10:55:06 +0000 @@ -160,6 +160,7 @@ typedef struct XTXLogCache { #define XT_LOG_ENT_END_OF_LOG 37 /* This is a record that indicates the end of the log, and * fills to the end of a 512 byte block. */ +#define XT_LOG_ENT_PREPARE 39 /* XA prepare log entry. */ #define XT_LOG_FILE_MAGIC 0xAE88FE12 #define XT_LOG_VERSION_NO 1 @@ -201,6 +202,14 @@ typedef struct XTXactEndEntry { XTDiskValue4 xe_not_used_4; /* Was the end sequence number (no longer used - v1.0.04+), set to zero). */ } XTXactEndEntryDRec, *XTXactEndEntryDPtr; +typedef struct XTXactPrepareEntry { + xtWord1 xp_status_1; /* XT_LOG_ENT_PREPARE */ + XTDiskValue2 xp_checksum_2; + XTDiskValue4 xp_xact_id_4; /* The transaction. */ + xtWord1 xp_xa_len_1; /* The length of the XA data. */ + xtWord1 xp_xa_data[XT_MAX_XA_DATA_SIZE]; +} XTXactPrepareEntryDRec, *XTXactPrepareEntryDPtr; + typedef struct XTXactCleanupEntry { xtWord1 xc_status_1; /* XT_LOG_ENT_CLEANUP */ xtWord1 xc_checksum_1; @@ -344,6 +353,7 @@ typedef union XTXactLogBuffer { XTactOpSyncEntryDRec os; XTactExtRecEntryDRec er; XTactNoOpEntryDRec no; + XTXactPrepareEntryDRec xp; } XTXactLogBufferDRec, *XTXactLogBufferDPtr; /* ---------------------------------------- */ @@ -453,9 +463,9 @@ private: xtBool xlog_open_log(xtLogID log_id, off_t curr_eof, struct XTThread *thread); } XTDatabaseLogRec, *XTDatabaseLogPtr; -xtBool xt_xlog_flush_log(struct XTThread *thread); +xtBool xt_xlog_flush_log(struct XTDatabase *db, struct XTThread *thread); xtBool xt_xlog_log_data(struct XTThread *thread, size_t len, XTXactLogBufferDPtr log_entry, xtBool commit); -xtBool xt_xlog_modify_table(struct XTOpenTable *ot, u_int status, xtOpSeqNo op_seq, xtRecordID free_list, xtRecordID address, size_t size, xtWord1 *data); +xtBool xt_xlog_modify_table(xtTableID tab_id, u_int status, xtOpSeqNo op_seq, xtRecordID free_list, xtRecordID address, size_t size, xtWord1 *data, struct XTThread *thread); void xt_xlog_init(struct XTThread *self, size_t cache_size); void xt_xlog_exit(struct XTThread *self); === modified file 'storage/pbxt/src/xt_config.h' --- a/storage/pbxt/src/xt_config.h 2009-08-31 11:07:44 +0000 +++ b/storage/pbxt/src/xt_config.h 2009-11-24 10:55:06 +0000 @@ -50,7 +50,9 @@ const int max_connections = 500; /* * Make sure we use the thread safe version of the library. */ +#ifndef _THREAD_SAFE // Seems to be defined by some Drizzle header #define _THREAD_SAFE +#endif /* * This causes things to be defined like stuff in inttypes.h @@ -72,12 +74,12 @@ const int max_connections = 500; #define XT_MAC #endif -#if defined(MSDOS) || defined(__WIN__) +#if defined(MSDOS) || defined(__WIN__) || defined(_WIN64) #define XT_WIN #endif #ifdef XT_WIN -#ifdef _DEBUG +#if defined(_DEBUG) && !defined(DEBUG) #define DEBUG #endif // _DEBUG #else @@ -101,8 +103,13 @@ const int max_connections = 500; * Definition of which atomic operations to use: */ #ifdef XT_WIN +#ifdef _WIN64 +/* 64-bit Windows atomic ops are not yet supported: */ +#define XT_NO_ATOMICS +#else /* MS Studio style embedded assembler for x86 */ #define XT_ATOMIC_WIN32_X86 +#endif #elif defined(__GNUC__) && (defined(__x86_64__) || defined(__i386__)) /* Use GNU style embedded assembler for x86 */ #define XT_ATOMIC_GNUC_X86 @@ -115,4 +122,10 @@ const int max_connections = 500; #define XT_NO_ATOMICS #endif +#ifndef DRIZZLED +#if MYSQL_VERSION_ID >= 50404 +#define MYSQL_SUPPORTS_BACKUP +#endif +#endif + #endif === modified file 'storage/pbxt/src/xt_defs.h' --- a/storage/pbxt/src/xt_defs.h 2009-08-17 11:12:36 +0000 +++ b/storage/pbxt/src/xt_defs.h 2009-11-24 10:55:06 +0000 @@ -379,6 +379,19 @@ typedef struct XTPathStr { */ //#define XT_IMPLEMENT_NO_ACTION +/* Define this value if online-backup should be supported. + * Note that, online backup is currently only supported + * by MySQL 6.0.9 or later + */ +#define XT_ENABLE_ONLINE_BACKUP + +/* Define this switch if you don't want to use atomic + * synchronisation. + */ +#ifndef XT_NO_ATOMICS +//#define XT_NO_ATOMICS +#endif + /* ---------------------------------------------------------------------- * GLOBAL CONSTANTS */ @@ -406,6 +419,8 @@ typedef struct XTPathStr { #define XT_ADD_PTR(p, l) ((void *) ((char *) (p) + (l))) +#define XT_MAX_XA_DATA_SIZE (3*4 + 128) /* Corresponds to the maximum size of struct xid_t in handler.h. */ + /* ---------------------------------------------------------------------- * DEFINES DEPENDENT ON CONSTANTS */ @@ -744,6 +759,7 @@ extern xtBool pbxt_crash_debug; #define MYSQL_PLUGIN_VAR_HEADER DRIZZLE_PLUGIN_VAR_HEADER #define MYSQL_SYSVAR_STR DRIZZLE_SYSVAR_STR #define MYSQL_SYSVAR_INT DRIZZLE_SYSVAR_INT +#define MYSQL_SYSVAR_BOOL DRIZZLE_SYSVAR_BOOL #define MYSQL_SYSVAR DRIZZLE_SYSVAR #define MYSQL_STORAGE_ENGINE_PLUGIN DRIZZLE_STORAGE_ENGINE_PLUGIN #define MYSQL_INFORMATION_SCHEMA_PLUGIN DRIZZLE_INFORMATION_SCHEMA_PLUGIN @@ -752,9 +768,8 @@ extern xtBool pbxt_crash_debug; #define mx_tmp_use_all_columns(x, y) (x)->use_all_columns(y) #define mx_tmp_restore_column_map(x, y) (x)->restore_column_map(y) -#define MX_BIT_FAST_TEST_AND_SET(x, y) bitmap_test_and_set(x, y) -#define MX_TABLE_TYPES_T handler::Table_flags +#define MX_TABLE_TYPES_T Cursor::Table_flags #define MX_UINT8_T uint8_t #define MX_ULONG_T uint32_t #define MX_ULONGLONG_T uint64_t @@ -762,6 +777,10 @@ extern xtBool pbxt_crash_debug; #define MX_CHARSET_INFO struct charset_info_st #define MX_CONST_CHARSET_INFO const struct charset_info_st #define MX_CONST const +#define MX_BITMAP MyBitmap +#define MX_BIT_SIZE() numOfBitsInMap() +#define MX_BIT_SET(x, y) (x)->setBit(y) +#define MX_BIT_FAST_TEST_AND_SET(x, y) (x)->testAndSet(y) #define my_bool bool #define int16 int16_t @@ -771,6 +790,7 @@ extern xtBool pbxt_crash_debug; #define uchar unsigned char #define longlong int64_t #define ulonglong uint64_t +#define handler Cursor #define HAVE_LONG_LONG @@ -823,10 +843,13 @@ extern xtBool pbxt_crash_debug; class PBXTStorageEngine; typedef PBXTStorageEngine handlerton; +class Session; + +extern "C" void session_mark_transaction_to_rollback(Session *session, bool all); #else // DRIZZLED /* The MySQL case: */ -#if MYSQL_VERSION_ID >= 60008 +#if MYSQL_VERSION_ID >= 50404 #define STRUCT_TABLE struct TABLE #else #define STRUCT_TABLE struct st_table @@ -844,13 +867,13 @@ typedef PBXTStorageEngine handlerton; #define MX_CHARSET_INFO CHARSET_INFO #define MX_CONST_CHARSET_INFO struct charset_info_st #define MX_CONST +#define MX_BITMAP MY_BITMAP +#define MX_BIT_SIZE() n_bits +#define MX_BIT_SET(x, y) bitmap_set_bit(x, y) #endif // DRIZZLED -#define MX_BITMAP MY_BITMAP -#define MX_BIT_SIZE() n_bits #define MX_BIT_IS_SUBSET(x, y) bitmap_is_subset(x, y) -#define MX_BIT_SET(x, y) bitmap_set_bit(x, y) #ifndef XT_SCAN_CORE_DEFINED #define XT_SCAN_CORE_DEFINED === modified file 'storage/pbxt/src/xt_errno.h' --- a/storage/pbxt/src/xt_errno.h 2009-09-03 06:15:03 +0000 +++ b/storage/pbxt/src/xt_errno.h 2009-11-24 10:55:06 +0000 @@ -119,6 +119,9 @@ #define XT_ERR_FK_REF_TEMP_TABLE -95 #define XT_ERR_MYSQL_SHUTDOWN -98 #define XT_ERR_MYSQL_NO_THREAD -99 +#define XT_ERR_BUFFER_TOO_SMALL -100 +#define XT_ERR_BAD_BACKUP_FORMAT -101 +#define XT_ERR_PBXT_NOT_INSTALLED -102 #ifdef XT_WIN #define XT_ENOMEM ERROR_NOT_ENOUGH_MEMORY

1 0

[Maria-developers] Updated (by Alexi): Store in binlog text of statements that caused RBR events (47)
by worklog-noreply＠askmonty.org 02 Dec '09

02 Dec '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Store in binlog text of statements that caused RBR events CREATION DATE..: Sat, 15 Aug 2009, 23:48 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: Knielsen CATEGORY.......: Server-RawIdeaBin TASK ID........: 47 (http://askmonty.org/worklog/?tid=47) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Alexi - Wed, 02 Dec 2009, 13:32)=-=- Low Level Design modified. --- /tmp/wklog.47.old.17456 2009-12-02 13:32:18.000000000 +0200 +++ /tmp/wklog.47.new.17456 2009-12-02 13:32:18.000000000 +0200 @@ -1,8 +1 @@ -mysql_binlog_send() [sql/sql_repl.cc] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -1. When sending events to a slave, master should simply skip - Annotate_rows events (they are not needed for replication). - [ ??? Multi-master - currently not clear ] -2. When sending events to mysqlbinlog (remote case), master - must send Annotate_rows events as well. -=-=(Alexi - Wed, 02 Dec 2009, 13:31)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.17414 2009-12-02 11:31:59.000000000 +0000 +++ /tmp/wklog.47.new.17414 2009-12-02 11:31:59.000000000 +0000 @@ -104,3 +104,10 @@ ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ ... +When master sends Annotate rows events +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +1. Master always sends Annotate_rows events to mysqlbinlog (in + remote case). +2. Master sends Annotate_rows events to a slave only if the slave has + both log-slave-updates and binlog-annotate-row-events options set. + -=-=(Knielsen - Mon, 30 Nov 2009, 11:21)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.18210 2009-11-30 11:21:33.000000000 +0200 +++ /tmp/wklog.47.new.18210 2009-11-30 11:21:33.000000000 +0200 @@ -28,7 +28,14 @@ Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query -text in its data part. Example: +text in its data part. + +The numeric code for this event must be assigned carefully. It should be +coordinated with MySQL/Sun, otherwise we can get into a situation where MySQL +uses the same numeric code for one event that MariaDB uses for +ANNOTATE_ROWS_EVENT, which would make merging the two impossible. + +Example: ... ************************ -=-=(Alexi - Mon, 30 Nov 2009, 10:33)=-=- Low Level Design modified. --- /tmp/wklog.47.old.16188 2009-11-30 10:33:44.000000000 +0200 +++ /tmp/wklog.47.new.16188 2009-11-30 10:33:44.000000000 +0200 @@ -2,6 +2,7 @@ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. When sending events to a slave, master should simply skip Annotate_rows events (they are not needed for replication). + [ ??? Multi-master - currently not clear ] 2. When sending events to mysqlbinlog (remote case), master must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 13:00)=-=- Low Level Design modified. --- /tmp/wklog.47.old.32047 2009-11-29 13:00:21.000000000 +0200 +++ /tmp/wklog.47.new.32047 2009-11-29 13:00:21.000000000 +0200 @@ -1 +1,7 @@ +mysql_binlog_send() [sql/sql_repl.cc] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +1. When sending events to a slave, master should simply skip + Annotate_rows events (they are not needed for replication). +2. When sending events to mysqlbinlog (remote case), master + must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 09:50)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.24993 2009-11-29 07:50:36.000000000 +0000 +++ /tmp/wklog.47.new.24993 2009-11-29 07:50:36.000000000 +0000 @@ -4,3 +4,96 @@ > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). +New server option +~~~~~~~~~~~~~~~~~ + --binlog-annotate-row-events + +Setting this option makes RBR (row-) events in the binary log to be +preceded by Annotate rows events (see below). The corresponding +'binlog_annotate_row_events' system variable is dynamic and has both +global and session values. Default global value is OFF. + +Note. Session values are usefull to make it possible to annotate only + some selected statements: + + ... + SET SESSION binlog_annotate_row_events=ON; + ... statements to be annotated ... + SET SESSION binlog_annotate_row_events=OFF; + ... statements not to be annotated ... + +New binlog event type +~~~~~~~~~~~~~~~~~~~~~ + Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] + +Describes the query which caused the corresponding row event. In binary log, +precedes each Table_map_log_event. Contains empty post-header and the query +text in its data part. Example: + + ... + ************************ + ANNOTATE_ROWS_EVENT [51] + ************************ + 000000C7 | 54 1B 12 4B | time_when = 1259477844 + 000000CB | 33 | event_type = 51 + 000000CC | 64 00 00 00 | server_id = 100 + 000000D0 | 2C 00 00 00 | event_len = 44 + 000000D4 | F3 00 00 00 | log_pos = 000000F3 + 000000D8 | 00 00 | flags = <none> + ------------------------ + 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" + 000000DE | 72 74 20 69 | + 000000E2 | 6E 74 6F 20 | + 000000E6 | 74 31 20 76 | + 000000EA | 61 6C 75 65 | + 000000EE | 73 20 28 31 | + 000000F2 | 29 | + ************************ + TABLE_MAP_EVENT [19] + ************************ + 000000F3 | 54 1B 12 4B | time_when = 1259477844 + 000000F7 | 13 | event_type = 19 + 000000F8 | 64 00 00 00 | server_id = 100 + 000000FC | 29 00 00 00 | event_len = 41 + 00000100 | 1C 01 00 00 | log_pos = 0000011C + 00000104 | 00 00 | flags = <none> + ------------------------ + ... + ************************ + WRITE_ROWS_EVENT [23] + ************************ + 0000011C | 54 1B 12 4B | time_when = 1259477844 + 00000120 | 17 | event_type = 23 + 00000121 | 64 00 00 00 | server_id = 100 + 00000125 | 22 00 00 00 | event_len = 34 + 00000129 | 3E 01 00 00 | log_pos = 0000013E + 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F + ------------------------ + 0000012F | 0F 00 00 00 | table_id = 15 + ... + +mysqlbinlog output +~~~~~~~~~~~~~~~~~~ +Something like this: + + ... + # at 199 + # at 243 + # at 284 + #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values +(1)` + #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped +to number 15 + #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 +flags: STMT_END_F + + BINLOG ' + VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= + VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= + VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== + '/*!*/; + ### INSERT INTO test.t1 + ### SET + ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ + ... + -=-=(Knielsen - Fri, 27 Nov 2009, 13:30)=-=- Observers changed: Knielsen -=-=(Psergey - Sun, 16 Aug 2009, 11:08)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.12485 2009-08-16 11:08:33.000000000 +0300 +++ /tmp/wklog.47.new.12485 2009-08-16 11:08:33.000000000 +0300 @@ -1 +1,6 @@ +First suggestion: + +> I think for this we would actually need a new binlog event type +> (Comment_log_event?). Unless we want to log an empty statement Query_log_event +> containing only a comment (a bit of a hack). -=-=(Psergey - Sun, 16 Aug 2009, 00:02)=-=- Dependency created: 39 now depends on 47 DESCRIPTION: Store in binlog (and show in mysqlbinlog output) texts of statements that caused RBR events This is needed for (list from Monty): - Easier to understand why updates happened - Would make it easier to find out where in application things went wrong (as you can search for exact strings) - Allow one to filter things based on comments in the statement. The cost of this can be that the binlog will be approximately 2x in size (especially insert of big blob's would be a bit painful), so this should be an optional feature. HIGH-LEVEL SPECIFICATION: First suggestion: > I think for this we would actually need a new binlog event type > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). New server option ~~~~~~~~~~~~~~~~~ --binlog-annotate-row-events Setting this option makes RBR (row-) events in the binary log to be preceded by Annotate rows events (see below). The corresponding 'binlog_annotate_row_events' system variable is dynamic and has both global and session values. Default global value is OFF. Note. Session values are usefull to make it possible to annotate only some selected statements: ... SET SESSION binlog_annotate_row_events=ON; ... statements to be annotated ... SET SESSION binlog_annotate_row_events=OFF; ... statements not to be annotated ... New binlog event type ~~~~~~~~~~~~~~~~~~~~~ Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query text in its data part. The numeric code for this event must be assigned carefully. It should be coordinated with MySQL/Sun, otherwise we can get into a situation where MySQL uses the same numeric code for one event that MariaDB uses for ANNOTATE_ROWS_EVENT, which would make merging the two impossible. Example: ... ************************ ANNOTATE_ROWS_EVENT [51] ************************ 000000C7 | 54 1B 12 4B | time_when = 1259477844 000000CB | 33 | event_type = 51 000000CC | 64 00 00 00 | server_id = 100 000000D0 | 2C 00 00 00 | event_len = 44 000000D4 | F3 00 00 00 | log_pos = 000000F3 000000D8 | 00 00 | flags = <none> ------------------------ 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" 000000DE | 72 74 20 69 | 000000E2 | 6E 74 6F 20 | 000000E6 | 74 31 20 76 | 000000EA | 61 6C 75 65 | 000000EE | 73 20 28 31 | 000000F2 | 29 | ************************ TABLE_MAP_EVENT [19] ************************ 000000F3 | 54 1B 12 4B | time_when = 1259477844 000000F7 | 13 | event_type = 19 000000F8 | 64 00 00 00 | server_id = 100 000000FC | 29 00 00 00 | event_len = 41 00000100 | 1C 01 00 00 | log_pos = 0000011C 00000104 | 00 00 | flags = <none> ------------------------ ... ************************ WRITE_ROWS_EVENT [23] ************************ 0000011C | 54 1B 12 4B | time_when = 1259477844 00000120 | 17 | event_type = 23 00000121 | 64 00 00 00 | server_id = 100 00000125 | 22 00 00 00 | event_len = 34 00000129 | 3E 01 00 00 | log_pos = 0000013E 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F ------------------------ 0000012F | 0F 00 00 00 | table_id = 15 ... mysqlbinlog output ~~~~~~~~~~~~~~~~~~ Something like this: ... # at 199 # at 243 # at 284 #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values (1)` #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped to number 15 #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 flags: STMT_END_F BINLOG ' VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== '/*!*/; ### INSERT INTO test.t1 ### SET ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ ... When master sends Annotate rows events ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. Master always sends Annotate_rows events to mysqlbinlog (in remote case). 2. Master sends Annotate_rows events to a slave only if the slave has both log-slave-updates and binlog-annotate-row-events options set. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Alexi): Store in binlog text of statements that caused RBR events (47)
by worklog-noreply＠askmonty.org 02 Dec '09

02 Dec '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Store in binlog text of statements that caused RBR events CREATION DATE..: Sat, 15 Aug 2009, 23:48 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: Knielsen CATEGORY.......: Server-RawIdeaBin TASK ID........: 47 (http://askmonty.org/worklog/?tid=47) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Alexi - Wed, 02 Dec 2009, 13:32)=-=- Low Level Design modified. --- /tmp/wklog.47.old.17456 2009-12-02 13:32:18.000000000 +0200 +++ /tmp/wklog.47.new.17456 2009-12-02 13:32:18.000000000 +0200 @@ -1,8 +1 @@ -mysql_binlog_send() [sql/sql_repl.cc] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -1. When sending events to a slave, master should simply skip - Annotate_rows events (they are not needed for replication). - [ ??? Multi-master - currently not clear ] -2. When sending events to mysqlbinlog (remote case), master - must send Annotate_rows events as well. -=-=(Alexi - Wed, 02 Dec 2009, 13:31)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.17414 2009-12-02 11:31:59.000000000 +0000 +++ /tmp/wklog.47.new.17414 2009-12-02 11:31:59.000000000 +0000 @@ -104,3 +104,10 @@ ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ ... +When master sends Annotate rows events +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +1. Master always sends Annotate_rows events to mysqlbinlog (in + remote case). +2. Master sends Annotate_rows events to a slave only if the slave has + both log-slave-updates and binlog-annotate-row-events options set. + -=-=(Knielsen - Mon, 30 Nov 2009, 11:21)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.18210 2009-11-30 11:21:33.000000000 +0200 +++ /tmp/wklog.47.new.18210 2009-11-30 11:21:33.000000000 +0200 @@ -28,7 +28,14 @@ Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query -text in its data part. Example: +text in its data part. + +The numeric code for this event must be assigned carefully. It should be +coordinated with MySQL/Sun, otherwise we can get into a situation where MySQL +uses the same numeric code for one event that MariaDB uses for +ANNOTATE_ROWS_EVENT, which would make merging the two impossible. + +Example: ... ************************ -=-=(Alexi - Mon, 30 Nov 2009, 10:33)=-=- Low Level Design modified. --- /tmp/wklog.47.old.16188 2009-11-30 10:33:44.000000000 +0200 +++ /tmp/wklog.47.new.16188 2009-11-30 10:33:44.000000000 +0200 @@ -2,6 +2,7 @@ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. When sending events to a slave, master should simply skip Annotate_rows events (they are not needed for replication). + [ ??? Multi-master - currently not clear ] 2. When sending events to mysqlbinlog (remote case), master must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 13:00)=-=- Low Level Design modified. --- /tmp/wklog.47.old.32047 2009-11-29 13:00:21.000000000 +0200 +++ /tmp/wklog.47.new.32047 2009-11-29 13:00:21.000000000 +0200 @@ -1 +1,7 @@ +mysql_binlog_send() [sql/sql_repl.cc] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +1. When sending events to a slave, master should simply skip + Annotate_rows events (they are not needed for replication). +2. When sending events to mysqlbinlog (remote case), master + must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 09:50)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.24993 2009-11-29 07:50:36.000000000 +0000 +++ /tmp/wklog.47.new.24993 2009-11-29 07:50:36.000000000 +0000 @@ -4,3 +4,96 @@ > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). +New server option +~~~~~~~~~~~~~~~~~ + --binlog-annotate-row-events + +Setting this option makes RBR (row-) events in the binary log to be +preceded by Annotate rows events (see below). The corresponding +'binlog_annotate_row_events' system variable is dynamic and has both +global and session values. Default global value is OFF. + +Note. Session values are usefull to make it possible to annotate only + some selected statements: + + ... + SET SESSION binlog_annotate_row_events=ON; + ... statements to be annotated ... + SET SESSION binlog_annotate_row_events=OFF; + ... statements not to be annotated ... + +New binlog event type +~~~~~~~~~~~~~~~~~~~~~ + Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] + +Describes the query which caused the corresponding row event. In binary log, +precedes each Table_map_log_event. Contains empty post-header and the query +text in its data part. Example: + + ... + ************************ + ANNOTATE_ROWS_EVENT [51] + ************************ + 000000C7 | 54 1B 12 4B | time_when = 1259477844 + 000000CB | 33 | event_type = 51 + 000000CC | 64 00 00 00 | server_id = 100 + 000000D0 | 2C 00 00 00 | event_len = 44 + 000000D4 | F3 00 00 00 | log_pos = 000000F3 + 000000D8 | 00 00 | flags = <none> + ------------------------ + 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" + 000000DE | 72 74 20 69 | + 000000E2 | 6E 74 6F 20 | + 000000E6 | 74 31 20 76 | + 000000EA | 61 6C 75 65 | + 000000EE | 73 20 28 31 | + 000000F2 | 29 | + ************************ + TABLE_MAP_EVENT [19] + ************************ + 000000F3 | 54 1B 12 4B | time_when = 1259477844 + 000000F7 | 13 | event_type = 19 + 000000F8 | 64 00 00 00 | server_id = 100 + 000000FC | 29 00 00 00 | event_len = 41 + 00000100 | 1C 01 00 00 | log_pos = 0000011C + 00000104 | 00 00 | flags = <none> + ------------------------ + ... + ************************ + WRITE_ROWS_EVENT [23] + ************************ + 0000011C | 54 1B 12 4B | time_when = 1259477844 + 00000120 | 17 | event_type = 23 + 00000121 | 64 00 00 00 | server_id = 100 + 00000125 | 22 00 00 00 | event_len = 34 + 00000129 | 3E 01 00 00 | log_pos = 0000013E + 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F + ------------------------ + 0000012F | 0F 00 00 00 | table_id = 15 + ... + +mysqlbinlog output +~~~~~~~~~~~~~~~~~~ +Something like this: + + ... + # at 199 + # at 243 + # at 284 + #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values +(1)` + #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped +to number 15 + #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 +flags: STMT_END_F + + BINLOG ' + VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= + VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= + VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== + '/*!*/; + ### INSERT INTO test.t1 + ### SET + ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ + ... + -=-=(Knielsen - Fri, 27 Nov 2009, 13:30)=-=- Observers changed: Knielsen -=-=(Psergey - Sun, 16 Aug 2009, 11:08)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.12485 2009-08-16 11:08:33.000000000 +0300 +++ /tmp/wklog.47.new.12485 2009-08-16 11:08:33.000000000 +0300 @@ -1 +1,6 @@ +First suggestion: + +> I think for this we would actually need a new binlog event type +> (Comment_log_event?). Unless we want to log an empty statement Query_log_event +> containing only a comment (a bit of a hack). -=-=(Psergey - Sun, 16 Aug 2009, 00:02)=-=- Dependency created: 39 now depends on 47 DESCRIPTION: Store in binlog (and show in mysqlbinlog output) texts of statements that caused RBR events This is needed for (list from Monty): - Easier to understand why updates happened - Would make it easier to find out where in application things went wrong (as you can search for exact strings) - Allow one to filter things based on comments in the statement. The cost of this can be that the binlog will be approximately 2x in size (especially insert of big blob's would be a bit painful), so this should be an optional feature. HIGH-LEVEL SPECIFICATION: First suggestion: > I think for this we would actually need a new binlog event type > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). New server option ~~~~~~~~~~~~~~~~~ --binlog-annotate-row-events Setting this option makes RBR (row-) events in the binary log to be preceded by Annotate rows events (see below). The corresponding 'binlog_annotate_row_events' system variable is dynamic and has both global and session values. Default global value is OFF. Note. Session values are usefull to make it possible to annotate only some selected statements: ... SET SESSION binlog_annotate_row_events=ON; ... statements to be annotated ... SET SESSION binlog_annotate_row_events=OFF; ... statements not to be annotated ... New binlog event type ~~~~~~~~~~~~~~~~~~~~~ Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query text in its data part. The numeric code for this event must be assigned carefully. It should be coordinated with MySQL/Sun, otherwise we can get into a situation where MySQL uses the same numeric code for one event that MariaDB uses for ANNOTATE_ROWS_EVENT, which would make merging the two impossible. Example: ... ************************ ANNOTATE_ROWS_EVENT [51] ************************ 000000C7 | 54 1B 12 4B | time_when = 1259477844 000000CB | 33 | event_type = 51 000000CC | 64 00 00 00 | server_id = 100 000000D0 | 2C 00 00 00 | event_len = 44 000000D4 | F3 00 00 00 | log_pos = 000000F3 000000D8 | 00 00 | flags = <none> ------------------------ 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" 000000DE | 72 74 20 69 | 000000E2 | 6E 74 6F 20 | 000000E6 | 74 31 20 76 | 000000EA | 61 6C 75 65 | 000000EE | 73 20 28 31 | 000000F2 | 29 | ************************ TABLE_MAP_EVENT [19] ************************ 000000F3 | 54 1B 12 4B | time_when = 1259477844 000000F7 | 13 | event_type = 19 000000F8 | 64 00 00 00 | server_id = 100 000000FC | 29 00 00 00 | event_len = 41 00000100 | 1C 01 00 00 | log_pos = 0000011C 00000104 | 00 00 | flags = <none> ------------------------ ... ************************ WRITE_ROWS_EVENT [23] ************************ 0000011C | 54 1B 12 4B | time_when = 1259477844 00000120 | 17 | event_type = 23 00000121 | 64 00 00 00 | server_id = 100 00000125 | 22 00 00 00 | event_len = 34 00000129 | 3E 01 00 00 | log_pos = 0000013E 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F ------------------------ 0000012F | 0F 00 00 00 | table_id = 15 ... mysqlbinlog output ~~~~~~~~~~~~~~~~~~ Something like this: ... # at 199 # at 243 # at 284 #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values (1)` #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped to number 15 #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 flags: STMT_END_F BINLOG ' VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== '/*!*/; ### INSERT INTO test.t1 ### SET ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ ... When master sends Annotate rows events ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. Master always sends Annotate_rows events to mysqlbinlog (in remote case). 2. Master sends Annotate_rows events to a slave only if the slave has both log-slave-updates and binlog-annotate-row-events options set. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Alexi): Store in binlog text of statements that caused RBR events (47)
by worklog-noreply＠askmonty.org 02 Dec '09

02 Dec '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Store in binlog text of statements that caused RBR events CREATION DATE..: Sat, 15 Aug 2009, 23:48 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: Knielsen CATEGORY.......: Server-RawIdeaBin TASK ID........: 47 (http://askmonty.org/worklog/?tid=47) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Alexi - Wed, 02 Dec 2009, 13:32)=-=- Low Level Design modified. --- /tmp/wklog.47.old.17456 2009-12-02 13:32:18.000000000 +0200 +++ /tmp/wklog.47.new.17456 2009-12-02 13:32:18.000000000 +0200 @@ -1,8 +1 @@ -mysql_binlog_send() [sql/sql_repl.cc] -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -1. When sending events to a slave, master should simply skip - Annotate_rows events (they are not needed for replication). - [ ??? Multi-master - currently not clear ] -2. When sending events to mysqlbinlog (remote case), master - must send Annotate_rows events as well. -=-=(Alexi - Wed, 02 Dec 2009, 13:31)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.17414 2009-12-02 11:31:59.000000000 +0000 +++ /tmp/wklog.47.new.17414 2009-12-02 11:31:59.000000000 +0000 @@ -104,3 +104,10 @@ ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ ... +When master sends Annotate rows events +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +1. Master always sends Annotate_rows events to mysqlbinlog (in + remote case). +2. Master sends Annotate_rows events to a slave only if the slave has + both log-slave-updates and binlog-annotate-row-events options set. + -=-=(Knielsen - Mon, 30 Nov 2009, 11:21)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.18210 2009-11-30 11:21:33.000000000 +0200 +++ /tmp/wklog.47.new.18210 2009-11-30 11:21:33.000000000 +0200 @@ -28,7 +28,14 @@ Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query -text in its data part. Example: +text in its data part. + +The numeric code for this event must be assigned carefully. It should be +coordinated with MySQL/Sun, otherwise we can get into a situation where MySQL +uses the same numeric code for one event that MariaDB uses for +ANNOTATE_ROWS_EVENT, which would make merging the two impossible. + +Example: ... ************************ -=-=(Alexi - Mon, 30 Nov 2009, 10:33)=-=- Low Level Design modified. --- /tmp/wklog.47.old.16188 2009-11-30 10:33:44.000000000 +0200 +++ /tmp/wklog.47.new.16188 2009-11-30 10:33:44.000000000 +0200 @@ -2,6 +2,7 @@ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. When sending events to a slave, master should simply skip Annotate_rows events (they are not needed for replication). + [ ??? Multi-master - currently not clear ] 2. When sending events to mysqlbinlog (remote case), master must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 13:00)=-=- Low Level Design modified. --- /tmp/wklog.47.old.32047 2009-11-29 13:00:21.000000000 +0200 +++ /tmp/wklog.47.new.32047 2009-11-29 13:00:21.000000000 +0200 @@ -1 +1,7 @@ +mysql_binlog_send() [sql/sql_repl.cc] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +1. When sending events to a slave, master should simply skip + Annotate_rows events (they are not needed for replication). +2. When sending events to mysqlbinlog (remote case), master + must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 09:50)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.24993 2009-11-29 07:50:36.000000000 +0000 +++ /tmp/wklog.47.new.24993 2009-11-29 07:50:36.000000000 +0000 @@ -4,3 +4,96 @@ > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). +New server option +~~~~~~~~~~~~~~~~~ + --binlog-annotate-row-events + +Setting this option makes RBR (row-) events in the binary log to be +preceded by Annotate rows events (see below). The corresponding +'binlog_annotate_row_events' system variable is dynamic and has both +global and session values. Default global value is OFF. + +Note. Session values are usefull to make it possible to annotate only + some selected statements: + + ... + SET SESSION binlog_annotate_row_events=ON; + ... statements to be annotated ... + SET SESSION binlog_annotate_row_events=OFF; + ... statements not to be annotated ... + +New binlog event type +~~~~~~~~~~~~~~~~~~~~~ + Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] + +Describes the query which caused the corresponding row event. In binary log, +precedes each Table_map_log_event. Contains empty post-header and the query +text in its data part. Example: + + ... + ************************ + ANNOTATE_ROWS_EVENT [51] + ************************ + 000000C7 | 54 1B 12 4B | time_when = 1259477844 + 000000CB | 33 | event_type = 51 + 000000CC | 64 00 00 00 | server_id = 100 + 000000D0 | 2C 00 00 00 | event_len = 44 + 000000D4 | F3 00 00 00 | log_pos = 000000F3 + 000000D8 | 00 00 | flags = <none> + ------------------------ + 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" + 000000DE | 72 74 20 69 | + 000000E2 | 6E 74 6F 20 | + 000000E6 | 74 31 20 76 | + 000000EA | 61 6C 75 65 | + 000000EE | 73 20 28 31 | + 000000F2 | 29 | + ************************ + TABLE_MAP_EVENT [19] + ************************ + 000000F3 | 54 1B 12 4B | time_when = 1259477844 + 000000F7 | 13 | event_type = 19 + 000000F8 | 64 00 00 00 | server_id = 100 + 000000FC | 29 00 00 00 | event_len = 41 + 00000100 | 1C 01 00 00 | log_pos = 0000011C + 00000104 | 00 00 | flags = <none> + ------------------------ + ... + ************************ + WRITE_ROWS_EVENT [23] + ************************ + 0000011C | 54 1B 12 4B | time_when = 1259477844 + 00000120 | 17 | event_type = 23 + 00000121 | 64 00 00 00 | server_id = 100 + 00000125 | 22 00 00 00 | event_len = 34 + 00000129 | 3E 01 00 00 | log_pos = 0000013E + 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F + ------------------------ + 0000012F | 0F 00 00 00 | table_id = 15 + ... + +mysqlbinlog output +~~~~~~~~~~~~~~~~~~ +Something like this: + + ... + # at 199 + # at 243 + # at 284 + #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values +(1)` + #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped +to number 15 + #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 +flags: STMT_END_F + + BINLOG ' + VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= + VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= + VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== + '/*!*/; + ### INSERT INTO test.t1 + ### SET + ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ + ... + -=-=(Knielsen - Fri, 27 Nov 2009, 13:30)=-=- Observers changed: Knielsen -=-=(Psergey - Sun, 16 Aug 2009, 11:08)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.12485 2009-08-16 11:08:33.000000000 +0300 +++ /tmp/wklog.47.new.12485 2009-08-16 11:08:33.000000000 +0300 @@ -1 +1,6 @@ +First suggestion: + +> I think for this we would actually need a new binlog event type +> (Comment_log_event?). Unless we want to log an empty statement Query_log_event +> containing only a comment (a bit of a hack). -=-=(Psergey - Sun, 16 Aug 2009, 00:02)=-=- Dependency created: 39 now depends on 47 DESCRIPTION: Store in binlog (and show in mysqlbinlog output) texts of statements that caused RBR events This is needed for (list from Monty): - Easier to understand why updates happened - Would make it easier to find out where in application things went wrong (as you can search for exact strings) - Allow one to filter things based on comments in the statement. The cost of this can be that the binlog will be approximately 2x in size (especially insert of big blob's would be a bit painful), so this should be an optional feature. HIGH-LEVEL SPECIFICATION: First suggestion: > I think for this we would actually need a new binlog event type > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). New server option ~~~~~~~~~~~~~~~~~ --binlog-annotate-row-events Setting this option makes RBR (row-) events in the binary log to be preceded by Annotate rows events (see below). The corresponding 'binlog_annotate_row_events' system variable is dynamic and has both global and session values. Default global value is OFF. Note. Session values are usefull to make it possible to annotate only some selected statements: ... SET SESSION binlog_annotate_row_events=ON; ... statements to be annotated ... SET SESSION binlog_annotate_row_events=OFF; ... statements not to be annotated ... New binlog event type ~~~~~~~~~~~~~~~~~~~~~ Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query text in its data part. The numeric code for this event must be assigned carefully. It should be coordinated with MySQL/Sun, otherwise we can get into a situation where MySQL uses the same numeric code for one event that MariaDB uses for ANNOTATE_ROWS_EVENT, which would make merging the two impossible. Example: ... ************************ ANNOTATE_ROWS_EVENT [51] ************************ 000000C7 | 54 1B 12 4B | time_when = 1259477844 000000CB | 33 | event_type = 51 000000CC | 64 00 00 00 | server_id = 100 000000D0 | 2C 00 00 00 | event_len = 44 000000D4 | F3 00 00 00 | log_pos = 000000F3 000000D8 | 00 00 | flags = <none> ------------------------ 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" 000000DE | 72 74 20 69 | 000000E2 | 6E 74 6F 20 | 000000E6 | 74 31 20 76 | 000000EA | 61 6C 75 65 | 000000EE | 73 20 28 31 | 000000F2 | 29 | ************************ TABLE_MAP_EVENT [19] ************************ 000000F3 | 54 1B 12 4B | time_when = 1259477844 000000F7 | 13 | event_type = 19 000000F8 | 64 00 00 00 | server_id = 100 000000FC | 29 00 00 00 | event_len = 41 00000100 | 1C 01 00 00 | log_pos = 0000011C 00000104 | 00 00 | flags = <none> ------------------------ ... ************************ WRITE_ROWS_EVENT [23] ************************ 0000011C | 54 1B 12 4B | time_when = 1259477844 00000120 | 17 | event_type = 23 00000121 | 64 00 00 00 | server_id = 100 00000125 | 22 00 00 00 | event_len = 34 00000129 | 3E 01 00 00 | log_pos = 0000013E 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F ------------------------ 0000012F | 0F 00 00 00 | table_id = 15 ... mysqlbinlog output ~~~~~~~~~~~~~~~~~~ Something like this: ... # at 199 # at 243 # at 284 #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values (1)` #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped to number 15 #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 flags: STMT_END_F BINLOG ' VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== '/*!*/; ### INSERT INTO test.t1 ### SET ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ ... When master sends Annotate rows events ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. Master always sends Annotate_rows events to mysqlbinlog (in remote case). 2. Master sends Annotate_rows events to a slave only if the slave has both log-slave-updates and binlog-annotate-row-events options set. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Alexi): Store in binlog text of statements that caused RBR events (47)
by worklog-noreply＠askmonty.org 02 Dec '09

02 Dec '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Store in binlog text of statements that caused RBR events CREATION DATE..: Sat, 15 Aug 2009, 23:48 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: Knielsen CATEGORY.......: Server-RawIdeaBin TASK ID........: 47 (http://askmonty.org/worklog/?tid=47) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Alexi - Wed, 02 Dec 2009, 13:31)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.17414 2009-12-02 11:31:59.000000000 +0000 +++ /tmp/wklog.47.new.17414 2009-12-02 11:31:59.000000000 +0000 @@ -104,3 +104,10 @@ ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ ... +When master sends Annotate rows events +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +1. Master always sends Annotate_rows events to mysqlbinlog (in + remote case). +2. Master sends Annotate_rows events to a slave only if the slave has + both log-slave-updates and binlog-annotate-row-events options set. + -=-=(Knielsen - Mon, 30 Nov 2009, 11:21)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.18210 2009-11-30 11:21:33.000000000 +0200 +++ /tmp/wklog.47.new.18210 2009-11-30 11:21:33.000000000 +0200 @@ -28,7 +28,14 @@ Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query -text in its data part. Example: +text in its data part. + +The numeric code for this event must be assigned carefully. It should be +coordinated with MySQL/Sun, otherwise we can get into a situation where MySQL +uses the same numeric code for one event that MariaDB uses for +ANNOTATE_ROWS_EVENT, which would make merging the two impossible. + +Example: ... ************************ -=-=(Alexi - Mon, 30 Nov 2009, 10:33)=-=- Low Level Design modified. --- /tmp/wklog.47.old.16188 2009-11-30 10:33:44.000000000 +0200 +++ /tmp/wklog.47.new.16188 2009-11-30 10:33:44.000000000 +0200 @@ -2,6 +2,7 @@ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. When sending events to a slave, master should simply skip Annotate_rows events (they are not needed for replication). + [ ??? Multi-master - currently not clear ] 2. When sending events to mysqlbinlog (remote case), master must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 13:00)=-=- Low Level Design modified. --- /tmp/wklog.47.old.32047 2009-11-29 13:00:21.000000000 +0200 +++ /tmp/wklog.47.new.32047 2009-11-29 13:00:21.000000000 +0200 @@ -1 +1,7 @@ +mysql_binlog_send() [sql/sql_repl.cc] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +1. When sending events to a slave, master should simply skip + Annotate_rows events (they are not needed for replication). +2. When sending events to mysqlbinlog (remote case), master + must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 09:50)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.24993 2009-11-29 07:50:36.000000000 +0000 +++ /tmp/wklog.47.new.24993 2009-11-29 07:50:36.000000000 +0000 @@ -4,3 +4,96 @@ > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). +New server option +~~~~~~~~~~~~~~~~~ + --binlog-annotate-row-events + +Setting this option makes RBR (row-) events in the binary log to be +preceded by Annotate rows events (see below). The corresponding +'binlog_annotate_row_events' system variable is dynamic and has both +global and session values. Default global value is OFF. + +Note. Session values are usefull to make it possible to annotate only + some selected statements: + + ... + SET SESSION binlog_annotate_row_events=ON; + ... statements to be annotated ... + SET SESSION binlog_annotate_row_events=OFF; + ... statements not to be annotated ... + +New binlog event type +~~~~~~~~~~~~~~~~~~~~~ + Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] + +Describes the query which caused the corresponding row event. In binary log, +precedes each Table_map_log_event. Contains empty post-header and the query +text in its data part. Example: + + ... + ************************ + ANNOTATE_ROWS_EVENT [51] + ************************ + 000000C7 | 54 1B 12 4B | time_when = 1259477844 + 000000CB | 33 | event_type = 51 + 000000CC | 64 00 00 00 | server_id = 100 + 000000D0 | 2C 00 00 00 | event_len = 44 + 000000D4 | F3 00 00 00 | log_pos = 000000F3 + 000000D8 | 00 00 | flags = <none> + ------------------------ + 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" + 000000DE | 72 74 20 69 | + 000000E2 | 6E 74 6F 20 | + 000000E6 | 74 31 20 76 | + 000000EA | 61 6C 75 65 | + 000000EE | 73 20 28 31 | + 000000F2 | 29 | + ************************ + TABLE_MAP_EVENT [19] + ************************ + 000000F3 | 54 1B 12 4B | time_when = 1259477844 + 000000F7 | 13 | event_type = 19 + 000000F8 | 64 00 00 00 | server_id = 100 + 000000FC | 29 00 00 00 | event_len = 41 + 00000100 | 1C 01 00 00 | log_pos = 0000011C + 00000104 | 00 00 | flags = <none> + ------------------------ + ... + ************************ + WRITE_ROWS_EVENT [23] + ************************ + 0000011C | 54 1B 12 4B | time_when = 1259477844 + 00000120 | 17 | event_type = 23 + 00000121 | 64 00 00 00 | server_id = 100 + 00000125 | 22 00 00 00 | event_len = 34 + 00000129 | 3E 01 00 00 | log_pos = 0000013E + 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F + ------------------------ + 0000012F | 0F 00 00 00 | table_id = 15 + ... + +mysqlbinlog output +~~~~~~~~~~~~~~~~~~ +Something like this: + + ... + # at 199 + # at 243 + # at 284 + #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values +(1)` + #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped +to number 15 + #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 +flags: STMT_END_F + + BINLOG ' + VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= + VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= + VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== + '/*!*/; + ### INSERT INTO test.t1 + ### SET + ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ + ... + -=-=(Knielsen - Fri, 27 Nov 2009, 13:30)=-=- Observers changed: Knielsen -=-=(Psergey - Sun, 16 Aug 2009, 11:08)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.12485 2009-08-16 11:08:33.000000000 +0300 +++ /tmp/wklog.47.new.12485 2009-08-16 11:08:33.000000000 +0300 @@ -1 +1,6 @@ +First suggestion: + +> I think for this we would actually need a new binlog event type +> (Comment_log_event?). Unless we want to log an empty statement Query_log_event +> containing only a comment (a bit of a hack). -=-=(Psergey - Sun, 16 Aug 2009, 00:02)=-=- Dependency created: 39 now depends on 47 DESCRIPTION: Store in binlog (and show in mysqlbinlog output) texts of statements that caused RBR events This is needed for (list from Monty): - Easier to understand why updates happened - Would make it easier to find out where in application things went wrong (as you can search for exact strings) - Allow one to filter things based on comments in the statement. The cost of this can be that the binlog will be approximately 2x in size (especially insert of big blob's would be a bit painful), so this should be an optional feature. HIGH-LEVEL SPECIFICATION: First suggestion: > I think for this we would actually need a new binlog event type > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). New server option ~~~~~~~~~~~~~~~~~ --binlog-annotate-row-events Setting this option makes RBR (row-) events in the binary log to be preceded by Annotate rows events (see below). The corresponding 'binlog_annotate_row_events' system variable is dynamic and has both global and session values. Default global value is OFF. Note. Session values are usefull to make it possible to annotate only some selected statements: ... SET SESSION binlog_annotate_row_events=ON; ... statements to be annotated ... SET SESSION binlog_annotate_row_events=OFF; ... statements not to be annotated ... New binlog event type ~~~~~~~~~~~~~~~~~~~~~ Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query text in its data part. The numeric code for this event must be assigned carefully. It should be coordinated with MySQL/Sun, otherwise we can get into a situation where MySQL uses the same numeric code for one event that MariaDB uses for ANNOTATE_ROWS_EVENT, which would make merging the two impossible. Example: ... ************************ ANNOTATE_ROWS_EVENT [51] ************************ 000000C7 | 54 1B 12 4B | time_when = 1259477844 000000CB | 33 | event_type = 51 000000CC | 64 00 00 00 | server_id = 100 000000D0 | 2C 00 00 00 | event_len = 44 000000D4 | F3 00 00 00 | log_pos = 000000F3 000000D8 | 00 00 | flags = <none> ------------------------ 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" 000000DE | 72 74 20 69 | 000000E2 | 6E 74 6F 20 | 000000E6 | 74 31 20 76 | 000000EA | 61 6C 75 65 | 000000EE | 73 20 28 31 | 000000F2 | 29 | ************************ TABLE_MAP_EVENT [19] ************************ 000000F3 | 54 1B 12 4B | time_when = 1259477844 000000F7 | 13 | event_type = 19 000000F8 | 64 00 00 00 | server_id = 100 000000FC | 29 00 00 00 | event_len = 41 00000100 | 1C 01 00 00 | log_pos = 0000011C 00000104 | 00 00 | flags = <none> ------------------------ ... ************************ WRITE_ROWS_EVENT [23] ************************ 0000011C | 54 1B 12 4B | time_when = 1259477844 00000120 | 17 | event_type = 23 00000121 | 64 00 00 00 | server_id = 100 00000125 | 22 00 00 00 | event_len = 34 00000129 | 3E 01 00 00 | log_pos = 0000013E 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F ------------------------ 0000012F | 0F 00 00 00 | table_id = 15 ... mysqlbinlog output ~~~~~~~~~~~~~~~~~~ Something like this: ... # at 199 # at 243 # at 284 #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values (1)` #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped to number 15 #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 flags: STMT_END_F BINLOG ' VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== '/*!*/; ### INSERT INTO test.t1 ### SET ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ ... When master sends Annotate rows events ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. Master always sends Annotate_rows events to mysqlbinlog (in remote case). 2. Master sends Annotate_rows events to a slave only if the slave has both log-slave-updates and binlog-annotate-row-events options set. LOW-LEVEL DESIGN: mysql_binlog_send() [sql/sql_repl.cc] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. When sending events to a slave, master should simply skip Annotate_rows events (they are not needed for replication). [ ??? Multi-master - currently not clear ] 2. When sending events to mysqlbinlog (remote case), master must send Annotate_rows events as well. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Alexi): Store in binlog text of statements that caused RBR events (47)
by worklog-noreply＠askmonty.org 02 Dec '09

02 Dec '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Store in binlog text of statements that caused RBR events CREATION DATE..: Sat, 15 Aug 2009, 23:48 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: Knielsen CATEGORY.......: Server-RawIdeaBin TASK ID........: 47 (http://askmonty.org/worklog/?tid=47) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Alexi - Wed, 02 Dec 2009, 13:31)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.17414 2009-12-02 11:31:59.000000000 +0000 +++ /tmp/wklog.47.new.17414 2009-12-02 11:31:59.000000000 +0000 @@ -104,3 +104,10 @@ ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ ... +When master sends Annotate rows events +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +1. Master always sends Annotate_rows events to mysqlbinlog (in + remote case). +2. Master sends Annotate_rows events to a slave only if the slave has + both log-slave-updates and binlog-annotate-row-events options set. + -=-=(Knielsen - Mon, 30 Nov 2009, 11:21)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.18210 2009-11-30 11:21:33.000000000 +0200 +++ /tmp/wklog.47.new.18210 2009-11-30 11:21:33.000000000 +0200 @@ -28,7 +28,14 @@ Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query -text in its data part. Example: +text in its data part. + +The numeric code for this event must be assigned carefully. It should be +coordinated with MySQL/Sun, otherwise we can get into a situation where MySQL +uses the same numeric code for one event that MariaDB uses for +ANNOTATE_ROWS_EVENT, which would make merging the two impossible. + +Example: ... ************************ -=-=(Alexi - Mon, 30 Nov 2009, 10:33)=-=- Low Level Design modified. --- /tmp/wklog.47.old.16188 2009-11-30 10:33:44.000000000 +0200 +++ /tmp/wklog.47.new.16188 2009-11-30 10:33:44.000000000 +0200 @@ -2,6 +2,7 @@ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. When sending events to a slave, master should simply skip Annotate_rows events (they are not needed for replication). + [ ??? Multi-master - currently not clear ] 2. When sending events to mysqlbinlog (remote case), master must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 13:00)=-=- Low Level Design modified. --- /tmp/wklog.47.old.32047 2009-11-29 13:00:21.000000000 +0200 +++ /tmp/wklog.47.new.32047 2009-11-29 13:00:21.000000000 +0200 @@ -1 +1,7 @@ +mysql_binlog_send() [sql/sql_repl.cc] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +1. When sending events to a slave, master should simply skip + Annotate_rows events (they are not needed for replication). +2. When sending events to mysqlbinlog (remote case), master + must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 09:50)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.24993 2009-11-29 07:50:36.000000000 +0000 +++ /tmp/wklog.47.new.24993 2009-11-29 07:50:36.000000000 +0000 @@ -4,3 +4,96 @@ > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). +New server option +~~~~~~~~~~~~~~~~~ + --binlog-annotate-row-events + +Setting this option makes RBR (row-) events in the binary log to be +preceded by Annotate rows events (see below). The corresponding +'binlog_annotate_row_events' system variable is dynamic and has both +global and session values. Default global value is OFF. + +Note. Session values are usefull to make it possible to annotate only + some selected statements: + + ... + SET SESSION binlog_annotate_row_events=ON; + ... statements to be annotated ... + SET SESSION binlog_annotate_row_events=OFF; + ... statements not to be annotated ... + +New binlog event type +~~~~~~~~~~~~~~~~~~~~~ + Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] + +Describes the query which caused the corresponding row event. In binary log, +precedes each Table_map_log_event. Contains empty post-header and the query +text in its data part. Example: + + ... + ************************ + ANNOTATE_ROWS_EVENT [51] + ************************ + 000000C7 | 54 1B 12 4B | time_when = 1259477844 + 000000CB | 33 | event_type = 51 + 000000CC | 64 00 00 00 | server_id = 100 + 000000D0 | 2C 00 00 00 | event_len = 44 + 000000D4 | F3 00 00 00 | log_pos = 000000F3 + 000000D8 | 00 00 | flags = <none> + ------------------------ + 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" + 000000DE | 72 74 20 69 | + 000000E2 | 6E 74 6F 20 | + 000000E6 | 74 31 20 76 | + 000000EA | 61 6C 75 65 | + 000000EE | 73 20 28 31 | + 000000F2 | 29 | + ************************ + TABLE_MAP_EVENT [19] + ************************ + 000000F3 | 54 1B 12 4B | time_when = 1259477844 + 000000F7 | 13 | event_type = 19 + 000000F8 | 64 00 00 00 | server_id = 100 + 000000FC | 29 00 00 00 | event_len = 41 + 00000100 | 1C 01 00 00 | log_pos = 0000011C + 00000104 | 00 00 | flags = <none> + ------------------------ + ... + ************************ + WRITE_ROWS_EVENT [23] + ************************ + 0000011C | 54 1B 12 4B | time_when = 1259477844 + 00000120 | 17 | event_type = 23 + 00000121 | 64 00 00 00 | server_id = 100 + 00000125 | 22 00 00 00 | event_len = 34 + 00000129 | 3E 01 00 00 | log_pos = 0000013E + 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F + ------------------------ + 0000012F | 0F 00 00 00 | table_id = 15 + ... + +mysqlbinlog output +~~~~~~~~~~~~~~~~~~ +Something like this: + + ... + # at 199 + # at 243 + # at 284 + #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values +(1)` + #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped +to number 15 + #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 +flags: STMT_END_F + + BINLOG ' + VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= + VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= + VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== + '/*!*/; + ### INSERT INTO test.t1 + ### SET + ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ + ... + -=-=(Knielsen - Fri, 27 Nov 2009, 13:30)=-=- Observers changed: Knielsen -=-=(Psergey - Sun, 16 Aug 2009, 11:08)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.12485 2009-08-16 11:08:33.000000000 +0300 +++ /tmp/wklog.47.new.12485 2009-08-16 11:08:33.000000000 +0300 @@ -1 +1,6 @@ +First suggestion: + +> I think for this we would actually need a new binlog event type +> (Comment_log_event?). Unless we want to log an empty statement Query_log_event +> containing only a comment (a bit of a hack). -=-=(Psergey - Sun, 16 Aug 2009, 00:02)=-=- Dependency created: 39 now depends on 47 DESCRIPTION: Store in binlog (and show in mysqlbinlog output) texts of statements that caused RBR events This is needed for (list from Monty): - Easier to understand why updates happened - Would make it easier to find out where in application things went wrong (as you can search for exact strings) - Allow one to filter things based on comments in the statement. The cost of this can be that the binlog will be approximately 2x in size (especially insert of big blob's would be a bit painful), so this should be an optional feature. HIGH-LEVEL SPECIFICATION: First suggestion: > I think for this we would actually need a new binlog event type > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). New server option ~~~~~~~~~~~~~~~~~ --binlog-annotate-row-events Setting this option makes RBR (row-) events in the binary log to be preceded by Annotate rows events (see below). The corresponding 'binlog_annotate_row_events' system variable is dynamic and has both global and session values. Default global value is OFF. Note. Session values are usefull to make it possible to annotate only some selected statements: ... SET SESSION binlog_annotate_row_events=ON; ... statements to be annotated ... SET SESSION binlog_annotate_row_events=OFF; ... statements not to be annotated ... New binlog event type ~~~~~~~~~~~~~~~~~~~~~ Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query text in its data part. The numeric code for this event must be assigned carefully. It should be coordinated with MySQL/Sun, otherwise we can get into a situation where MySQL uses the same numeric code for one event that MariaDB uses for ANNOTATE_ROWS_EVENT, which would make merging the two impossible. Example: ... ************************ ANNOTATE_ROWS_EVENT [51] ************************ 000000C7 | 54 1B 12 4B | time_when = 1259477844 000000CB | 33 | event_type = 51 000000CC | 64 00 00 00 | server_id = 100 000000D0 | 2C 00 00 00 | event_len = 44 000000D4 | F3 00 00 00 | log_pos = 000000F3 000000D8 | 00 00 | flags = <none> ------------------------ 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" 000000DE | 72 74 20 69 | 000000E2 | 6E 74 6F 20 | 000000E6 | 74 31 20 76 | 000000EA | 61 6C 75 65 | 000000EE | 73 20 28 31 | 000000F2 | 29 | ************************ TABLE_MAP_EVENT [19] ************************ 000000F3 | 54 1B 12 4B | time_when = 1259477844 000000F7 | 13 | event_type = 19 000000F8 | 64 00 00 00 | server_id = 100 000000FC | 29 00 00 00 | event_len = 41 00000100 | 1C 01 00 00 | log_pos = 0000011C 00000104 | 00 00 | flags = <none> ------------------------ ... ************************ WRITE_ROWS_EVENT [23] ************************ 0000011C | 54 1B 12 4B | time_when = 1259477844 00000120 | 17 | event_type = 23 00000121 | 64 00 00 00 | server_id = 100 00000125 | 22 00 00 00 | event_len = 34 00000129 | 3E 01 00 00 | log_pos = 0000013E 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F ------------------------ 0000012F | 0F 00 00 00 | table_id = 15 ... mysqlbinlog output ~~~~~~~~~~~~~~~~~~ Something like this: ... # at 199 # at 243 # at 284 #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values (1)` #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped to number 15 #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 flags: STMT_END_F BINLOG ' VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== '/*!*/; ### INSERT INTO test.t1 ### SET ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ ... When master sends Annotate rows events ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. Master always sends Annotate_rows events to mysqlbinlog (in remote case). 2. Master sends Annotate_rows events to a slave only if the slave has both log-slave-updates and binlog-annotate-row-events options set. LOW-LEVEL DESIGN: mysql_binlog_send() [sql/sql_repl.cc] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. When sending events to a slave, master should simply skip Annotate_rows events (they are not needed for replication). [ ??? Multi-master - currently not clear ] 2. When sending events to mysqlbinlog (remote case), master must send Annotate_rows events as well. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Alexi): Store in binlog text of statements that caused RBR events (47)
by worklog-noreply＠askmonty.org 02 Dec '09

02 Dec '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Store in binlog text of statements that caused RBR events CREATION DATE..: Sat, 15 Aug 2009, 23:48 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: Knielsen CATEGORY.......: Server-RawIdeaBin TASK ID........: 47 (http://askmonty.org/worklog/?tid=47) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Alexi - Wed, 02 Dec 2009, 13:31)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.17414 2009-12-02 11:31:59.000000000 +0000 +++ /tmp/wklog.47.new.17414 2009-12-02 11:31:59.000000000 +0000 @@ -104,3 +104,10 @@ ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ ... +When master sends Annotate rows events +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +1. Master always sends Annotate_rows events to mysqlbinlog (in + remote case). +2. Master sends Annotate_rows events to a slave only if the slave has + both log-slave-updates and binlog-annotate-row-events options set. + -=-=(Knielsen - Mon, 30 Nov 2009, 11:21)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.18210 2009-11-30 11:21:33.000000000 +0200 +++ /tmp/wklog.47.new.18210 2009-11-30 11:21:33.000000000 +0200 @@ -28,7 +28,14 @@ Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query -text in its data part. Example: +text in its data part. + +The numeric code for this event must be assigned carefully. It should be +coordinated with MySQL/Sun, otherwise we can get into a situation where MySQL +uses the same numeric code for one event that MariaDB uses for +ANNOTATE_ROWS_EVENT, which would make merging the two impossible. + +Example: ... ************************ -=-=(Alexi - Mon, 30 Nov 2009, 10:33)=-=- Low Level Design modified. --- /tmp/wklog.47.old.16188 2009-11-30 10:33:44.000000000 +0200 +++ /tmp/wklog.47.new.16188 2009-11-30 10:33:44.000000000 +0200 @@ -2,6 +2,7 @@ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. When sending events to a slave, master should simply skip Annotate_rows events (they are not needed for replication). + [ ??? Multi-master - currently not clear ] 2. When sending events to mysqlbinlog (remote case), master must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 13:00)=-=- Low Level Design modified. --- /tmp/wklog.47.old.32047 2009-11-29 13:00:21.000000000 +0200 +++ /tmp/wklog.47.new.32047 2009-11-29 13:00:21.000000000 +0200 @@ -1 +1,7 @@ +mysql_binlog_send() [sql/sql_repl.cc] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +1. When sending events to a slave, master should simply skip + Annotate_rows events (they are not needed for replication). +2. When sending events to mysqlbinlog (remote case), master + must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 09:50)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.24993 2009-11-29 07:50:36.000000000 +0000 +++ /tmp/wklog.47.new.24993 2009-11-29 07:50:36.000000000 +0000 @@ -4,3 +4,96 @@ > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). +New server option +~~~~~~~~~~~~~~~~~ + --binlog-annotate-row-events + +Setting this option makes RBR (row-) events in the binary log to be +preceded by Annotate rows events (see below). The corresponding +'binlog_annotate_row_events' system variable is dynamic and has both +global and session values. Default global value is OFF. + +Note. Session values are usefull to make it possible to annotate only + some selected statements: + + ... + SET SESSION binlog_annotate_row_events=ON; + ... statements to be annotated ... + SET SESSION binlog_annotate_row_events=OFF; + ... statements not to be annotated ... + +New binlog event type +~~~~~~~~~~~~~~~~~~~~~ + Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] + +Describes the query which caused the corresponding row event. In binary log, +precedes each Table_map_log_event. Contains empty post-header and the query +text in its data part. Example: + + ... + ************************ + ANNOTATE_ROWS_EVENT [51] + ************************ + 000000C7 | 54 1B 12 4B | time_when = 1259477844 + 000000CB | 33 | event_type = 51 + 000000CC | 64 00 00 00 | server_id = 100 + 000000D0 | 2C 00 00 00 | event_len = 44 + 000000D4 | F3 00 00 00 | log_pos = 000000F3 + 000000D8 | 00 00 | flags = <none> + ------------------------ + 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" + 000000DE | 72 74 20 69 | + 000000E2 | 6E 74 6F 20 | + 000000E6 | 74 31 20 76 | + 000000EA | 61 6C 75 65 | + 000000EE | 73 20 28 31 | + 000000F2 | 29 | + ************************ + TABLE_MAP_EVENT [19] + ************************ + 000000F3 | 54 1B 12 4B | time_when = 1259477844 + 000000F7 | 13 | event_type = 19 + 000000F8 | 64 00 00 00 | server_id = 100 + 000000FC | 29 00 00 00 | event_len = 41 + 00000100 | 1C 01 00 00 | log_pos = 0000011C + 00000104 | 00 00 | flags = <none> + ------------------------ + ... + ************************ + WRITE_ROWS_EVENT [23] + ************************ + 0000011C | 54 1B 12 4B | time_when = 1259477844 + 00000120 | 17 | event_type = 23 + 00000121 | 64 00 00 00 | server_id = 100 + 00000125 | 22 00 00 00 | event_len = 34 + 00000129 | 3E 01 00 00 | log_pos = 0000013E + 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F + ------------------------ + 0000012F | 0F 00 00 00 | table_id = 15 + ... + +mysqlbinlog output +~~~~~~~~~~~~~~~~~~ +Something like this: + + ... + # at 199 + # at 243 + # at 284 + #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values +(1)` + #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped +to number 15 + #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 +flags: STMT_END_F + + BINLOG ' + VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= + VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= + VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== + '/*!*/; + ### INSERT INTO test.t1 + ### SET + ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ + ... + -=-=(Knielsen - Fri, 27 Nov 2009, 13:30)=-=- Observers changed: Knielsen -=-=(Psergey - Sun, 16 Aug 2009, 11:08)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.12485 2009-08-16 11:08:33.000000000 +0300 +++ /tmp/wklog.47.new.12485 2009-08-16 11:08:33.000000000 +0300 @@ -1 +1,6 @@ +First suggestion: + +> I think for this we would actually need a new binlog event type +> (Comment_log_event?). Unless we want to log an empty statement Query_log_event +> containing only a comment (a bit of a hack). -=-=(Psergey - Sun, 16 Aug 2009, 00:02)=-=- Dependency created: 39 now depends on 47 DESCRIPTION: Store in binlog (and show in mysqlbinlog output) texts of statements that caused RBR events This is needed for (list from Monty): - Easier to understand why updates happened - Would make it easier to find out where in application things went wrong (as you can search for exact strings) - Allow one to filter things based on comments in the statement. The cost of this can be that the binlog will be approximately 2x in size (especially insert of big blob's would be a bit painful), so this should be an optional feature. HIGH-LEVEL SPECIFICATION: First suggestion: > I think for this we would actually need a new binlog event type > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). New server option ~~~~~~~~~~~~~~~~~ --binlog-annotate-row-events Setting this option makes RBR (row-) events in the binary log to be preceded by Annotate rows events (see below). The corresponding 'binlog_annotate_row_events' system variable is dynamic and has both global and session values. Default global value is OFF. Note. Session values are usefull to make it possible to annotate only some selected statements: ... SET SESSION binlog_annotate_row_events=ON; ... statements to be annotated ... SET SESSION binlog_annotate_row_events=OFF; ... statements not to be annotated ... New binlog event type ~~~~~~~~~~~~~~~~~~~~~ Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query text in its data part. The numeric code for this event must be assigned carefully. It should be coordinated with MySQL/Sun, otherwise we can get into a situation where MySQL uses the same numeric code for one event that MariaDB uses for ANNOTATE_ROWS_EVENT, which would make merging the two impossible. Example: ... ************************ ANNOTATE_ROWS_EVENT [51] ************************ 000000C7 | 54 1B 12 4B | time_when = 1259477844 000000CB | 33 | event_type = 51 000000CC | 64 00 00 00 | server_id = 100 000000D0 | 2C 00 00 00 | event_len = 44 000000D4 | F3 00 00 00 | log_pos = 000000F3 000000D8 | 00 00 | flags = <none> ------------------------ 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" 000000DE | 72 74 20 69 | 000000E2 | 6E 74 6F 20 | 000000E6 | 74 31 20 76 | 000000EA | 61 6C 75 65 | 000000EE | 73 20 28 31 | 000000F2 | 29 | ************************ TABLE_MAP_EVENT [19] ************************ 000000F3 | 54 1B 12 4B | time_when = 1259477844 000000F7 | 13 | event_type = 19 000000F8 | 64 00 00 00 | server_id = 100 000000FC | 29 00 00 00 | event_len = 41 00000100 | 1C 01 00 00 | log_pos = 0000011C 00000104 | 00 00 | flags = <none> ------------------------ ... ************************ WRITE_ROWS_EVENT [23] ************************ 0000011C | 54 1B 12 4B | time_when = 1259477844 00000120 | 17 | event_type = 23 00000121 | 64 00 00 00 | server_id = 100 00000125 | 22 00 00 00 | event_len = 34 00000129 | 3E 01 00 00 | log_pos = 0000013E 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F ------------------------ 0000012F | 0F 00 00 00 | table_id = 15 ... mysqlbinlog output ~~~~~~~~~~~~~~~~~~ Something like this: ... # at 199 # at 243 # at 284 #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values (1)` #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped to number 15 #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 flags: STMT_END_F BINLOG ' VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== '/*!*/; ### INSERT INTO test.t1 ### SET ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ ... When master sends Annotate rows events ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. Master always sends Annotate_rows events to mysqlbinlog (in remote case). 2. Master sends Annotate_rows events to a slave only if the slave has both log-slave-updates and binlog-annotate-row-events options set. LOW-LEVEL DESIGN: mysql_binlog_send() [sql/sql_repl.cc] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. When sending events to a slave, master should simply skip Annotate_rows events (they are not needed for replication). [ ??? Multi-master - currently not clear ] 2. When sending events to mysqlbinlog (remote case), master must send Annotate_rows events as well. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] [Branch ~maria-captains/maria/5.1] Rev 2774: Fixed link problem when compiling without uca
by noreply＠launchpad.net 01 Dec '09

01 Dec '09

------------------------------------------------------------ revno: 2774 committer: Michael Widenius <monty(a)askmonty.org> branch nick: mariadb-5.1 timestamp: Tue 2009-12-01 18:09:02 +0200 message: Fixed link problem when compiling without uca modified: cmd-line-utils/readline/config_readline.h strings/ctype-uca.c -- lp:maria https://code.launchpad.net/~maria-captains/maria/5.1 Your team Maria developers is subscribed to branch lp:maria. To unsubscribe from this branch go to https://code.launchpad.net/~maria-captains/maria/5.1/+edit-subscription.

1 0

[Maria-developers] [Branch ~maria-captains/maria/5.1] Rev 2773: Automerge Daniel's README fixes.
by noreply＠launchpad.net 01 Dec '09

01 Dec '09

Merge authors: Daniel Bartholomew (dbart) Related merge proposals: https://code.launchpad.net/~dbart/maria/maria-5.1-docmods/+merge/14679 proposed by: Daniel Bartholomew (dbart) ------------------------------------------------------------ revno: 2773 [merge] committer: knielsen(a)knielsen-hq.org branch nick: work-5.1 timestamp: Tue 2009-12-01 08:42:58 +0100 message: Automerge Daniel's README fixes. modified: README -- lp:maria https://code.launchpad.net/~maria-captains/maria/5.1 Your team Maria developers is subscribed to branch lp:maria. To unsubscribe from this branch go to https://code.launchpad.net/~maria-captains/maria/5.1/+edit-subscription.

1 0

[Maria-developers] [Branch ~maria-captains/maria/5.1] Rev 2772: Merge free documentation from MySQL 5.1.41 source tarball into MariaDB.
by noreply＠launchpad.net 01 Dec '09

01 Dec '09

Merge authors: Kristian Nielsen (knielsen) ------------------------------------------------------------ revno: 2772 [merge] committer: knielsen(a)knielsen-hq.org branch nick: work-5.1 timestamp: Tue 2009-12-01 08:38:15 +0100 message: Merge free documentation from MySQL 5.1.41 source tarball into MariaDB. modified: Docs/INSTALL-BINARY INSTALL-SOURCE INSTALL-WIN-SOURCE man/comp_err.1 man/innochecksum.1 man/make_win_bin_dist.1 man/msql2mysql.1 man/my_print_defaults.1 man/myisam_ftdump.1 man/myisamchk.1 man/myisamlog.1 man/myisampack.1 man/mysql-stress-test.pl.1 man/mysql-test-run.pl.1 man/mysql.1 man/mysql.server.1 man/mysql_client_test.1 man/mysql_config.1 man/mysql_convert_table_format.1 man/mysql_find_rows.1 man/mysql_fix_extensions.1 man/mysql_fix_privilege_tables.1 man/mysql_install_db.1 man/mysql_secure_installation.1 man/mysql_setpermission.1 man/mysql_tzinfo_to_sql.1 man/mysql_upgrade.1 man/mysql_waitpid.1 man/mysql_zap.1 man/mysqlaccess.1 man/mysqladmin.1 man/mysqlbinlog.1 man/mysqlbug.1 man/mysqlcheck.1 man/mysqld.8 man/mysqld_multi.1 man/mysqld_safe.1 man/mysqldump.1 man/mysqldumpslow.1 man/mysqlhotcopy.1 man/mysqlimport.1 man/mysqlmanager.8 man/mysqlshow.1 man/mysqlslap.1 man/mysqltest.1 man/ndbd.8 man/ndbd_redo_log_reader.1 man/ndbmtd.8 man/perror.1 man/replace.1 man/resolve_stack_dump.1 man/resolveip.1 scripts/fill_help_tables.sql support-files/MacOSX/ReadMe.txt The size of the diff (15484 lines) is larger than your specified limit of 5000 lines -- lp:maria https://code.launchpad.net/~maria-captains/maria/5.1 Your team Maria developers is subscribed to branch lp:maria. To unsubscribe from this branch go to https://code.launchpad.net/~maria-captains/maria/5.1/+edit-subscription.

1 0

[Maria-developers] [Branch ~maria-captains/maria/5.1] Rev 2771: Merge MySQL 5.1.41 into MariaDB trunk, including a number of after-merge fixes.
by noreply＠launchpad.net 01 Dec '09

01 Dec '09

Merge authors: <Dao-Gang.Qu(a)sun.com> <Li-Bing.Song(a)sun.com> Alexander Barkov <bar(a)mysql.com> Alexander Nozdrin <alik(a)sun.com> Alexey Botchkov <holyfoot(a)mysql.com>... ------------------------------------------------------------ revno: 2771 [merge] committer: knielsen(a)knielsen-hq.org branch nick: work-5.1 timestamp: Mon 2009-11-30 22:37:27 +0100 message: Merge MySQL 5.1.41 into MariaDB trunk, including a number of after-merge fixes. Also merge charset patch. removed: mysql-test/include/have_dynamic_loading.inc mysys/mf_strip.c storage/innodb_plugin/README storage/innodb_plugin/handler/handler0vars.h storage/innodb_plugin/handler/win_delay_loader.cc storage/innodb_plugin/win-plugin/ storage/innodb_plugin/win-plugin/README storage/innodb_plugin/win-plugin/win-plugin.diff added: mysql-test/extra/binlog_tests/binlog_failure_mixing_engines.test mysql-test/extra/rpl_tests/rpl_auto_increment_insert_view.test mysql-test/extra/rpl_tests/rpl_auto_increment_invoke_trigger.test mysql-test/extra/rpl_tests/rpl_autoinc_func_invokes_trigger.test mysql-test/include/have_case_insensitive_fs.inc mysql-test/include/have_debug_sync.inc mysql-test/include/have_dynamic_loading.inc mysql-test/include/have_mysql_upgrade.inc mysql-test/include/have_not_innodb_plugin.inc mysql-test/include/not_windows_embedded.inc mysql-test/lib/v1/incompatible.tests mysql-test/r/bug46760.result mysql-test/r/case_insensitive_fs.require mysql-test/r/debug_sync.result mysql-test/r/grant_lowercase_fs.result mysql-test/r/have_debug_sync.require mysql-test/r/innodb_bug44369.result mysql-test/r/innodb_bug46000.result mysql-test/r/innodb_bug47777.result mysql-test/r/locale.result mysql-test/r/lowercase_mixed_tmpdir_innodb.result mysql-test/r/not_true.require mysql-test/r/partition_innodb_builtin.result mysql-test/r/partition_innodb_plugin.result mysql-test/r/partition_open_files_limit.result mysql-test/r/sp-bugs.result mysql-test/r/subselect4.result mysql-test/std_data/binlog_transaction.000001 mysql-test/std_data/latin1.xml mysql-test/suite/binlog/r/binlog_delete_and_flush_index.result mysql-test/suite/binlog/r/binlog_mixed_failure_mixing_engines.result mysql-test/suite/binlog/r/binlog_row_failure_mixing_engines.result mysql-test/suite/binlog/r/binlog_row_mysqlbinlog_verbose.result mysql-test/suite/binlog/r/binlog_stm_do_db.result mysql-test/suite/binlog/std_data/update-full-row.binlog mysql-test/suite/binlog/std_data/update-partial-row.binlog mysql-test/suite/binlog/std_data/write-full-row.binlog mysql-test/suite/binlog/std_data/write-partial-row.binlog mysql-test/suite/binlog/t/binlog_delete_and_flush_index.test mysql-test/suite/binlog/t/binlog_mixed_failure_mixing_engines.test mysql-test/suite/binlog/t/binlog_row_failure_mixing_engines.test mysql-test/suite/binlog/t/binlog_row_mysqlbinlog_verbose.test mysql-test/suite/binlog/t/binlog_stm_do_db-master.opt mysql-test/suite/binlog/t/binlog_stm_do_db.test mysql-test/suite/federated/federated_debug-master.opt mysql-test/suite/federated/federated_debug.result mysql-test/suite/federated/federated_debug.test mysql-test/suite/innodb/r/innodb-consistent.result mysql-test/suite/innodb/r/innodb_bug44571.result mysql-test/suite/innodb/t/innodb-consistent-master.opt mysql-test/suite/innodb/t/innodb-consistent.test mysql-test/suite/innodb/t/innodb_bug44571.test mysql-test/suite/rpl/r/rpl_auto_increment_update_failure.result mysql-test/suite/rpl/r/rpl_mysql_upgrade.result mysql-test/suite/rpl/r/rpl_row_disabled_slave_key.result mysql-test/suite/rpl/t/rpl_auto_increment_update_failure.test mysql-test/suite/rpl/t/rpl_get_master_version_and_clock-slave.opt mysql-test/suite/rpl/t/rpl_mysql_upgrade.test mysql-test/suite/rpl/t/rpl_row_disabled_slave_key.test mysql-test/t/bug46760-master.opt mysql-test/t/bug46760.test mysql-test/t/debug_sync.test mysql-test/t/grant_lowercase_fs.test mysql-test/t/innodb_bug44369.test mysql-test/t/innodb_bug46000.test mysql-test/t/innodb_bug47777.test mysql-test/t/locale.test mysql-test/t/lowercase_mixed_tmpdir_innodb-master.opt mysql-test/t/lowercase_mixed_tmpdir_innodb-master.sh mysql-test/t/lowercase_mixed_tmpdir_innodb.test mysql-test/t/partition_innodb_builtin.test mysql-test/t/partition_innodb_plugin.test mysql-test/t/partition_open_files_limit-master.opt mysql-test/t/partition_open_files_limit.test mysql-test/t/sp-bugs.test mysql-test/t/status-master.opt mysql-test/t/subselect4.test sql/debug_sync.cc sql/debug_sync.h storage/innodb_plugin/mysql-test/innodb-consistent-master.opt storage/innodb_plugin/mysql-test/innodb-consistent.result storage/innodb_plugin/mysql-test/innodb-consistent.test storage/innodb_plugin/mysql-test/innodb_bug44369.result storage/innodb_plugin/mysql-test/innodb_bug44369.test storage/innodb_plugin/mysql-test/innodb_bug44571.result storage/innodb_plugin/mysql-test/innodb_bug44571.test storage/innodb_plugin/mysql-test/innodb_bug46000.result storage/innodb_plugin/mysql-test/innodb_bug46000.test storage/innodb_plugin/revert_gen.sh storage/innodb_plugin/scripts/export.sh storage/innodb_plugin/ut/ut0auxconf_have_gcc_atomics.c renamed: mysql-test/r/bug40113.result => mysql-test/r/innodb_lock_wait_timeout_1.result mysql-test/t/bug40113-master.opt => mysql-test/t/innodb_lock_wait_timeout_1-master.opt mysql-test/t/bug40113.test => mysql-test/t/innodb_lock_wait_timeout_1.test modified: .bzrignore CMakeLists.txt Makefile.am client/mysql.cc client/mysql_upgrade.c client/mysqlbinlog.cc client/mysqlcheck.c client/mysqlimport.c client/mysqlslap.c client/mysqltest.cc cmd-line-utils/readline/config_readline.h cmd-line-utils/readline/display.c configure.in extra/yassl/include/yassl_int.hpp extra/yassl/taocrypt/src/random.cpp include/m_ctype.h include/my_dbug.h include/my_sys.h include/myisamchk.h include/mysql.h include/mysql.h.pp include/violite.h libmysql/libmysql.c libmysql/libmysql.def libmysqld/CMakeLists.txt libmysqld/Makefile.am libmysqld/lib_sql.cc libmysqld/libmysqld.c libmysqld/libmysqld.def mysql-test/collections/README.experimental mysql-test/collections/default.experimental mysql-test/extra/binlog_tests/binlog.test mysql-test/extra/binlog_tests/drop_temp_table.test mysql-test/extra/rpl_tests/rpl_auto_increment.test mysql-test/extra/rpl_tests/rpl_extraSlave_Col.test mysql-test/extra/rpl_tests/rpl_failed_optimize.test mysql-test/extra/rpl_tests/rpl_loaddata.test mysql-test/extra/rpl_tests/rpl_row_sp006.test mysql-test/extra/rpl_tests/rpl_stm_000001.test mysql-test/include/check-warnings.test mysql-test/include/concurrent.inc mysql-test/include/have_example_plugin.inc mysql-test/include/have_simple_parser.inc mysql-test/include/have_udf.inc mysql-test/include/mix1.inc mysql-test/include/mtr_warnings.sql mysql-test/lib/My/ConfigFactory.pm mysql-test/lib/My/Platform.pm mysql-test/lib/My/SafeProcess/safe_kill_win.cc mysql-test/lib/My/SafeProcess/safe_process_win.cc mysql-test/lib/mtr_cases.pm mysql-test/lib/mtr_report.pm mysql-test/lib/v1/mtr_cases.pl mysql-test/mysql-stress-test.pl mysql-test/mysql-test-run.pl mysql-test/r/almost_full.result mysql-test/r/alter_table.result mysql-test/r/analyse.result mysql-test/r/archive.result mysql-test/r/bug46080.result mysql-test/r/create.result mysql-test/r/ctype_ldml.result mysql-test/r/ctype_uca.result mysql-test/r/delete.result mysql-test/r/distinct.result mysql-test/r/explain.result mysql-test/r/func_group.result mysql-test/r/func_in.result mysql-test/r/func_str.result mysql-test/r/gis-rtree.result mysql-test/r/gis.result mysql-test/r/grant.result mysql-test/r/grant3.result mysql-test/r/group_min_max.result mysql-test/r/information_schema_db.result mysql-test/r/innodb-autoinc.result mysql-test/r/innodb-index.result mysql-test/r/innodb_file_format.result mysql-test/r/innodb_mysql.result mysql-test/r/insert_select.result mysql-test/r/join.result mysql-test/r/lowercase_fs_off.result mysql-test/r/lowercase_table3.result mysql-test/r/myisam.result mysql-test/r/myisam_crash_before_flush_keys.result mysql-test/r/mysqlbinlog.result mysql-test/r/mysqltest.result mysql-test/r/olap.result mysql-test/r/order_by.result mysql-test/r/partition.result mysql-test/r/partition_csv.result mysql-test/r/partition_innodb.result mysql-test/r/partition_pruning.result mysql-test/r/ps_grant.result mysql-test/r/query_cache.result mysql-test/r/range.result mysql-test/r/select.result mysql-test/r/sp-error.result mysql-test/r/sp.result mysql-test/r/subselect.result mysql-test/r/subselect3.result mysql-test/r/system_mysql_db.result mysql-test/r/trigger_notembedded.result mysql-test/r/type_bit.result mysql-test/r/type_newdecimal.result mysql-test/r/udf.result mysql-test/r/update.result mysql-test/r/upgrade.result mysql-test/r/view_grant.result mysql-test/r/warnings.result mysql-test/r/windows.result mysql-test/r/xa.result mysql-test/std_data/Index.xml mysql-test/suite/binlog/r/binlog_killed_simulate.result mysql-test/suite/binlog/r/binlog_row_binlog.result mysql-test/suite/binlog/r/binlog_row_drop_tmp_tbl.result mysql-test/suite/binlog/r/binlog_row_mix_innodb_myisam.result mysql-test/suite/binlog/r/binlog_stm_binlog.result mysql-test/suite/binlog/r/binlog_stm_blackhole.result mysql-test/suite/binlog/r/binlog_stm_drop_tmp_tbl.result mysql-test/suite/binlog/r/binlog_stm_mix_innodb_myisam.result mysql-test/suite/binlog/t/binlog_row_mysqlbinlog_db_filter.test mysql-test/suite/binlog/t/binlog_stm_unsafe_warning.test mysql-test/suite/federated/my.cnf mysql-test/suite/funcs_1/r/is_columns_mysql.result mysql-test/suite/funcs_1/r/is_statistics.result mysql-test/suite/innodb/r/innodb-zip.result mysql-test/suite/innodb/t/innodb-zip.test mysql-test/suite/innodb/t/innodb_information_schema.test mysql-test/suite/parts/inc/partition_auto_increment.inc mysql-test/suite/parts/r/partition_auto_increment_innodb.result mysql-test/suite/parts/r/partition_auto_increment_maria.result mysql-test/suite/parts/r/partition_auto_increment_memory.result mysql-test/suite/parts/r/partition_auto_increment_myisam.result mysql-test/suite/parts/r/partition_auto_increment_ndb.result mysql-test/suite/parts/r/partition_recover_myisam.result mysql-test/suite/parts/t/partition_auto_increment_archive.test mysql-test/suite/parts/t/partition_auto_increment_blackhole.test mysql-test/suite/parts/t/partition_recover_myisam.test mysql-test/suite/pbxt/r/func_group.result mysql-test/suite/pbxt/r/grant.result mysql-test/suite/pbxt/r/group_min_max.result mysql-test/suite/pbxt/r/join_nested.result mysql-test/suite/pbxt/r/negation_elimination.result mysql-test/suite/pbxt/r/ps_grant.result mysql-test/suite/pbxt/r/skip_grants.result mysql-test/suite/pbxt/r/subselect.result mysql-test/suite/pbxt/r/view_grant.result mysql-test/suite/pbxt/t/grant.test mysql-test/suite/pbxt/t/ps_grant.test mysql-test/suite/pbxt/t/subselect.test mysql-test/suite/rpl/r/rpl_auto_increment.result mysql-test/suite/rpl/r/rpl_bug33931.result mysql-test/suite/rpl/r/rpl_do_grant.result mysql-test/suite/rpl/r/rpl_extraCol_innodb.result mysql-test/suite/rpl/r/rpl_extraCol_myisam.result mysql-test/suite/rpl/r/rpl_get_master_version_and_clock.result mysql-test/suite/rpl/r/rpl_idempotency.result mysql-test/suite/rpl/r/rpl_init_slave_errors.result mysql-test/suite/rpl/r/rpl_innodb_mixed_dml.result mysql-test/suite/rpl/r/rpl_loaddata.result mysql-test/suite/rpl/r/rpl_loaddata_fatal.result mysql-test/suite/rpl/r/rpl_loaddata_map.result mysql-test/suite/rpl/r/rpl_loaddatalocal.result mysql-test/suite/rpl/r/rpl_log_pos.result mysql-test/suite/rpl/r/rpl_packet.result mysql-test/suite/rpl/r/rpl_row_create_table.result mysql-test/suite/rpl/r/rpl_row_sp006_InnoDB.result mysql-test/suite/rpl/r/rpl_slave_load_remove_tmpfile.result mysql-test/suite/rpl/r/rpl_stm_log.result mysql-test/suite/rpl/r/rpl_temporary_errors.result mysql-test/suite/rpl/t/disabled.def mysql-test/suite/rpl/t/rpl_bug33931.test mysql-test/suite/rpl/t/rpl_do_grant.test mysql-test/suite/rpl/t/rpl_drop_temp.test mysql-test/suite/rpl/t/rpl_err_ignoredtable.test mysql-test/suite/rpl/t/rpl_get_master_version_and_clock.test mysql-test/suite/rpl/t/rpl_idempotency.test mysql-test/suite/rpl/t/rpl_init_slave_errors.test mysql-test/suite/rpl/t/rpl_loaddatalocal.test mysql-test/suite/rpl/t/rpl_log_pos.test mysql-test/suite/rpl/t/rpl_packet.test mysql-test/suite/rpl/t/rpl_slave_load_remove_tmpfile.test mysql-test/suite/rpl/t/rpl_temporary_errors.test mysql-test/suite/rpl_ndb/r/rpl_ndb_circular_simplex.result mysql-test/suite/rpl_ndb/r/rpl_ndb_extraCol.result mysql-test/suite/rpl_ndb/r/rpl_ndb_sp006.result mysql-test/suite/rpl_ndb/t/rpl_ndb_circular.test mysql-test/suite/rpl_ndb/t/rpl_ndb_circular_simplex.test mysql-test/t/almost_full.test mysql-test/t/alter_table.test mysql-test/t/analyse.test mysql-test/t/archive.test mysql-test/t/bug46080.test mysql-test/t/create.test mysql-test/t/ctype_ldml.test mysql-test/t/ctype_uca.test mysql-test/t/delete.test mysql-test/t/disabled.def mysql-test/t/distinct.test mysql-test/t/explain.test mysql-test/t/flush_read_lock_kill.test mysql-test/t/func_group.test mysql-test/t/func_in.test mysql-test/t/func_str.test mysql-test/t/gis-rtree.test mysql-test/t/gis.test mysql-test/t/grant3.test mysql-test/t/information_schema_db.test mysql-test/t/innodb-autoinc.test mysql-test/t/innodb-index.test mysql-test/t/innodb_bug39438.test mysql-test/t/innodb_file_format.test mysql-test/t/innodb_mysql.test mysql-test/t/insert_select.test mysql-test/t/join.test mysql-test/t/kill.test mysql-test/t/lowercase_fs_off.test mysql-test/t/lowercase_table3.test mysql-test/t/myisam-system.test mysql-test/t/myisam.test mysql-test/t/myisam_crash_before_flush_keys.test mysql-test/t/mysqlbinlog.test mysql-test/t/mysqltest.test mysql-test/t/named_pipe.test mysql-test/t/olap.test mysql-test/t/order_by.test mysql-test/t/partition.test mysql-test/t/partition_csv.test mysql-test/t/partition_innodb.test mysql-test/t/plugin.test mysql-test/t/plugin_load.test mysql-test/t/ps_not_windows.test mysql-test/t/query_cache.test mysql-test/t/range.test mysql-test/t/select.test mysql-test/t/sp-error.test mysql-test/t/sp.test mysql-test/t/subselect.test mysql-test/t/subselect3.test mysql-test/t/type_bit.test mysql-test/t/type_newdecimal.test mysql-test/t/udf.test mysql-test/t/update.test mysql-test/t/upgrade.test mysql-test/t/view_grant.test mysql-test/t/warnings.test mysql-test/t/windows.test mysql-test/t/xa.test mysql-test/valgrind.supp mysys/CMakeLists.txt mysys/Makefile.am mysys/charset-def.c mysys/hash.c mysys/mf_keycache.c mysys/my_copy.c mysys/my_getopt.c mysys/my_largepage.c mysys/my_static.c mysys/my_thr_init.c mysys/my_wincond.c mysys/thr_lock.c mysys/typelib.c regex/CMakeLists.txt scripts/make_win_bin_dist scripts/mysql_system_tables.sql scripts/mysql_system_tables_fix.sql sql-common/client.c sql-common/my_time.c sql/CMakeLists.txt sql/Makefile.am sql/event_data_objects.cc sql/events.cc sql/field.cc sql/field.h sql/ha_ndbcluster.cc sql/ha_ndbcluster_binlog.cc sql/ha_partition.cc sql/ha_partition.h sql/handler.cc sql/handler.h sql/item.cc sql/item.h sql/item_cmpfunc.cc sql/item_cmpfunc.h sql/item_func.cc sql/item_func.h sql/item_geofunc.cc sql/item_strfunc.cc sql/item_subselect.cc sql/item_sum.cc sql/item_timefunc.cc sql/item_xmlfunc.cc sql/log.cc sql/log_event.cc sql/log_event.h sql/log_event_old.cc sql/my_decimal.h sql/mysql_priv.h sql/mysqld.cc sql/opt_range.cc sql/opt_sum.cc sql/partition_info.cc sql/records.cc sql/repl_failsafe.cc sql/rpl_filter.cc sql/set_var.cc sql/set_var.h sql/share/errmsg.txt sql/slave.cc sql/slave.h sql/sp.cc sql/sp_head.cc sql/sql_acl.cc sql/sql_base.cc sql/sql_binlog.cc sql/sql_cache.cc sql/sql_class.cc sql/sql_class.h sql/sql_db.cc sql/sql_delete.cc sql/sql_handler.cc sql/sql_insert.cc sql/sql_lex.h sql/sql_load.cc sql/sql_locale.cc sql/sql_parse.cc sql/sql_partition.cc sql/sql_plugin.cc sql/sql_prepare.cc sql/sql_rename.cc sql/sql_repl.cc sql/sql_select.cc sql/sql_select.h sql/sql_show.cc sql/sql_table.cc sql/sql_tablespace.cc sql/sql_trigger.cc sql/sql_udf.cc sql/sql_update.cc sql/sql_view.cc sql/sql_yacc.yy sql/structs.h sql/table.cc sql/table.h sql/time.cc sql/udf_example.c sql/unireg.cc storage/archive/ha_archive.cc storage/blackhole/ha_blackhole.cc storage/csv/ha_tina.cc storage/federatedx/ha_federatedx.cc storage/heap/hp_write.c storage/innobase/dict/dict0dict.c storage/innobase/handler/ha_innodb.cc storage/innobase/handler/ha_innodb.h storage/innobase/os/os0proc.c storage/innobase/row/row0mysql.c storage/innodb_plugin/CMakeLists.txt storage/innodb_plugin/ChangeLog storage/innodb_plugin/Makefile.am storage/innodb_plugin/btr/btr0btr.c storage/innodb_plugin/btr/btr0sea.c storage/innodb_plugin/buf/buf0buf.c storage/innodb_plugin/buf/buf0flu.c storage/innodb_plugin/buf/buf0lru.c storage/innodb_plugin/buf/buf0rea.c storage/innodb_plugin/dict/dict0crea.c storage/innodb_plugin/dict/dict0dict.c storage/innodb_plugin/fil/fil0fil.c storage/innodb_plugin/fsp/fsp0fsp.c storage/innodb_plugin/handler/ha_innodb.cc storage/innodb_plugin/handler/ha_innodb.h storage/innodb_plugin/handler/handler0alter.cc storage/innodb_plugin/include/buf0buf.h storage/innodb_plugin/include/buf0buf.ic storage/innodb_plugin/include/buf0lru.h storage/innodb_plugin/include/buf0rea.h storage/innodb_plugin/include/buf0types.h storage/innodb_plugin/include/dict0crea.h storage/innodb_plugin/include/dict0dict.h storage/innodb_plugin/include/dict0mem.h storage/innodb_plugin/include/fsp0fsp.h storage/innodb_plugin/include/lock0lock.h storage/innodb_plugin/include/log0log.h storage/innodb_plugin/include/log0log.ic storage/innodb_plugin/include/log0recv.h storage/innodb_plugin/include/mtr0mtr.h storage/innodb_plugin/include/os0file.h storage/innodb_plugin/include/os0sync.h storage/innodb_plugin/include/page0page.h storage/innodb_plugin/include/page0page.ic storage/innodb_plugin/include/page0zip.h storage/innodb_plugin/include/rem0cmp.h storage/innodb_plugin/include/rem0rec.ic storage/innodb_plugin/include/row0ins.h storage/innodb_plugin/include/row0mysql.h storage/innodb_plugin/include/srv0srv.h storage/innodb_plugin/include/trx0rec.h storage/innodb_plugin/include/trx0rec.ic storage/innodb_plugin/include/trx0roll.h storage/innodb_plugin/include/trx0sys.ic storage/innodb_plugin/include/trx0trx.h storage/innodb_plugin/include/univ.i storage/innodb_plugin/include/ut0auxconf.h storage/innodb_plugin/include/ut0byte.h storage/innodb_plugin/include/ut0byte.ic storage/innodb_plugin/include/ut0ut.h storage/innodb_plugin/lock/lock0lock.c storage/innodb_plugin/log/log0log.c storage/innodb_plugin/log/log0recv.c storage/innodb_plugin/mem/mem0mem.c storage/innodb_plugin/mtr/mtr0mtr.c storage/innodb_plugin/mysql-test/innodb-analyze.test storage/innodb_plugin/mysql-test/innodb-zip.result storage/innodb_plugin/mysql-test/innodb-zip.test storage/innodb_plugin/mysql-test/innodb_bug34300.test storage/innodb_plugin/mysql-test/innodb_bug36169.test storage/innodb_plugin/mysql-test/innodb_bug36172.test storage/innodb_plugin/mysql-test/innodb_file_format.result storage/innodb_plugin/mysql-test/innodb_file_format.test storage/innodb_plugin/os/os0file.c storage/innodb_plugin/os/os0proc.c storage/innodb_plugin/page/page0cur.c storage/innodb_plugin/page/page0page.c storage/innodb_plugin/page/page0zip.c storage/innodb_plugin/plug.in.disabled storage/innodb_plugin/rem/rem0cmp.c storage/innodb_plugin/row/row0ins.c storage/innodb_plugin/row/row0merge.c storage/innodb_plugin/row/row0mysql.c storage/innodb_plugin/srv/srv0srv.c storage/innodb_plugin/srv/srv0start.c storage/innodb_plugin/sync/sync0rw.c storage/innodb_plugin/sync/sync0sync.c storage/innodb_plugin/thr/thr0loc.c storage/innodb_plugin/trx/trx0rec.c storage/innodb_plugin/trx/trx0roll.c storage/innodb_plugin/trx/trx0trx.c storage/innodb_plugin/ut/ut0auxconf_atomic_pthread_t_solaris.c storage/innodb_plugin/ut/ut0ut.c storage/maria/ha_maria.cc storage/myisam/ha_myisam.cc storage/myisam/mi_check.c storage/myisam/mi_search.c storage/myisam/mi_write.c storage/myisam/myisamchk.c storage/myisam/sort.c storage/myisammrg/myrg_open.c storage/mysql_storage_engine.cmake storage/ndb/src/kernel/blocks/dblqh/DblqhMain.cpp storage/ndb/src/kernel/blocks/suma/Suma.cpp storage/pbxt/src/discover_xt.cc storage/xtradb/dict/dict0dict.c storage/xtradb/handler/ha_innodb.cc storage/xtradb/handler/ha_innodb.h storage/xtradb/handler/handler0alter.cc strings/ctype-mb.c strings/ctype-simple.c strings/ctype-uca.c strings/ctype-ucs2.c support-files/binary-configure.sh tests/mysql_client_test.c unittest/mysys/Makefile.am vio/vio.c vio/vio_priv.h vio/viosocket.c vio/viosslfactories.c mysql-test/r/innodb_lock_wait_timeout_1.result mysql-test/t/innodb_lock_wait_timeout_1.test The size of the diff (41758 lines) is larger than your specified limit of 5000 lines -- lp:maria https://code.launchpad.net/~maria-captains/maria/5.1 Your team Maria developers is subscribed to branch lp:maria. To unsubscribe from this branch go to https://code.launchpad.net/~maria-captains/maria/5.1/+edit-subscription.

1 0

[Maria-developers] [Branch ~maria-captains/maria/5.1] Rev 2770: Fixes after comments from last push:
by noreply＠launchpad.net 01 Dec '09

01 Dec '09

------------------------------------------------------------ revno: 2770 committer: Michael Widenius <monty(a)askmonty.org> branch nick: maria-5.1 timestamp: Mon 2009-11-30 15:36:06 +0200 message: Fixes after comments from last push: - Removed some not needed casts - Change plugin.h to be 'binary compatible' with old versions - Added mysql_ft_size_t typedef to plugin.h to make it trivial to change string lengths to size_t on next ABI change - Made some fixes suggested by Kristian to make things more portable and future safe (when it comes to strict aliasing) modified: include/ft_global.h include/mysql/plugin.h include/mysql/plugin.h.pp mysql-test/t/information_schema.test sql/sp_head.cc sql/sql_select.cc sql/table.cc storage/maria/ma_ft_boolean_search.c storage/maria/ma_ft_nlq_search.c storage/maria/ma_ft_parser.c storage/maria/ma_ftdefs.h storage/maria/maria_ftdump.c storage/myisam/ft_boolean_search.c storage/myisam/ft_nlq_search.c storage/myisam/ft_parser.c storage/myisam/myisam_ftdump.c -- lp:maria https://code.launchpad.net/~maria-captains/maria/5.1 Your team Maria developers is subscribed to branch lp:maria. To unsubscribe from this branch go to https://code.launchpad.net/~maria-captains/maria/5.1/+edit-subscription.

1 0

[Maria-developers] [Branch ~maria-captains/maria/5.1] Rev 2769: Automatic merge
by noreply＠launchpad.net 01 Dec '09

01 Dec '09

Merge authors: Michael Widenius (monty) ------------------------------------------------------------ revno: 2769 [merge] committer: Michael Widenius <monty(a)askmonty.org> branch nick: maria-5.1 timestamp: Mon 2009-11-30 13:13:02 +0200 message: Automatic merge modified: .bzrignore BUILD/compile-pentium client/mysql_upgrade.c cmd-line-utils/readline/config_readline.h cmd-line-utils/readline/display.c cmd-line-utils/readline/history.c cmd-line-utils/readline/rlmbutil.h cmd-line-utils/readline/text.c cmd-line-utils/readline/xmalloc.c configure.in include/mysql/plugin.h include/mysql/plugin.h.pp libmysql/libmysql.c mysql-test/t/information_schema.test mysql-test/t/not_partition.test mysys/lf_hash.c mysys/my_redel.c regex/engine.c regex/engine.ih sql/log_event.cc sql/mysqld.cc sql/scheduler.cc sql/sp_head.cc sql/sql_base.cc sql/sql_builtin.cc.in sql/sql_class.cc sql/sql_class.h sql/sql_insert.cc sql/sql_parse.cc sql/sql_select.cc sql/sql_show.cc sql/table.cc storage/maria/ha_maria.cc storage/maria/lockman.c storage/maria/ma_check.c storage/maria/ma_check_standalone.h storage/maria/ma_ft_boolean_search.c storage/maria/ma_ft_nlq_search.c storage/maria/ma_ft_parser.c storage/maria/ma_ftdefs.h storage/maria/ma_sort.c storage/maria/ma_state.c storage/maria/maria_def.h storage/maria/maria_ftdump.c storage/maria/trnman.c storage/myisam/ft_boolean_search.c storage/myisam/ft_nlq_search.c storage/myisam/ft_parser.c storage/myisam/ft_stopwords.c storage/myisam/ftdefs.h storage/myisam/ha_myisam.cc storage/myisam/mi_check.c storage/myisam/myisam_ftdump.c storage/myisam/myisamchk.c storage/myisam/myisamdef.h storage/myisam/myisamlog.c storage/myisam/sort.c storage/xtradb/fil/fil0fil.c storage/xtradb/trx/trx0i_s.c -- lp:maria https://code.launchpad.net/~maria-captains/maria/5.1 Your team Maria developers is subscribed to branch lp:maria. To unsubscribe from this branch go to https://code.launchpad.net/~maria-captains/maria/5.1/+edit-subscription.

1 0

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (monty:2774)
by Michael Widenius 01 Dec '09

01 Dec '09

#At lp:maria based on revid:knielsen@knielsen-hq.org-20091201074258-sbl29hdqa39h6tgz 2774 Michael Widenius 2009-12-01 Fixed link problem when compiling without uca modified: cmd-line-utils/readline/config_readline.h strings/ctype-uca.c per-file messages: cmd-line-utils/readline/config_readline.h Add back defines needed for linux to get rid of compiler warnings, but hide them for now until we have a solution for compile failure on MacOSX strings/ctype-uca.c Always define UCA contraction functions used by ctype-mb.c === modified file 'cmd-line-utils/readline/config_readline.h' --- a/cmd-line-utils/readline/config_readline.h 2009-11-30 21:37:27 +0000 +++ b/cmd-line-utils/readline/config_readline.h 2009-12-01 16:09:02 +0000 @@ -7,6 +7,13 @@ # include <config.h> #endif +#ifdef NOT_YET /* causes problem on MacOSX */ +/* to get wcwidth() defined */ +#define _XOPEN_SOURCE 600 +#define _XOPEN_SOURCE_EXTENDED +#define _XOPEN_ +#endif + /* Ultrix botches type-ahead when switching from canonical to non-canonical mode, at least through version 4.3 === modified file 'strings/ctype-uca.c' --- a/strings/ctype-uca.c 2009-11-30 12:42:24 +0000 +++ b/strings/ctype-uca.c 2009-12-01 16:09:02 +0000 @@ -36,6 +36,12 @@ #include "m_string.h" #include "m_ctype.h" + +#define MY_UCA_CNT_FLAG_SIZE 4096 +#define MY_UCA_CNT_FLAG_MASK 4095 +#define MY_UCA_CNT_HEAD 1 +#define MY_UCA_CNT_TAIL 2 + #ifdef HAVE_UCA_COLLATIONS #define MY_UCA_NPAGES 256 @@ -6756,16 +6762,6 @@ typedef struct my_uca_scanner_handler_st static uint16 nochar[]= {0,0}; - -#define MY_UCA_CNT_FLAG_SIZE 4096 -#define MY_UCA_CNT_FLAG_MASK 4095 - -#define MY_UCA_CNT_HEAD 1 -#define MY_UCA_CNT_TAIL 2 - - - - /********** Helper functions to handle contraction ************/ @@ -6836,85 +6832,6 @@ my_uca_alloc_contractions(CHARSET_INFO * return 0; } - -/** - Check if UCA data has contractions (public version) - - @cs Pointer to CHARSET_INFO data - @retval 0 - no contraction, 1 - have contractions. -*/ - -my_bool -my_uca_have_contractions(CHARSET_INFO *cs) -{ - return cs->contractions != NULL; -} - - -/** - Check if a character can be contraction head - - @cs Pointer to CHARSET_INFO data - @wc Code point - - @retval 0 - cannot be contraction head - @retval 1 - can be contraction head -*/ - -my_bool -my_uca_can_be_contraction_head(CHARSET_INFO *cs, my_wc_t wc) -{ - return cs->contractions->flags[wc & MY_UCA_CNT_FLAG_MASK] & MY_UCA_CNT_HEAD; -} - - -/** - Check if a character can be contraction tail - - @cs Pointer to CHARSET_INFO data - @wc Code point - - @retval 0 - cannot be contraction tail - @retval 1 - can be contraction tail -*/ - -my_bool -my_uca_can_be_contraction_tail(CHARSET_INFO *cs, my_wc_t wc) -{ - return cs->contractions->flags[wc & MY_UCA_CNT_FLAG_MASK] & MY_UCA_CNT_TAIL; -} - - -/** - Find a contraction and return its weight array - - @cs Pointer to CHARSET data - @wc1 First character - @wc2 Second character - - @return Weight array - @retval NULL - no contraction found - @retval ptr - contraction weight array -*/ - -uint16 * -my_uca_contraction2_weight(CHARSET_INFO *cs, my_wc_t wc1, my_wc_t wc2) -{ - MY_CONTRACTIONS *list= cs->contractions; - MY_CONTRACTION *c, *last; - for (c= list->item, last= &list->item[list->nitems]; c < last; c++) - { - if (c->ch[0] == wc1 && c->ch[1] == wc2) - { - return c->weight; - } - } - return NULL; -} - - - - #ifdef HAVE_CHARSET_ucs2 /* Initialize collation weight scanner @@ -9607,3 +9524,78 @@ CHARSET_INFO my_charset_utf8_croatian_uc #endif /* HAVE_CHARSET_utf8 */ #endif /* HAVE_UCA_COLLATIONS */ + +/** + Check if UCA data has contractions (public version) + + @cs Pointer to CHARSET_INFO data + @retval 0 - no contraction, 1 - have contractions. +*/ + +my_bool +my_uca_have_contractions(CHARSET_INFO *cs) +{ + return cs->contractions != NULL; +} + +/** + Check if a character can be contraction head + + @cs Pointer to CHARSET_INFO data + @wc Code point + + @retval 0 - cannot be contraction head + @retval 1 - can be contraction head +*/ + +my_bool +my_uca_can_be_contraction_head(CHARSET_INFO *cs, my_wc_t wc) +{ + return cs->contractions->flags[wc & MY_UCA_CNT_FLAG_MASK] & MY_UCA_CNT_HEAD; +} + + +/** + Check if a character can be contraction tail + + @cs Pointer to CHARSET_INFO data + @wc Code point + + @retval 0 - cannot be contraction tail + @retval 1 - can be contraction tail +*/ + +my_bool +my_uca_can_be_contraction_tail(CHARSET_INFO *cs, my_wc_t wc) +{ + return cs->contractions->flags[wc & MY_UCA_CNT_FLAG_MASK] & MY_UCA_CNT_TAIL; +} + + +/** + Find a contraction and return its weight array + + @cs Pointer to CHARSET data + @wc1 First character + @wc2 Second character + + @return Weight array + @retval NULL - no contraction found + @retval ptr - contraction weight array +*/ + +uint16 * +my_uca_contraction2_weight(CHARSET_INFO *cs, my_wc_t wc1, my_wc_t wc2) +{ + MY_CONTRACTIONS *list= cs->contractions; + MY_CONTRACTION *c, *last; + for (c= list->item, last= &list->item[list->nitems]; c < last; c++) + { + if (c->ch[0] == wc1 && c->ch[1] == wc2) + { + return c->weight; + } + } + return NULL; +} +

1 0

[Maria-developers] Updated (by Guest): Provide key cache statistics (58)
by worklog-noreply＠askmonty.org 01 Dec '09

01 Dec '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Provide key cache statistics CREATION DATE..: Thu, 22 Oct 2009, 12:28 SUPERVISOR.....: Bothorsen IMPLEMENTOR....: COPIES TO......: CATEGORY.......: Server-RawIdeaBin TASK ID........: 58 (http://askmonty.org/worklog/?tid=58) VERSION........: WorkLog-3.4 STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Monty - Fri, 27 Nov 2009, 12:36)=-=- High Level Description modified. --- /tmp/wklog.58.old.15108 2009-11-27 10:36:13.000000000 +0000 +++ /tmp/wklog.58.new.15108 2009-11-27 10:36:13.000000000 +0000 @@ -1,8 +1,7 @@ -Provide the key cache -statistics available for the default cache (key blocks used, -unflushed, etc.) for all named caches. +Provide the same key cache statistics for all MyISAM key caches as we provide +for the default key cache. -Monty answers in an email: +Background: Monty answers in an email to customer: This is a much simpler task and something that is important to get done. We would have to introduce a 'show keycache statistics' command @@ -11,3 +10,44 @@ The raw coding is probably 1-2 days, but on top if this we need testing, a test environment and building of biniaries for you. Henrik/Bo and Igor will come back to you with a more exact estimate. + +----------------- +Currently we have: + +show status like "key%"; ++------------------------+-------+ +| Variable_name | Value | ++------------------------+-------+ +| Key_blocks_not_flushed | 0 | +| Key_blocks_unused | 13389 | +| Key_blocks_used | 7 | +| Key_read_requests | 22 | +| Key_reads | 7 | +| Key_write_requests | 0 | +| Key_writes | 0 | ++------------------------+-------+ + +Sergei suggested we introduce for each new key cache a set of variables prefixed +with the keycache name: + +show status like "keycache1.%"; +Which should show: + ++----------------------------------+-------+ +| Variable_name | Value | ++----------------------------------+-------+ +| keycache1.Key_blocks_not_flushed | 0 | +| keycache1.Key_blocks_unused | 13389 | +| keycache1.Key_blocks_used | 7 | +| keycache1.Key_read_requests | 22 | +| keycache1.Key_reads | 7 | +| keycache1.Key_write_requests | 0 | +| keycache1.Key_writes | 0 | ++----------------------------------+-------+ + +The task would thus to be to automatically introduce new variables when we +create a new key cache and automatically remove these variables when the cache +is deleted. + +The other option would be to add an information schema where we populate the +information schema with data from all existing key caches. DESCRIPTION: Provide the same key cache statistics for all MyISAM key caches as we provide for the default key cache. Background: Monty answers in an email to customer: This is a much simpler task and something that is important to get done. We would have to introduce a 'show keycache statistics' command that would iterate over all keycaches and provide the statistics. The raw coding is probably 1-2 days, but on top if this we need testing, a test environment and building of biniaries for you. Henrik/Bo and Igor will come back to you with a more exact estimate. ----------------- Currently we have: show status like "key%"; +------------------------+-------+ | Variable_name | Value | +------------------------+-------+ | Key_blocks_not_flushed | 0 | | Key_blocks_unused | 13389 | | Key_blocks_used | 7 | | Key_read_requests | 22 | | Key_reads | 7 | | Key_write_requests | 0 | | Key_writes | 0 | +------------------------+-------+ Sergei suggested we introduce for each new key cache a set of variables prefixed with the keycache name: show status like "keycache1.%"; Which should show: +----------------------------------+-------+ | Variable_name | Value | +----------------------------------+-------+ | keycache1.Key_blocks_not_flushed | 0 | | keycache1.Key_blocks_unused | 13389 | | keycache1.Key_blocks_used | 7 | | keycache1.Key_read_requests | 22 | | keycache1.Key_reads | 7 | | keycache1.Key_write_requests | 0 | | keycache1.Key_writes | 0 | +----------------------------------+-------+ The task would thus to be to automatically introduce new variables when we create a new key cache and automatically remove these variables when the cache is deleted. The other option would be to add an information schema where we populate the information schema with data from all existing key caches. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] need maria dev team help to repro a bug
by Vladimir Kolesnikov 01 Dec '09

01 Dec '09

Hi guys, I'm working on this problem: https://bugs.launchpad.net/pbxt/+bug/489091 Unfortunately I cannot rperoduce plan diffs on the latest lp:~maria-captains/maria/5.1-merge. Is there a way to get access to any of your machines where it is reproducible. Thanks, Vladimir

2 1

[Maria-developers] Why 5.1.39 is not on launchpad?
by Sergey Petrunya 01 Dec '09

01 Dec '09

Hi! I noticed that although 5.1.39 release is done and published, it's not on launchpad: when one goes to https://launchpad.net/maria/, one - sees 5.1.38 binaries offered for download on the right of the page - "Series and milestones" chart doesn't list either 5.1.38 or 5.1.39. This may cause some confusion. Is there any reason that we don't put releases on Launchpad? BR Sergey -- Sergey Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

5 6

[Maria-developers] WL40: Current status & issues to discuss
by Alexi1952 01 Dec '09

01 Dec '09

Bo suggested me to publish my last WL#40 status report on the dev list. Meaning to get a feedback (critical notes, tips, suggestions, etc) I rewrote the report to make it more readable and included some thoughts for discussion. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ WL40. Add a mysqlbinlog option to filter updates to certain tables ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Selected option : "2.5 Extend Query Events With Tables Info". 1. Extending Query Events With Tables Info 2. Example of the Extended Query Event Binary Format 3. Adding binlog-with-tables-info Option 4. How Tables Info is Formed 4.1. CREATE TABLE statement 4.2. RENAME TABLE Statement 5. TODO: Is Tables Info Enough? 6. TODO: How Renames are to be filtered? 7. Other TODO's 1. Extending Query Events With Tables Info ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Reason: to do filtering based on "*-table" options we need to have a list of all tables refered to by the query described by Query event. The Query event binary format is extended with a tables info, i.e. the list of tables involved in the query execution. More precisely, two following fields are added to the Query event header: * Query length (2 bytes). (Reason: in usual format, Query event doesn't include this field because the query is considered to extent to the end of the event what is not the case for the extended format). * Tables info length (2 bytes). Length of tables info contained in the data part of the event. and the following data is added to the Query event data part: * Tables info (tables_info_len bytes). List of 4-tuples (db_len, db, table_name_len, table_name) for each table involved in the execution of the query described by this event. db_len may be zero meaning that this is the default db. Note. In the extended format the default db is written without trailing zero which is redundant since its length is already known from the event header. 2. Example of the Extended Query Event Binary Format ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ************************ QUERY_EVENT [2] ************************ 00000294 | 2F A1 00 4B | time_when = 1258332463 00000298 | 02 | event_type = 2 00000299 | 64 00 00 00 | server_id = 100 0000029D | 75 00 00 00 | event_len = 117 000002A1 | 09 03 00 00 | log_pos = 00000309 000002A5 | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F ------------------------ 000002A7 | 01 00 00 00 | proxy_id = 1 000002AB | C9 00 00 00 | exec_time = 201 000002AF | 04 | dbase_len = 4 000002B0 | 00 00 | error_code = 0 000002B2 | 1A 00 | status_vars_len = 26 000002B4 | 2B 00 | query_len = 43 (*) 000002B6 | 08 00 | tables_info_len = 8 (*) ------------------------ Status Variables ------------------------ 000002B8 | 00 | FLAGS2: 000002B9 | 00 40 00 00 | 0x00004000 000002BD | 01 | SQL_MODE: 000002BE | 00 00 00 00 | 0x0000000000000000 000002C2 | 00 00 00 00 | 000002C6 | 06 | CATALOG_NZ: 000002C7 | 03 | catalog_len = 3 000002C8 | 73 74 64 | catalog = "std" 000002CB | 04 | CHARSET_CODE: 000002CC | 08 00 | charset_client = 8 000002CE | 08 00 | collation_connect = 8 000002D0 | 08 00 | collation_server = 8 ------------------------ 000002D2 | 74 65 73 74 | dbase = "test" (**) 000002D6 | 52 45 4E 41 | query = "RENAME TABLE t1 TO tmp, t2 TO t1, tmp TO t2" 000002DA | 4D 45 20 54 | 000002DE | 41 42 4C 45 | 000002E2 | 20 74 31 20 | 000002E6 | 54 4F 20 74 | 000002EA | 6D 70 2C 20 | 000002EE | 74 32 20 54 | 000002F2 | 4F 20 74 31 | 000002F6 | 2C 20 74 6D | 000002FA | 70 20 54 4F | 000002FE | 20 74 32 | ------------------------ Tables info (*) ------------------------ 00000301 | 00 | dbase_len = 0 (***) 00000302 | 02 | table_len = 2 00000303 | 74 31 | table_str = "t1" 00000305 | 00 | dbase_len = 0 (***) 00000306 | 02 | table_len = 2 00000307 | 74 32 | table_str = "t2" ************************ (*) - Added to the new Query event binary format. (**) - Modified: in the new Query event binary format database name does not contain trailing zero. (***) - Database length = 0 means that it is the default database. 3. Adding binlog-with-tables-info Option ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ New --binlog-with-tables-info option is added. When set, all Query events are to be written to the extended format (with tables info). Reasons for adding the option: * Adding tables info leads binary log to grow in size what is superfluous if filtering within mysqlbinlog is not supposed to be used. * If the current implementation of the extended format will appear to have bugs switching off this option will guarantee that replication will work as before that. The option is defined as GLOBAL and read-only. Reason: * The Query events format used in the binary log can be recognized only by its length contained in the Format description event at the beginning of the binary log. Because of that, two different Query event formats can not be mixed within one binary log. 4. How Tables Info is Formed ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Forming tables info for some types of statement is not quite straightforward. Below is two examples of those "not-straightforward" cases. 4.1. CREATE TABLE statement ~~~~~~~~~~~~~~~~~~~~~~~~~~~ CREATE [TEMPORARY] TABLE [IF NOT EXISTS] t (...) CREATE [TEMPORARY] TABLE [IF NOT EXISTS] t LIKE t1 CREATE [TEMPORARY] TABLE [IF NOT EXISTS] t SELECT (t1, ..., tn) Implementation of forming tables info in this case is based on the following rules of binlogging CREATE TABLE statements: SBR. The original CREATE TABLE statement is always binlogged as is. In this case tables_info = { t }, { t, t1 }, or { t, t1, ..., tn } respectively. RBR. In case of RBR we have the following specifics of binlogging CREATE TABLE statement: * temporary tables are not binlogged; * CREATE TABLE t LIKE t1 with temporary t1 is transformed to the ordinary CREATE TABLE (...); * CREATE TABLE SELECT is always transformed to the ordinary CREATE TABLE (...) (with possible subsequent INSERT statements). So for RBR tables info is formed as follows: IF t is temporary // the statement will not be binlogged tables_info = empty IF CREATE TABLE t (...) // the statement will be binlogged as is tables_ info = {t} IF CREATE TABLE t LIKE t1 IF t1 is temporary // the statement will be binlogged as CREATE TABLE t (...) tables_info = {t} ELSE // the statement will be binlogged as is tables_info = {t, t1} IF CREATE TABLE t SELECT (t1, ..., tn) // the statement will be binlogged as CREATE TABLE (...) tables_info = {t} Note. In case of RBR, CREATE TABLE SELECT is binlogged as the following sequence of events: BEGIN: Query event CREATE TABLE: Query event INSERT's: Write row events COMMIT Query event BEGIN and COMMIT in this case are supplied with the same tables_info = {t} as the CREATE TABLE statement. Also INSERT's are preceded by the Table map event refering to the same table t. So filtering of all this events (including BEGIN and COMMIT) will be based on the same tables info and either all of them will pass the filter or all of them will be filtered out. 4.2. RENAME TABLE Statement ~~~~~~~~~~~~~~~~~~~~~~~~~~~ RENAME TABLE tl TO s1, ..., tn TO sn For this type of statements tables info is formed, roughly speaking, of tables standing left to "TO", i.e. t1, ..., tn. But including all t1, ..., tn is not always correct. Consider renaming t1 <-> t2: RENAME TABLE t1 TO tmp, t2 TO t1, tmp TO t2 Setting tables_info = ( t1, t2, tmp } is not good: mysqlbinlog being executed with do-tables = { t1, t2 } will filter this statement away because tmp does not belong to do-tables. So before adding ti to the tables_info we have to check that ti != sj for all j < i. See tables_info in the example of Section 2. 5. TODO: Is Tables Info Enough? ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Assume we run mysqlbinlog with ignore_tables = { t1 } for the following sequence of statements: (1) CREATE TABLE t1 (...); ... (2) CREATE TABLE t2 LIKE t1; ... (3) INSERT INTO t2 VALUES ...; ... In this case: (1) will be filtered away - directly stated by the ignore rules; (2) will be filtered away - can not be executed because refers to t1 (3) will be left - though can not be executed It is not good to have erroneous statements in mysqlbinlog output. We can overcome this in two ways: 1. To explain to the user that in those cases, to avoid erroneous statements, he/she should add t2 to the ignore_tables list as well. 2. To make the filtering bit clever. E.g., additionally to tables info, we may extend the Query event format with a Statement type field (BTW, see also WL#41 "Add a mysqlbinlog option to filter certain kinds of statements"). In this case, when filtering, if we reject a "CREATE TABLE t ..." statement because its LIKE or SELECT part refers to an ignored table, we add table t to the ignore_tables list ourself. 6. TODO: How Renames are to be filtered? ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Consider the RENAME TABLE statement. If we have do_tables = { t } and t is renamed to s: (1) RENAME TABLE t TO s; <smth for s> what the filter should do? Should we consider t and s to be the same table with different names so that <smth for s> is not to be filtered away (in this case during filtering process we should add s to the do_tables list - like we do it in the case of CREATE TABLE)? Or should we treat do- ignore-table rules as talking about *table names* so that --do-table=t means "do only table with name t" - then we filter <smth for s> away because it refers to the table with another name s? Note that in the last case if we rename s back to t: (2) RENAME TABLE s TO t; <smth for t> the RENAME statement will be filtered away (it renames a table which does not belong to the do_tables list) while <smth for t> will pass the filter and its subsequent execution will result with error (we can resolve this by creating a rename_list during the filtering process to trace all renamings specified by the processed events - similar to what is described in Section 4.2). Note. With the second treatment of renaming, we can do the following (looks like attractive but not sure is usefull :) <smth for t> RENAME TABLE t TO s <smth for s that can be filtered away> RENAME TABLE s TO t <smth for t> 7. Other TODO's ~~~~~~~~~~~~~~~ * Continue with forming tables_info for different types of query. * Filtering algorithm and *-table options for mysqlbinlog. Should it be "clever" or "stupid" (see Sections 5 & 6) ? * Add Statement type field to the Query event format (?) Needed for "clever' filtering and for WL#41 * Other stuff (testing, documenting, etc)

2 1

[Maria-developers] Reviewing bug reports: feature needed
by Hakan Kuecuekyilmaz 01 Dec '09

01 Dec '09

Hi all, after reviewing some MariaDB related bugs on bugs.launchpad.net/maria it occurred to me, that there is no link between the bug state "Fix Committed" and the corresponding patch, which fixed the issue. Is there a way to link a bzr patch/push to a launchpad bug report? If not, where can I do a feature request about it? Best regards, Hakan -- Hakan Küçükyılmaz, QA/Benchmark Engineer, Stuttgart/Germany Monty Program Ab, http://askmonty.org/ Skype: hakank_ Phone: +49 171 1919839

5 5

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (knielsen:2773)
by knielsen＠knielsen-hq.org 01 Dec '09

01 Dec '09

#At lp:maria 2773 knielsen(a)knielsen-hq.org 2009-12-01 [merge] Automerge Daniel's README fixes. modified: README === modified file 'README' --- a/README 2009-10-23 16:48:54 +0000 +++ b/README 2009-11-03 04:59:48 +0000 @@ -1,60 +1,57 @@ -This is a release of MariaDB, a branch of MySQL. +This is a release of MariaDB. -MariaDB is a drop-in replacement of MySQL, with more features, less -bugs and better performance. +MariaDB is designed as a drop-in replacement of MySQL(R) with more +features, new storage engines, fewer bugs, and better performance. -MariaDB is brought to you by many of the original developers of MySQL, -that now work for Monty Program Ab, and by many people in the +MariaDB is brought to you by many of the original developers of MySQL +who now work for Monty Program Ab, and by many people in the community. -MySQL, which is the base of MariaDB, is brought to you by Sun. +MySQL, which is the base of MariaDB, is a product and trademark of Sun +Microsystems, Inc. For a list of developers and other contributors, +see the Credits appendix. You can also do 'SHOW authors' to get a +list of active contributors. -License information can be found in these files: -- For GPL (free) distributions, see the COPYING file and - the EXCEPTIONS-CLIENT file. - -A description of the MariaDB project can be found at: +A description of the MariaDB project and a manual can be found at: http://askmonty.org/wiki/index.php/MariaDB - -GPLv2 Disclaimer -For the avoidance of doubt, except that if any license choice -other than GPL or LGPL is available it will apply instead, Sun -elects to use only the General Public License version 2 (GPLv2) -at this time for any software where a choice of GPL license versions -is made available with the language indicating that GPLv2 or any -later version may be used, or where a choice of which version of -the GPL is applied is otherwise unspecified. - -The differences between MariaDB and MySQL can be found at: http://askmonty.org/wiki/index.php/MariaDB_versus_MySQL +http://askmonty.org/wiki/index.php/Manual:Contents -Documentation about MySQL can be found at: -http://dev.mysql.com/doc - -For further information about MySQL documentation, see: -- The current MySQL documentation: +As MariaDB is a full replacement of MySQL, the MySQL manual at +http://dev.mysql.com/doc is generally applicable. -Some manual sections of special interest: +More help is available from the Maria Discuss mailing list +https://launchpad.net/~maria-discuss +and the #maria IRC channel on Freenode. +*************************************************************************** +NOTE: -- For a list of developers and other contributors, see the Credits - appendix. +MariaDB is specifically available only under version 2 of the GNU +General Public License (GPLv2). (I.e. Without the "any later version" +clause.) This is inherited from MySQL. Please see the README file in +the MySQL distribution for more information. -A local copy of the MySQL Reference Manual can be found in the Docs -directory in GNU Info format. You can also browse the manual online or -download it in any of several formats from -http://dev.mysql.com/doc +License information can be found in the COPYING file and the +EXCEPTIONS-CLIENT file. -************************************************************ +*************************************************************************** IMPORTANT: -Bug or error reports regarding MariaDB should be sent to +Bug and/or error reports regarding MariaDB should be submitted at https://bugs.launchpad.net/maria -Bugs in the MySQL code can also be sent to http://bugs.mysql.com + +Bugs in the MySQL code can also be submitted at http://bugs.mysql.com + *************************************************************************** + + + + +*************************************************************************** %%The following software may be included in this product: Fred Fish's Dbug Library

1 0

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (knielsen:2772)
by knielsen＠knielsen-hq.org 01 Dec '09

01 Dec '09

#At lp:maria 2772 knielsen(a)knielsen-hq.org 2009-12-01 [merge] Merge free documentation from MySQL 5.1.41 source tarball into MariaDB. modified: Docs/INSTALL-BINARY INSTALL-SOURCE INSTALL-WIN-SOURCE man/comp_err.1 man/innochecksum.1 man/make_win_bin_dist.1 man/msql2mysql.1 man/my_print_defaults.1 man/myisam_ftdump.1 man/myisamchk.1 man/myisamlog.1 man/myisampack.1 man/mysql-stress-test.pl.1 man/mysql-test-run.pl.1 man/mysql.1 man/mysql.server.1 man/mysql_client_test.1 man/mysql_config.1 man/mysql_convert_table_format.1 man/mysql_find_rows.1 man/mysql_fix_extensions.1 man/mysql_fix_privilege_tables.1 man/mysql_install_db.1 man/mysql_secure_installation.1 man/mysql_setpermission.1 man/mysql_tzinfo_to_sql.1 man/mysql_upgrade.1 man/mysql_waitpid.1 man/mysql_zap.1 man/mysqlaccess.1 man/mysqladmin.1 man/mysqlbinlog.1 man/mysqlbug.1 man/mysqlcheck.1 man/mysqld.8 man/mysqld_multi.1 man/mysqld_safe.1 man/mysqldump.1 man/mysqldumpslow.1 man/mysqlhotcopy.1 man/mysqlimport.1 man/mysqlmanager.8 man/mysqlshow.1 man/mysqlslap.1 man/mysqltest.1 man/ndbd.8 man/ndbd_redo_log_reader.1 man/ndbmtd.8 man/perror.1 man/replace.1 man/resolve_stack_dump.1 man/resolveip.1 scripts/fill_help_tables.sql support-files/MacOSX/ReadMe.txt === modified file 'Docs/INSTALL-BINARY' --- a/Docs/INSTALL-BINARY 2009-10-23 16:48:54 +0000 +++ b/Docs/INSTALL-BINARY 2009-12-01 07:38:15 +0000 @@ -3,7 +3,7 @@ describe how to install MariaDB; However also applies. -2.9. Installing MariaDB from tar.gz Packages on Other Unix-Like Systems +2.2. Installing MariaDB from Generic Binaries on Unix/Linux This section covers the installation of MariaDB binary distributions that are provided for various platforms in the form of compressed @@ -22,11 +22,13 @@ also applies. * A reasonable tar to unpack the distribution. GNU tar is known to work. Some operating systems come with a preinstalled version of tar that is known to have problems. For example, - the tar provided with early versions of Mac OS X, SunOS 4.x - and Solaris 8 and earlier are known to have problems with long - file names. On Mac OS X, you can use the preinstalled gnutar - program. On other systems with a deficient tar, you should - install GNU tar first. + the tar provided with early versions of Mac OS X, SunOS 4.x, + Solaris 8, Solaris 9, Solaris 10 and OpenSolaris, and HP-UX + are known to have problems with long file names. On Mac OS X, + you can use the preinstalled gnutar program. On Solaris 10 and + OpenSolaris you can use the preinstalled gtar. On other + systems with a deficient tar, you should install GNU tar + first. If you run into problems and need to file a bug report, please report them to: http://bugs.launchpad.net/maria @@ -52,7 +54,7 @@ shell> bin/mysqld_safe --user=mysql & Note This procedure does not set up any passwords for MariaDB accounts. - After following the procedure, proceed to Section 2.11, + After following the procedure, proceed to Section 2.13, "Post-Installation Setup and Testing." A more detailed version of the preceding description for @@ -147,7 +149,7 @@ shell> chown -R mysql data machine, you can copy support-files/mysql.server to the location where your system has its startup files. More information can be found in the support-files/mysql.server - script itself and in Section 2.11.2.2, "Starting and Stopping + script itself and in Section 2.13.1.2, "Starting and Stopping MariaDB Automatically." 10. You can set up new accounts using the bin/mysql_setpermission script if you install the DBI and DBD::mysql Perl modules. See @@ -186,5 +188,5 @@ Note The accounts that are listed in the MariaDB grant tables initially have no passwords. After starting the server, you should set up - passwords for them using the instructions in Section 2.11, + passwords for them using the instructions in Section 2.13, "Post-Installation Setup and Testing." === modified file 'INSTALL-SOURCE' --- a/INSTALL-SOURCE 2009-09-16 12:03:18 +0000 +++ b/INSTALL-SOURCE 2009-12-01 07:24:05 +0000 @@ -5,9 +5,8 @@ Chapter 2. Installing and Upgrading MySQ of the procedure follows and later sections provide the details. If you plan to upgrade an existing version of MySQL to a newer version rather than install MySQL for the first time, see Section - 2.12.1, "Upgrading MySQL," for information about upgrade - procedures and about issues that you should consider before - upgrading. + 2.4.1, "Upgrading MySQL," for information about upgrade procedures + and about issues that you should consider before upgrading. If you are interested in migrating to MySQL from another database system, you may wish to read Section A.8, "MySQL 5.1 FAQ --- @@ -15,96 +14,61 @@ Chapter 2. Installing and Upgrading MySQ concerning migration issues. 1. Determine whether MySQL runs and is supported on your - platform. Please note that not all platforms are equally - suitable for running MySQL, and that not all platforms on - which MySQL is known to run are officially supported by Sun - Microsystems, Inc.: - - + For MySQL Enterprise Server, the officially supported - platforms are listed at - http://www.mysql.com/support/supportedplatforms.html. - - + MySQL Community Server runs on the platforms listed at - Section 2.1.1, "Operating Systems Supported by MySQL - Community Server." - - 2. Choose which distribution to install. Several versions of - MySQL are available, and most are available in several - distribution formats. You can choose from pre-packaged - distributions containing binary (precompiled) programs or - source code. When in doubt, use a binary distribution. We also - provide public access to our current source tree for those who - want to see our most recent developments and help us test new - code. To determine which version and type of distribution you - should use, see Section 2.1.2, "Choosing Which MySQL - Distribution to Install." - - 3. Download the distribution that you want to install. For - instructions, see Section 2.1.3, "How to Get MySQL." To verify - the integrity of the distribution, use the instructions in - Section 2.1.4, "Verifying Package Integrity Using MD5 + platform. + Please note that not all platforms are equally suitable for + running MySQL, and that not all platforms on which MySQL is + known to run are officially supported by Sun Microsystems, + Inc.: + + 2. Choose which distribution to install. + Several versions of MySQL are available, and most are + available in several distribution formats. You can choose from + pre-packaged distributions containing binary (precompiled) + programs or source code. When in doubt, use a binary + distribution. We also provide public access to our current + source tree for those who want to see our most recent + developments and help us test new code. To determine which + version and type of distribution you should use, see Section + 2.1.2, "Choosing Which MySQL Distribution to Install." + + 3. Download the distribution that you want to install. + For instructions, see Section 2.1.3, "How to Get MySQL." To + verify the integrity of the distribution, use the instructions + in Section 2.1.4, "Verifying Package Integrity Using MD5 Checksums or GnuPG." - 4. Install the distribution. To install MySQL from a binary - distribution, use the instructions in Section 2.2, "Standard - MySQL Installation Using a Binary Distribution." To install - MySQL from a source distribution or from the current - development source tree, use the instructions in Section 2.10, - "MySQL Installation Using a Source Distribution." - If you encounter installation difficulties, see Section 2.13, - "Operating System-Specific Notes," for information on solving - problems for particular platforms. - - 5. Perform any necessary post-installation setup. After - installing MySQL, read Section 2.11, "Post-Installation Setup - and Testing." This section contains important information - about making sure the MySQL server is working properly. It - also describes how to secure the initial MySQL user accounts, - which have no passwords until you assign passwords. The - section applies whether you install MySQL using a binary or - source distribution. + 4. Install the distribution. + To install MySQL from a binary distribution, use the + instructions in Section 2.2, "Installing MySQL from Generic + Binaries on Unix/Linux." + To install MySQL from a source distribution or from the + current development source tree, use the instructions in + Section 2.3, "MySQL Installation Using a Source Distribution." + + 5. Perform any necessary post-installation setup. + After installing MySQL, read Section 2.13, "Post-Installation + Setup and Testing." This section contains important + information about making sure the MySQL server is working + properly. It also describes how to secure the initial MySQL + user accounts, which have no passwords until you assign + passwords. The section applies whether you install MySQL using + a binary or source distribution. 6. If you want to run the MySQL benchmark scripts, Perl support for MySQL must be available. See Section 2.15, "Perl Installation Notes." -2.1. General Installation Issues - - The MySQL installation procedure depends on whether you will - install MySQL Enterprise Server or MySQL Community Server. The set - of applicable platforms depends on which distribution you will - install: - - * For MySQL Enterprise Server, the officially supported - platforms are listed at - http://www.mysql.com/support/supportedplatforms.html. - - * MySQL Community Server runs on the platforms listed at Section - 2.1.1, "Operating Systems Supported by MySQL Community - Server." - - For MySQL Enterprise Server, install the main distribution plus - any service packs or hotfixes that you wish to apply using the - Enterprise Installer. For platforms that do not yet have an - Enterprise Installer, use the Community Server instructions. - - For MySQL Community Server, install the main distribution plus any - hotfixes and updates: - - * Download a binary release, or download a source release and - build MySQL yourself from the source code. - - * Retrieve MySQL from the Bazaar tree and build it from source. - The Bazaar tree contains the latest developer code. +2.1. General Installation Guidance The immediately following sections contain the information necessary to choose, download, and verify your distribution. The instructions in later sections of the chapter describe how to install the distribution that you choose. For binary - distributions, see the instructions at Section 2.2, "Standard - MySQL Installation Using a Binary Distribution." To build MySQL - from source, use the instructions at Section 2.10, "MySQL - Installation Using a Source Distribution." + distributions, see the instructions at Section 2.2, "Installing + MySQL from Generic Binaries on Unix/Linux" or the corresponding + section for your platform if available. To build MySQL from + source, use the instructions in Section 2.3, "MySQL Installation + Using a Source Distribution." 2.1.1. Operating Systems Supported by MySQL Community Server @@ -129,58 +93,31 @@ Important MySQL has been reported to compile successfully on the following combinations of operating system and thread package. - * AIX 4.x, 5.x with native threads. See Section 2.13.5.3, - "IBM-AIX notes." - - * Amiga. + * AIX 4.x, 5.x with native threads. See Section 2.12, + "Installing MySQL on AIX." AIX 5.3 should be upgraded to + technology level 7 (5300-07). - * FreeBSD 5.x and up with native threads. + * FreeBSD 5.x and up with native threads. See Section 2.10, + "Installing MySQL on FreeBSD." - * HP-UX 11.x with the native threads. See Section 2.13.5.2, - "HP-UX Version 11.x Notes." + * HP-UX 11.x with the native threads. See Section 2.11, + "Installing MySQL on HP-UX." * Linux, builds on all fairly recent Linux distributions with - glibc 2.3. See Section 2.13.1, "Linux Notes." - - * Mac OS X. See Section 2.13.2, "Mac OS X Notes." - - * NetBSD 1.3/1.4 Intel and NetBSD 1.3 Alpha. See Section - 2.13.4.2, "NetBSD Notes." + glibc 2.3. See Section 2.6, "Installing MySQL on Linux." - * Novell NetWare 6.0 and 6.5. See Section 2.8, "Installing MySQL - on NetWare." + * Mac OS X. See Section 2.7, "Installing MySQL on Mac OS X." - * OpenBSD 2.5 and with native threads. OpenBSD earlier than 2.5 - with the MIT-pthreads package. See Section 2.13.4.3, "OpenBSD - 2.5 Notes." - - * SCO OpenServer 5.0.X with a recent port of the FSU Pthreads - package. See Section 2.13.5.8, "SCO UNIX and OpenServer 5.0.x - Notes." - - * SCO Openserver 6.0.x. See Section 2.13.5.9, "SCO OpenServer - 6.0.x Notes." - - * SCO UnixWare 7.1.x. See Section 2.13.5.10, "SCO UnixWare 7.1.x - and OpenUNIX 8.0.0 Notes." - - * SGI Irix 6.x with native threads. See Section 2.13.5.7, "SGI - Irix Notes." - - * Solaris 2.5 and above with native threads on SPARC and x86. - See Section 2.13.3, "Solaris Notes." - - * Tru64 Unix. See Section 2.13.5.5, "Alpha-DEC-UNIX Notes - (Tru64)." + * Solaris 2.8 on SPARC and x86, including support for native + threads. See Section 2.8.1, "Solaris Notes." * Windows 2000, Windows XP, Windows Vista, Windows Server 2003, - and Windows Server 2008. See Section 2.3, "Installing MySQL on + and Windows Server 2008. See Section 2.5, "Installing MySQL on Windows." MySQL has also been known to run on other systems in the past. See - Section 2.13, "Operating System-Specific Notes." Some porting - effort might be required for current versions of MySQL on these - systems. + Section 2.1, "General Installation Guidance." Some porting effort + might be required for current versions of MySQL on these systems. Not all platforms are equally well-suited for running MySQL. How well a certain platform is suited for a high-load mission-critical @@ -208,8 +145,8 @@ Important * General file system stability and performance. * Table size. If your tables are large, performance is affected - by the ability of the file system to deal with large files at - all and to deal with them efficiently. + by the ability of the file system to deal with large files and + dealing with them efficiently. * Our level of expertise here at Sun Microsystems, Inc. with the platform. If we know a platform well, we enable @@ -240,7 +177,7 @@ Important development process, multiple release series co-exist, each at a different stage of maturity: - * MySQL 5.4 and 6.0 are the current development release series. + * MySQL 5.5 is the current development release series. * MySQL 5.1 is the current General Availability (Production) release series. New releases are issued for bugfixes only; no @@ -255,9 +192,9 @@ Important has ended. Extended support for MySQL 4.1 remains available. According to the MySQL Lifecycle Policy - (http://www.mysql.com/company/legal/lifecycle/#policy) only - Security and Severity Level 1 issues are still being fixed for - MySQL 4.1. + (http://www.mysql.com/about/legal/lifecycle/) only Security + and Severity Level 1 issues are still being fixed for MySQL + 4.1. We do not believe in a complete code freeze because this prevents us from making bugfixes and other fixes that must be done. By @@ -367,13 +304,13 @@ Important * The MySQL benchmark suite This suite runs a range of common queries. It is also a test to determine whether the latest batch of optimizations - actually made the code faster. See Section 7.1.4, "The MySQL + actually made the code faster. See Section 7.1.3, "The MySQL Benchmark Suite." * The crash-me test This test tries to determine what features the database supports and what its capabilities and limitations are. See - Section 7.1.4, "The MySQL Benchmark Suite." + Section 7.1.3, "The MySQL Benchmark Suite." We also test the newest MySQL version in our internal production environment, on at least one machine. We have more than 100GB of @@ -492,21 +429,6 @@ Important as soon as possible. (We would like other companies to do this, too!) -2.1.2.4. MySQL Binaries Compiled by Sun Microsystems, Inc. - - Sun Microsystems, Inc. provides a set of binary distributions of - MySQL. In addition to binaries provided in platform-specific - package formats, we offer binary distributions for a number of - platforms in the form of compressed tar files (.tar.gz files). See - Section 2.2, "Standard MySQL Installation Using a Binary - Distribution." For Windows distributions, see Section 2.3, - "Installing MySQL on Windows." - - If you want to compile a debug version of MySQL from a source - distribution, you should add --with-debug or --with-debug=full to - the configure command used to configure the distribution and - remove any -fomit-frame-pointer options. - 2.1.3. How to Get MySQL Check our downloads page at http://dev.mysql.com/downloads/ for @@ -553,8 +475,8 @@ Important shell> md5sum package_name Example: -shell> md5sum mysql-standard-5.1.39-linux-i686.tar.gz -aaab65abbec64d5e907dcd41b8699945 mysql-standard-5.1.39-linux-i686.ta +shell> md5sum mysql-standard-5.1.41-linux-i686.tar.gz +aaab65abbec64d5e907dcd41b8699945 mysql-standard-5.1.41-linux-i686.ta r.gz You should verify that the resulting checksum (the string of @@ -728,8 +650,8 @@ pg-signature.html signature, which also is available from the download page. The signature file has the same name as the distribution file with an .asc extension, as shown by the examples in the following table. - Distribution file mysql-standard-5.1.39-linux-i686.tar.gz - Signature file mysql-standard-5.1.39-linux-i686.tar.gz.asc + Distribution file mysql-standard-5.1.41-linux-i686.tar.gz + Signature file mysql-standard-5.1.41-linux-i686.tar.gz.asc Make sure that both files are stored in the same directory and then run the following command to verify the signature for the @@ -737,7 +659,7 @@ pg-signature.html shell> gpg --verify package_name.asc Example: -shell> gpg --verify mysql-standard-5.1.39-linux-i686.tar.gz.asc +shell> gpg --verify mysql-standard-5.1.41-linux-i686.tar.gz.asc gpg: Signature made Tue 12 Jul 2005 23:35:41 EST using DSA key ID 507 2E1F5 gpg: Good signature from "MySQL Package signing key (www.mysql.com) < @@ -757,8 +679,8 @@ build(a)mysql.com>" shell> rpm --checksig package_name.rpm Example: -shell> rpm --checksig MySQL-server-5.1.39-0.glibc23.i386.rpm -MySQL-server-5.1.39-0.glibc23.i386.rpm: md5 gpg OK +shell> rpm --checksig MySQL-server-5.1.41-0.glibc23.i386.rpm +MySQL-server-5.1.41-0.glibc23.i386.rpm: md5 gpg OK Note @@ -786,22 +708,6 @@ shell> rpm --import mysql_pubkey.asc Sun Microsystems, Inc. A distribution provided by another vendor might use a layout different from those shown here. - For MySQL 5.1 on Windows, the default installation directory is - C:\Program Files\MySQL\MySQL Server 5.1. (Some Windows users - prefer to install in C:\mysql, the directory that formerly was - used as the default. However, the layout of the subdirectories - remains the same.) The installation directory has the following - subdirectories. - Directory Contents of Directory - bin Client programs and the mysqld server - data Log files, databases - Docs Manual in CHM format - examples Example programs and scripts - include Include (header) files - lib Libraries - scripts Utility scripts - share Error message files - Installations created from our Linux RPM distributions result in files under the following system directories. Directory Contents of Directory @@ -863,8100 +769,7171 @@ shell> rpm --import mysql_pubkey.asc distribution by executing the scripts/make_binary_distribution script from the top directory of the source distribution. -2.2. Standard MySQL Installation Using a Binary Distribution +2.2. Installing MySQL from Generic Binaries on Unix/Linux - The next several sections cover the installation of MySQL on - platforms where we offer packages using the native packaging - format of the respective platform. (This is also known as - performing a "binary install.") However, binary distributions of - MySQL are available for many other platforms as well. See Section - 2.9, "Installing MySQL from tar.gz Packages on Other Unix-Like - Systems," for generic installation instructions for these packages - that apply to all platforms. - - See Section 2.1, "General Installation Issues," for more - information on what other binary distributions are available and - how to obtain them. - -2.3. Installing MySQL on Windows - - A native Windows distribution of MySQL has been available since - version 3.21 and represents a sizable percentage of the daily - downloads of MySQL. This section describes the process for - installing MySQL on Windows. + This section covers the installation of MySQL binary distributions + that are provided for various platforms in the form of compressed + tar files (files with a .tar.gz extension). See Section 2.2, + "Installing MySQL from Generic Binaries on Unix/Linux," for a + detailed list. -Note + To obtain MySQL, see Section 2.1.3, "How to Get MySQL." - If you are upgrading MySQL from an existing installation older - than MySQL 4.1.5, you must first perform the procedure described - in Section 2.3.14, "Upgrading MySQL on Windows." + Sun Microsystems, Inc. provides a set of binary distributions of + MySQL. In addition to binaries provided in platform-specific + package formats, we offer binary distributions for a number of + platforms in the form of compressed tar files (.tar.gz files). For + Windows distributions, see Section 2.5, "Installing MySQL on + Windows." - To run MySQL on Windows, you need the following: + If you want to compile a debug version of MySQL from a source + distribution, you should add --with-debug or --with-debug=full to + the configure command used to configure the distribution and + remove any -fomit-frame-pointer options. - * A Windows operating system such as Windows 2000, Windows XP, - Windows Vista, Windows Server 2003, or Windows Server 2008. - Both 32-bit and 64-bit versions are supported. - A Windows operating system permits you to run the MySQL server - as a service. See Section 2.3.11, "Starting MySQL as a Windows - Service." - Generally, you should install MySQL on Windows using an - account that has administrator rights. Otherwise, you may - encounter problems with certain operations such as editing the - PATH environment variable or accessing the Service Control - Manager. Once installed, MySQL does not need to be executed - using a user with Administrator privileges. + MySQL tar file binary distributions have names of the form + mysql-VERSION-OS.tar.gz, where VERSION is a number (for example, + 5.1.41), and OS indicates the type of operating system for which + the distribution is intended (for example, pc-linux-i686). - * TCP/IP protocol support. + In addition to these generic packages, we also offer binaries in + platform-specific package formats for selected platforms. See the + platform specific sections for more information, for more + information on how to install these. - * Enough space on the hard drive to unpack, install, and create - the databases in accordance with your requirements (generally - a minimum of 200 megabytes is recommended.) + You need the following tools to install a MySQL tar file binary + distribution: - For a list of limitations within the Windows version of MySQL, see - Section D.7.3, "Windows Platform Limitations." + * GNU gunzip to uncompress the distribution. - There may also be other requirements, depending on how you plan to - use MySQL: + * A reasonable tar to unpack the distribution. GNU tar is known + to work. Some operating systems come with a preinstalled + version of tar that is known to have problems. For example, + the tar provided with early versions of Mac OS X, SunOS 4.x, + Solaris 8, Solaris 9, Solaris 10 and OpenSolaris, and HP-UX + are known to have problems with long file names. On Mac OS X, + you can use the preinstalled gnutar program. On Solaris 10 and + OpenSolaris you can use the preinstalled gtar. On other + systems with a deficient tar, you should install GNU tar + first. - * If you plan to connect to the MySQL server via ODBC, you need - a Connector/ODBC driver. See Section 21.1, "MySQL - Connector/ODBC." + If you run into problems and need to file a bug report, please use + the instructions in Section 1.6, "How to Report Bugs or Problems." - * If you plan to use MySQL server with ADO.NET applications, you - need the Connector/NET driver. See Section 21.2, "MySQL - Connector/NET." + The basic commands that you must execute to install and use a + MySQL binary distribution are: +shell> groupadd mysql +shell> useradd -g mysql mysql +shell> cd /usr/local +shell> gunzip < /path/to/mysql-VERSION-OS.tar.gz | tar xvf - +shell> ln -s full-path-to-mysql-VERSION-OS mysql +shell> cd mysql +shell> chown -R mysql . +shell> chgrp -R mysql . +shell> scripts/mysql_install_db --user=mysql +shell> chown -R root . +shell> chown -R mysql data +shell> bin/mysqld_safe --user=mysql & - * If you need tables with a size larger than 4GB, install MySQL - on an NTFS or newer file system. Don't forget to use MAX_ROWS - and AVG_ROW_LENGTH when you create tables. See Section - 12.1.17, "CREATE TABLE Syntax." +Note - MySQL for Windows is available in several distribution formats: + This procedure does not set up any passwords for MySQL accounts. + After following the procedure, proceed to Section 2.13, + "Post-Installation Setup and Testing." - * Binary distributions are available that contain a setup - program that installs everything you need so that you can - start the server immediately. Another binary distribution - format contains an archive that you simply unpack in the - installation location and then configure yourself. For - details, see Section 2.3.1, "Choosing An Installation - Package." + A more detailed version of the preceding description for + installing a binary distribution follows: - * The source distribution contains all the code and support - files for building the executables using the Visual Studio - compiler system. + 1. Add a login user and group for mysqld to run as: +shell> groupadd mysql +shell> useradd -g mysql mysql + These commands add the mysql group and the mysql user. The + syntax for useradd and groupadd may differ slightly on + different versions of Unix, or they may have different names + such as adduser and addgroup. + You might want to call the user and group something else + instead of mysql. If so, substitute the appropriate name in + the following steps. - Generally speaking, you should use a binary distribution that - includes an installer. It is simpler to use than the others, and - you need no additional tools to get MySQL up and running. The - installer for the Windows version of MySQL, combined with a GUI - Configuration Wizard, automatically installs MySQL, creates an - option file, starts the server, and secures the default user - accounts. + 2. Pick the directory under which you want to unpack the + distribution and change location into it. In the following + example, we unpack the distribution under /usr/local. (The + instructions, therefore, assume that you have permission to + create files and directories in /usr/local. If that directory + is protected, you must perform the installation as root.) +shell> cd /usr/local -Caution + 3. Obtain a distribution file using the instructions in Section + 2.1.3, "How to Get MySQL." For a given release, binary + distributions for all platforms are built from the same MySQL + source distribution. - Using virus scanning software such as Norton/Symantec Anti-Virus - on directories containing MySQL data and temporary tables can - cause issues, both in terms of the performance of MySQL and the - virus-scanning software mis-identifying the contents of the files - as containing spam. This is because of the fingerprinting - mechanism used by the virus scanning software, and the way in - which MySQL rapidly updates different files, which may be - identified as a potential security risk. - - After installing MySQL Server, it is recommended that you disable - virus scanning on the main directory (datadir) being used to store - your MySQL table data. There is usually a system built into the - virus scanning software to allow certain directories to be - specifically ignored during virus scanning. - - In addition, by default, MySQL creates temporary files in the - standard Windows temporary directory. To prevent the temporary - files also being scanned, you should configure a separate - temporary directory for MySQL temporary files and add this to the - virus scanning exclusion list. To do this, add a configuration - option for the tmpdir parameter to your my.ini configuration file. - For more information, see Section 2.3.7, "Creating an Option - File." - - The following section describes how to install MySQL on Windows - using a binary distribution. To use an installation package that - does not include an installer, follow the procedure described in - Section 2.3.5, "Installing MySQL from a Noinstall Zip Archive." To - install using a source distribution, see Section 2.10.6, - "Installing MySQL from Source on Windows." + 4. Unpack the distribution, which creates the installation + directory. Then create a symbolic link to that directory: +shell> gunzip < /path/to/mysql-VERSION-OS.tar.gz | tar xvf - +shell> ln -s full-path-to-mysql-VERSION-OS mysql + The tar command creates a directory named mysql-VERSION-OS. + The ln command makes a symbolic link to that directory. This + lets you refer more easily to the installation directory as + /usr/local/mysql. + With GNU tar, no separate invocation of gunzip is necessary. + You can replace the first line with the following alternative + command to uncompress and extract the distribution: +shell> tar zxvf /path/to/mysql-VERSION-OS.tar.gz - MySQL distributions for Windows can be downloaded from - http://dev.mysql.com/downloads/. See Section 2.1.3, "How to Get - MySQL." + 5. Change location into the installation directory: +shell> cd mysql + You will find several files and subdirectories in the mysql + directory. The most important for installation purposes are + the bin and scripts subdirectories: -2.3.1. Choosing An Installation Package + + The bin directory contains client programs and the + server. You should add the full path name of this + directory to your PATH environment variable so that your + shell finds the MySQL programs properly. See Section + 2.14, "Environment Variables." - For MySQL 5.1, there are three installation packages to choose - from when installing MySQL on Windows: + + The scripts directory contains the mysql_install_db + script used to initialize the mysql database containing + the grant tables that store the server access + permissions. - * The Essentials Package: This package has a file name similar - to mysql-essential-5.1.39-win32.msi and contains the minimum - set of files needed to install MySQL on Windows, including the - Configuration Wizard. This package does not include optional - components such as the embedded server and benchmark suite. - - * The Complete Package: This package has a file name similar to - mysql-5.1.39-win32.zip and contains all files needed for a - complete Windows installation, including the Configuration - Wizard. This package includes optional components such as the - embedded server and benchmark suite. + 6. Ensure that the distribution contents are accessible to mysql. + If you unpacked the distribution as mysql, no further action + is required. If you unpacked the distribution as root, its + contents will be owned by root. Change its ownership to mysql + by executing the following commands as root in the + installation directory: +shell> chown -R mysql . +shell> chgrp -R mysql . + The first command changes the owner attribute of the files to + the mysql user. The second changes the group attribute to the + mysql group. - * The Noinstall Archive: This package has a file name similar to - mysql-noinstall-5.1.39-win32.zip and contains all the files - found in the Complete install package, with the exception of - the Configuration Wizard. This package does not include an - automated installer, and must be manually installed and - configured. - - The Essentials package is recommended for most users. It is - provided as an .msi file for use with the Windows Installer. The - Complete and Noinstall distributions are packaged as Zip archives. - To use them, you must have a tool that can unpack .zip files. + 7. If you have not installed MySQL before, you must create the + MySQL data directory and initialize the grant tables: +shell> scripts/mysql_install_db --user=mysql + If you run the command as root, include the --user option as + shown. If you run the command while logged in as that user, + you can omit the --user option. + The command should create the data directory and its contents + with mysql as the owner. + After creating or updating the grant tables, you need to + restart the server manually. - Your choice of install package affects the installation process - you must follow. If you choose to install either the Essentials or - Complete install packages, see Section 2.3.2, "Installing MySQL - with the Automated Installer." If you choose to install MySQL from - the Noinstall archive, see Section 2.3.5, "Installing MySQL from a - Noinstall Zip Archive." - -2.3.2. Installing MySQL with the Automated Installer - - New MySQL users can use the MySQL Installation Wizard and MySQL - Configuration Wizard to install MySQL on Windows. These are - designed to install and configure MySQL in such a way that new - users can immediately get started using MySQL. + 8. Most of the MySQL installation can be owned by root if you + like. The exception is that the data directory must be owned + by mysql. To accomplish this, run the following commands as + root in the installation directory: +shell> chown -R root . +shell> chown -R mysql data - The MySQL Installation Wizard and MySQL Configuration Wizard are - available in the Essentials and Complete install packages. They - are recommended for most standard MySQL installations. Exceptions - include users who need to install multiple instances of MySQL on a - single server host and advanced users who want complete control of - server configuration. + 9. If you want MySQL to start automatically when you boot your + machine, you can copy support-files/mysql.server to the + location where your system has its startup files. More + information can be found in the support-files/mysql.server + script itself and in Section 2.13.1.2, "Starting and Stopping + MySQL Automatically." + 10. You can set up new accounts using the bin/mysql_setpermission + script if you install the DBI and DBD::mysql Perl modules. See + Section 4.6.14, "mysql_setpermission --- Interactively Set + Permissions in Grant Tables." For Perl module installation + instructions, see Section 2.15, "Perl Installation Notes." + 11. If you would like to use mysqlaccess and have the MySQL + distribution in some nonstandard location, you must change the + location where mysqlaccess expects to find the mysql client. + Edit the bin/mysqlaccess script at approximately line 18. + Search for a line that looks like this: +$MYSQL = '/usr/local/bin/mysql'; # path to mysql executable + Change the path to reflect the location where mysql actually + is stored on your system. If you do not do this, a Broken pipe + error will occur when you run mysqlaccess. -2.3.3. Using the MySQL Installation Wizard + After everything has been unpacked and installed, you should test + your distribution. To start the MySQL server, use the following + command: +shell> bin/mysqld_safe --user=mysql & - MySQL Installation Wizard is an installer for the MySQL server - that uses the latest installer technologies for Microsoft Windows. - The MySQL Installation Wizard, in combination with the MySQL - Configuration Wizard, allows a user to install and configure a - MySQL server that is ready for use immediately after installation. + If you run the command as root, you must use the --user option as + shown. The value of the option is the name of the login account + that you created in the first step to use for running the server. + If you run the command while logged in as mysql, you can omit the + --user option. - The MySQL Installation Wizard is the standard installer for all - MySQL server distributions, version 4.1.5 and higher. Users of - previous versions of MySQL need to shut down and remove their - existing MySQL installations manually before installing MySQL with - the MySQL Installation Wizard. See Section 2.3.3.6, "Upgrading - MySQL with the Installation Wizard," for more information on - upgrading from a previous version. + If the command fails immediately and prints mysqld ended, you can + find some information in the host_name.err file in the data + directory. - Microsoft has included an improved version of their Microsoft - Windows Installer (MSI) in the recent versions of Windows. MSI has - become the de-facto standard for application installations on - Windows 2000, Windows XP, and Windows Server 2003. The MySQL - Installation Wizard makes use of this technology to provide a - smoother and more flexible installation process. + More information about mysqld_safe is given in Section 4.3.2, + "mysqld_safe --- MySQL Server Startup Script." - The Microsoft Windows Installer Engine was updated with the - release of Windows XP; those using a previous version of Windows - can reference this Microsoft Knowledge Base article - (http://support.microsoft.com/default.aspx?scid=kb;EN-US;292539) - for information on upgrading to the latest version of the Windows - Installer Engine. +Note - In addition, Microsoft has introduced the WiX (Windows Installer - XML) toolkit recently. This is the first highly acknowledged Open - Source project from Microsoft. We have switched to WiX because it - is an Open Source project and it allows us to handle the complete - Windows installation process in a flexible manner using scripts. + The accounts that are listed in the MySQL grant tables initially + have no passwords. After starting the server, you should set up + passwords for them using the instructions in Section 2.13, + "Post-Installation Setup and Testing." - Improving the MySQL Installation Wizard depends on the support and - feedback of users like you. If you find that the MySQL - Installation Wizard is lacking some feature important to you, or - if you discover a bug, please report it in our bugs database using - the instructions given in Section 1.6, "How to Report Bugs or - Problems." +2.3. MySQL Installation Using a Source Distribution -2.3.3.1. Downloading and Starting the MySQL Installation Wizard + Before you proceed with an installation from source, first check + whether our binary is available for your platform and whether it + works for you. We put a great deal of effort into ensuring that + our binaries are built with the best possible options. - The MySQL installation packages can be downloaded from - http://dev.mysql.com/downloads/. If the package you download is - contained within a Zip archive, you need to extract the archive - first. - -Note + To obtain a source distribution for MySQL, Section 2.1.3, "How to + Get MySQL." If you want to build MySQL from source on Windows, see + Section 2.5.10, "Installing MySQL from Source on Windows." - If you are installing on Windows Vista it is best to open a - network port before beginning the installation. To do this, first - ensure that you are logged in as an Administrator, go to the - Control Panel, and double click the Windows Firewall icon. Choose - the Allow a program through Windows Firewall option and click the - Add port button. Enter MySQL into the Name text box and 3306 (or - the port of your choice) into the Port number text box. Also - ensure that the TCP protocol radio button is selected. If you - wish, you can also limit access to the MySQL server by choosing - the Change scope button. Confirm your choices by clicking the OK - button. If you do not open a port prior to installation, you - cannot configure the MySQL server immediately after installation. - Additionally, when running the MySQL Installation Wizard on - Windows Vista, ensure that you are logged in as a user with - administrative rights. + MySQL source distributions are provided as compressed tar archives + and have names of the form mysql-VERSION.tar.gz, where VERSION is + a number like 5.1.41. - The process for starting the wizard depends on the contents of the - installation package you download. If there is a setup.exe file - present, double-click it to start the installation process. If - there is an .msi file present, double-click it to start the - installation process. + You need the following tools to build and install MySQL from + source: -2.3.3.2. Choosing an Install Type + * GNU gunzip to uncompress the distribution. - There are three installation types available: Typical, Complete, - and Custom. + * A reasonable tar to unpack the distribution. GNU tar is known + to work. Some operating systems come with a preinstalled + version of tar that is known to have problems. For example, + the tar provided with early versions of Mac OS X, SunOS 4.x, + Solaris 8, Solaris 9, Solaris 10 and OpenSolaris, and HP-UX + are known to have problems with long file names. On Mac OS X, + you can use the preinstalled gnutar program. On Solaris 10 and + OpenSolaris you can use the preinstalled gtar. On other + systems with a deficient tar, you should install GNU tar + first. - The Typical installation type installs the MySQL server, the mysql - command-line client, and the command-line utilities. The - command-line clients and utilities include mysqldump, myisamchk, - and several other tools to help you manage the MySQL server. + * A working ANSI C++ compiler. gcc 2.95.2 or later, SGI C++, and + SunPro C++ are some of the compilers that are known to work. + libg++ is not needed when using gcc. gcc 2.7.x has a bug that + makes it impossible to compile some perfectly legal C++ files, + such as sql/sql_base.cc. If you have only gcc 2.7.x, you must + upgrade your gcc to be able to compile MySQL. gcc 2.8.1 is + also known to have problems on some platforms, so it should be + avoided if a newer compiler exists for the platform. gcc + 2.95.2 or later is recommended. - The Complete installation type installs all components included in - the installation package. The full installation package includes - components such as the embedded server library, the benchmark - suite, support scripts, and documentation. + * A good make program. GNU make is always recommended and is + sometimes required. (BSD make fails, and vendor-provided make + implementations may fail as well.) If you have problems, use + GNU make 3.75 or newer. - The Custom installation type gives you complete control over which - packages you wish to install and the installation path that is - used. See Section 2.3.3.3, "The Custom Install Dialog," for more - information on performing a custom install. + * libtool 1.5.24 or later is also recommended. - If you choose the Typical or Complete installation types and click - the Next button, you advance to the confirmation screen to verify - your choices and begin the installation. If you choose the Custom - installation type and click the Next button, you advance to the - custom installation dialog, described in Section 2.3.3.3, "The - Custom Install Dialog." + If you are using a version of gcc recent enough to understand the + -fno-exceptions option, it is very important that you use this + option. Otherwise, you may compile a binary that crashes randomly. + Also use -felide-constructors and -fno-rtti along with + -fno-exceptions. When in doubt, do the following: +CFLAGS="-O3" CXX=gcc CXXFLAGS="-O3 -felide-constructors \ + -fno-exceptions -fno-rtti" ./configure \ + --prefix=/usr/local/mysql --enable-assembler \ + --with-mysqld-ldflags=-all-static -2.3.3.3. The Custom Install Dialog + On most systems, this gives you a fast and stable binary. - If you wish to change the installation path or the specific - components that are installed by the MySQL Installation Wizard, - choose the Custom installation type. + If you run into problems and need to file a bug report, please use + the instructions in Section 1.6, "How to Report Bugs or Problems." - A tree view on the left side of the custom install dialog lists - all available components. Components that are not installed have a - red X icon; components that are installed have a gray icon. To - change whether a component is installed, click on that component's - icon and choose a new option from the drop-down list that appears. +2.3.1. Source Installation Overview - You can change the default installation path by clicking the - Change... button to the right of the displayed installation path. + The basic commands that you must execute to install a MySQL source + distribution are: +shell> groupadd mysql +shell> useradd -g mysql mysql +shell> gunzip < mysql-VERSION.tar.gz | tar -xvf - +shell> cd mysql-VERSION +shell> ./configure --prefix=/usr/local/mysql +shell> make +shell> make install +shell> cp support-files/my-medium.cnf /etc/my.cnf +shell> cd /usr/local/mysql +shell> chown -R mysql . +shell> chgrp -R mysql . +shell> bin/mysql_install_db --user=mysql +shell> chown -R root . +shell> chown -R mysql var +shell> bin/mysqld_safe --user=mysql & - After choosing your installation components and installation path, - click the Next button to advance to the confirmation dialog. + If you start from a source RPM, do the following: +shell> rpmbuild --rebuild --clean MySQL-VERSION.src.rpm -2.3.3.4. The Confirmation Dialog + This makes a binary RPM that you can install. For older versions + of RPM, you may have to replace the command rpmbuild with rpm + instead. - Once you choose an installation type and optionally choose your - installation components, you advance to the confirmation dialog. - Your installation type and installation path are displayed for you - to review. +Note - To install MySQL if you are satisfied with your settings, click - the Install button. To change your settings, click the Back - button. To exit the MySQL Installation Wizard without installing - MySQL, click the Cancel button. + This procedure does not set up any passwords for MySQL accounts. + After following the procedure, proceed to Section 2.13, + "Post-Installation Setup and Testing," for post-installation setup + and testing. - After installation is complete, you have the option of registering - with the MySQL web site. Registration gives you access to post in - the MySQL forums at forums.mysql.com (http://forums.mysql.com) - along with the ability to report bugs at bugs.mysql.com - (http://bugs.mysql.com) and to subscribe to our newsletter. The - final screen of the installer provides a summary of the - installation and gives you the option to launch the MySQL - Configuration Wizard, which you can use to create a configuration - file, install the MySQL service, and configure security settings. + A more detailed version of the preceding description for + installing MySQL from a source distribution follows: -2.3.3.5. Changes Made by MySQL Installation Wizard + 1. Add a login user and group for mysqld to run as: +shell> groupadd mysql +shell> useradd -g mysql mysql + These commands add the mysql group and the mysql user. The + syntax for useradd and groupadd may differ slightly on + different versions of Unix, or they may have different names + such as adduser and addgroup. + You might want to call the user and group something else + instead of mysql. If so, substitute the appropriate name in + the following steps. - Once you click the Install button, the MySQL Installation Wizard - begins the installation process and makes certain changes to your - system which are described in the sections that follow. + 2. Perform the following steps as the mysql user, except as + noted. - Changes to the Registry + 3. Pick the directory under which you want to unpack the + distribution and change location into it. - The MySQL Installation Wizard creates one Windows registry key in - a typical install situation, located in - HKEY_LOCAL_MACHINE\SOFTWARE\MySQL AB. + 4. Obtain a distribution file using the instructions in Section + 2.1.3, "How to Get MySQL." - The MySQL Installation Wizard creates a key named after the major - version of the server that is being installed, such as MySQL - Server 5.1. It contains two string values, Location and Version. - The Location string contains the path to the installation - directory. In a default installation it contains C:\Program - Files\MySQL\MySQL Server 5.1\. The Version string contains the - release number. For example, for an installation of MySQL Server - 5.1.39, the key contains a value of 5.1.39. + 5. Unpack the distribution into the current directory: +shell> gunzip < /path/to/mysql-VERSION.tar.gz | tar xvf - + This command creates a directory named mysql-VERSION. + With GNU tar, no separate invocation of gunzip is necessary. + You can use the following alternative command to uncompress + and extract the distribution: +shell> tar zxvf /path/to/mysql-VERSION-OS.tar.gz - These registry keys are used to help external tools identify the - installed location of the MySQL server, preventing a complete scan - of the hard-disk to determine the installation path of the MySQL - server. The registry keys are not required to run the server, and - if you install MySQL using the noinstall Zip archive, the registry - keys are not created. + 6. Change location into the top-level directory of the unpacked + distribution: +shell> cd mysql-VERSION + Note that currently you must configure and build MySQL from + this top-level directory. You cannot build it in a different + directory. - Changes to the Start Menu + 7. Configure the release and compile everything: +shell> ./configure --prefix=/usr/local/mysql +shell> make + When you run configure, you might want to specify other + options. Run ./configure --help for a list of options. Section + 2.3.2, "Typical configure Options," discusses some of the more + useful options. + If configure fails and you are going to send mail to a MySQL + mailing list to ask for assistance, please include any lines + from config.log that you think can help solve the problem. + Also include the last couple of lines of output from + configure. To file a bug report, please use the instructions + in Section 1.6, "How to Report Bugs or Problems." + If the compile fails, see Section 2.3.4, "Dealing with + Problems Compiling MySQL," for help. - The MySQL Installation Wizard creates a new entry in the Windows - Start menu under a common MySQL menu heading named after the major - version of MySQL that you have installed. For example, if you - install MySQL 5.1, the MySQL Installation Wizard creates a MySQL - Server 5.1 section in the Start menu. + 8. Install the distribution: +shell> make install + You might need to run this command as root. + If you want to set up an option file, use one of those present + in the support-files directory as a template. For example: +shell> cp support-files/my-medium.cnf /etc/my.cnf + You might need to run this command as root. + If you want to configure support for InnoDB tables, you should + edit the /etc/my.cnf file, remove the # character before the + option lines that start with innodb_..., and modify the option + values to be what you want. See Section 4.2.3.3, "Using Option + Files," and Section 13.6.2, "InnoDB Configuration." - The following entries are created within the new Start menu - section: + 9. Change location into the installation directory: +shell> cd /usr/local/mysql + 10. If you ran the make install command as root, the installed + files will be owned by root. Ensure that the installation is + accessible to mysql by executing the following commands as + root in the installation directory: +shell> chown -R mysql . +shell> chgrp -R mysql . + The first command changes the owner attribute of the files to + the mysql user. The second changes the group attribute to the + mysql group. + 11. If you have not installed MySQL before, you must create the + MySQL data directory and initialize the grant tables: +shell> bin/mysql_install_db --user=mysql + If you run the command as root, include the --user option as + shown. If you run the command while logged in as mysql, you + can omit the --user option. + The command should create the data directory and its contents + with mysql as the owner. + After using mysql_install_db to create the grant tables for + MySQL, you must restart the server manually. The mysqld_safe + command to do this is shown in a later step. + 12. Most of the MySQL installation can be owned by root if you + like. The exception is that the data directory must be owned + by mysql. To accomplish this, run the following commands as + root in the installation directory: +shell> chown -R root . +shell> chown -R mysql var + 13. If you want MySQL to start automatically when you boot your + machine, you can copy support-files/mysql.server to the + location where your system has its startup files. More + information can be found in the support-files/mysql.server + script itself; see also Section 2.13.1.2, "Starting and + Stopping MySQL Automatically." + 14. You can set up new accounts using the bin/mysql_setpermission + script if you install the DBI and DBD::mysql Perl modules. See + Section 4.6.14, "mysql_setpermission --- Interactively Set + Permissions in Grant Tables." For Perl module installation + instructions, see Section 2.15, "Perl Installation Notes." - * MySQL Command Line Client: This is a shortcut to the mysql - command-line client and is configured to connect as the root - user. The shortcut prompts for a root user password when you - connect. + After everything has been installed, you should test your + distribution. To start the MySQL server, use the following + command: +shell> /usr/local/mysql/bin/mysqld_safe --user=mysql & - * MySQL Server Instance Config Wizard: This is a shortcut to the - MySQL Configuration Wizard. Use this shortcut to configure a - newly installed server, or to reconfigure an existing server. + If you run the command as root, you should use the --user option + as shown. The value of the option is the name of the login account + that you created in the first step to use for running the server. + If you run the command while logged in as that user, you can omit + the --user option. - * MySQL Documentation: This is a link to the MySQL server - documentation that is stored locally in the MySQL server - installation directory. This option is not available when the - MySQL server is installed using the Essentials installation - package. - - Changes to the File System - - The MySQL Installation Wizard by default installs the MySQL 5.1 - server to C:\Program Files\MySQL\MySQL Server 5.1, where Program - Files is the default location for applications in your system, and - 5.1 is the major version of your MySQL server. This is the - recommended location for the MySQL server, replacing the former - default location C:\mysql. + If the command fails immediately and prints mysqld ended, you can + find some information in the host_name.err file in the data + directory. - By default, all MySQL applications are stored in a common - directory at C:\Program Files\MySQL, where Program Files is the - default location for applications in your Windows installation. A - typical MySQL installation on a developer machine might look like - this: -C:\Program Files\MySQL\MySQL Server 5.1 -C:\Program Files\MySQL\MySQL Workbench 5.1 OSS + More information about mysqld_safe is given in Section 4.3.2, + "mysqld_safe --- MySQL Server Startup Script." - This approach makes it easier to manage and maintain all MySQL - applications installed on a particular system. +Note - In MySQL 5.1.23 and earlier, the default location for the data - files used by MySQL is located within the corresponding MySQL - Server installation directory. For MySQL 5.1.24 and later, the - default location of the data directory is the AppData directory - configured for the user that installed the MySQL application. + The accounts that are listed in the MySQL grant tables initially + have no passwords. After starting the server, you should set up + passwords for them using the instructions in Section 2.13, + "Post-Installation Setup and Testing." -2.3.3.6. Upgrading MySQL with the Installation Wizard +2.3.2. Typical configure Options - The MySQL Installation Wizard can perform server upgrades - automatically using the upgrade capabilities of MSI. That means - you do not need to remove a previous installation manually before - installing a new release. The installer automatically shuts down - and removes the previous MySQL service before installing the new - version. + The configure script gives you a great deal of control over how + you configure a MySQL source distribution. Typically you do this + using options on the configure command line. You can also affect + configure using certain environment variables. See Section 2.14, + "Environment Variables." For a full list of options supported by + configure, run this command: +shell> ./configure --help - Automatic upgrades are available only when upgrading between - installations that have the same major and minor version numbers. - For example, you can upgrade automatically from MySQL 4.1.5 to - MySQL 4.1.6, but not from MySQL 5.0 to MySQL 5.1. + A list of the available configure options is provided in the table + below. - See Section 2.3.14, "Upgrading MySQL on Windows." + Table 2.1. Build (configure) Reference + Formats Description Default Introduced Removed + --bindir=DIR User executables EPREFIX/bin + --build=BUILD Configure for building on BUILD guessed + --cache-file=FILE Cache test results in FILE disabled + -C Alias for `--cache-file=config.cache' + --config-cache + --datadir=DIR Read-only architecture-independent data PREFIX/share -2.3.4. MySQL Server Instance Configuration Wizard + --disable-FEATURE Do not include FEATURE + --disable-dependency-tracking Disable dependency tracking + --disable-grant-options Disable GRANT options + --disable-largefile Omit support for large files + --disable-libtool-lock Disable libtool lock + --disable-thread-safe-client Compile the client without threads + 5.1.7 + --enable-FEATURE Enable FEATURE + --enable-assembler Use assembler versions of some string functions + if available + --enable-debug-sync Compile in Debug Sync facility 5.1.41 + --enable-dependency-tracking Do not reject slow dependency + extractors + --enable-fast-install Optimize for fast installation yes + --enable-local-infile Enable LOAD DATA LOCAL INFILE disabled + --enable-shared Build shared libraries yes + --enable-static Build static libraries yes + --enable-thread-safe-client Compile the client with threads + --exec-prefix=EPREFIX Install architecture-dependent files in + EPREFIX + -h Display this help and exit + --help + --help=short Display options specific to this package + --help=recursive Display the short help of all the included + packages + --host=HOST Cross-compile to build programs to run on HOST + --includedir=DIR C header files PREFIX/include + --infodir=DIR Info documentation PREFIX/info + --libdir=DIR Object code libraries EPREFIX/lib + --libexecdir=DIR Program executables EPREFIX/libexec + --localstatedir=DIR Modifiable single-machine data PREFIX/var + --mandir=DIR man documentation PREFIX/man + -n Do not create output files + --no-create + --oldincludedir=DIR C header files for non-gcc /usr/include + --prefix=PREFIX Install architecture-independent files in PREFIX - The MySQL Server Instance Configuration Wizard helps automate the - process of configuring your server. It creates a custom MySQL - configuration file (my.ini or my.cnf) by asking you a series of - questions and then applying your responses to a template to - generate the configuration file that is tuned to your - installation. + --program-prefix=PREFIX Prepend PREFIX to installed program names - The MySQL Server Instance Configuration Wizard is included with - the MySQL 5.1 server. The MySQL Server Instance Configuration - Wizard is only available for Windows. - -2.3.4.1. Starting the MySQL Server Instance Configuration Wizard - - The MySQL Server Instance Configuration Wizard is normally started - as part of the installation process. You should only need to run - the MySQL Server Instance Configuration Wizard again when you need - to change the configuration parameters of your server. + --program-suffix=SUFFIX Append SUFFIX to installed program names - If you chose not to open a port prior to installing MySQL on - Windows Vista, you can choose to use the MySQL Server - Configuration Wizard after installation. However, you must open a - port in the Windows Firewall. To do this see the instructions - given in Section 2.3.3.1, "Downloading and Starting the MySQL - Installation Wizard." Rather than opening a port, you also have - the option of adding MySQL as a program that bypasses the Windows - Firewall. One or the other option is sufficient --- you need not - do both. Additionally, when running the MySQL Server Configuration - Wizard on Windows Vista ensure that you are logged in as a user - with administrative rights. - MySQL Server Instance Configuration Wizard - - You can launch the MySQL Configuration Wizard by clicking the - MySQL Server Instance Config Wizard entry in the MySQL section of - the Windows Start menu. + --program-transform-name=PROGRAM run sed PROGRAM on installed + program names + -q Do not print `checking...' messages + --quiet + --sbindir=DIR System admin executables EPREFIX/sbin + --sharedstatedir=DIR Modifiable architecture-independent data + PREFIX/com + --srcdir=DIR Find the sources in DIR configure directory or .. + --sysconfdir=DIR Read-only single-machine data PREFIX/etc + --target=TARGET Configure for building compilers for TARGET + -V Display version information and exit + --version + --with-PACKAGE Use PACKAGE + --with-archive-storage-engine Enable the Archive Storage Engine no - Alternatively, you can navigate to the bin directory of your MySQL - installation and launch the MySQLInstanceConfig.exe file directly. + --with-atomic-ops Implement atomic operations using pthread + rwlocks or atomic CPU instructions for multi-processor 5.1.12 + --with-berkeley-db Use BerkeleyDB located in DIR no + --with-berkeley-db-includes Find Berkeley DB headers in DIR + --with-berkeley-db-libs Find Berkeley DB libraries in DIR + --with-big-tables Support tables with more than 4 G rows even on + 32 bit platforms + --with-blackhole-storage-engine Enable the Blackhole Storage + Engine no + --with-charset Default character set + --with-client-ldflags Extra linking arguments for clients + --with-collation Default collation + --with-comment Comment about compilation environment + --with-csv-storage-engine Enable the CSV Storage Engine yes + --with-darwin-mwcc Use Metrowerks CodeWarrior wrappers on OS + X/Darwin + --with-debug Add debug code 5.1.7 + --with-debug=full Add debug code (adds memory checker, very slow) - The MySQL Server Instance Configuration Wizard places the my.ini - file in the installation directory for the MySQL server. This - helps associate configuration files with particular server - instances. + --with-embedded-privilege-control Build parts to check user's + privileges (only affects embedded library) + --with-embedded-server Build the embedded server + --with-error-inject Enable error injection in MySQL Server + 5.1.11 + --with-example-storage-engine Enable the Example Storage Engine no - To ensure that the MySQL server knows where to look for the my.ini - file, an argument similar to this is passed to the MySQL server as - part of the service installation: ---defaults-file="C:\Program Files\MySQL\MySQL Server 5.1\my.ini" + --with-extra-charsets Use charsets in addition to default + --with-fast-mutexes Compile with fast mutexes enabled 5.1.5 + --with-federated-storage-engine Enable federated storage engine no + 5.1.3 5.1.9 + --with-gnu-ld Assume the C compiler uses GNU ld no + --with-innodb Enable innobase storage engine no 5.1.3 5.1.9 + --with-lib-ccflags Extra CC options for libraries + --with-libwrap=DIR Compile in libwrap (tcp_wrappers) support + --with-low-memory Try to use less memory to compile to avoid + memory limitations + --with-machine-type Set the machine type, like "powerpc" + --with-max-indexes=N Sets the maximum number of indexes per table + 64 + --with-mysqld-ldflags Extra linking arguments for mysqld + --with-mysqld-libs Extra libraries to link with for mysqld + --with-mysqld-user What user the mysqld daemon shall be run as - Here, C:\Program Files\MySQL\MySQL Server 5.1 is replaced with the - installation path to the MySQL Server. The --defaults-file option - instructs the MySQL server to read the specified file for - configuration options when it starts. + --with-mysqlmanager Build the mysqlmanager binary Build if server + is built + --with-named-curses-libs Use specified curses libraries + --with-named-thread-libs Use specified thread libraries + --with-ndb-ccflags Extra CC options for ndb compile + --with-ndb-docs Include the NDB Cluster ndbapi and mgmapi + documentation + --with-ndb-port Port for NDB Cluster management server + --with-ndb-port-base Port for NDB Cluster management server + --with-ndb-sci=DIR Provide MySQL with a custom location of sci + library + --with-ndb-test Include the NDB Cluster ndbapi test programs + --with-ndbcluster Include the NDB Cluster table handler no + --with-openssl=DIR Include the OpenSSL support + --with-openssl-includes Find OpenSSL headers in DIR + --with-openssl-libs Find OpenSSL libraries in DIR + --with-other-libc=DIR Link against libc and other standard + libraries installed in the specified nonstandard location + --with-pic Try to use only PIC/non-PIC objects Use both + --with-plugin-PLUGIN Forces the named plugin to be linked into + mysqld statically 5.1.11 + --with-plugins Plugins to include in mysqld none 5.1.11 + --with-pstack Use the pstack backtrace library + --with-pthread Force use of pthread library + --with-row-based-replication Include row-based replication 5.1.5 + 5.1.6 + --with-server-suffix Append value to the version string + --with-ssl=DIR Include SSL support 5.1.11 + --with-system-type Set the system type, like "sun-solaris10" + --with-tags Include additional configurations automatic + --with-tcp-port Which port to use for MySQL services 3306 + --with-unix-socket-path Where to put the unix-domain socket + --with-yassl Include the yaSSL support + --with-zlib-dir=no|bundled|DIR Provide MySQL with a custom + location of compression library + --without-PACKAGE Do not use PACKAGE + --without-bench Skip building of the benchmark suite + --without-debug Build a production version without debugging code - Apart from making changes to the my.ini file by running the MySQL - Server Instance Configuration Wizard again, you can modify it by - opening it with a text editor and making any necessary changes. - You can also modify the server configuration with the MySQL - Administrator (http://www.mysql.com/products/administrator/) - utility. For more information about server configuration, see - Section 5.1.2, "Server Command Options." + --without-docs Skip building of the documentation + --without-extra-tools Skip building utilities in the tools + directory + --without-geometry Do not build geometry-related parts + --without-libedit Use system libedit instead of bundled copy + --without-man Skip building of the man pages + --without-ndb-binlog Disable ndb binlog 5.1.6 + --without-ndb-debug Disable special ndb debug features + --without-plugin-PLUGIN Exclude PLUGIN 5.1.11 + --without-query-cache Do not build query cache + --without-readline Use system readline instead of bundled copy - MySQL clients and utilities such as the mysql and mysqldump - command-line clients are not able to locate the my.ini file - located in the server installation directory. To configure the - client and utility applications, create a new my.ini file in the - Windows installation directory (for example, C:\WINDOWS). + --without-row-based-replication Don't include row-based + replication 5.1.7 5.1.14 + --without-server Only build the client + --without-uca Skip building of the national Unicode collations - Under Windows Server 2003, Windows Server 2000, Windows XP, and - Windows Vista MySQL Server Instance Configuration Wizard will - configure MySQL to work as a Windows service. To start and stop - MySQL you use the Services application that is supplied as part of - the Windows Administrator Tools. - -2.3.4.2. Choosing a Maintenance Option - - If the MySQL Server Instance Configuration Wizard detects an - existing configuration file, you have the option of either - reconfiguring your existing server, or removing the server - instance by deleting the configuration file and stopping and - removing the MySQL service. + Some of the configure options available are described here. For + options that may be of use if you have difficulties building + MySQL, see Section 2.3.4, "Dealing with Problems Compiling MySQL." - To reconfigure an existing server, choose the Re-configure - Instance option and click the Next button. Any existing - configuration file is not overwritten, but renamed (within the - same directory) using a timestamp (Windows) or sequential number - (Linux). To remove the existing server instance, choose the Remove - Instance option and click the Next button. + * To compile just the MySQL client libraries and client programs + and not the server, use the --without-server option: +shell> ./configure --without-server + If you have no C++ compiler, some client programs such as + mysql cannot be compiled because they require C++.. In this + case, you can remove the code in configure that tests for the + C++ compiler and then run ./configure with the + --without-server option. The compile step should still try to + build all clients, but you can ignore any warnings about files + such as mysql.cc. (If make stops, try make -k to tell it to + continue with the rest of the build even if errors occur.) - If you choose the Remove Instance option, you advance to a - confirmation window. Click the Execute button. The MySQL Server - Configuration Wizard stops and removes the MySQL service, and then - deletes the configuration file. The server installation and its - data folder are not removed. + * If you want to build the embedded MySQL library (libmysqld.a), + use the --with-embedded-server option. - If you choose the Re-configure Instance option, you advance to the - Configuration Type dialog where you can choose the type of - installation that you wish to configure. - -2.3.4.3. Choosing a Configuration Type - - When you start the MySQL Server Instance Configuration Wizard for - a new MySQL installation, or choose the Re-configure Instance - option for an existing installation, you advance to the - Configuration Type dialog. - MySQL Server Instance Configuration Wizard: Configuration Type - - There are two configuration types available: Detailed - Configuration and Standard Configuration. The Standard - Configuration option is intended for new users who want to get - started with MySQL quickly without having to make many decisions - about server configuration. The Detailed Configuration option is - intended for advanced users who want more fine-grained control - over server configuration. - - If you are new to MySQL and need a server configured as a - single-user developer machine, the Standard Configuration should - suit your needs. Choosing the Standard Configuration option causes - the MySQL Configuration Wizard to set all configuration options - automatically with the exception of Service Options and Security - Options. + * If you don't want your log files and database directories + located under /usr/local/var, use a configure command + something like one of these: +shell> ./configure --prefix=/usr/local/mysql +shell> ./configure --prefix=/usr/local \ + --localstatedir=/usr/local/mysql/data + The first command changes the installation prefix so that + everything is installed under /usr/local/mysql rather than the + default of /usr/local. The second command preserves the + default installation prefix, but overrides the default + location for database directories (normally /usr/local/var) + and changes it to /usr/local/mysql/data. + You can also specify the installation directory and data + directory locations at server startup time by using the + --basedir and --datadir options. These can be given on the + command line or in an MySQL option file, although it is more + common to use an option file. See Section 4.2.3.3, "Using + Option Files." - The Standard Configuration sets options that may be incompatible - with systems where there are existing MySQL installations. If you - have an existing MySQL installation on your system in addition to - the installation you wish to configure, the Detailed Configuration - option is recommended. + * This option specifies the port number on which the server + listens for TCP/IP connections. The default is port 3306. To + listen on a different port, use a configure command like this: +shell> ./configure --with-tcp-port=3307 - To complete the Standard Configuration, please refer to the - sections on Service Options and Security Options in Section - 2.3.4.10, "The Service Options Dialog," and Section 2.3.4.11, "The - Security Options Dialog," respectively. + * If you are using Unix and you want the MySQL socket file + location to be somewhere other than the default location + (normally in the directory /tmp or /var/run), use a configure + command like this: +shell> ./configure \ + --with-unix-socket-path=/usr/local/mysql/tmp/mysql.sock + The socket file name must be an absolute path name. You can + also change the location of mysql.sock at server startup by + using a MySQL option file. See Section B.5.4.5, "How to + Protect or Change the MySQL Unix Socket File." -2.3.4.4. The Server Type Dialog + * If you want to compile statically linked programs (for + example, to make a binary distribution, to get better + performance, or to work around problems with some Red Hat + Linux distributions), run configure like this: +shell> ./configure --with-client-ldflags=-all-static \ + --with-mysqld-ldflags=-all-static - There are three different server types available to choose from. - The server type that you choose affects the decisions that the - MySQL Server Instance Configuration Wizard makes with regard to - memory, disk, and processor usage. - MySQL Server Instance Configuration Wizard: Server Type + * If you are using gcc and don't have libg++ or libstdc++ + installed, you can tell configure to use gcc as your C++ + compiler: +shell> CC=gcc CXX=gcc ./configure + When you use gcc as your C++ compiler, it does not attempt to + link in libg++ or libstdc++. This may be a good thing to do + even if you have those libraries installed. Some versions of + them have caused strange problems for MySQL users in the past. + The following list indicates some compilers and environment + variable settings that are commonly used with each one. - * Developer Machine: Choose this option for a typical desktop - workstation where MySQL is intended only for personal use. It - is assumed that many other desktop applications are running. - The MySQL server is configured to use minimal system - resources. + + gcc 2.7.2: +CC=gcc CXX=gcc CXXFLAGS="-O3 -felide-constructors" - * Server Machine: Choose this option for a server machine where - the MySQL server is running alongside other server - applications such as FTP, email, and Web servers. The MySQL - server is configured to use a moderate portion of the system - resources. + + gcc 2.95.2: +CFLAGS="-O3 -mpentiumpro" CXX=gcc CXXFLAGS="-O3 -mpentiumpro \ +-felide-constructors -fno-exceptions -fno-rtti" - * Dedicated MySQL Server Machine: Choose this option for a - server machine that is intended to run only the MySQL server. - It is assumed that no other applications are running. The - MySQL server is configured to use all available system - resources. + + pgcc 2.90.29 or newer: +CFLAGS="-O3 -mpentiumpro -mstack-align-double" CXX=gcc \ +CXXFLAGS="-O3 -mpentiumpro -mstack-align-double \ +-felide-constructors -fno-exceptions -fno-rtti" + In most cases, you can get a reasonably optimized MySQL binary + by using the options from the preceding list and adding the + following options to the configure line: +--prefix=/usr/local/mysql --enable-assembler \ +--with-mysqld-ldflags=-all-static + The full configure line would, in other words, be something + like the following for all recent gcc versions: +CFLAGS="-O3 -mpentiumpro" CXX=gcc CXXFLAGS="-O3 -mpentiumpro \ +-felide-constructors -fno-exceptions -fno-rtti" ./configure \ +--prefix=/usr/local/mysql --enable-assembler \ +--with-mysqld-ldflags=-all-static + The binaries we provide on the MySQL Web site at + http://dev.mysql.com/downloads/ are all compiled with full + optimization and should be perfect for most users. See Section + 2.2, "Installing MySQL from Generic Binaries on Unix/Linux." + There are some configuration settings you can tweak to build + an even faster binary, but these are only for advanced users. + See Section 7.5.1, "How Compiling and Linking Affects the + Speed of MySQL." + If the build fails and produces errors about your compiler or + linker not being able to create the shared library + libmysqlclient.so.N (where N is a version number), you can + work around this problem by giving the --disable-shared option + to configure. In this case, configure does not build a shared + libmysqlclient.so.N library. -Note + * By default, MySQL uses the latin1 (cp1252 West European) + character set. To change the default set, use the + --with-charset option: +shell> ./configure --with-charset=CHARSET + CHARSET may be one of binary, armscii8, ascii, big5, cp1250, + cp1251, cp1256, cp1257, cp850, cp852, cp866, cp932, dec8, + eucjpms, euckr, gb2312, gbk, geostd8, greek, hebrew, hp8, + keybcs2, koi8r, koi8u, latin1, latin2, latin5, latin7, macce, + macroman, sjis, swe7, tis620, ucs2, ujis, utf8. See Section + 9.2, "The Character Set Used for Data and Sorting." + (Additional character sets might be available. Check the + output from ./configure --help for the current list.) + The default collation may also be specified. MySQL uses the + latin1_swedish_ci collation by default. To change this, use + the --with-collation option: +shell> ./configure --with-collation=COLLATION + To change both the character set and the collation, use both + the --with-charset and --with-collation options. The collation + must be a legal collation for the character set. (Use the SHOW + COLLATION statement to determine which collations are + available for each character set.) + With the configure option --with-extra-charsets=LIST, you can + define which additional character sets should be compiled into + the server. LIST is one of the following: - By selecting one of the preconfigured configurations, the values - and settings of various options in your my.cnf or my.ini will be - altered accordingly. The default values and options as described - in the reference manual may therefore be different to the options - and values that were created during the execution of the - configuration wizard. + + A list of character set names separated by spaces -2.3.4.5. The Database Usage Dialog + + complex to include all character sets that can't be + dynamically loaded - The Database Usage dialog allows you to indicate the storage - engines that you expect to use when creating MySQL tables. The - option you choose determines whether the InnoDB storage engine is - available and what percentage of the server resources are - available to InnoDB. - MySQL Server Instance Configuration Wizard: Usage Dialog + + all to include all character sets into the binaries + Clients that want to convert characters between the server and + the client should use the SET NAMES statement. See Section + 5.1.5, "Session System Variables," and Section 9.1.4, + "Connection Character Sets and Collations." - * Multifunctional Database: This option enables both the InnoDB - and MyISAM storage engines and divides resources evenly - between the two. This option is recommended for users who use - both storage engines on a regular basis. + * To configure MySQL with debugging code, use the --with-debug + option: +shell> ./configure --with-debug + This causes a safe memory allocator to be included that can + find some errors and that provides output about what is + happening. See MySQL Internals: Porting + (http://forge.mysql.com/wiki/MySQL_Internals_Porting) + As of MySQL 5.1.12, using --with-debug to configure MySQL with + debugging support enables you to use the + --debug="d,parser_debug" option when you start the server. + This causes the Bison parser that is used to process SQL + statements to dump a parser trace to the server's standard + error output. Typically, this output is written to the error + log. - * Transactional Database Only: This option enables both the - InnoDB and MyISAM storage engines, but dedicates most server - resources to the InnoDB storage engine. This option is - recommended for users who use InnoDB almost exclusively and - make only minimal use of MyISAM. + * To cause the Debug Sync facility to be compiled into the + server, use the --enable-debug-sync option. This facility is + used for testing and debugging. When compiled in, Debug Sync + is disabled by default. To enable it, start mysqld with the + --debug-sync-timeout=N option, where N is a timeout value + greater than 0. (The default value is 0, which disables Debug + Sync.) N becomes the default timeout for individual + synchronization points. + Debug Sync is also compiled in if you configure with the + --with-debug option (which implies --enable-debug-sync), + unless you also use the --disable-debug-sync option. + For a description of the Debug Sync facility and how to use + synchronization points, see MySQL Internals: Test + Synchronization + (http://forge.mysql.com/wiki/MySQL_Internals_Test_Synchronizat + ion). + The --enable-debug-sync and --disable-debug-sync options were + added in MySQL 5.1.41. - * Non-Transactional Database Only: This option disables the - InnoDB storage engine completely and dedicates all server - resources to the MyISAM storage engine. This option is - recommended for users who do not use InnoDB. + * If your client programs are using threads, you must compile a + thread-safe version of the MySQL client library with the + --enable-thread-safe-client configure option. This creates a + libmysqlclient_r library with which you should link your + threaded applications. See Section 21.9.16.2, "How to Make a + Threaded Client." - The Configuration Wizard uses a template to generate the server - configuration file. The Database Usage dialog sets one of the - following option strings: -Multifunctional Database: MIXED -Transactional Database Only: INNODB -Non-Transactional Database Only: MYISAM + * Some features require that the server be built with + compression library support, such as the COMPRESS() and + UNCOMPRESS() functions, and compression of the client/server + protocol. The --with-zlib-dir=no|bundled|DIR option provides + control over compression library support. The value no + explicitly disables compression support. bundled causes the + zlib library bundled in the MySQL sources to be used. A DIR + path name specifies the directory in which to find the + compression library sources. - When these options are processed through the default template - (my-template.ini) the result is: -Multifunctional Database: -default-storage-engine=InnoDB -_myisam_pct=50 + * It is possible to build MySQL with large table support using + the --with-big-tables option. + This option causes the variables that store table row counts + to be declared as unsigned long long rather than unsigned + long. This enables tables to hold up to approximately + 1.844E+19 ((2^32)^2) rows rather than 2^32 (~4.295E+09) rows. + Previously it was necessary to pass -DBIG_TABLES to the + compiler manually in order to enable this feature. -Transactional Database Only: -default-storage-engine=InnoDB -_myisam_pct=5 + * Run configure with the --disable-grant-options option to cause + the --bootstrap, --skip-grant-tables, and --init-file options + for mysqld to be disabled. For Windows, the configure.js + script recognizes the DISABLE_GRANT_OPTIONS flag, which has + the same effect. The capability is available as of MySQL + 5.1.15. -Non-Transactional Database Only: -default-storage-engine=MyISAM -_myisam_pct=100 -skip-innodb + * This option allows MySQL Community Server features to be + enabled. Additional options may be required for individual + features, such as --enable-profiling to enable statement + profiling. This option was added in MySQL 5.1.24. It is + enabled by default as of MySQL 5.1.28; to disable it, use + --disable-community-features. - The _myisam_pct value is used to calculate the percentage of - resources dedicated to MyISAM. The remaining resources are - allocated to InnoDB. + * When given with --enable-community-features, the + --enable-profiling option enables the statement profiling + capability exposed by the SHOW PROFILE and SHOW PROFILES + statements. (See Section 12.5.5.33, "SHOW PROFILES Syntax.") + This option was added in MySQL 5.1.24. It is enabled by + default as of MySQL 5.1.28; to disable it, use + --disable-profiling. -2.3.4.6. The InnoDB Tablespace Dialog + * See Section 2.1, "General Installation Guidance," for options + that pertain to particular operating systems. - Some users may want to locate the InnoDB tablespace files in a - different location than the MySQL server data directory. Placing - the tablespace files in a separate location can be desirable if - your system has a higher capacity or higher performance storage - device available, such as a RAID storage system. - MySQL Server Instance Configuration Wizard: InnoDB Data Tablespace + * See Section 5.5.7.2, "Using SSL Connections," for options that + pertain to configuring MySQL to support secure (encrypted) + connections. - To change the default location for the InnoDB tablespace files, - choose a new drive from the drop-down list of drive letters and - choose a new path from the drop-down list of paths. To create a - custom path, click the ... button. - - If you are modifying the configuration of an existing server, you - must click the Modify button before you change the path. In this - situation you must move the existing tablespace files to the new - location manually before starting the server. - -2.3.4.7. The Concurrent Connections Dialog - - To prevent the server from running out of resources, it is - important to limit the number of concurrent connections to the - MySQL server that can be established. The Concurrent Connections - dialog allows you to choose the expected usage of your server, and - sets the limit for concurrent connections accordingly. It is also - possible to set the concurrent connection limit manually. - MySQL Server Instance Configuration Wizard: Connections - - * Decision Support (DSS)/OLAP: Choose this option if your server - does not require a large number of concurrent connections. The - maximum number of connections is set at 100, with an average - of 20 concurrent connections assumed. - - * Online Transaction Processing (OLTP): Choose this option if - your server requires a large number of concurrent connections. - The maximum number of connections is set at 500. + * Several configure options apply to plugin selection and + building: +--with-plugins=PLUGIN[,PLUGIN]... +--with-plugins=GROUP +--with-plugin-PLUGIN +--without-plugin-PLUGIN + PLUGIN is an individual plugin name such as csv or archive. + As shorthand, GROUP is a configuration group name such as none + (select no plugins) or all (select all plugins). + You can build a plugin as static (compiled into the server) or + dynamic (built as a dynamic library that must be installed + using the INSTALL PLUGIN statement before it can be used). + Some plugins might not support static or dynamic build. + configure --help shows the following information pertaining to + plugins: - * Manual Setting: Choose this option to set the maximum number - of concurrent connections to the server manually. Choose the - number of concurrent connections from the drop-down box - provided, or enter the maximum number of connections into the - drop-down box if the number you desire is not listed. + + The plugin-related options -2.3.4.8. The Networking and Strict Mode Options Dialog + + The names of all available plugins - Use the Networking Options dialog to enable or disable TCP/IP - networking and to configure the port number that is used to - connect to the MySQL server. - MySQL Server Instance Configuration Wizard: Network Configuration + + For each plugin, a description of its purpose, which + build types it supports (static or dynamic), and which + plugin groups it is a part of. + --with-plugins can take a list of one or more plugin names + separated by commas, or a plugin group name. The named plugins + are configured to be built as static plugins. + --with-plugin-PLUGIN configures the given plugin to be built + as a static plugin. + --without-plugin-PLUGIN disables the given plugin from being + built. + If a plugin is named both with a --with and --without option, + the result is undefined. + For any plugin that is not explicitly selected or disabled, it + is selected to be built dynamically if it supports dynamic + build, and not built if it does not support dynamic build. + (Thus, in the case that no plugin options are given, all + plugins that support dynamic build are selected to be built as + dynamic plugins. Plugins that do not support dynamic build are + not built.) - TCP/IP networking is enabled by default. To disable TCP/IP - networking, uncheck the box next to the Enable TCP/IP Networking - option. +2.3.3. Installing from the Development Source Tree - Port 3306 is used by default. To change the port used to access - MySQL, choose a new port number from the drop-down box or type a - new port number directly into the drop-down box. If the port - number you choose is in use, you are prompted to confirm your - choice of port number. +Caution - Set the Server SQL Mode to either enable or disable strict mode. - Enabling strict mode (default) makes MySQL behave more like other - database management systems. If you run applications that rely on - MySQL's old "forgiving" behavior, make sure to either adapt those - applications or to disable strict mode. For more information about - strict mode, see Section 5.1.8, "Server SQL Modes." + You should read this section only if you are interested in helping + us test our new code. If you just want to get MySQL up and running + on your system, you should use a standard release distribution + (either a binary or source distribution). -2.3.4.9. The Character Set Dialog + To obtain the most recent development source tree, you must have + Bazaar installed. You can obtain Bazaar from the Bazaar VCS + Website (http://bazaar-vcs.org) Bazaar is supported by any + platform that supports Python, and is therefore compatible with + any Linux, Unix, Windows or Mac OS X host. Instructions for + downloading and installing Bazaar on the different platforms are + available on the Bazaar website. - The MySQL server supports multiple character sets and it is - possible to set a default server character set that is applied to - all tables, columns, and databases unless overridden. Use the - Character Set dialog to change the default character set of the - MySQL server. - MySQL Server Instance Configuration Wizard: Character Set + All MySQL projects are hosted on Launchpad + (http://launchpad.net/) MySQL projects, including MySQL server, + MySQL Workbench, and others are available from the Sun/MySQL + Engineering (http://launchpad.net/~mysql) page. For the + repositories related only to MySQL server, see the MySQL Server + (http://launchpad.net/mysql-server) page. - * Standard Character Set: Choose this option if you want to use - latin1 as the default server character set. latin1 is used for - English and many Western European languages. + To build under Unix/Linux, you must have the following tools + installed: - * Best Support For Multilingualism: Choose this option if you - want to use utf8 as the default server character set. This is - a Unicode character set that can store characters from many - different languages. + * GNU make, available from http://www.gnu.org/software/make/. + Although some platforms come with their own make + implementations, it is highly recommended that you use GNU + make. It may already be available on your system as gmake. - * Manual Selected Default Character Set / Collation: Choose this - option if you want to pick the server's default character set - manually. Choose the desired character set from the provided - drop-down list. + * autoconf 2.58 (or newer), available from + http://www.gnu.org/software/autoconf/. -2.3.4.10. The Service Options Dialog + * automake 1.8.1, available from + http://www.gnu.org/software/automake/. - On Windows platforms, the MySQL server can be installed as a - Windows service. When installed this way, the MySQL server can be - started automatically during system startup, and even restarted - automatically by Windows in the event of a service failure. + * libtool 1.5, available from + http://www.gnu.org/software/libtool/. - The MySQL Server Instance Configuration Wizard installs the MySQL - server as a service by default, using the service name MySQL. If - you do not wish to install the service, uncheck the box next to - the Install As Windows Service option. You can change the service - name by picking a new service name from the drop-down box provided - or by entering a new service name into the drop-down box. + * m4, available from http://www.gnu.org/software/m4/. -Note + * bison, available from http://www.gnu.org/software/bison/. You + should use the latest version of bison where possible. Version + 1.75 and version 2.1 are known to work. There have been + reported problems with bison 1.875. If you experience + problems, upgrade to a later, rather than earlier, version. + Versions of bison older than 1.75 may report this error: +sql_yacc.yy:#####: fatal error: maximum table size (32767) exceeded + The maximum table size is not actually exceeded; the error is + caused by bugs in older versions of bison. - Service names can include any legal character except forward (/) - or backward (\) slashes, and must be less than 256 characters - long. + To build under Windows you must have Microsoft Visual C++ 2005 + Express Edition, Visual Studio .Net 2003 (7.1), or Visual Studio + 2005 (8.0) compiler system. -Warning + Once the necessary tools are installed, you must create a local + branch of the MySQL source code on your machine: - If you are installing multiple versions of MySQL onto the same - machine, you must choose a different service name for each version - that you install. If you do not choose a different service for - each installed version then the service manager information will - be inconsistent and this will cause problems when you try to - uninstall a previous version. + 1. To obtain a copy of the MySQL source code, you must create a + new Bazaar branch. If you do not already have a Bazaar + repository directory set up, you need to initialize a new + directory: +shell> mkdir mysql-server +shell> bzr init-repo --trees mysql-server - If you have already installed multiple versions using the same - service name, you must manually edit the contents of the - HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services parameters - within the Windows registry to update the association of the - service name with the correct server version. + 2. Once you have an initialized directory, you can branch from + the public MySQL server repositories to create a local source + tree. To create a branch of a specific version: +shell> cd mysql-server +shell> bzr branch lp:mysql-server/5.1 mysql-5.1 - Typically, when installing multiple versions you create a service - name based on the version information. For example, you might - install MySQL 5.x as mysql5, or specific versions such as MySQL - 5.1.30 as mysql50130. + 3. The initial download will take some time to complete, + depending on the speed of your connection. Please be patient. + Once you have downloaded the first tree, additional trees + should take significantly less time to download. - To install the MySQL server as a service but not have it started - automatically at startup, uncheck the box next to the Launch the - MySQL Server Automatically option. + 4. When building from the Bazaar branch, you may want to create a + copy of your active branch so that you can make configuration + and other changes without affecting the original branch + contents. You can achieve this by branching from the original + branch: +shell> bzr branch mysql-5.1 mysql-5.1-build -2.3.4.11. The Security Options Dialog + 5. To obtain changes made after you have set up the branch + initially, update it using the pull option periodically. Use + this command in the top-level directory of the local copy: +shell> bzr pull + You can examine the changeset comments for the tree by using + the log option to bzr: +shell> bzr log + You can also browse changesets, comments, and source code + online. To browse this information for MySQL 5.1, go to the + Launchpad MySQL Server (http://launchpad.net/mysql-server) + page. + If you see diffs (changes) or code that you have a question + about, do not hesitate to send email to the MySQL internals + mailing list. See Section 1.5.1, "MySQL Mailing Lists." Also, + if you think you have a better idea on how to do something, + send an email message to the list with a patch. + + After you have the local branch, you can build MySQL server from + the source code. On Windows, the build process is different from + Unix/Linux: see Section 2.5.10, "Installing MySQL from Source on + Windows." - It is strongly recommended that you set a root password for your - MySQL server, and the MySQL Server Instance Configuration Wizard - requires by default that you do so. If you do not wish to set a - root password, uncheck the box next to the Modify Security - Settings option. - MySQL Server Instance Configuration Wizard: Security + On Unix/Linux, use the autoconf system to create the configure + script so that you can configure the build environment before + building. The following example shows the typical commands + required to build MySQL from a source tree. - To set the root password, enter the desired password into both the - New root password and Confirm boxes. If you are reconfiguring an - existing server, you need to enter the existing root password into - the Current root password box. + 1. Change location to the top-level directory of the source tree; + replace mysql-5.1 with the appropriate directory name. +shell> cd mysql-5.1 - To prevent root logins from across the network, check the box next - to the Root may only connect from localhost option. This increases - the security of your root account. + 2. Prepare the source tree for configuration. + Prior to MySQL 5.1.12, you must separately configure the + InnoDB storage engine. Run the following command from the main + source directory: +shell> (cd storage/innobase; autoreconf --force --install) + You can omit the previous command for MySQL 5.1.12 and later, + or if you do not require InnoDB support. + Prepare the remainder of the source tree: +shell> autoreconf --force --install + As an alternative to the preceding autoreconf command, you can + use BUILD/autorun.sh, which acts as a shortcut for the + following sequence of commands: +shell> aclocal; autoheader +shell> libtoolize --automake --force +shell> automake --force --add-missing; autoconf + If you get some strange errors during this stage, verify that + you have the correct version of libtool installed. - To create an anonymous user account, check the box next to the - Create An Anonymous Account option. Creating an anonymous account - can decrease server security and cause login and permission - difficulties. For this reason, it is not recommended. + 3. Configure the source tree and compile MySQL: +shell> ./configure # Add your favorite options here +shell> make + For a description of some configure options, see Section + 2.3.2, "Typical configure Options." + A collection of our standard configuration scripts is located + in the BUILD/ subdirectory. For example, you may find it more + convenient to use the BUILD/compile-pentium-debug script than + the preceding set of shell commands. To compile on a different + architecture, modify the script by removing flags that are + Pentium-specific, or use another script that may be more + appropriate. These scripts are provided on an "as-is" basis. + They are not officially maintained and their contents may + change from release to release. -2.3.4.12. The Confirmation Dialog + 4. When the build is done, run make install. Be careful with this + on a production machine; the command may overwrite your live + release installation. If you already have MySQL installed and + do not want to overwrite it, run ./configure with values for + the --prefix, --with-tcp-port, and --with-unix-socket-path + options different from those used for your production server. - The final dialog in the MySQL Server Instance Configuration Wizard - is the Confirmation Dialog. To start the configuration process, - click the Execute button. To return to a previous dialog, click - the Back button. To exit the MySQL Server Instance Configuration - Wizard without configuring the server, click the Cancel button. - MySQL Server Instance Configuration Wizard: Confirmation + 5. Play hard with your new installation and try to make the new + features crash. Start by running make test. See Section + 22.1.2, "MySQL Test Suite." - After you click the Execute button, the MySQL Server Instance - Configuration Wizard performs a series of tasks and displays the - progress onscreen as the tasks are performed. + 6. If you have gotten to the make stage, but the distribution + does not compile, please enter the problem into our bugs + database using the instructions given in Section 1.6, "How to + Report Bugs or Problems." If you have installed the latest + versions of the required GNU tools, and they crash trying to + process our configuration files, please report that also. + However, if you get a command not found error or a similar + problem for aclocal, configure, or other required tools, do + not report it. Instead, make sure that all the required tools + are installed and that your PATH variable is set correctly so + that your shell can find them. - The MySQL Server Instance Configuration Wizard first determines - configuration file options based on your choices using a template - prepared by MySQL developers and engineers. This template is named - my-template.ini and is located in your server installation - directory. +2.3.4. Dealing with Problems Compiling MySQL - The MySQL Configuration Wizard then writes these options to the - corresponding configuration file. + All MySQL programs compile cleanly for us with no warnings on + Solaris or Linux using gcc. On other systems, warnings may occur + due to differences in system include files. See Section 2.3.5, + "MIT-pthreads Notes," for warnings that may occur when using + MIT-pthreads. For other problems, check the following list. - If you chose to create a service for the MySQL server, the MySQL - Server Instance Configuration Wizard creates and starts the - service. If you are reconfiguring an existing service, the MySQL - Server Instance Configuration Wizard restarts the service to apply - your configuration changes. - - If you chose to set a root password, the MySQL Configuration - Wizard connects to the server, sets your new root password, and - applies any other security settings you may have selected. - - After the MySQL Server Instance Configuration Wizard has completed - its tasks, it displays a summary. Click the Finish button to exit - the MySQL Server Configuration Wizard. + The solution to many problems involves reconfiguring. If you do + need to reconfigure, take note of the following: -2.3.5. Installing MySQL from a Noinstall Zip Archive + * If configure is run after it has previously been run, it may + use information that was gathered during its previous + invocation. This information is stored in config.cache. When + configure starts up, it looks for that file and reads its + contents if it exists, on the assumption that the information + is still correct. That assumption is invalid when you + reconfigure. - Users who are installing from the Noinstall package can use the - instructions in this section to manually install MySQL. The - process for installing MySQL from a Zip archive is as follows: + * Each time you run configure, you must run make again to + recompile. However, you may want to remove old object files + from previous builds first because they were compiled using + different configuration options. - 1. Extract the archive to the desired install directory + To prevent old configuration information or object files from + being used, run these commands before re-running configure: +shell> rm config.cache +shell> make clean - 2. Create an option file + Alternatively, you can run make distclean. - 3. Choose a MySQL server type + The following list describes some of the problems when compiling + MySQL that have been found to occur most often: - 4. Start the MySQL server + * If you get errors such as the ones shown here when compiling + sql_yacc.cc, you probably have run out of memory or swap + space: +Internal compiler error: program cc1plus got fatal signal 11 +Out of virtual memory +Virtual memory exhausted + The problem is that gcc requires a huge amount of memory to + compile sql_yacc.cc with inline functions. Try running + configure with the --with-low-memory option: +shell> ./configure --with-low-memory + This option causes -fno-inline to be added to the compile line + if you are using gcc and -O0 if you are using something else. + You should try the --with-low-memory option even if you have + so much memory and swap space that you think you can't + possibly have run out. This problem has been observed to occur + even on systems with generous hardware configurations, and the + --with-low-memory option usually fixes it. - 5. Secure the default user accounts + * By default, configure picks c++ as the compiler name and GNU + c++ links with -lg++. If you are using gcc, that behavior can + cause problems during configuration such as this: +configure: error: installation or configuration problem: +C++ compiler cannot create executables. + You might also observe problems during compilation related to + g++, libg++, or libstdc++. + One cause of these problems is that you may not have g++, or + you may have g++ but not libg++, or libstdc++. Take a look at + the config.log file. It should contain the exact reason why + your C++ compiler didn't work. To work around these problems, + you can use gcc as your C++ compiler. Try setting the + environment variable CXX to "gcc -O3". For example: +shell> CXX="gcc -O3" ./configure + This works because gcc compiles C++ source files as well as + g++ does, but does not link in libg++ or libstdc++ by default. + Another way to fix these problems is to install g++, libg++, + and libstdc++. However, do not use libg++ or libstdc++ with + MySQL because this only increases the binary size of mysqld + without providing any benefits. Some versions of these + libraries have also caused strange problems for MySQL users in + the past. - This process is described in the sections that follow. + * If your compile fails with errors such as any of the + following, you must upgrade your version of make to GNU make: +making all in mit-pthreads +make: Fatal error in reader: Makefile, line 18: +Badly formed macro assignment + Or: +make: file `Makefile' line 18: Must be a separator (: + Or: +pthread.h: No such file or directory + Solaris and FreeBSD are known to have troublesome make + programs. + GNU make 3.75 is known to work. -2.3.6. Extracting the Install Archive + * If you want to define flags to be used by your C or C++ + compilers, do so by adding the flags to the CFLAGS and + CXXFLAGS environment variables. You can also specify the + compiler names this way using CC and CXX. For example: +shell> CC=gcc +shell> CFLAGS=-O3 +shell> CXX=gcc +shell> CXXFLAGS=-O3 +shell> export CC CFLAGS CXX CXXFLAGS + See Section 2.2, "Installing MySQL from Generic Binaries on + Unix/Linux," for a list of flag definitions that have been + found to be useful on various systems. - To install MySQL manually, do the following: + * If you get errors such as those shown here when compiling + mysqld, configure did not correctly detect the type of the + last argument to accept(), getsockname(), or getpeername(): +cxx: Error: mysqld.cc, line 645: In this statement, the referenced + type of the pointer value ''length'' is ''unsigned long'', + which is not compatible with ''int''. +new_sock = accept(sock, (struct sockaddr *)&cAddr, &length); + To fix this, edit the config.h file (which is generated by + configure). Look for these lines: +/* Define as the base type of the last arg to accept */ +#define SOCKET_SIZE_TYPE XXX + Change XXX to size_t or int, depending on your operating + system. (You must do this each time you run configure because + configure regenerates config.h.) - 1. If you are upgrading from a previous version please refer to - Section 2.3.14, "Upgrading MySQL on Windows," before beginning - the upgrade process. + * The sql_yacc.cc file is generated from sql_yacc.yy. Normally, + the build process does not need to create sql_yacc.cc because + MySQL comes with a pre-generated copy. However, if you do need + to re-create it, you might encounter this error: +"sql_yacc.yy", line xxx fatal: default action causes potential... + This is a sign that your version of yacc is deficient. You + probably need to install bison (the GNU version of yacc) and + use that instead. - 2. Make sure that you are logged in as a user with administrator - privileges. + * On Debian Linux 3.0, you need to install gawk instead of the + default mawk. - 3. Choose an installation location. Traditionally, the MySQL - server is installed in C:\mysql. The MySQL Installation Wizard - installs MySQL under C:\Program Files\MySQL. If you do not - install MySQL at C:\mysql, you must specify the path to the - install directory during startup or in an option file. See - Section 2.3.7, "Creating an Option File." + * If you need to debug mysqld or a MySQL client, run configure + with the --with-debug option, and then recompile and link your + clients with the new client library. See MySQL Internals: + Porting (http://forge.mysql.com/wiki/MySQL_Internals_Porting) - 4. Extract the install archive to the chosen installation - location using your preferred Zip archive tool. Some tools may - extract the archive to a folder within your chosen - installation location. If this occurs, you can move the - contents of the subfolder into the chosen installation - location. + * If you get a compilation error on Linux (for example, SuSE + Linux 8.1 or Red Hat Linux 7.3) similar to the following one, + you probably do not have g++ installed: +libmysql.c:1329: warning: passing arg 5 of `gethostbyname_r' from +incompatible pointer type +libmysql.c:1329: too few arguments to function `gethostbyname_r' +libmysql.c:1329: warning: assignment makes pointer from integer +without a cast +make[2]: *** [libmysql.lo] Error 1 + By default, the configure script attempts to determine the + correct number of arguments by using g++ (the GNU C++ + compiler). This test yields incorrect results if g++ is not + installed. There are two ways to work around this problem: -2.3.7. Creating an Option File + + Make sure that the GNU C++ g++ is installed. On some + Linux distributions, the required package is called gpp; + on others, it is named gcc-c++. - If you need to specify startup options when you run the server, - you can indicate them on the command line or place them in an - option file. For options that are used every time the server - starts, you may find it most convenient to use an option file to - specify your MySQL configuration. This is particularly true under - the following circumstances: + + Use gcc as your C++ compiler by setting the CXX + environment variable to gcc: +export CXX="gcc" + You must run configure again after making either of those + changes. - * The installation or data directory locations are different - from the default locations (C:\Program Files\MySQL\MySQL - Server 5.1 and C:\Program Files\MySQL\MySQL Server 5.1\data). +2.3.5. MIT-pthreads Notes - * You need to tune the server settings, such as memory, cache, - or InnoDB configuration information. + This section describes some of the issues involved in using + MIT-pthreads. - When the MySQL server starts on Windows, it looks for options in - two files: the my.ini file in the Windows directory, and the - C:\my.cnf file. The Windows directory typically is named something - like C:\WINDOWS. You can determine its exact location from the - value of the WINDIR environment variable using the following - command: -C:\> echo %WINDIR% + On Linux, you should not use MIT-pthreads. Use the installed + LinuxThreads implementation instead. See Section 2.6, "Installing + MySQL on Linux." - MySQL looks for options first in the my.ini file, and then in the - my.cnf file. However, to avoid confusion, it is best if you use - only one file. If your PC uses a boot loader where C: is not the - boot drive, your only option is to use the my.ini file. Whichever - option file you use, it must be a plain text file. + If your system does not provide native thread support, you should + build MySQL using the MIT-pthreads package. This includes older + FreeBSD systems, SunOS 4.x, Solaris 2.4 and earlier, and some + others. See Section 2.1, "General Installation Guidance." - You can also make use of the example option files included with - your MySQL distribution; see Section 4.2.3.3.2, "Preconfigured - Option Files." + MIT-pthreads is not part of the MySQL 5.1 source distribution. If + you require this package, you need to download it separately from + http://dev.mysql.com/Downloads/Contrib/pthreads-1_60_beta6-mysql.t + ar.gz - An option file can be created and modified with any text editor, - such as Notepad. For example, if MySQL is installed in E:\mysql - and the data directory is in E:\mydata\data, you can create an - option file containing a [mysqld] section to specify values for - the basedir and datadir options: -[mysqld] -# set basedir to your installation path -basedir=E:/mysql -# set datadir to the location of your data directory -datadir=E:/mydata/data + After downloading, extract this source archive into the top level + of the MySQL source directory. It creates a new subdirectory named + mit-pthreads. - Note that Windows path names are specified in option files using - (forward) slashes rather than backslashes. If you do use - backslashes, you must double them: -[mysqld] -# set basedir to your installation path -basedir=E:\\mysql -# set datadir to the location of your data directory -datadir=E:\\mydata\\data + * On most systems, you can force MIT-pthreads to be used by + running configure with the --with-mit-threads option: +shell> ./configure --with-mit-threads + Building in a nonsource directory is not supported when using + MIT-pthreads because we want to minimize our changes to this + code. - MySQL Enterprise For expert advice on the start-up options - appropriate to your circumstances, subscribe to the MySQL - Enterprise Monitor. For more information, see - http://www.mysql.com/products/enterprise/advisors.html. + * The checks that determine whether to use MIT-pthreads occur + only during the part of the configuration process that deals + with the server code. If you have configured the distribution + using --without-server to build only the client code, clients + do not know whether MIT-pthreads is being used and use Unix + socket file connections by default. Because Unix socket files + do not work under MIT-pthreads on some platforms, this means + you need to use -h or --host with a value other than localhost + when you run client programs. - In MySQL 5.1.23 and earlier, the MySQL installer places the data - directory directly under the directory where you install MySQL. On - MySQL 5.1.24 and later, the data directory is located within the - AppData directory for the user running MySQL. + * When MySQL is compiled using MIT-pthreads, system locking is + disabled by default for performance reasons. You can tell the + server to use system locking with the --external-locking + option. This is needed only if you want to be able to run two + MySQL servers against the same data files, but that is not + recommended, anyway. - If you would like to use a data directory in a different location, - you should copy the entire contents of the data directory to the - new location. For example, if you want to use E:\mydata as the - data directory instead, you must do two things: + * Sometimes the pthread bind() command fails to bind to a socket + without any error message (at least on Solaris). The result is + that all connections to the server fail. For example: +shell> mysqladmin version +mysqladmin: connect to server at '' failed; +error: 'Can't connect to mysql server on localhost (146)' + The solution to this problem is to kill the mysqld server and + restart it. This has happened to us only when we have forcibly + stopped the server and restarted it immediately. - 1. Move the entire data directory and all of its contents from - the default location (for example C:\Program Files\MySQL\MySQL - Server 5.1\data) to E:\mydata. + * With MIT-pthreads, the sleep() system call isn't interruptible + with SIGINT (break). This is noticeable only when you run + mysqladmin --sleep. You must wait for the sleep() call to + terminate before the interrupt is served and the process + stops. - 2. Use a --datadir option to specify the new data directory - location each time you start the server. + * When linking, you might receive warning messages like these + (at least on Solaris); they can be ignored: +ld: warning: symbol `_iob' has differing sizes: + (file /my/local/pthreads/lib/libpthread.a(findfp.o) value=0x4; +file /usr/lib/libc.so value=0x140); + /my/local/pthreads/lib/libpthread.a(findfp.o) definition taken +ld: warning: symbol `__iob' has differing sizes: + (file /my/local/pthreads/lib/libpthread.a(findfp.o) value=0x4; +file /usr/lib/libc.so value=0x140); + /my/local/pthreads/lib/libpthread.a(findfp.o) definition taken -2.3.8. Selecting a MySQL Server Type + * Some other warnings also can be ignored: +implicit declaration of function `int strtoll(...)' +implicit declaration of function `int strtoul(...)' - The following table shows the available servers for Windows in - MySQL 5.1.20 and earlier. - Binary Description - mysqld-nt Optimized binary with named-pipe support - mysqld Optimized binary without named-pipe support - mysqld-debug Like mysqld-nt, but compiled with full debugging and - automatic memory allocation checking + * We have not been able to make readline work with MIT-pthreads. + (This is not necessary, but may be of interest to some.) - The following table shows the available servers for Windows in - MySQL 5.1.21 and later. - Binary Description - mysqld Optimized binary with named-pipe support - mysqld-debug Like mysqld, but compiled with full debugging and - automatic memory allocation checking +2.4. Upgrading or Downgrading MySQL - All of the preceding binaries are optimized for modern Intel - processors, but should work on any Intel i386-class or higher - processor. +2.4.1. Upgrading MySQL - Each of the servers in a distribution support the same set of - storage engines. The SHOW ENGINES statement displays which engines - a given server supports. - - All Windows MySQL 5.1 servers have support for symbolic linking of - database directories. + As a general rule, to upgrade from one release series to another, + you should go to the next series rather than skipping a series. To + upgrade from a release series previous to MySQL 5.0, upgrade to + each successive release series in turn until you have reached + MySQL 5.0, and then proceed with the upgrade to MySQL 5.1. For + example, if you currently are running MySQL 4.0 and wish to + upgrade to a newer series, upgrade to MySQL 4.1 first before + upgrading to 5.0, and so forth. For information on upgrading to + MySQL 5.0, see the MySQL 5.0 Reference Manual; for earlier + releases, see the MySQL 3.23, 4.0, 4.1 Reference Manual. - MySQL supports TCP/IP on all Windows platforms. MySQL servers on - Windows support named pipes as indicated in the following list. - However, the default is to use TCP/IP regardless of platform. - (Named pipes are slower than TCP/IP in many Windows - configurations.) + If you perform a binary (in-place) upgrade without dumping and + reloading tables, you cannot upgrade directly from MySQL 4.1 to + 5.1. This occurs due to an incompatible change in the MyISAM table + index format in MySQL 5.0. Upgrade from MySQL 4.1 to 5.0 and + repair all MyISAM tables (see Section 2.4.4, "Rebuilding or + Repairing Tables or Indexes"). Then upgrade from MySQL 5.0 to 5.1 + and check and repair your tables. - Use of named pipes is subject to these conditions: + To upgrade from MySQL 5.0 to 5.1, use the items in the following + checklist as a guide: - * Named pipes are enabled only if you start the server with the - --enable-named-pipe option. It is necessary to use this option - explicitly because some users have experienced problems with - shutting down the MySQL server when named pipes were used. + * Before any upgrade, back up your databases, including the + mysql database that contains the grant tables. See Section + 6.1, "Database Backup Methods." - * For MySQL 5.1.20 and earlier, named-pipe connections are - allowed only by the mysqld-nt and mysqld-debug servers. For - MySQL 5.1.21 and later, the mysqld and mysqld-debug servers - both contain support for named-pipe connections. + * Read all the notes in Section 2.4.1.1, "Upgrading from MySQL + 5.0 to 5.1." These notes enable you to identify upgrade issues + that apply to your current MySQL installation. Some + incompatibilities discussed in that section require your + attention before upgrading. Others should be dealt with after + upgrading. -Note + * Read Appendix C, "MySQL Change History" as well, which + provides information about features that are new in MySQL 5.1 + or differ from those found in MySQL 5.0. - Most of the examples in this manual use mysqld as the server name. - If you choose to use a different server, such as mysqld-nt or - mysqld-debug, make the appropriate substitutions in the commands - that are shown in the examples. + * After you upgrade to a new version of MySQL, run mysql_upgrade + (see Section 4.4.8, "mysql_upgrade --- Check Tables for MySQL + Upgrade"). This program checks your tables, and attempts to + repair them if necessary. It also updates your grant tables to + make sure that they have the current structure so that you can + take advantage of any new capabilities. (Some releases of + MySQL introduce changes to the structure of the grant tables + to add new privileges or features.) -2.3.9. Starting the Server for the First Time + * If you are running MySQL Server on Windows, see Section 2.5.7, + "Upgrading MySQL on Windows." - This section gives a general overview of starting the MySQL - server. The following sections provide more specific information - for starting the MySQL server from the command line or as a - Windows service. + * If you are using replication, see Section 16.3.3, "Upgrading a + Replication Setup," for information on upgrading your + replication setup. - The information here applies primarily if you installed MySQL - using the Noinstall version, or if you wish to configure and test - MySQL manually rather than with the GUI tools. + * If you are upgrading an installation originally produced by + installing multiple RPM packages, it is best to upgrade all + the packages, not just some. For example, if you previously + installed the server and client RPMs, do not upgrade just the + server RPM. - The examples in these sections assume that MySQL is installed - under the default location of C:\Program Files\MySQL\MySQL Server - 5.1. Adjust the path names shown in the examples if you have MySQL - installed in a different location. + * As of MySQL 5.1.9, the mysqld-max server is included in binary + distributions. There is no separate MySQL-Max distribution. As + of MySQL 5.1.12, there is no mysqld-max server at all in + binary distributions. They contain a server that includes the + features previously included in mysqld-max. - Clients have two options. They can use TCP/IP, or they can use a - named pipe if the server supports named-pipe connections. + * If you have created a user-defined function (UDF) with a given + name and upgrade MySQL to a version that implements a new + built-in function with the same name, the UDF becomes + inaccessible. To correct this, use DROP FUNCTION to drop the + UDF, and then use CREATE FUNCTION to re-create the UDF with a + different nonconflicting name. The same is true if the new + version of MySQL implements a built-in function with the same + name as an existing stored function. See Section 8.2.4, + "Function Name Parsing and Resolution," for the rules + describing how the server interprets references to different + kinds of functions. - MySQL for Windows also supports shared-memory connections if the - server is started with the --shared-memory option. Clients can - connect through shared memory by using the --protocol=MEMORY - option. + You can always move the MySQL format files and data files between + different versions on systems with the same architecture as long + as you stay within versions for the same release series of MySQL. - For information about which server binary to run, see Section - 2.3.8, "Selecting a MySQL Server Type." + If you are cautious about using new versions, you can always + rename your old mysqld before installing a newer one. For example, + if you are using MySQL 5.0.13 and want to upgrade to 5.1.10, + rename your current server from mysqld to mysqld-5.0.13. If your + new mysqld then does something unexpected, you can simply shut it + down and restart with your old mysqld. - Testing is best done from a command prompt in a console window (or - "DOS window"). In this way you can have the server display status - messages in the window where they are easy to see. If something is - wrong with your configuration, these messages make it easier for - you to identify and fix any problems. + If, after an upgrade, you experience problems with recompiled + client programs, such as Commands out of sync or unexpected core + dumps, you probably have used old header or library files when + compiling your programs. In this case, you should check the date + for your mysql.h file and libmysqlclient.a library to verify that + they are from the new MySQL distribution. If not, recompile your + programs with the new headers and libraries. - To start the server, enter this command: -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" --console + If problems occur, such as that the new mysqld server does not + start or that you cannot connect without a password, verify that + you do not have an old my.cnf file from your previous + installation. You can check this with the --print-defaults option + (for example, mysqld --print-defaults). If this command displays + anything other than the program name, you have an active my.cnf + file that affects server or client operation. - For a server that includes InnoDB support, you should see the - messages similar to those following as it starts (the path names - and sizes may differ): -InnoDB: The first specified datafile c:\ibdata\ibdata1 did not exist: -InnoDB: a new database to be created! -InnoDB: Setting file c:\ibdata\ibdata1 size to 209715200 -InnoDB: Database physically writes the file full: wait... -InnoDB: Log file c:\iblogs\ib_logfile0 did not exist: new to be creat -ed -InnoDB: Setting log file c:\iblogs\ib_logfile0 size to 31457280 -InnoDB: Log file c:\iblogs\ib_logfile1 did not exist: new to be creat -ed -InnoDB: Setting log file c:\iblogs\ib_logfile1 size to 31457280 -InnoDB: Log file c:\iblogs\ib_logfile2 did not exist: new to be creat -ed -InnoDB: Setting log file c:\iblogs\ib_logfile2 size to 31457280 -InnoDB: Doublewrite buffer not found: creating new -InnoDB: Doublewrite buffer created -InnoDB: creating foreign key constraint system tables -InnoDB: foreign key constraint system tables created -011024 10:58:25 InnoDB: Started + If your MySQL installation contains a large amount of data that + might take a long time to convert after an in-place upgrade, you + might find it useful to create a "dummy" database instance for + assessing what conversions might be needed and the work involved + to perform them. Make a copy of your MySQL instance that contains + a full copy of the mysql database, plus all other databases + without data. Run your upgrade procedure on this dummy instance to + see what actions might be needed so that you can better evaluate + the work involved when performing actual data conversion on your + original database instance. - When the server finishes its startup sequence, you should see - something like this, which indicates that the server is ready to - service client connections: -mysqld: ready for connections -Version: '5.1.39' socket: '' port: 3306 + It is a good idea to rebuild and reinstall the Perl DBD::mysql + module whenever you install a new release of MySQL. The same + applies to other MySQL interfaces as well, such as PHP mysql + extensions and the Python MySQLdb module. - The server continues to write to the console any further - diagnostic output it produces. You can open a new console window - in which to run client programs. +2.4.1.1. Upgrading from MySQL 5.0 to 5.1 - If you omit the --console option, the server writes diagnostic - output to the error log in the data directory (C:\Program - Files\MySQL\MySQL Server 5.1\data by default). The error log is - the file with the .err extension. + After upgrading a 5.0 installation to 5.0.10 or above, it is + necessary to upgrade your grant tables. Otherwise, creating stored + procedures and functions might not work. To perform this upgrade, + run mysql_upgrade. Note - The accounts that are listed in the MySQL grant tables initially - have no passwords. After starting the server, you should set up - passwords for them using the instructions in Section 2.11, - "Post-Installation Setup and Testing." + It is good practice to back up your data before installing any new + version of software. Although MySQL works very hard to ensure a + high level of quality, you should protect your data by making a + backup. -2.3.10. Starting MySQL from the Windows Command Line + To upgrade to 5.1 from any previous version, MySQL recommends that + you dump your tables with mysqldump before upgrading and reload + the dump file after upgrading. - The MySQL server can be started manually from the command line. - This can be done on any version of Windows. + If you perform a binary (in-place) upgrade without dumping and + reloading tables, you cannot upgrade directly from MySQL 4.1 to + 5.1. This occurs due to an incompatible change in the MyISAM table + index format in MySQL 5.0. Upgrade from MySQL 4.1 to 5.0 and + repair all MyISAM tables (see Section 2.4.4, "Rebuilding or + Repairing Tables or Indexes"). Then upgrade from MySQL 5.0 to 5.1 + and check and repair your tables. - To start the mysqld server from the command line, you should start - a console window (or "DOS window") and enter this command: -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" + In general, you should do the following when upgrading from MySQL + 5.0 to 5.1: - The path to mysqld may vary depending on the install location of - MySQL on your system. + * Read all the items in the following sections to see whether + any of them might affect your applications: - You can stop the MySQL server by executing this command: -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqladmin" -u root - shutdown + + Section 2.4.1, "Upgrading MySQL," has general update + information. -Note + + The items in the change lists found later in this section + enable you to identify upgrade issues that apply to your + current MySQL installation. - If the MySQL root user account has a password, you need to invoke - mysqladmin with the -p option and supply the password when - prompted. + + The MySQL 5.1 change history describes significant new + features you can use in 5.1 or that differ from those + found in MySQL 5.0. Some of these changes may result in + incompatibilities. See Section C.1, "Changes in Release + 5.1.x (Production)." - This command invokes the MySQL administrative utility mysqladmin - to connect to the server and tell it to shut down. The command - connects as the MySQL root user, which is the default - administrative account in the MySQL grant system. Note that users - in the MySQL grant system are wholly independent from any login - users under Windows. + * Note particularly any changes that are marked Known issue or + Incompatible change. These incompatibilities with earlier + versions of MySQL may require your attention before you + upgrade. + Our aim is to avoid these changes, but occasionally they are + necessary to correct problems that would be worse than an + incompatibility between releases. If any upgrade issue + applicable to your installation involves an incompatibility + that requires special handling, follow the instructions given + in the incompatibility description. Often this will involve a + dump and reload, or use of a statement such as CHECK TABLE or + REPAIR TABLE. + For dump and reload instructions, see Section 2.4.4, + "Rebuilding or Repairing Tables or Indexes." Any procedure + that involves REPAIR TABLE with the USE_FRM option must be + done before upgrading. Use of this statement with a version of + MySQL different from the one used to create the table (that + is, using it after upgrading) may damage the table. See + Section 12.5.2.6, "REPAIR TABLE Syntax." - If mysqld doesn't start, check the error log to see whether the - server wrote any messages there to indicate the cause of the - problem. The error log is located in the C:\Program - Files\MySQL\MySQL Server 5.1\data directory. It is the file with a - suffix of .err. You can also try to start the server as mysqld - --console; in this case, you may get some useful information on - the screen that may help solve the problem. + * After you upgrade to a new version of MySQL, run mysql_upgrade + (see Section 4.4.8, "mysql_upgrade --- Check Tables for MySQL + Upgrade"). This program checks your tables, and attempts to + repair them if necessary. It also updates your grant tables to + make sure that they have the current structure so that you can + take advantage of any new capabilities. (Some releases of + MySQL introduce changes to the structure of the grant tables + to add new privileges or features.) - The last option is to start mysqld with the --standalone and - --debug options. In this case, mysqld writes a log file - C:\mysqld.trace that should contain the reason why mysqld doesn't - start. See MySQL Internals: Porting - (http://forge.mysql.com/wiki/MySQL_Internals_Porting) - - Use mysqld --verbose --help to display all the options that mysqld - supports. - -2.3.11. Starting MySQL as a Windows Service - - On Windows, the recommended way to run MySQL is to install it as a - Windows service, whereby MySQL starts and stops automatically when - Windows starts and stops. A MySQL server installed as a service - can also be controlled from the command line using NET commands, - or with the graphical Services utility. Generally, to install - MySQL as a Windows service you should be logged in using an - account that has administrator rights. + * Check Section 2.4.3, "Checking Whether Tables or Indexes Must + Be Rebuilt," to see whether changes to table formats or to + character sets or collations were made between your current + version of MySQL and the version to which you are upgrading. + If so and these changes result in an incompatibility between + MySQL versions, you will need to upgrade the affected tables + using the instructions in Section 2.4.4, "Rebuilding or + Repairing Tables or Indexes." - The Services utility (the Windows Service Control Manager) can be - found in the Windows Control Panel (under Administrative Tools on - Windows 2000, XP, Vista and Server 2003). To avoid conflicts, it - is advisable to close the Services utility while performing server - installation or removal operations from the command line. + * If you are running MySQL Server on Windows, see Section 2.5.7, + "Upgrading MySQL on Windows." - Before installing MySQL as a Windows service, you should first - stop the current server if it is running by using the following - command: -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqladmin" - -u root shutdown + * If you are using replication, see Section 16.3.3, "Upgrading a + Replication Setup," for information on upgrading your + replication setup. -Note + If your MySQL installation contains a large amount of data that + might take a long time to convert after an in-place upgrade, you + might find it useful to create a "dummy" database instance for + assessing what conversions might be needed and the work involved + to perform them. Make a copy of your MySQL instance that contains + a full copy of the mysql database, plus all other databases + without data. Run your upgrade procedure on this dummy instance to + see what actions might be needed so that you can better evaluate + the work involved when performing actual data conversion on your + original database instance. - If the MySQL root user account has a password, you need to invoke - mysqladmin with the -p option and supply the password when - prompted. + MySQL Enterprise MySQL Enterprise subscribers will find more + information about upgrading in the Knowledge Base articles found + at Upgrading + (https://kb.mysql.com/search.php?cat=search&category=41) Access + to the MySQL Knowledge Base collection of articles is one of the + advantages of subscribing to MySQL Enterprise. For more + information, see + http://www.mysql.com/products/enterprise/advisors.html. - This command invokes the MySQL administrative utility mysqladmin - to connect to the server and tell it to shut down. The command - connects as the MySQL root user, which is the default - administrative account in the MySQL grant system. Note that users - in the MySQL grant system are wholly independent from any login - users under Windows. + The following lists describe changes that may affect applications + and that you should watch out for when upgrading to MySQL 5.1. - Install the server as a service using this command: -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" --install + Configuration Changes: - The service-installation command does not start the server. - Instructions for that are given later in this section. + * Before MySQL 5.1.11, to build MySQL from source with SSL + support enabled, you would invoke configure with either the + --with-openssl or --with-yassl option. In MySQL 5.1.11, those + options both have been replaced by the --with-ssl option. By + default, --with-ssl causes the bundled yaSSL library to be + used. To select OpenSSL instead, give the option as + --with-ssl=path, where path is the directory where the OpenSSL + header files and libraries are located. - To make it easier to invoke MySQL programs, you can add the path - name of the MySQL bin directory to your Windows system PATH - environment variable: + Server Changes: - * On the Windows desktop, right-click on the My Computer icon, - and select Properties. + * Known issue: After a binary upgrade to MySQL 5.1 from a MySQL + 5.0 installation that contains ARCHIVE tables, accessing those + tables will cause the server to crash, even if you have run + mysql_upgrade or CHECK TABLE ... FOR UPGRADE. To work around + this problem, use mysqldump to dump all ARCHIVE tables before + upgrading, and reload them into MySQL 5.1 after upgrading. + + * Known issue: The fix for + Bug#23491: http://bugs.mysql.com/23491 introduced a problem + with SHOW CREATE VIEW, which is used by mysqldump. This causes + an incompatibility when upgrading from versions affected by + that bug fix (MySQL 5.0.40 through 5.0.43, MySQL 5.1.18 + through 5.1.19): If you use mysqldump before upgrading from an + affected version and reload the data after upgrading to a + higher version, you must drop and recreate your views. - * Next select the Advanced tab from the System Properties menu - that appears, and click the Environment Variables button. + * Known issue: Dumps performed by using mysqldump to generate a + dump file before the upgrade and reloading the file after + upgrading are subject to the following problem: + Before MySQL 5.0.40, mysqldump displays SPATIAL index + definitions using prefix lengths for the indexed columns. + These prefix lengths are accepted in MySQL 5.0, but not as of + MySQL 5.1. If you use mysqldump from versions of MySQL older + than 5.0.40, any table containing SPATIAL indexes will cause + an error when the dump file is reloaded into MySQL 5.1 or + higher. + For example, a table definition might look like this when + dumped in MySQL 5.0: +CREATE TABLE `t` ( + `g` geometry NOT NULL, + SPATIAL KEY `g` (`g`(32)) +) ENGINE=MyISAM DEFAULT CHARSET=latin1 + The SPATIAL index definition will not be accepted in MySQL + 5.1. To work around this, edit the dump file to remove the + prefix: +CREATE TABLE `t` ( + `g` geometry NOT NULL, + SPATIAL KEY `g` (`g`) +) ENGINE=MyISAM DEFAULT CHARSET=latin1 + Dump files can be large, so it may be preferable to dump table + definitions and data separately to make it easier to edit the + definitions: +shell> mysqldump --no-data other_args > definitions.sql +shell> mysqldump --no-create-info other_args > data.sql + Then edit definitions.sql before reloading definitions.sql and + data.sql, in that order. + If you upgrade to a version of MySQL 5.0 higher than 5.0.40 + before upgrading to MySQL 5.1, this problem does not occur. - * Under System Variables, select Path, and then click the Edit - button. The Edit System Variable dialogue should appear. + * Known issue: Before MySQL 5.1.30, the CHECK TABLE ... FOR + UPGRADE statement did not check for incompatible collation + changes made in MySQL 5.1.24. (This also affects mysqlcheck + and mysql_upgrade, which cause that statement to be executed.) + Prior to the fix made in 5.1.30, a binary upgrade (performed + without dumping tables with mysqldump before the upgrade and + reloading the dump file after the upgrade) would corrupt + tables. After the fix, CHECK TABLE ... FOR UPGRADE properly + detects the problem and warns about tables that need repair. + However, the fix is not backward compatible and can result in + a downgrading problem under these circumstances: - * Place your cursor at the end of the text appearing in the - space marked Variable Value. (Use the End key to ensure that - your cursor is positioned at the very end of the text in this - space.) Then enter the complete path name of your MySQL bin - directory (for example, C:\Program Files\MySQL\MySQL Server - 5.1\bin), Note that there should be a semicolon separating - this path from any values present in this field. Dismiss this - dialogue, and each dialogue in turn, by clicking OK until all - of the dialogues that were opened have been dismissed. You - should now be able to invoke any MySQL executable program by - typing its name at the DOS prompt from any directory on the - system, without having to supply the path. This includes the - servers, the mysql client, and all MySQL command-line - utilities such as mysqladmin and mysqldump. - You should not add the MySQL bin directory to your Windows - PATH if you are running multiple MySQL servers on the same - machine. + 1. Perform a binary upgrade to a version of MySQL that + includes the fix. -Warning + 2. Run CHECK TABLE ... FOR UPGRADE (or mysqlcheck or + mysql_upgrade) to upgrade tables. - You must exercise great care when editing your system PATH by - hand; accidental deletion or modification of any portion of the - existing PATH value can leave you with a malfunctioning or even - unusable system. + 3. Perform a binary downgrade to a version of MySQL that + does not include the fix. + The solution is to dump tables with mysqldump before the + downgrade and reload the dump file after the downgrade. + Alternatively, drop and recreate affected indexes. - The following additional arguments can be used in MySQL 5.1 when - installing the service: + * Known issue: MySQL introduces encoding for table names that + have non-ASCII characters (see Section 8.2.3, "Mapping of + Identifiers to File Names"). After a binary upgrade from MySQL + 5.0 to 5.1 or higher, the server recognizes names that have + non-ASCII characters and adds a #mysql50# prefix to them. + As of MySQL 5.1.31, mysql_upgrade encodes these names by + executing the following command: +mysqlcheck --all-databases --check-upgrade --fix-db-names --fix-table +-names + Prior to MySQL 5.1.31, mysql_upgrade does not execute this + command, so you should execute it manually if you have + database or table names that contain nonalphanumeric + characters. + Prior to MySQL 5.1.23, the mysqlcheck command does not perform + the name encoding for views. To work around this problem, drop + each affected view and recreate it. + mysqlcheck cannot fix names that contain literal instances of + the @ character that is used for encoding special characters. + If you have databases or tables that contain this character, + use mysqldump to dump them before upgrading to MySQL 5.1, and + then reload the dump file after upgrading. - * You can specify a service name immediately following the - --install option. The default service name is MySQL. + * Known issue: When upgrading from MySQL 5.0 to versions of 5.1 + prior to 5.1.23, running mysqlcheck (or mysql_upgrade, which + runs mysqlcheck) to upgrade tables fails for names that must + be written as quoted identifiers. To work around this problem, + rename each affected table to a name that does not require + quoting: +RENAME TABLE `tab``le_a` TO table_a; +RENAME TABLE `table b` TO table_b; + After renaming the tables, run the mysql_upgrade program. Then + rename the tables back to their original names: +RENAME TABLE table_a TO `tab``le_a`; +RENAME TABLE table_b TO `table b`; - * If a service name is given, it can be followed by a single - option. By convention, this should be - --defaults-file=file_name to specify the name of an option - file from which the server should read options when it starts. - The use of a single option other than --defaults-file is - possible but discouraged. --defaults-file is more flexible - because it enables you to specify multiple startup options for - the server by placing them in the named option file. + * Known issue: In connection with view creation, the server + created arc directories inside database directories and + maintained useless copies of .frm files there. Creation and + renaming procedures of those copies as well as creation of arc + directories has been discontinued in MySQL 5.1.29. + This change does cause a problem when downgrading to older + server versions which manifests itself under these + circumstances: - * You can also specify a --local-service option following the - service name. This causes the server to run using the - LocalService Windows account that has limited system - privileges. This account is available only for Windows XP or - newer. If both --defaults-file and --local-service are given - following the service name, they can be in any order. + 1. Create a view v_orig in MySQL 5.1.29 or higher. - For a MySQL server that is installed as a Windows service, the - following rules determine the service name and option files that - the server uses: + 2. Rename the view to v_new and then back to v_orig. - * If the service-installation command specifies no service name - or the default service name (MySQL) following the --install - option, the server uses the a service name of MySQL and reads - options from the [mysqld] group in the standard option files. + 3. Downgrade to an older 5.1.x server and run mysql_upgrade. - * If the service-installation command specifies a service name - other than MySQL following the --install option, the server - uses that service name. It reads options from the [mysqld] - group and the group that has the same name as the service in - the standard option files. This allows you to use the [mysqld] - group for options that should be used by all MySQL services, - and an option group with the service name for use by the - server installed with that service name. + 4. Try to rename v_orig to v_new again. This operation + fails. + As a workaround to avoid this problem, use either of these + approaches: - * If the service-installation command specifies a - --defaults-file option after the service name, the server - reads options only from the [mysqld] group of the named file - and ignores the standard option files. + + Dump your data using mysqldump before downgrading and + reload the dump file after downgrading. - As a more complex example, consider the following command: -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" - --install MySQL --defaults-file=C:\my-opts.cnf + + Instead of renaming a view after the downgrade, drop it + and recreate it. - Here, the default service name (MySQL) is given after the - --install option. If no --defaults-file option had been given, - this command would have the effect of causing the server to read - the [mysqld] group from the standard option files. However, - because the --defaults-file option is present, the server reads - options from the [mysqld] option group, and only from the named - file. + * Incompatible change: Character set or collation changes were + made in MySQL 5.1.21, 5.1.23, and 5.1.24 that may require + table indexes to be rebuilt. For details, see Section 2.4.3, + "Checking Whether Tables or Indexes Must Be Rebuilt." - You can also specify options as Start parameters in the Windows - Services utility before you start the MySQL service. + * Incompatible change: MySQL 5.1 implements support for a plugin + API that allows the loading and unloading of components at + runtime, without restarting the server. Section 22.2, "The + MySQL Plugin Interface." The plugin API requires the + mysql.plugin table. After upgrading from an older version of + MySQL, you should run the mysql_upgrade command to create this + table. See Section 4.4.8, "mysql_upgrade --- Check Tables for + MySQL Upgrade." + Plugins are installed in the directory named by the plugin_dir + system variable. This variable also controls the location from + which the server loads user-defined functions (UDFs), which is + a change from earlier versions of MySQL. That is, all UDF + library files now must be installed in the plugin directory. + When upgrading from an older version of MySQL, you must + migrate your UDF files to the plugin directory. - Once a MySQL server has been installed as a service, Windows - starts the service automatically whenever Windows starts. The - service also can be started immediately from the Services utility, - or by using a NET START MySQL command. The NET command is not case - sensitive. + * Incompatible change: The table_cache system variable has been + renamed to table_open_cache. Any scripts that refer to + table_cache must be updated to use the new name. - When run as a service, mysqld has no access to a console window, - so no messages can be seen there. If mysqld does not start, check - the error log to see whether the server wrote any messages there - to indicate the cause of the problem. The error log is located in - the MySQL data directory (for example, C:\Program - Files\MySQL\MySQL Server 5.1\data). It is the file with a suffix - of .err. + * Incompatible change: In MySQL 5.1.36, options for loading + plugins such as pluggable storage engines were changed from + boolean to tristate format. The implementations overlap, but + if you previously used options of the form --plugin_name=0 or + --plugin_name=1, you should instead use --plugin_name=OFF or + --plugin_name=ON, respectively. For details, see Section + 5.1.3, "Server Options for Loading Plugins." - When a MySQL server has been installed as a service, and the - service is running, Windows stops the service automatically when - Windows shuts down. The server also can be stopped manually by - using the Services utility, the NET STOP MySQL command, or the - mysqladmin shutdown command. + * Incompatible change: From MySQL 5.1.24 to 5.1.31, the UPDATE + statement was changed such that assigning NULL to a NOT NULL + column caused an error even when strict SQL mode was not + enabled. The original behavior before MySQL 5.1.24 was that + such assignments caused an error only in strict SQL mode, and + otherwise set the column to the implicit default value for the + column data type and generated a warning. (For information + about implicit default values, see Section 10.1.4, "Data Type + Default Values.") + The change caused compatibility problems for applications that + relied on the original behavior. It also caused replication + problems between servers that had the original behavior and + those that did not, for applications that assigned NULL to NOT + NULL columns in UPDATE statements without strict SQL mode + enabled. The change was reverted in MySQL 5.1.32 so that + UPDATE again had the original behavior. Problems can still + occur if you replicate between servers that have the modified + UPDATE behavior and those that do not. - You also have the choice of installing the server as a manual - service if you do not wish for the service to be started - automatically during the boot process. To do this, use the - --install-manual option rather than the --install option: -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" --install-m -anual + * Incompatible change: As of MySQL 5.1.29, the default binary + logging mode has been changed from MIXED to STATEMENT for + compatibility with MySQL 5.0. - To remove a server that is installed as a service, first stop it - if it is running by executing NET STOP MySQL. Then use the - --remove option to remove it: -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" --remove + * Incompatible change: In MySQL 5.1.25, a change was made to the + way that the server handles prepared statements. This affects + prepared statements processed at the SQL level (using the + PREPARE statement) and those processed using the binary + client-server protocol (using the mysql_stmt_prepare() C API + function). + Previously, changes to metadata of tables or views referred to + in a prepared statement could cause a server crash when the + statement was next executed, or perhaps an error at execute + time with a crash occurring later. For example, this could + happen after dropping a table and recreating it with a + different definition. + Now metadata changes to tables or views referred to by + prepared statements are detected and cause automatic + repreparation of the statement when it is next executed. + Metadata changes occur for DDL statements such as those that + create, drop, alter, rename, or truncate tables, or that + analyze, optimize, or repair tables. Repreparation also occurs + after referenced tables or views are flushed from the table + definition cache, either implicitly to make room for new + entries in the cache, or explicitly due to FLUSH TABLES. + Repreparation is automatic, but to the extent that it occurs, + performance of prepared statements is diminished. + Table content changes (for example, with INSERT or UPDATE) do + not cause repreparation, nor do SELECT statements. + An incompatibility with previous versions of MySQL is that a + prepared statement may now return a different set of columns + or different column types from one execution to the next. For + example, if the prepared statement is SELECT * FROM t1, + altering t1 to contain a different number of columns causes + the next execution to return a number of columns different + from the previous execution. + Older versions of the client library cannot handle this change + in behavior. For applications that use prepared statements + with the new server, an upgrade to the new client library is + strongly recommended. + Along with this change to statement repreparation, the default + value of the table_definition_cache system variable has been + increased from 128 to 256. The purpose of this increase is to + lessen the chance that prepared statements will need + repreparation due to referred-to tables/views having been + flushed from the cache to make room for new entries. + A new status variable, Com_stmt_reprepare, has been introduced + to track the number of repreparations. - If mysqld is not running as a service, you can start it from the - command line. For instructions, see Section 2.3.10, "Starting - MySQL from the Windows Command Line." + * Incompatible change: As of MySQL 5.1.23, within a stored + routine, it is no longer allowable to declare a cursor for a + SHOW or DESCRIBE statement. This happened to work in some + instances, but is no longer supported. In many cases, a + workaround for this change is to use the cursor with a SELECT + query to read from an INFORMATION_SCHEMA table that produces + the same information as the SHOW statement. - Please see Section 2.3.13, "Troubleshooting a MySQL Installation - Under Windows," if you encounter difficulties during installation. + * Incompatible change: SHOW CREATE VIEW displays view + definitions using an AS alias_name clause for each column. If + a column is created from an expression, the default alias is + the expression text, which can be quite long. As of MySQL + 5.1.23, aliases for column names in CREATE VIEW statements are + checked against the maximum column length of 64 characters + (not the maximum alias length of 256 characters). As a result, + views created from the output of SHOW CREATE VIEW fail if any + column alias exceeds 64 characters. This can cause problems + for replication or loading dump files. For additional + information and workarounds, see Section D.4, "Restrictions on + Views." -2.3.12. Testing The MySQL Installation + * Incompatible change: Several issues were identified for stored + programs (stored procedures and functions, triggers, and + events) and views containing non-ASCII symbols. These issues + involved conversion errors due to incomplete character set + information when translating these objects to and from stored + format. + To address these problems, the representation for these + objects was changed in MySQL 5.1.21. However, the fixes affect + all stored programs and views. (For example, you will see + warnings about "no creation context.") To avoid warnings from + the server about the use of old definitions from any release + prior to 5.1.21, you should dump stored programs and views + with mysqldump after upgrading to 5.1.21 or higher, and then + reload them to recreate them with new definitions. Invoke + mysqldump with a --default-character-set option that names the + non-ASCII character set that was used for the definitions when + the objects were originally defined. - You can test whether the MySQL server is working by executing any - of the following commands: -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqlshow" -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqlshow" -u root -mysql -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqladmin" version - status proc -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysql" test + * Incompatible change: As of MySQL 5.1.20, mysqld_safe supports + error logging to syslog on systems that support the logger + command. The new --syslog and --skip-syslog options can be + used instead of the --log-error option to control logging + behavior, as described in Section 4.3.2, "mysqld_safe --- + MySQL Server Startup Script." + In 5.1.21 and up, the default is --skip-syslog, which is + compatible with the default behavior of writing an error log + file for releases prior to 5.1.20. + In 5.1.20 only, the following conditions apply: 1) The default + is to use syslog, which is not compatible with releases prior + to 5.1.20. 2) Logging to syslog may fail to operate correctly + in some cases. For these reasons, avoid using MySQL 5.1.20. - If mysqld is slow to respond to TCP/IP connections from client - programs, there is probably a problem with your DNS. In this case, - start mysqld with the --skip-name-resolve option and use only - localhost and IP numbers in the Host column of the MySQL grant - tables. + * Incompatible change: As of MySQL 5.1.18, the plugin interface + and its handling of system variables was changed. Command-line + options such as --skip-innodb now cause an error if InnoDB is + not built-in or plugin-loaded. You should use + --loose-skip-innodb if you do not want any error even if + InnoDB is not available. The --loose prefix modifier should be + used for all command-line options where you are uncertain + whether the plugin exists and when you want the operation to + proceed even if the option is necessarily ignored due to the + absence of the plugin. (For a desecription of how --loose + works, see Section 4.2.3.1, "Using Options on the Command + Line.") - You can force a MySQL client to use a named-pipe connection rather - than TCP/IP by specifying the --pipe or --protocol=PIPE option, or - by specifying . (period) as the host name. Use the --socket option - to specify the name of the pipe if you do not want to use the - default pipe name. + * Incompatible change: As of MySQL 5.1.15, InnoDB rolls back + only the last statement on a transaction timeout. A new + option, --innodb_rollback_on_timeout, causes InnoDB to abort + and roll back the entire transaction if a transaction timeout + occurs (the same behavior as in MySQL 4.1). - Note that if you have set a password for the root account, deleted - the anonymous account, or created a new user account, then you - must use the appropriate -u and -p options with the commands shown - above in order to connect with the MySQL Server. See Section - 4.2.2, "Connecting to the MySQL Server." + * Incompatible change: As of MySQL 5.1.15, the following + conditions apply to enabling the read_only system variable: - For more information about mysqlshow, see Section 4.5.6, - "mysqlshow --- Display Database, Table, and Column Information." + + If you attempt to enable read_only while you have any + explicit locks (acquired with LOCK TABLES or have a + pending transaction, an error will occur. -2.3.13. Troubleshooting a MySQL Installation Under Windows + + If other clients hold explicit table locks or have + pending transactions, the attempt to enable read_only + blocks until the locks are released and the transactions + end. While the attempt to enable read_only is pending, + requests by other clients for table locks or to begin + transactions also block until read_only has been set. - When installing and running MySQL for the first time, you may - encounter certain errors that prevent the MySQL server from - starting. The purpose of this section is to help you diagnose and - correct some of these errors. + + read_only can be enabled while you hold a global read + lock (acquired with FLUSH TABLES WITH READ LOCK) because + that does not involve table locks. + Previously, the attempt to enable read_only would return + immediately even if explicit locks or transactions were + pending, so some data changes could occur for statements + executing in the server at the same time. - Your first resource when troubleshooting server issues is the - error log. The MySQL server uses the error log to record - information relevant to the error that prevents the server from - starting. The error log is located in the data directory specified - in your my.ini file. The default data directory location is - C:\Program Files\MySQL\MySQL Server 5.1\data. See Section 5.2.2, - "The Error Log." + * Incompatible change: The number of function names affected by + IGNORE_SPACE was reduced significantly in MySQL 5.1.13, from + about 200 to about 30. (For details about IGNORE_SPACE, see + Section 8.2.4, "Function Name Parsing and Resolution.") This + change improves the consistency of parser operation. However, + it also introduces the possibility of incompatibility for old + SQL code that relies on the following conditions: - Another source of information regarding possible errors is the - console messages displayed when the MySQL service is starting. Use - the NET START MySQL command from the command line after installing - mysqld as a service to see any error messages regarding the - starting of the MySQL server as a service. See Section 2.3.11, - "Starting MySQL as a Windows Service." + + IGNORE_SPACE is disabled. - The following examples show other common error messages you may - encounter when installing MySQL and starting the server for the - first time: + + The presence or absence of whitespace following a + function name is used to distinguish between a built-in + function and stored function that have the same name (for + example, PI() versus PI ()). + For functions that are no longer affected by IGNORE_SPACE as + of MySQL 5.1.13, that strategy no longer works. Either of the + following approaches can be used if you have code that is + subject to the preceding incompatibility: - * If the MySQL server cannot find the mysql privileges database - or other critical files, you may see these messages: -System error 1067 has occurred. -Fatal error: Can't open privilege tables: Table 'mysql.host' doesn't -exist - These messages often occur when the MySQL base or data - directories are installed in different locations than the - default locations (C:\Program Files\MySQL\MySQL Server 5.1 and - C:\Program Files\MySQL\MySQL Server 5.1\data, respectively). - This situation may occur when MySQL is upgraded and installed - to a new location, but the configuration file is not updated - to reflect the new location. In addition, there may be old and - new configuration files that conflict. Be sure to delete or - rename any old configuration files when upgrading MySQL. - If you have installed MySQL to a directory other than - C:\Program Files\MySQL\MySQL Server 5.1, you need to ensure - that the MySQL server is aware of this through the use of a - configuration (my.ini) file. The my.ini file needs to be - located in your Windows directory, typically C:\WINDOWS. You - can determine its exact location from the value of the WINDIR - environment variable by issuing the following command from the - command prompt: -C:\> echo %WINDIR% - An option file can be created and modified with any text - editor, such as Notepad. For example, if MySQL is installed in - E:\mysql and the data directory is D:\MySQLdata, you can - create the option file and set up a [mysqld] section to - specify values for the basedir and datadir options: -[mysqld] -# set basedir to your installation path -basedir=E:/mysql -# set datadir to the location of your data directory -datadir=D:/MySQLdata - Note that Windows path names are specified in option files - using (forward) slashes rather than backslashes. If you do use - backslashes, you must double them: -[mysqld] -# set basedir to your installation path -basedir=C:\\Program Files\\MySQL\\MySQL Server 5.1 -# set datadir to the location of your data directory -datadir=D:\\MySQLdata - If you change the datadir value in your MySQL configuration - file, you must move the contents of the existing MySQL data - directory before restarting the MySQL server. - See Section 2.3.7, "Creating an Option File." - - * If you reinstall or upgrade MySQL without first stopping and - removing the existing MySQL service and install MySQL using - the MySQL Configuration Wizard, you may see this error: -Error: Cannot create Windows service for MySql. Error: 0 - This occurs when the Configuration Wizard tries to install the - service and finds an existing service with the same name. - One solution to this problem is to choose a service name other - than mysql when using the configuration wizard. This allows - the new service to be installed correctly, but leaves the - outdated service in place. Although this is harmless, it is - best to remove old services that are no longer in use. - To permanently remove the old mysql service, execute the - following command as a user with administrative privileges, on - the command-line: -C:\> sc delete mysql -[SC] DeleteService SUCCESS - If the sc utility is not available for your version of - Windows, download the delsrv utility from - http://www.microsoft.com/windows2000/techinfo/reskit/tools/exi - sting/delsrv-o.asp and use the delsrv mysql syntax. - -2.3.14. Upgrading MySQL on Windows - - This section lists some of the steps you should take when - upgrading MySQL on Windows. - - 1. Review Section 2.12.1, "Upgrading MySQL," for additional - information on upgrading MySQL that is not specific to - Windows. - - 2. You should always back up your current MySQL installation - before performing an upgrade. See Section 6.1, "Database - Backups." - - 3. Download the latest Windows distribution of MySQL from - http://dev.mysql.com/downloads/. - - 4. Before upgrading MySQL, you must stop the server. If the - server is installed as a service, stop the service with the - following command from the command prompt: -C:\> NET STOP MySQL - If you are not running the MySQL server as a service, use the - following command to stop it: -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqladmin" -u root - shutdown - -Note - If the MySQL root user account has a password, you need to - invoke mysqladmin with the -p option and supply the password - when prompted. - - 5. When upgrading to MySQL 5.1 from a version previous to 4.1.5, - or when upgrading from a version of MySQL installed from a Zip - archive to a version of MySQL installed with the MySQL - Installation Wizard, you must manually remove the previous - installation and MySQL service (if the server is installed as - a service). - To remove the MySQL service, use the following command: -C:\> C:\mysql\bin\mysqld --remove - If you do not remove the existing service, the MySQL - Installation Wizard may fail to properly install the new MySQL - service. - - 6. When upgrading from MySQL 5.1.23 to MySQL 5.1.24, the change - in the default location of the data directory from a directory - within the MySQL installation to the AppData folder means that - you must manually copy the data files from your old - installation to the new location. - - 7. If you are using the MySQL Installation Wizard, start the - wizard as described in Section 2.3.3, "Using the MySQL - Installation Wizard." - - 8. If you are installing MySQL from a Zip archive, extract the - archive. You may either overwrite your existing MySQL - installation (usually located at C:\mysql), or install it into - a different directory, such as C:\mysql5. Overwriting the - existing installation is recommended. - - 9. If you were running MySQL as a Windows service and you had to - remove the service earlier in this procedure, reinstall the - service. (See Section 2.3.11, "Starting MySQL as a Windows - Service.") - 10. Restart the server. For example, use NET START MySQL if you - run MySQL as a service, or invoke mysqld directly otherwise. - 11. If you encounter errors, see Section 2.3.13, "Troubleshooting - a MySQL Installation Under Windows." - -2.3.15. MySQL on Windows Compared to MySQL on Unix - - MySQL for Windows has proven itself to be very stable. The Windows - version of MySQL has the same features as the corresponding Unix - version, with the following exceptions: - - * Limited number of ports - Windows systems have about 4,000 ports available for client - connections, and after a connection on a port closes, it takes - two to four minutes before the port can be reused. In - situations where clients connect to and disconnect from the - server at a high rate, it is possible for all available ports - to be used up before closed ports become available again. If - this happens, the MySQL server appears to be unresponsive even - though it is running. Note that ports may be used by other - applications running on the machine as well, in which case the - number of ports available to MySQL is lower. - For more information about this problem, see - http://support.microsoft.com/default.aspx?scid=kb;en-us;196271 - . - - * Concurrent reads - MySQL depends on the pread() and pwrite() system calls to be - able to mix INSERT and SELECT. Currently, we use mutexes to - emulate pread() and pwrite(). We intend to replace the file - level interface with a virtual interface in the future so that - we can use the readfile()/writefile() interface to get more - speed. The current implementation limits the number of open - files that MySQL 5.1 can use to 2,048, which means that you - cannot run as many concurrent threads on Windows as on Unix. - - * Blocking read - MySQL uses a blocking read for each connection. That has the - following implications if named-pipe connections are enabled: - - + A connection is not disconnected automatically after - eight hours, as happens with the Unix version of MySQL. - - + If a connection hangs, it is not possible to break it - without killing MySQL. - - + mysqladmin kill does not work on a sleeping connection. - - + mysqladmin shutdown cannot abort as long as there are - sleeping connections. - We plan to fix this problem in the future. - - * ALTER TABLE - While you are executing an ALTER TABLE statement, the table is - locked from being used by other threads. This has to do with - the fact that on Windows, you can't delete a file that is in - use by another thread. In the future, we may find some way to - work around this problem. - - * DROP TABLE - DROP TABLE on a table that is in use by a MERGE table does not - work on Windows because the MERGE handler does the table - mapping hidden from the upper layer of MySQL. Because Windows - does not allow dropping files that are open, you first must - flush all MERGE tables (with FLUSH TABLES) or drop the MERGE - table before dropping the table. - - * DATA DIRECTORY and INDEX DIRECTORY - The DATA DIRECTORY and INDEX DIRECTORY options for CREATE - TABLE are ignored on Windows, because Windows doesn't support - symbolic links. These options also are ignored on systems that - have a nonfunctional realpath() call. - - * DROP DATABASE - You cannot drop a database that is in use by some thread. - - * Case-insensitive names - File names are not case sensitive on Windows, so MySQL - database and table names are also not case sensitive on - Windows. The only restriction is that database and table names - must be specified using the same case throughout a given - statement. See Section 8.2.2, "Identifier Case Sensitivity." - - * Directory and file names - On Windows, MySQL Server supports only directory and file - names that are compatible with the current ANSI code pages. - For example, the following Japanese directory name will not - work in the Western locale (code page 1252): -datadir="C:/维基百科关于中文维基百科" - The same limitation applies to directory and file names - referred to in SQL statements, such as the data file path name - in LOAD DATA INFILE. - - * The "\" path name separator character - Path name components in Windows are separated by the "\" - character, which is also the escape character in MySQL. If you - are using LOAD DATA INFILE or SELECT ... INTO OUTFILE, use - Unix-style file names with "/" characters: -mysql> LOAD DATA INFILE 'C:/tmp/skr.txt' INTO TABLE skr; -mysql> SELECT * INTO OUTFILE 'C:/tmp/skr.txt' FROM skr; - Alternatively, you must double the "\" character: -mysql> LOAD DATA INFILE 'C:\\tmp\\skr.txt' INTO TABLE skr; -mysql> SELECT * INTO OUTFILE 'C:\\tmp\\skr.txt' FROM skr; - - * Problems with pipes - Pipes do not work reliably from the Windows command-line - prompt. If the pipe includes the character ^Z / CHAR(24), - Windows thinks that it has encountered end-of-file and aborts - the program. - This is mainly a problem when you try to apply a binary log as - follows: -C:\> mysqlbinlog binary_log_file | mysql --user=root - If you have a problem applying the log and suspect that it is - because of a ^Z / CHAR(24) character, you can use the - following workaround: -C:\> mysqlbinlog binary_log_file --result-file=/tmp/bin.sql -C:\> mysql --user=root --execute "source /tmp/bin.sql" - The latter command also can be used to reliably read in any - SQL file that may contain binary data. - - * Access denied for user error - If MySQL cannot resolve your host name properly, you may get - the following error when you attempt to run a MySQL client - program to connect to a server running on the same machine: -Access denied for user 'some_user'@'unknown' -to database 'mysql' - To fix this problem, you should create a file named - \windows\hosts containing the following information: -127.0.0.1 localhost - - Here are some open issues for anyone who might want to help us - improve MySQL on Windows: - - * Add macros to use the faster thread-safe increment/decrement - methods provided by Windows. - -2.4. Installing MySQL from RPM Packages on Linux - - The recommended way to install MySQL on RPM-based Linux - distributions is by using the RPM packages. The RPMs that we - provide to the community should work on all versions of Linux that - support RPM packages and use glibc 2.3. To obtain RPM packages, - see Section 2.1.3, "How to Get MySQL." - - For non-RPM Linux distributions, you can install MySQL using a - .tar.gz package. See Section 2.9, "Installing MySQL from tar.gz - Packages on Other Unix-Like Systems." - - We do provide some platform-specific RPMs; the difference between - a platform-specific RPM and a generic RPM is that a - platform-specific RPM is built on the targeted platform and is - linked dynamically whereas a generic RPM is linked statically with - LinuxThreads. - -Note - - RPM distributions of MySQL often are provided by other vendors. Be - aware that they may differ in features and capabilities from those - built by us, and that the instructions in this manual do not - necessarily apply to installing them. The vendor's instructions - should be consulted instead. - - If you have problems with an RPM file (for example, if you receive - the error Sorry, the host 'xxxx' could not be looked up), see - Section 2.13.1.2, "Linux Binary Distribution Notes." - - In most cases, you need to install only the MySQL-server and - MySQL-client packages to get a functional MySQL installation. The - other packages are not required for a standard installation. - - RPMs for MySQL Cluster. Beginning with MySQL 5.1.24, standard - MySQL server RPMs built by MySQL no longer provide support for the - NDBCLUSTER storage engine. MySQL Cluster users wanting to upgrade - MySQL 5.1.23 or earlier installations from RPMs built by MySQL - should upgrade to MySQL Cluster NDB 6.2 or MySQL Cluster NDB 6.3; - RPMs that should work with most Linux distributions are available - for both of these release series. - -Important - - When upgrading a MySQL Cluster RPM installation, you must upgrade - all installed RPMs, including the Server and Client RPMs. - - For more information about installing MySQL Cluster from RPMs, see - Section 17.2.2, "MySQL Cluster Multi-Computer Installation." - - For upgrades, if your installation was originally produced by - installing multiple RPM packages, it is best to upgrade all the - packages, not just some. For example, if you previously installed - the server and client RPMs, do not upgrade just the server RPM. - - If you get a dependency failure when trying to install MySQL - packages (for example, error: removing these packages would break - dependencies: libmysqlclient.so.10 is needed by ...), you should - also install the MySQL-shared-compat package, which includes both - the shared libraries for backward compatibility - (libmysqlclient.so.12 for MySQL 4.0 and libmysqlclient.so.10 for - MySQL 3.23). - - Some Linux distributions still ship with MySQL 3.23 and they - usually link applications dynamically to save disk space. If these - shared libraries are in a separate package (for example, - MySQL-shared), it is sufficient to simply leave this package - installed and just upgrade the MySQL server and client packages - (which are statically linked and do not depend on the shared - libraries). For distributions that include the shared libraries in - the same package as the MySQL server (for example, Red Hat Linux), - you could either install our 3.23 MySQL-shared RPM, or use the - MySQL-shared-compat package instead. (Do not install both.) - - The RPM packages shown in the following list are available. The - names shown here use a suffix of .glibc23.i386.rpm, but particular - packages can have different suffixes, as described later. - - * MySQL-server-VERSION.glibc23.i386.rpm - The MySQL server. You need this unless you only want to - connect to a MySQL server running on another machine. - - * MySQL-client-VERSION.glibc23.i386.rpm - The standard MySQL client programs. You probably always want - to install this package. - - * MySQL-devel-VERSION.glibc23.i386.rpm - The libraries and include files that are needed if you want to - compile other MySQL clients, such as the Perl modules. - - * MySQL-debuginfo-VERSION.glibc23.i386.rpm - This package contains debugging information. debuginfo RPMs - are never needed to use MySQL software; this is true both for - the server and for client programs. However, they contain - additional information that might be needed by a debugger to - analyze a crash. - - * MySQL-shared-VERSION.glibc23.i386.rpm - This package contains the shared libraries - (libmysqlclient.so*) that certain languages and applications - need to dynamically load and use MySQL. It contains - single-threaded and thread-safe libraries. If you install this - package, do not install the MySQL-shared-compat package. - - * MySQL-shared-compat-VERSION.glibc23.i386.rpm - This package includes the shared libraries for MySQL 3.23, - 4.0, and so on, up to the current release. It contains - single-threaded and thread-safe libraries. Install this - package instead of MySQL-shared if you have applications - installed that are dynamically linked against older versions - of MySQL but you want to upgrade to the current version - without breaking the library dependencies. - - * MySQL-embedded-VERSION.glibc23.i386.rpm - The embedded MySQL server library. - - * MySQL-ndb-management-VERSION.glibc23.i386.rpm, - MySQL-ndb-storage-VERSION.glibc23.i386.rpm, - MySQL-ndb-tools-VERSION.glibc23.i386.rpm, - MySQL-ndb-extra-VERSION.glibc23.i386.rpm - Packages that contain additional files for MySQL Cluster - installations. - -Note - The MySQL-ndb-tools RPM requires a working installation of - perl. Prior to MySQL 5.1.18, the DBI and HTML::Template - packages were also required. See Section 2.15, "Perl - Installation Notes," and Section 17.6.21, "ndb_size.pl --- - NDBCLUSTER Size Requirement Estimator," for more information. - - * MySQL-test-VERSION.glibc23.i386.rpm - This package includes the MySQL test suite. - - * MySQL-VERSION.src.rpm - This contains the source code for all of the previous - packages. It can also be used to rebuild the RPMs on other - architectures (for example, Alpha or SPARC). - - The suffix of RPM package names (following the VERSION value) has - the following syntax: -.PLATFORM.CPU.rpm - - The PLATFORM and CPU values indicate the type of system for which - the package is built. PLATFORM indicates the platform and CPU - indicates the processor type or family. - - All packages are dynamically linked against glibc 2.3. The - PLATFORM value indicates whether the package is platform - independent or intended for a specific platform, as shown in the - following table. - glibc23 Platform independent, should run on any Linux distribution - that supports glibc 2.3 - rhel3, rhel4 Red Hat Enterprise Linux 3 or 4 - sles9, sles10 SuSE Linux Enterprise Server 9 or 10 - - In MySQL 5.1, only glibc23 packages are available currently. - - The CPU value indicates the processor type or family for which the - package is built. - i386 x86 processor, 386 and up - i586 x86 processor, Pentium and up - x86_64 64-bit x86 processor - ia64 Itanium (IA-64) processor - - To see all files in an RPM package (for example, a MySQL-server - RPM), run a command like this: -shell> rpm -qpl MySQL-server-VERSION.glibc23.i386.rpm - - To perform a standard minimal installation, install the server and - client RPMs: -shell> rpm -i MySQL-server-VERSION.glibc23.i386.rpm -shell> rpm -i MySQL-client-VERSION.glibc23.i386.rpm - - To install only the client programs, install just the client RPM: -shell> rpm -i MySQL-client-VERSION.glibc23.i386.rpm - - RPM provides a feature to verify the integrity and authenticity of - packages before installing them. If you would like to learn more - about this feature, see Section 2.1.4, "Verifying Package - Integrity Using MD5 Checksums or GnuPG." - - The server RPM places data under the /var/lib/mysql directory. The - RPM also creates a login account for a user named mysql (if one - does not exist) to use for running the MySQL server, and creates - the appropriate entries in /etc/init.d/ to start the server - automatically at boot time. (This means that if you have performed - a previous installation and have made changes to its startup - script, you may want to make a copy of the script so that you - don't lose it when you install a newer RPM.) See Section 2.11.2.2, - "Starting and Stopping MySQL Automatically," for more information - on how MySQL can be started automatically on system startup. - - If you want to install the MySQL RPM on older Linux distributions - that do not support initialization scripts in /etc/init.d - (directly or via a symlink), you should create a symbolic link - that points to the location where your initialization scripts - actually are installed. For example, if that location is - /etc/rc.d/init.d, use these commands before installing the RPM to - create /etc/init.d as a symbolic link that points there: -shell> cd /etc -shell> ln -s rc.d/init.d . - - However, all current major Linux distributions should support the - new directory layout that uses /etc/init.d, because it is required - for LSB (Linux Standard Base) compliance. - - If the RPM files that you install include MySQL-server, the mysqld - server should be up and running after installation. You should be - able to start using MySQL. - - If something goes wrong, you can find more information in the - binary installation section. See Section 2.9, "Installing MySQL - from tar.gz Packages on Other Unix-Like Systems." - -Note - - The accounts that are listed in the MySQL grant tables initially - have no passwords. After starting the server, you should set up - passwords for them using the instructions in Section 2.11, - "Post-Installation Setup and Testing." - - During RPM installation, a user named mysql and a group named - mysql are created on the system. This is done using the useradd, - groupadd, and usermod commands. Those commands require appropriate - administrative privileges, which is ensured for locally managed - users and groups (as listed in the /etc/passwd and /etc/group - files) by the RPM installation process being run by root. - - For nonlocal user management (LDAP, NIS, and so forth), the - administrative tools may require additional authentication (such - as a password), and will fail if the installing user does not - provide this authentication. Even if they fail, the RPM - installation will not abort but succeed, and this is intentional. - If they failed, some of the intended transfer of ownership may be - missing, and it is recommended that the system administrator then - manually ensures some appropriate user andgroup exists and - manually transfers ownership following the actions in the RPM spec - file. - -2.5. Installing MySQL on Mac OS X - - You can install MySQL on Mac OS X 10.3.x ("Panther") or newer - using a Mac OS X binary package in PKG format instead of the - binary tarball distribution. Please note that older versions of - Mac OS X (for example, 10.1.x or 10.2.x) are not supported by this - package. - - The package is located inside a disk image (.dmg) file that you - first need to mount by double-clicking its icon in the Finder. It - should then mount the image and display its contents. - - To obtain MySQL, see Section 2.1.3, "How to Get MySQL." - -Note - - Before proceeding with the installation, be sure to shut down all - running MySQL server instances by either using the MySQL Manager - Application (on Mac OS X Server) or via mysqladmin shutdown on the - command line. - - To actually install the MySQL PKG file, double-click on the - package icon. This launches the Mac OS X Package Installer, which - guides you through the installation of MySQL. - - Due to a bug in the Mac OS X package installer, you may see this - error message in the destination disk selection dialog: -You cannot install this software on this disk. (null) + + If a stored function has a name that conflicts with a + built-in function, refer to the stored function with a + schema name qualifier, regardless of whether whitespace + is present. For example, write schema_name.PI() or + schema_name.PI (). - If this error occurs, simply click the Go Back button once to - return to the previous screen. Then click Continue to advance to - the destination disk selection again, and you should be able to - choose the destination disk correctly. We have reported this bug - to Apple and it is investigating this problem. - - The Mac OS X PKG of MySQL installs itself into - /usr/local/mysql-VERSION and also installs a symbolic link, - /usr/local/mysql, that points to the new location. If a directory - named /usr/local/mysql exists, it is renamed to - /usr/local/mysql.bak first. Additionally, the installer creates - the grant tables in the mysql database by executing - mysql_install_db. - - The installation layout is similar to that of a tar file binary - distribution; all MySQL binaries are located in the directory - /usr/local/mysql/bin. The MySQL socket file is created as - /tmp/mysql.sock by default. See Section 2.1.5, "Installation - Layouts." - - MySQL installation requires a Mac OS X user account named mysql. A - user account with this name should exist by default on Mac OS X - 10.2 and up. + + Alternatively, rename the stored function to use a + nonconflicting name and change invocations of the + function to use the new name. - If you are running Mac OS X Server, a version of MySQL should - already be installed. The following table shows the versions of - MySQL that ship with Mac OS X Server versions. - Mac OS X Server Version MySQL Version - 10.2-10.2.2 3.23.51 - 10.2.3-10.2.6 3.23.53 - 10.3 4.0.14 - 10.3.2 4.0.16 - 10.4.0 4.1.10a + * Incompatible change: For utf8 columns, the full-text parser + incorrectly considered several nonword punctuation and + whitespace characters as word characters, causing some + searches to return incorrect results. The fix involves a + change to the full-text parser in MySQL 5.1.12, so as of + 5.1.12, any tables that have FULLTEXT indexes on utf8 columns + must be repaired with REPAIR TABLE: +REPAIR TABLE tbl_name QUICK; - This manual section covers the installation of the official MySQL - Mac OS X PKG only. Make sure to read Apple's help information - about installing MySQL: Run the "Help View" application, select - "Mac OS X Server" help, do a search for "MySQL," and read the item - entitled "Installing MySQL." - - If you previously used Marc Liyanage's MySQL packages for Mac OS X - from http://www.entropy.ch, you can simply follow the update - instructions for packages using the binary installation layout as - given on his pages. - - If you are upgrading from Marc's 3.23.x versions or from the Mac - OS X Server version of MySQL to the official MySQL PKG, you also - need to convert the existing MySQL privilege tables to the current - format, because some new security privileges have been added. See - Section 4.4.8, "mysql_upgrade --- Check Tables for MySQL Upgrade." - - If you want MySQL to start automatically during system startup, - you also need to install the MySQL Startup Item. It is part of the - Mac OS X installation disk images as a separate installation - package. Simply double-click the MySQLStartupItem.pkg icon and - follow the instructions to install it. The Startup Item need be - installed only once. There is no need to install it each time you - upgrade the MySQL package later. + * Incompatible change: Storage engines can be pluggable at + runtime, so the distinction between disabled and invalid + storage engines no longer applies. As of MySQL 5.1.12, this + affects the NO_ENGINE_SUBSTITUTION SQL mode, as described in + Section 5.1.8, "Server SQL Modes." - The Startup Item for MySQL is installed into - /Library/StartupItems/MySQLCOM. (Before MySQL 4.1.2, the location - was /Library/StartupItems/MySQL, but that collided with the MySQL - Startup Item installed by Mac OS X Server.) Startup Item - installation adds a variable MYSQLCOM=-YES- to the system - configuration file /etc/hostconfig. If you want to disable the - automatic startup of MySQL, simply change this variable to - MYSQLCOM=-NO-. - - On Mac OS X Server, the default MySQL installation uses the - variable MYSQL in the /etc/hostconfig file. The MySQL Startup Item - installer disables this variable by setting it to MYSQL=-NO-. This - avoids boot time conflicts with the MYSQLCOM variable used by the - MySQL Startup Item. However, it does not shut down a running MySQL - server. You should do that yourself. + * Incompatible change: The structure of FULLTEXT indexes has + been changed in MySQL 5.1.6. After upgrading to MySQL 5.1.6 or + greater, any tables that have FULLTEXT indexes must be + repaired with REPAIR TABLE: +REPAIR TABLE tbl_name QUICK; - After the installation, you can start up MySQL by running the - following commands in a terminal window. You must have - administrator privileges to perform this task. + * Incompatible change: In MySQL 5.1.6, when log tables were + implemented, the default log destination for the general query + and slow query log was TABLE. As of MySQL 5.1.21, this default + has been changed to FILE, which is compatible with MySQL 5.0, + but incompatible with earlier releases of MySQL 5.1. If you + are upgrading from MySQL 5.0 to 5.1.21 or higher, no logging + option changes should be necessary. However, if you are + upgrading from 5.1.6 through 5.1.20 to 5.1.21 or higher and + were using TABLE logging, use the --log-output=TABLE option + explicitly to preserve your server's table-logging behavior. - If you have installed the Startup Item, use this command: -shell> sudo /Library/StartupItems/MySQLCOM/MySQLCOM start -(Enter your password, if necessary) -(Press Control-D or enter "exit" to exit the shell) + * Incompatible change: For ENUM columns that had enumeration + values containing commas, the commas were mapped to 0xff + internally. However, this rendered the commas + indistinguishable from true 0xff characters in the values. + This no longer occurs. However, the fix requires that you dump + and reload any tables that have ENUM columns containing true + 0xff in their values: Dump the tables using mysqldump with the + current server before upgrading from a version of MySQL 5.1 + older than 5.1.15 to version 5.1.15 or newer. - If you don't use the Startup Item, enter the following command - sequence: -shell> cd /usr/local/mysql -shell> sudo ./bin/mysqld_safe -(Enter your password, if necessary) -(Press Control-Z) -shell> bg -(Press Control-D or enter "exit" to exit the shell) + * As of MySQL 5.1.12, the lc_time_names system variable + specifies the locale that controls the language used to + display day and month names and abbreviations. This variable + affects the output from the DATE_FORMAT(), DAYNAME() and + MONTHNAME() functions. See Section 9.8, "MySQL Server Locale + Support." - You should be able to connect to the MySQL server, for example, by - running /usr/local/mysql/bin/mysql. + * As of MySQL 5.1.9, mysqld_safe no longer implicitly invokes + mysqld-max if it exists. Instead, it invokes mysqld unless a + --mysqld or --mysqld-version option is given to specify + another server explicitly. If you previously relied on the + implicit invocation of mysqld-max, you should use an + appropriate option now. As of MySQL 5.1.12, there is no longer + any separate mysqld-max server, so no change should be + necessary. -Note + SQL Changes: - The accounts that are listed in the MySQL grant tables initially - have no passwords. After starting the server, you should set up - passwords for them using the instructions in Section 2.11, - "Post-Installation Setup and Testing." + * Known issue: Prior to MySQL 5.1.17, the parser accepted + invalid code in SQL condition handlers, leading to server + crashes or unexpected execution behavior in stored programs. + Specifically, the parser allowed a condition handler to refer + to labels for blocks that enclose the handler declaration. + This was incorrect because block label scope does not include + the code for handlers declared within the labeled block. + As of 5.1.17, the parser rejects this invalid construct, but + if you perform a binary upgrade (without dumping and reloading + your databases), existing handlers that contain the construct + still are invalid and should be rewritten even if they appear + to function as you expect. + To find affected handlers, use mysqldump to dump all stored + procedures and functions, triggers, and events. Then attempt + to reload them into an upgraded server. Handlers that contain + illegal label references will be rejected. + For more information about condition handlers and writing them + to avoid invalid jumps, see Section 12.8.4.2, "DECLARE for + Handlers." - You might want to add aliases to your shell's resource file to - make it easier to access commonly used programs such as mysql and - mysqladmin from the command line. The syntax for bash is: -alias mysql=/usr/local/mysql/bin/mysql -alias mysqladmin=/usr/local/mysql/bin/mysqladmin + * Incompatible change: The parser accepted statements that + contained /* ... */ that were not properly closed with */, + such as SELECT 1 /* + 2. As of MySQL 5.1.23, statements that + contain unclosed /*-comments now are rejected with a syntax + error. + This fix has the potential to cause incompatibilities. Because + of Bug#26302: http://bugs.mysql.com/26302, which caused the + trailing */ to be truncated from comments in views, stored + routines, triggers, and events, it is possible that objects of + those types may have been stored with definitions that now + will be rejected as syntactically invalid. Such objects should + be dropped and re-created so that their definitions do not + contain truncated comments. - For tcsh, use: -alias mysql /usr/local/mysql/bin/mysql -alias mysqladmin /usr/local/mysql/bin/mysqladmin + * Incompatible change: Multiple-table DELETE statements + containing ambiguous aliases could have unintended side + effects such as deleting rows from the wrong table. Example: +DELETE FROM t1 AS a2 USING t1 AS a1 INNER JOIN t2 AS a2; + As of MySQL 5.1.23, alias declarations can be declared only in + the table_references part. Elsewhere in the statement, alias + references are allowed but not alias declarations. Statements + containing aliases that are no longer allowed must be + rewritten. - Even better, add /usr/local/mysql/bin to your PATH environment - variable. You can do this by modifying the appropriate startup - file for your shell. For more information, see Section 4.2.1, - "Invoking MySQL Programs." - - If you are upgrading an existing installation, note that - installing a new MySQL PKG does not remove the directory of an - older installation. Unfortunately, the Mac OS X Installer does not - yet offer the functionality required to properly upgrade - previously installed packages. - - To use your existing databases with the new installation, you'll - need to copy the contents of the old data directory to the new - data directory. Make sure that neither the old server nor the new - one is running when you do this. After you have copied over the - MySQL database files from the previous installation and have - successfully started the new server, you should consider removing - the old installation files to save disk space. Additionally, you - should also remove older versions of the Package Receipt - directories located in /Library/Receipts/mysql-VERSION.pkg. + * Incompatible change: As of MySQL 5.1.8, TYPE = engine_name is + still accepted as a synonym for the ENGINE = engine_name table + option but generates a warning. You should note that this + option is not available in MySQL 5.1.7, and is removed + altogether as of MySQL 5.4 and produces a syntax error. + TYPE has been deprecated since MySQL 4.0. -2.6. Installing MySQL on Solaris + * Incompatible change: The namespace for triggers changed in + MySQL 5.0.10. Previously, trigger names had to be unique per + table. Now they must be unique within the schema (database). + An implication of this change is that DROP TRIGGER syntax now + uses a schema name instead of a table name (schema name is + optional and, if omitted, the current schema will be used). + When upgrading from a version of MySQL 5 older than 5.0.10 to + MySQL 5.0.10 or newer, you must drop all triggers and + re-create them or DROP TRIGGER will not work after the + upgrade. Here is a suggested procedure for doing this: - To obtain a binary MySQL distribution for Solaris in tarball or - PKG format, http://dev.mysql.com/downloads/mysql/5.1.html. + 1. Upgrade to MySQL 5.0.10 or later to be able to access + trigger information in the INFORMATION_SCHEMA.TRIGGERS + table. (This should work even for pre-5.0.10 triggers.) - If you install MySQL using a binary tarball distribution on - Solaris, you may run into trouble even before you get the MySQL - distribution unpacked, as the Solaris tar cannot handle long file - names. This means that you may see errors when you try to unpack - MySQL. + 2. Dump all trigger definitions using the following SELECT + statement: +SELECT CONCAT('CREATE TRIGGER ', t.TRIGGER_SCHEMA, '.', t.TRIGGER_NAM +E, + ' ', t.ACTION_TIMING, ' ', t.EVENT_MANIPULATION, ' ON ' +, + t.EVENT_OBJECT_SCHEMA, '.', t.EVENT_OBJECT_TABLE, + ' FOR EACH ROW ', t.ACTION_STATEMENT, '//' ) +INTO OUTFILE '/tmp/triggers.sql' +FROM INFORMATION_SCHEMA.TRIGGERS AS t; + The statement uses INTO OUTFILE, so you must have the + FILE privilege. The file will be created on the server + host. Use a different file name if you like. To be 100% + safe, inspect the trigger definitions in the triggers.sql + file, and perhaps make a backup of the file. - If this occurs, you must use GNU tar (gtar) to unpack the - distribution. + 3. Stop the server and drop all triggers by removing all + .TRG files in your database directories. Change location + to your data directory and issue this command: +shell> rm */*.TRG - You can install MySQL on Solaris using a binary package in PKG - format instead of the binary tarball distribution. Before - installing using the binary PKG format, you should create the - mysql user and group, for example: -groupadd mysql -useradd -g mysql mysql + 4. Start the server and re-create all triggers using the + triggers.sql file: +mysql> delimiter // ; +mysql> source /tmp/triggers.sql // - Some basic PKG-handling commands follow: + 5. Check that all triggers were successfully created using + the SHOW TRIGGERS statement. - * To add a package: -pkgadd -d package_name.pkg + * Incompatible change: MySQL 5.1.6 introduces the TRIGGER + privilege. Previously, the SUPER privilege was needed to + create or drop triggers. Now those operations require the + TRIGGER privilege. This is a security improvement because you + no longer need to grant users the SUPER privilege to enable + them to create triggers. However, the requirement that the + account named in a trigger's DEFINER clause must have the + SUPER privilege has changed to a requirement for the TRIGGER + privilege. When upgrading from a previous version of MySQL 5.0 + or 5.1 to MySQL 5.1.6 or newer, be sure to update your grant + tables by running mysql_upgrade. This will assign the TRIGGER + privilege to all accounts that had the SUPER privilege. If you + fail to update the grant tables, triggers may fail when + activated. After updating the grant tables, you can revoke the + SUPER privilege from those accounts that no longer otherwise + require it. - * To remove a package: -pkgrm package_name + * Some keywords may be reserved in MySQL 5.1 that were not + reserved in MySQL 5.0. See Section 8.3, "Reserved Words." - * To get a full list of installed packages: -pkginfo + * The BACKUP TABLE, and RESTORE TABLE statements are deprecated. + mysqldump or mysqlhotcopy can be used as alternatives. - * To get detailed information for a package: -pkginfo -l package_name + * The LOAD DATA FROM MASTER and LOAD TABLE FROM MASTER + statements are deprecated. See Section 12.6.2.2, "LOAD DATA + FROM MASTER Syntax," for recommended alternatives. - * To list the files belonging to a package: -pkgchk -v package_name + * The INSTALL PLUGIN and UNINSTALL PLUGIN statements that are + used for the plugin API are new. So is the WITH PARSER clause + for FULLTEXT index creation that associates a parser plugin + with a full-text index. Section 22.2, "The MySQL Plugin + Interface." - * To get packaging information for an arbitrary file: -pkgchk -l -p file_name + C API Changes: - For additional information about installing MySQL on Solaris, see - Section 2.13.3, "Solaris Notes." + * Incompatible change: As of MySQL 5.1.7, the + mysql_stmt_attr_get() C API function returns a boolean rather + than an unsigned int for STMT_ATTR_UPDATE_MAX_LENGTH. + (Bug#16144: http://bugs.mysql.com/16144) -2.7. Installing MySQL on i5/OS +2.4.2. Downgrading MySQL - The i5/OS POWER MySQL package was created in cooperation with IBM. - MySQL works within the Portable Application Solution Environment - (PASE) on the System i series of hardware and will also provide - database services for the Zend Core for i5/OS. + This section describes what you should do to downgrade to an older + MySQL version in the unlikely case that the previous version + worked better than the new one. - MySQL for i5/OS is provided as a save file (.savf) package that - can be downloaded and installed directly without any additional - installation steps required. + If you are downgrading within the same release series (for + example, from 5.0.13 to 5.0.12) the general rule is that you just + have to install the new binaries on top of the old ones. There is + no need to do anything with the databases. As always, however, it + is always a good idea to make a backup. - MySQL is only supported on i5/OS V5R4 or later releases. The i5/OS - PASE must be installed for MySQL to operate. You must be able to - login as a user in *SECOFR class. + The following items form a checklist of things you should do + whenever you perform a downgrade: - You should the installation notes and tips for i5/OS before - starting installation. See i5/OS Installation Notes. + * Read the upgrading section for the release series from which + you are downgrading to be sure that it does not have any + features you really need. See Section 2.4.1, "Upgrading + MySQL." -Note + * If there is a downgrading section for that version, you should + read that as well. - The installation package will use an existing configuration if you - have previously installed MySQL (which is identified by looking - for the file /etc/my.cnf). The values for the data directory - (DATADIR) and owner of the MySQL files (USRPRF) specified during - the installation will be ignored, and the values determined from - the /etc/my.cnf will be used instead. + * To see which new features were added between the version to + which you are downgrading and your current version, see the + change logs (Appendix C, "MySQL Change History"). - If you want to change these parameters during a new install, you - should temporarily rename /etc/my.cnf, install MySQL using the new - parameters you want to use, and then merge your previous - /etc/my.cnf configuration settings with the new /etc/my.cnf file - that is created during installation. + * Check Section 2.4.3, "Checking Whether Tables or Indexes Must + Be Rebuilt," to see whether changes to table formats or to + character sets or collations were made between your current + version of MySQL and the version to which you are downgrading. + If so and these changes result in an incompatibility between + MySQL versions, you will need to downgrade the affected tables + using the instructions in Section 2.4.4, "Rebuilding or + Repairing Tables or Indexes." - To install MySQL on i5/OS, follow these steps: + In most cases, you can move the MySQL format files and data files + between different versions on the same architecture as long as you + stay within versions for the same release series of MySQL. - 1. Create a user profile MYSQL. The MYSQL user profile will own - all the MySQL files and databases and be the active user used - when the MySQL server is running. The profile should be - disabled so that you cannot log in as the MySQL user. To - create a user profile, use CRTUSRPRF: -CRTUSRPRF USRPRF(MYSQL) STATUS(*DISABLED) TEXT('MySQL user id') + If you downgrade from one release series to another, there may be + incompatibilities in table storage formats. In this case, use + mysqldump to dump your tables before downgrading. After + downgrading, reload the dump file using mysql or mysqlimport to + re-create your tables. For examples, see Section 2.4.5, "Copying + MySQL Databases to Another Machine." - 2. On the System i machine, create a save file that will be used - to receive the downloaded installation save file. The file - should be located within the General Purpose Library (QGPL): -CRTSAVF FILE(QGPL/MYSQLINST) + A typical symptom of a downward-incompatible table format change + when you downgrade is that you cannot open tables. In that case, + use the following procedure: - 3. Download the MySQL installation save file in 32-bit - (mysql-5.0.42-i5os-power-32bit.savf) or 64-bit - (mysql-5.0.42-i5os-power-64bit.savf) from MySQL Downloads - (http://dev.mysql.com/downloads) + 1. Stop the older MySQL server that you are downgrading to. - 4. You need to FTP the downloaded .savf file directly into the - QGPL/MYSQLINST file on the System i server. You can do this - through FTP using the following steps after logging in to the - System i machine: -ftp> bin -ftp> cd qgpl -ftp> put mysql-5.0.42-i5os-power.savf mysqlinst + 2. Restart the newer MySQL server you are downgrading from. - 5. Log into the System i server using a user in the *SECOFR - class, such as the QSECOFR user ID. + 3. Dump any tables that were inaccessible to the older server by + using mysqldump to create a dump file. - 6. You need to restore the installation library stored in the - .savf save file: -RSTLIB MYSQLINST DEV(*SAVF) SAVF(QGPL/MYSQLINST) + 4. Stop the newer MySQL server and restart the older one. - 7. You need to execute the installation command, - MYSQLINST/INSMYSQL. You can specify three parameter settings - during installation: + 5. Reload the dump file into the older server. Your tables should + be accessible. - + DIR('/opt/mysql') sets the installation location for the - MySQL files. The directory will be created if it does not - already exist. - - + DATADIR('/QOpenSys/mysal/data') sets the location of the - directory that will be used to store the database files - and binary logs. The default setting is - /QOpenSys/mysql/data. Note that if the installer detects - an existing installation (due to the existence of - /etc/my.cnf), then this parameter will be ignored. + It might also be the case that the structure of the system tables + in the mysql database has changed and that downgrading introduces + some loss of functionality or requires some adjustments. Here are + some examples: - + USRPRF(MYSQL) sets the user profile that will own the - files that are installed. The profile will be created if - it does not already exist. - MySQL can be installed anywhere, for this example we will - assume MySQL has been installed into /opt/mysql. The MYSQL - user profile that was created earlier in this sequence should - be used for the profile: -MYSQLINST/INSMYSQL DIR('/opt/mysql') DATADIR('/opt/mysqldata') USRPRF -(MYSQL) - If you are updating an installation over an existing MySQL - installation, you should use the same parameter values that - were used when MySQL was originally installed. - The installation copies all the necessary files into a - directory matching the package version (for example - mysql-5.0.42-i5os-power-32bit), sets the ownership on those - files, sets up the MySQL environment and creates the MySQL - configuration file (in /etc/my.cnf) completing all the steps - in a typical binary installation process automatically. If - this is a new installation of MySQL, or if the installer - detects that this is a new version (because the /etc/my.cnf - file does not exist), then the initial core MySQL databases - will also be created during installation. + * Trigger creation requires the TRIGGER privilege as of MySQL + 5.1. In MySQL 5.0, there is no TRIGGER privilege and SUPER is + required instead. If you downgrade from MySQL 5.1 to 5.0, you + will need to give the SUPER privilege to those accounts that + had the TRIGGER privilege in 5.1. - 8. Once the installation has completed, you can delete the - installation file: -DLTLIB LIB(MYSQLINST) + * Triggers were added in MySQL 5.0, so if you downgrade from 5.0 + to 4.1, you cannot use triggers at all. - To start MySQL: +2.4.2.1. Downgrading to MySQL 5.0 - 1. Log into the System i server using a user within the *SECOFR - class, such as the QSECOFR user ID. + When downgrading to MySQL 5.0 from MySQL 5.1, you should keep in + mind the following issues relating to features found in MySQL 5.1, + but not in MySQL 5.0: -Note - You should start mysqld_safe using a user that in the PASE - environment has the id=0 (the equivalent of the standard Unix - root user). If you do not use a user with this ID then the - system will be unable to change the user when executing mysqld - as set using --user option. If this happens, mysqld may be - unable to read the files located within the MySQL data - directory and the execution will fail. + * Partitioning. MySQL 5.0 does not support user-defined + partitioning. If a table was created as a partitioned table in + 5.1 (or if an table created in a previous version of MySQL was + altered to include partitions after an upgrade to 5.1), the + table is accessible after downgrade only if you do one of the + following: - 2. Enter the PASE environment using call qp2term. + + Export the table using mysqldump and then drop it in + MySQL 5.1; import the table again following the downgrade + to MySQL 5.0. - 3. Start the MySQL server by changing to the installation - directory and running mysqld_safe, specifying the user name - used to install the server. The installer conveniently - installs a symbolic link to the installation directory - (mysql-5.0.42-i5os-power-32bit) as /opt/mysql/mysql: -> cd /opt/mysql/mysql -> bin/mysqld_safe --user=mysql & - You should see a message similar to the following: -Starting mysqld daemon with databases » - from /opt/mysql/mysql-enterprise-5.0.42-i5os-power-32bit/data + + Prior to the downgrade, remove the table's partitioning + using ALTER TABLE table_name REMOVE PARTITIONING. - If you are having problems starting MySQL server, see Section - 2.11.2.3, "Starting and Troubleshooting the MySQL Server." + * Event Scheduler. MySQL 5.0 does not support scheduled events. + If your databases contain scheduled event definitions, you + should prevent them from being dumped when you use mysqldump + by using the --skip-events option. (See Section 4.5.4, + "mysqldump --- A Database Backup Program.") - To stop MySQL: + * Stored routines. MySQL 5.1.21 added a number of new columns + to the mysql.proc table in which stored routine definitions + are stored. If you are downgrading from MySQL 5.1.21 or later + to MySQL 5.0, you cannot import the MySQL 5.1 routine + definitions into MySQL 5.0.46 or earlier using the dump of + mysql.proc created by mysqldump (such as when using the + --all-databases option). Instead, you should run mysqldump + --routines prior to performing the downgrade and run the + stored routines DDL statements following the downgrade. + See Bug#11986: http://bugs.mysql.com/11986, + Bug#30029: http://bugs.mysql.com/30029, and + Bug#30660: http://bugs.mysql.com/30660, for more information. - 1. Log into the System i server using the *SECOFR class, such as - the QSECOFR user ID. + * Triggers. Trigger creation requires the TRIGGER privilege as + of MySQL 5.1. In MySQL 5.0, there is no TRIGGER privilege and + SUPER is required instead. If you downgrade from MySQL 5.1 to + 5.0, you will need to give the SUPER privilege to those + accounts that had the TRIGGER privilege in 5.1. - 2. Enter the PASE environment using call qp2term. +2.4.3. Checking Whether Tables or Indexes Must Be Rebuilt - 3. Stop the MySQL server by changing into the installation - directory and running mysqladmin, specifying the user name - used to install the server: -> cd /opt/mysql/mysql -> bin/mysqladmin -u root shutdown - If the session that you started and stopped MySQL are the - same, you may get the log output from mysqld: - STOPPING server from pid file » - /opt/mysql/mysql-enterprise-5.0.42-i5os-power-32bit/data/I5DBX.R -CHLAND.IBM.COM.pid - 070718 10:34:20 mysqld ended - If the sessions used to start and stop MySQL are different, - you will not receive any confirmation of the shutdown. + A binary upgrade or downgrade is one that installs one version of + MySQL "in place" over an existing version, without dumping and + reloading tables: - Note and tips + 1. Stop the server for the existing version if it is running. - * A problem has been identified with the installation process on - DBCS systems. If you are having problems install MySQL on a - DBCS system, you need to change your job's coded character set - identifier (CSSID) to 37 (EBCDIC) before executing the install - command, INSMYSQL. To do this, determine your existing CSSID - (using DSPJOB and selecting option 2), execute CHGJOB - CSSID(37), run INSMYSQL to install MySQL and then execute - CHGJOB again with your original CSSID. + 2. Install a different version of MySQL. This is an upgrade if + the new version is higher than the original version, a + downgrade if the version is lower. - * If you want to use the Perl scripts that are included with - MySQL, you need to download the iSeries Tools for Developers - (5799-PTL). See - http://www-03.ibm.com/servers/enable/site/porting/tools/. + 3. Start the server for the new version. -2.8. Installing MySQL on NetWare + In many cases, the tables from the previous version of MySQL can + be used without problem by the new version. However, sometimes + changes occur that require tables or table indexes to be rebuilt, + as described in this section. If you have tables that are affected + by any of the issues described here, rebuild the tables or indexes + as necessary using the instructions given in Section 2.4.4, + "Rebuilding or Repairing Tables or Indexes." - Porting MySQL to NetWare was an effort spearheaded by Novell. - Novell customers should be pleased to note that NetWare 6.5 ships - with bundled MySQL binaries, complete with an automatic commercial - use license for all servers running that version of NetWare. - - MySQL for NetWare is compiled using a combination of Metrowerks - CodeWarrior for NetWare and special cross-compilation versions of - the GNU autotools. + Table Incompatibilities - The latest binary packages for NetWare can be obtained at - http://dev.mysql.com/downloads/. See Section 2.1.3, "How to Get - MySQL." + After a binary upgrade to MySQL 5.1 from a MySQL 5.0 installation + that contains ARCHIVE tables, accessing those tables causes the + server to crash, even if you have run mysql_upgrade or CHECK TABLE + ... FOR UPGRADE. To work around this problem, use mysqldump to + dump all ARCHIVE tables before upgrading, and reload them into + MySQL 5.1 after upgrading. The same problem occurs for binary + downgrades from MySQL 5.1 to 5.0. + + Index Incompatibilities + + If you perform a binary upgrade without dumping and reloading + tables, you cannot upgrade directly from MySQL 4.1 to 5.1 or + higher. This occurs due to an incompatible change in the MyISAM + table index format in MySQL 5.0. Upgrade from MySQL 4.1 to 5.0 and + repair all MyISAM tables. Then upgrade from MySQL 5.0 to 5.1 and + check and repair your tables. + + Modifications to the handling of character sets or collations + might change the character sort order, which causes the ordering + of entries in any index that uses an affected character set or + collation to be incorrect. Such changes result in several possible + problems: - To host MySQL, the NetWare server must meet these requirements: + * Comparison results that differ from previous results - * The latest Support Pack of NetWare 6.5 - (http://support.novell.com/filefinder/18197/index.html) must - be installed. + * Inability to find some index values due to misordered index + entries - * The system must meet Novell's minimum requirements to run the - respective version of NetWare. + * Misordered ORDER BY results - * MySQL data and the program binaries must be installed on an - NSS volume; traditional volumes are not supported. + * Tables that CHECK TABLE reports as being in need of repair - To install MySQL for NetWare, use the following procedure: + The solution to these problems is to rebuild any indexes that use + an affected character set or collation, either by dropping and + re-creating the indexes, or by dumping and reloading the entire + table. For information about rebuilding indexes, see Section + 2.4.4, "Rebuilding or Repairing Tables or Indexes." - 1. If you are upgrading from a prior installation, stop the MySQL - server. This is done from the server console, using the - following command: -SERVER: mysqladmin -u root shutdown + To check whether a table has indexes that must be rebuilt, consult + the following list. It indicates which versions of MySQL + introduced character set or collation changes that require indexes + to be rebuilt. Each entry indicates the version in which the + change occurred and the character sets or collations that the + change affects. If the change is associated with a particular bug + report, the bug number is given. -Note - If the MySQL root user account has a password, you need to - invoke mysqladmin with the -p option and supply the password - when prompted. + The list applies both for binary upgrades and downgrades. For + example, Bug#27877: http://bugs.mysql.com/27877 was fixed in MySQL + 5.1.24 and 5.4.0, so it applies to upgrades from versions older + than 5.1.24 to 5.1.24 or newer, and to downgrades from 5.1.24 or + newer to versions older than 5.1.24. - 2. Log on to the target server from a client machine with access - to the location where you are installing MySQL. + In many cases, you can use CHECK TABLE ... FOR UPGRADE to identify + tables for which index rebuilding is required. (It will report: + Table upgrade required. Please do "REPAIR TABLE `tbl_name`" or + dump/reload to fix it!) In these cases, you can also use + mysqlcheck --check-upgrade or mysql_upgrade, which execute CHECK + TABLE. However, the use of CHECK TABLE applies only after + upgrades, not downgrades. Also, CHECK TABLE is not applicable to + all storage engines. For details about which storage engines CHECK + TABLE supports, see Section 12.5.2.3, "CHECK TABLE Syntax." - 3. Extract the binary package Zip file onto the server. Be sure - to allow the paths in the Zip file to be used. It is safe to - simply extract the file to SYS:\. - If you are upgrading from a prior installation, you may need - to copy the data directory (for example, SYS:MYSQL\DATA), as - well as my.cnf, if you have customized it. You can then delete - the old copy of MySQL. - - 4. You might want to rename the directory to something more - consistent and easy to use. The examples in this manual use - SYS:MYSQL to refer to the installation directory. - Note that MySQL installation on NetWare does not detect if a - version of MySQL is already installed outside the NetWare - release. Therefore, if you have installed the latest MySQL - version from the Web (for example, MySQL 4.1 or later) in - SYS:\MYSQL, you must rename the folder before upgrading the - NetWare server; otherwise, files in SYS:\MySQL are overwritten - by the MySQL version present in NetWare Support Pack. - - 5. At the server console, add a search path for the directory - containing the MySQL NLMs. For example: -SERVER: SEARCH ADD SYS:MYSQL\BIN - - 6. Initialize the data directory and the grant tables, if - necessary, by executing mysql_install_db at the server - console. - - 7. Start the MySQL server using mysqld_safe at the server - console. - - 8. To finish the installation, you should also add the following - commands to autoexec.ncf. For example, if your MySQL - installation is in SYS:MYSQL and you want MySQL to start - automatically, you could add these lines: -#Starts the MySQL 5.1.x database server -SEARCH ADD SYS:MYSQL\BIN -MYSQLD_SAFE - If you are running MySQL on NetWare 6.0, we strongly suggest - that you use the --skip-external-locking option on the command - line: -#Starts the MySQL 5.1.x database server -SEARCH ADD SYS:MYSQL\BIN -MYSQLD_SAFE --skip-external-locking - It is also necessary to use CHECK TABLE and REPAIR TABLE - instead of myisamchk, because myisamchk makes use of external - locking. External locking is known to have problems on NetWare - 6.0; the problem has been eliminated in NetWare 6.5. Note that - the use of MySQL on Netware 6.0 is not officially supported. - mysqld_safe on NetWare provides a screen presence. When you - unload (shut down) the mysqld_safe NLM, the screen does not go - away by default. Instead, it prompts for user input: -*<NLM has terminated; Press any key to close the screen>* - If you want NetWare to close the screen automatically instead, - use the --autoclose option to mysqld_safe. For example: -#Starts the MySQL 5.1.x database server -SEARCH ADD SYS:MYSQL\BIN -MYSQLD_SAFE --autoclose - The behavior of mysqld_safe on NetWare is described further in - Section 4.3.2, "mysqld_safe --- MySQL Server Startup Script." - - 9. When installing MySQL, either for the first time or upgrading - from a previous version, download and install the latest and - appropriate Perl module and PHP extensions for NetWare: - - + Perl: - http://forge.novell.com/modules/xfcontent/downloads.php/p - erl/Modules/ - - + PHP: - http://forge.novell.com/modules/xfcontent/downloads.php/p - hp/Modules/ - - If there was an existing installation of MySQL on the NetWare - server, be sure to check for existing MySQL startup commands in - autoexec.ncf, and edit or delete them as necessary. + Changes that cause index rebuilding to be necessary: -Note + * MySQL 5.0.48, 5.1.21 (Bug#29461: http://bugs.mysql.com/29461) + Affects indexes for columns that use any of these character + sets: eucjpms, euc_kr, gb2312, latin7, macce, ujis + Affected tables can be detected by CHECK TABLE ... FOR UPGRADE + as of MySQL 5.1.29, 5.4.0 (see + Bug#39585: http://bugs.mysql.com/39585) - The accounts that are listed in the MySQL grant tables initially - have no passwords. After starting the server, you should set up - passwords for them using the instructions in Section 2.11, - "Post-Installation Setup and Testing." + * MySQL 5.0.48, 5.1.23 (Bug#27562: http://bugs.mysql.com/27562) + Affects indexes that use the ascii_general_ci collation for + columns that contain any of these characters: '`' GRAVE + ACCENT, '[' LEFT SQUARE BRACKET, '\' REVERSE SOLIDUS, ']' + RIGHT SQUARE BRACKET, '~' TILDE + Affected tables can be detected by CHECK TABLE ... FOR UPGRADE + as of MySQL 5.1.29, 5.4.0 (see + Bug#39585: http://bugs.mysql.com/39585) -2.9. Installing MySQL from tar.gz Packages on Other Unix-Like Systems + * MySQL 5.1.24, 5.4.0 (Bug#27877: http://bugs.mysql.com/27877) + Affects indexes that use the utf8_general_ci or + ucs2_general_ci collation for columns that contain 'ß' LATIN + SMALL LETTER SHARP S (German). + Affected tables can be detected by CHECK TABLE ... FOR UPGRADE + as of MySQL 5.1.30, 5.4.0 (see + Bug#40053: http://bugs.mysql.com/40053) - This section covers the installation of MySQL binary distributions - that are provided for various platforms in the form of compressed - tar files (files with a .tar.gz extension). See Section 2.1.2.4, - "MySQL Binaries Compiled by Sun Microsystems, Inc.," for a - detailed list. +2.4.4. Rebuilding or Repairing Tables or Indexes - To obtain MySQL, see Section 2.1.3, "How to Get MySQL." + This section describes how to rebuild a table. This can be + necessitated by changes to MySQL such as how data types are + handled or changes to character set handling. For example, an + error in a collation might have been corrected, necessitating a + table rebuild to rebuild the indexes for character columns that + use the collation. It might also be that a table repair or upgrade + should be done as indicated by a table check operation such as + that performed by CHECK TABLE, mysqlcheck, or mysql_upgrade. - MySQL tar file binary distributions have names of the form - mysql-VERSION-OS.tar.gz, where VERSION is a number (for example, - 5.1.39), and OS indicates the type of operating system for which - the distribution is intended (for example, pc-linux-i686). + Methods for rebuilding a table include dumping and reloading it, + or using ALTER TABLE or REPAIR TABLE. - In addition to these generic packages, we also offer binaries in - platform-specific package formats for selected platforms. See - Section 2.2, "Standard MySQL Installation Using a Binary - Distribution," for more information on how to install these. +Note - You need the following tools to install a MySQL tar file binary - distribution: + If you are rebuilding tables because a different version of MySQL + will not handle them after a binary (in-place) upgrade or + downgrade, you must use the dump-and-reload method. Dump the + tables before upgrading or downgrading (using your original + version of MySQL), and reload the tables after upgrading or + downgrading (after installing the new version). - * GNU gunzip to uncompress the distribution. + If you use the dump-and-reload method of rebuilding tables only + for the purpose of rebuilding indexes, you can perform the dump + either before or after upgrading or downgrading. Reloading still + must be done afterward. - * A reasonable tar to unpack the distribution. GNU tar is known - to work. Some operating systems come with a preinstalled - version of tar that is known to have problems. For example, - the tar provided with early versions of Mac OS X, SunOS 4.x - and Solaris 8 and earlier are known to have problems with long - file names. On Mac OS X, you can use the preinstalled gnutar - program. On other systems with a deficient tar, you should - install GNU tar first. + To re-create a table by dumping and reloading it, use mysqldump to + create a dump file and mysql to reload the file: +shell> mysqldump db_name t1 > dump.sql +shell> mysql db_name < dump.sql - If you run into problems and need to file a bug report, please use - the instructions in Section 1.6, "How to Report Bugs or Problems." + To recreate all the tables in a single database, specify the + database name without any following table name: +shell> mysqldump db_name > dump.sql +shell> mysql db_name < dump.sql - The basic commands that you must execute to install and use a - MySQL binary distribution are: -shell> groupadd mysql -shell> useradd -g mysql mysql -shell> cd /usr/local -shell> gunzip < /path/to/mysql-VERSION-OS.tar.gz | tar xvf - -shell> ln -s full-path-to-mysql-VERSION-OS mysql -shell> cd mysql -shell> chown -R mysql . -shell> chgrp -R mysql . -shell> scripts/mysql_install_db --user=mysql -shell> chown -R root . -shell> chown -R mysql data -shell> bin/mysqld_safe --user=mysql & + To recreate all tables in all databases, use the --all-databases + option: +shell> mysqldump --all-databases > dump.sql +shell> mysql < dump.sql -Note + To rebuild a table with ALTER TABLE, use a "null" alteration; that + is, an ALTER TABLE statement that "changes" the table to use the + storage engine that it already has. For example, if t1 is a MyISAM + table, use this statement: +mysql> ALTER TABLE t1 ENGINE = MyISAM; - This procedure does not set up any passwords for MySQL accounts. - After following the procedure, proceed to Section 2.11, - "Post-Installation Setup and Testing." + If you are not sure which storage engine to specify in the ALTER + TABLE statement, use SHOW CREATE TABLE to display the table + definition. - A more detailed version of the preceding description for - installing a binary distribution follows: + If you must rebuild a table because a table checking operation + indicates that the table is corrupt or needs an upgrade, you can + use REPAIR TABLE if that statement supports the table's storage + engine. For example, to repair a MyISAM table, use this statement: +mysql> REPAIR TABLE t1; - 1. Add a login user and group for mysqld to run as: -shell> groupadd mysql -shell> useradd -g mysql mysql - These commands add the mysql group and the mysql user. The - syntax for useradd and groupadd may differ slightly on - different versions of Unix, or they may have different names - such as adduser and addgroup. - You might want to call the user and group something else - instead of mysql. If so, substitute the appropriate name in - the following steps. + For storage engines such as InnoDB that REPAIR TABLE does not + support, use mysqldump to create a dump file and mysql to reload + the file, as described earlier. - 2. Pick the directory under which you want to unpack the - distribution and change location into it. In the following - example, we unpack the distribution under /usr/local. (The - instructions, therefore, assume that you have permission to - create files and directories in /usr/local. If that directory - is protected, you must perform the installation as root.) -shell> cd /usr/local + For specifics about which storage engines REPAIR TABLE supports, + see Section 12.5.2.6, "REPAIR TABLE Syntax." - 3. Obtain a distribution file using the instructions in Section - 2.1.3, "How to Get MySQL." For a given release, binary - distributions for all platforms are built from the same MySQL - source distribution. + mysqlcheck --repair provides command-line access to the REPAIR + TABLE statement. This can be a more convenient means of repairing + tables because you can use the --databases or --all-databases + option to repair all tables in specific databases or all + databases, respectively: +shell> mysqlcheck --repair --databases db_name ... +shell> mysqlcheck --repair --all-databases - 4. Unpack the distribution, which creates the installation - directory. Then create a symbolic link to that directory: -shell> gunzip < /path/to/mysql-VERSION-OS.tar.gz | tar xvf - -shell> ln -s full-path-to-mysql-VERSION-OS mysql - The tar command creates a directory named mysql-VERSION-OS. - The ln command makes a symbolic link to that directory. This - lets you refer more easily to the installation directory as - /usr/local/mysql. - With GNU tar, no separate invocation of gunzip is necessary. - You can replace the first line with the following alternative - command to uncompress and extract the distribution: -shell> tar zxvf /path/to/mysql-VERSION-OS.tar.gz +2.4.5. Copying MySQL Databases to Another Machine - 5. Change location into the installation directory: -shell> cd mysql - You will find several files and subdirectories in the mysql - directory. The most important for installation purposes are - the bin and scripts subdirectories: + You can copy the .frm, .MYI, and .MYD files for MyISAM tables + between different architectures that support the same + floating-point format. (MySQL takes care of any byte-swapping + issues.) See Section 13.5, "The MyISAM Storage Engine." - + The bin directory contains client programs and the - server. You should add the full path name of this - directory to your PATH environment variable so that your - shell finds the MySQL programs properly. See Section - 2.14, "Environment Variables." + In cases where you need to transfer databases between different + architectures, you can use mysqldump to create a file containing + SQL statements. You can then transfer the file to the other + machine and feed it as input to the mysql client. - + The scripts directory contains the mysql_install_db - script used to initialize the mysql database containing - the grant tables that store the server access - permissions. + Use mysqldump --help to see what options are available. - 6. Ensure that the distribution contents are accessible to mysql. - If you unpacked the distribution as mysql, no further action - is required. If you unpacked the distribution as root, its - contents will be owned by root. Change its ownership to mysql - by executing the following commands as root in the - installation directory: -shell> chown -R mysql . -shell> chgrp -R mysql . - The first command changes the owner attribute of the files to - the mysql user. The second changes the group attribute to the - mysql group. + The easiest (although not the fastest) way to move a database + between two machines is to run the following commands on the + machine on which the database is located: +shell> mysqladmin -h 'other_hostname' create db_name +shell> mysqldump db_name | mysql -h 'other_hostname' db_name - 7. If you have not installed MySQL before, you must create the - MySQL data directory and initialize the grant tables: -shell> scripts/mysql_install_db --user=mysql - If you run the command as root, include the --user option as - shown. If you run the command while logged in as that user, - you can omit the --user option. - The command should create the data directory and its contents - with mysql as the owner. - After creating or updating the grant tables, you need to - restart the server manually. + If you want to copy a database from a remote machine over a slow + network, you can use these commands: +shell> mysqladmin create db_name +shell> mysqldump -h 'other_hostname' --compress db_name | mysql db_na +me - 8. Most of the MySQL installation can be owned by root if you - like. The exception is that the data directory must be owned - by mysql. To accomplish this, run the following commands as - root in the installation directory: -shell> chown -R root . -shell> chown -R mysql data + You can also store the dump in a file, transfer the file to the + target machine, and then load the file into the database there. + For example, you can dump a database to a compressed file on the + source machine like this: +shell> mysqldump --quick db_name | gzip > db_name.gz - 9. If you want MySQL to start automatically when you boot your - machine, you can copy support-files/mysql.server to the - location where your system has its startup files. More - information can be found in the support-files/mysql.server - script itself and in Section 2.11.2.2, "Starting and Stopping - MySQL Automatically." - 10. You can set up new accounts using the bin/mysql_setpermission - script if you install the DBI and DBD::mysql Perl modules. See - Section 4.6.14, "mysql_setpermission --- Interactively Set - Permissions in Grant Tables." For Perl module installation - instructions, see Section 2.15, "Perl Installation Notes." - 11. If you would like to use mysqlaccess and have the MySQL - distribution in some nonstandard location, you must change the - location where mysqlaccess expects to find the mysql client. - Edit the bin/mysqlaccess script at approximately line 18. - Search for a line that looks like this: -$MYSQL = '/usr/local/bin/mysql'; # path to mysql executable - Change the path to reflect the location where mysql actually - is stored on your system. If you do not do this, a Broken pipe - error will occur when you run mysqlaccess. + Transfer the file containing the database contents to the target + machine and run these commands there: +shell> mysqladmin create db_name +shell> gunzip < db_name.gz | mysql db_name - After everything has been unpacked and installed, you should test - your distribution. To start the MySQL server, use the following - command: -shell> bin/mysqld_safe --user=mysql & + You can also use mysqldump and mysqlimport to transfer the + database. For large tables, this is much faster than simply using + mysqldump. In the following commands, DUMPDIR represents the full + path name of the directory you use to store the output from + mysqldump. - If you run the command as root, you must use the --user option as - shown. The value of the option is the name of the login account - that you created in the first step to use for running the server. - If you run the command while logged in as mysql, you can omit the - --user option. + First, create the directory for the output files and dump the + database: +shell> mkdir DUMPDIR +shell> mysqldump --tab=DUMPDIR db_name - If the command fails immediately and prints mysqld ended, you can - find some information in the host_name.err file in the data - directory. + Then transfer the files in the DUMPDIR directory to some + corresponding directory on the target machine and load the files + into MySQL there: +shell> mysqladmin create db_name # create database +shell> cat DUMPDIR/*.sql | mysql db_name # create tables in databas +e +shell> mysqlimport db_name DUMPDIR/*.txt # load data into tables - More information about mysqld_safe is given in Section 4.3.2, - "mysqld_safe --- MySQL Server Startup Script." + Do not forget to copy the mysql database because that is where the + grant tables are stored. You might have to run commands as the + MySQL root user on the new machine until you have the mysql + database in place. -Note + After you import the mysql database on the new machine, execute + mysqladmin flush-privileges so that the server reloads the grant + table information. - The accounts that are listed in the MySQL grant tables initially - have no passwords. After starting the server, you should set up - passwords for them using the instructions in Section 2.11, - "Post-Installation Setup and Testing." +2.5. Installing MySQL on Windows -2.10. MySQL Installation Using a Source Distribution + This section describes the process for installing MySQL on + Windows. - Before you proceed with an installation from source, first check - whether our binary is available for your platform and whether it - works for you. We put a great deal of effort into ensuring that - our binaries are built with the best possible options. + To run MySQL on Windows, you need the following: - To obtain a source distribution for MySQL, Section 2.1.3, "How to - Get MySQL." If you want to build MySQL from source on Windows, see - Section 2.10.6, "Installing MySQL from Source on Windows." + * A Windows operating system such as Windows 2000, Windows XP, + Windows Vista, Windows Server 2003, or Windows Server 2008. + Both 32-bit and 64-bit versions are supported. + In addition to running MySQL as a standard application, you + can also run the MySQL server as a Windows service. By using a + service you can monitor and control the operation of the + server through the standard Windows service management tools. + For more information, see Section 2.5.5.6, "Starting MySQL as + a Windows Service." + Generally, you should install MySQL on Windows using an + account that has administrator rights. Otherwise, you may + encounter problems with certain operations such as editing the + PATH environment variable or accessing the Service Control + Manager. Once installed, MySQL does not need to be executed + using a user with Administrator privileges. - MySQL source distributions are provided as compressed tar archives - and have names of the form mysql-VERSION.tar.gz, where VERSION is - a number like 5.1.39. + * TCP/IP protocol support. - You need the following tools to build and install MySQL from - source: + * Enough space on the hard drive to unpack, install, and create + the databases in accordance with your requirements (generally + a minimum of 200 megabytes is recommended.) - * GNU gunzip to uncompress the distribution. + For a list of limitations within the Windows version of MySQL, see + Section D.7.3, "Windows Platform Limitations." - * A reasonable tar to unpack the distribution. GNU tar is known - to work. Some operating systems come with a preinstalled - version of tar that is known to have problems. For example, - the tar provided with early versions of Mac OS X, SunOS 4.x - and Solaris 8 and earlier are known to have problems with long - file names. On Mac OS X, you can use the preinstalled gnutar - program. On other systems with a deficient tar, you should - install GNU tar first. + In addition to the MySQL Server package, you may need or want + additional components to use MySQL with your application or + development environment. These include, but are not limited to: - * A working ANSI C++ compiler. gcc 2.95.2 or later, SGI C++, and - SunPro C++ are some of the compilers that are known to work. - libg++ is not needed when using gcc. gcc 2.7.x has a bug that - makes it impossible to compile some perfectly legal C++ files, - such as sql/sql_base.cc. If you have only gcc 2.7.x, you must - upgrade your gcc to be able to compile MySQL. gcc 2.8.1 is - also known to have problems on some platforms, so it should be - avoided if a newer compiler exists for the platform. gcc - 2.95.2 or later is recommended. + * If you plan to connect to the MySQL server via ODBC, you need + a Connector/ODBC driver. For more information, including + installation and configuration instructions, see Section 21.1, + "MySQL Connector/ODBC." + + * If you plan to use MySQL server with .NET applications, you + need the Connector/NET driver. For more information, including + installation and configuration instructions, see Section 21.2, + "MySQL Connector/NET." - * A good make program. GNU make is always recommended and is - sometimes required. (BSD make fails, and vendor-provided make - implementations may fail as well.) If you have problems, use - GNU make 3.75 or newer. + MySQL distributions for Windows can be downloaded from + http://dev.mysql.com/downloads/. See Section 2.1.3, "How to Get + MySQL." - * libtool 1.5.24 or later is also recommended. + MySQL for Windows is available in several distribution formats, + detailed below. Generally speaking, you should use a binary + distribution that includes an installer. It is simpler to use than + the others, and you need no additional tools to get MySQL up and + running. The installer for the Windows version of MySQL, combined + with a GUI Config Wizard, automatically installs MySQL, creates an + option file, starts the server, and secures the default user + accounts. - If you are using a version of gcc recent enough to understand the - -fno-exceptions option, it is very important that you use this - option. Otherwise, you may compile a binary that crashes randomly. - Also use -felide-constructors and -fno-rtti along with - -fno-exceptions. When in doubt, do the following: -CFLAGS="-O3" CXX=gcc CXXFLAGS="-O3 -felide-constructors \ - -fno-exceptions -fno-rtti" ./configure \ - --prefix=/usr/local/mysql --enable-assembler \ - --with-mysqld-ldflags=-all-static + * Binary installer distribution. The installable distribution + comes packaged as a Microsoft Windows Installer (MSI) package + that you can install manually or automatically on your + systems. Two formats are available, an essentials package that + contains all the files you need to install and configure + MySQL, but no additional components, and a complete package + that includes MySQL, configuration tools, benchmarks and other + components. For more information on the specific differences, + see Section 2.5.2, "Choosing An Installation Package" + For instructions on installing MySQL using one of the MSI + installation packages, see Section 2.5.3, "Installing MySQL + with the MSI Package." + + * Standard binary distribution format packaged as a Zip file + containing all of the necessary files that you unpack into + your chosen location. This package contains all of the files + in the full Windows MSI Installer package, but does not + including an installation program. + For instructions on installing MySQL using the Zip file, see + Section 2.5.5, "Installing MySQL from a noinstall Zip + Archive." - On most systems, this gives you a fast and stable binary. + * The source distribution contains all the code and support + files for building the executables using the Visual Studio + compiler system. + For instructions on building MySQL from source on Windows, see + Section 2.5.10, "Installing MySQL from Source on Windows." - If you run into problems and need to file a bug report, please use - the instructions in Section 1.6, "How to Report Bugs or Problems." + MySQL on Windows considerations: -2.10.1. Source Installation Overview + * Large Table Support + If you need tables with a size larger than 4GB, install MySQL + on an NTFS or newer file system. Don't forget to use MAX_ROWS + and AVG_ROW_LENGTH when you create tables. See Section + 12.1.17, "CREATE TABLE Syntax." - The basic commands that you must execute to install a MySQL source - distribution are: -shell> groupadd mysql -shell> useradd -g mysql mysql -shell> gunzip < mysql-VERSION.tar.gz | tar -xvf - -shell> cd mysql-VERSION -shell> ./configure --prefix=/usr/local/mysql -shell> make -shell> make install -shell> cp support-files/my-medium.cnf /etc/my.cnf -shell> cd /usr/local/mysql -shell> chown -R mysql . -shell> chgrp -R mysql . -shell> bin/mysql_install_db --user=mysql -shell> chown -R root . -shell> chown -R mysql var -shell> bin/mysqld_safe --user=mysql & + * MySQL and Virus Checking Software + Using virus scanning software such as Norton/Symantec + Anti-Virus on directories containing MySQL data and temporary + tables can cause issues, both in terms of the performance of + MySQL and the virus-scanning software mis-identifying the + contents of the files as containing spam. This is because of + the fingerprinting mechanism used by the virus scanning + software, and the way in which MySQL rapidly updates different + files, which may be identified as a potential security risk. + After installing MySQL Server, it is recommended that you + disable virus scanning on the main directory (datadir) being + used to store your MySQL table data. There is usually a system + built into the virus scanning software to allow certain + directories to be specifically ignored during virus scanning. + In addition, by default, MySQL creates temporary files in the + standard Windows temporary directory. To prevent the temporary + files also being scanned, you should configure a separate + temporary directory for MySQL temporary files and add this to + the virus scanning exclusion list. To do this, add a + configuration option for the tmpdir parameter to your my.ini + configuration file. For more information, see Section 2.5.5.2, + "Creating an Option File." - If you start from a source RPM, do the following: -shell> rpmbuild --rebuild --clean MySQL-VERSION.src.rpm +2.5.1. Windows Installation Layout - This makes a binary RPM that you can install. For older versions - of RPM, you may have to replace the command rpmbuild with rpm - instead. + For MySQL 5.1 on Windows, the default installation directory is + C:\Program Files\MySQL\MySQL Server 5.1. Some Windows users prefer + to install in C:\mysql, the directory that formerly was used as + the default. However, the layout of the subdirectories remains the + same. + + For MySQL 5.1.23 and earlier, all of the files are located within + this parent directory, using the following structure: + + Table 2.2. Installation Layout for Windows using MySQL 5.1.23 and + earlier + Directory Contents of Directory + bin Client programs and the mysqld server + data Log files, databases + Docs Manual in CHM format + examples Example programs and scripts + include Include (header) files + lib Libraries + scripts Utility scripts + share Error message files -Note + For MySQL 5.1.24 and later, the default location of data directory + was changed. The remainder of the directory structure remains the + same: - This procedure does not set up any passwords for MySQL accounts. - After following the procedure, proceed to Section 2.11, - "Post-Installation Setup and Testing," for post-installation setup - and testing. + Table 2.3. Installation Layout for Windows using MySQL 5.1.24 and + later + Directory Contents of Directory + bin Client programs and the mysqld server + C:\Documents and Settings\All Users\Application Data\MySQL Log + files, databases + Docs Manual in CHM format + examples Example programs and scripts + include Include (header) files + lib Libraries + scripts Utility scripts + share Error message files - A more detailed version of the preceding description for - installing MySQL from a source distribution follows: +2.5.2. Choosing An Installation Package - 1. Add a login user and group for mysqld to run as: -shell> groupadd mysql -shell> useradd -g mysql mysql - These commands add the mysql group and the mysql user. The - syntax for useradd and groupadd may differ slightly on - different versions of Unix, or they may have different names - such as adduser and addgroup. - You might want to call the user and group something else - instead of mysql. If so, substitute the appropriate name in - the following steps. + For MySQL 5.1, there are three installation packages to choose + from when installing MySQL on Windows: + Packaging + Feature Essentials Complete Zip (No-install) + Installer Yes Yes No + Directory-only + MySQL Server Instance Config Wizard Yes Yes No + Test Suite No Yes Yes + MySQL Server Yes Yes Yes + MySQL Client Programs Yes Yes Yes + C Headers/Libraries Yes Yes Yes + Embedded Server No Optional Yes + Scripts and Examples No Optional Yes + + In the above table: + + * Yes indiciates that the component is installed by default. + + * No indicates that the component is not installed or included. + + * Optional indicates that the component is included with the + package, but not installed unless explicitly requested using + the Custom installation mode. + + The workflow for installing using the MSI installer is shown + below: - 2. Perform the following steps as the mysql user, except as - noted. + Figure 2.1. Installation Workflow for Windows using MSI + Installation Workflow for Windows using MSI - 3. Pick the directory under which you want to unpack the - distribution and change location into it. + The workflow for installing using the MSI installer is shown + below: - 4. Obtain a distribution file using the instructions in Section - 2.1.3, "How to Get MySQL." + Figure 2.2. Installation Workflow for Windows using Zip + Installation Workflow for Windows using Zip - 5. Unpack the distribution into the current directory: -shell> gunzip < /path/to/mysql-VERSION.tar.gz | tar xvf - - This command creates a directory named mysql-VERSION. - With GNU tar, no separate invocation of gunzip is necessary. - You can use the following alternative command to uncompress - and extract the distribution: -shell> tar zxvf /path/to/mysql-VERSION-OS.tar.gz +Note - 6. Change location into the top-level directory of the unpacked - distribution: -shell> cd mysql-VERSION - Note that currently you must configure and build MySQL from - this top-level directory. You cannot build it in a different - directory. + For the Essentials and Complete packages in the MSI installer, you + can select individual components to be installed by using the + Custom mode, including disable the components confiurated for + installation by default. + + Full details on the components are suggested uses are provided + below for reference: + + * Windows Essentials --- this package has a file name similar to + mysql-essential-5.1.41-win32.msi and is supplied as a + Microsoft Installer (MSI) package. The package includes the + minimum set of files needed to install MySQL on Windows, + including the MySQL Server Instance Config Wizard. This + package does not include optional components such as the + embedded server, developer headers and libraries or benchmark + suite. + To install using this package, see Section 2.5.3, "Installing + MySQL with the MSI Package." + + * Windows MSI Installer (Complete) --- this package has a file + name similar to mysql-5.1.41-win32.zip and contains all files + needed for a complete Windows installation, including the + MySQL Server Instance Config Wizard. This package includes + optional components such as the embedded server and benchmark + suite. + To install using this package, see Section 2.5.3, "Installing + MySQL with the MSI Package." - 7. Configure the release and compile everything: -shell> ./configure --prefix=/usr/local/mysql -shell> make - When you run configure, you might want to specify other - options. Run ./configure --help for a list of options. Section - 2.10.2, "Typical configure Options," discusses some of the - more useful options. - If configure fails and you are going to send mail to a MySQL - mailing list to ask for assistance, please include any lines - from config.log that you think can help solve the problem. - Also include the last couple of lines of output from - configure. To file a bug report, please use the instructions - in Section 1.6, "How to Report Bugs or Problems." - If the compile fails, see Section 2.10.4, "Dealing with - Problems Compiling MySQL," for help. + * Without installer --- this package has a file name similar to + mysql-noinstall-5.1.41-win32.zip and contains all the files + found in the Complete install package, with the exception of + the MySQL Server Instance Config Wizard. This package does not + include an automated installer, and must be manually installed + and configured. + + The Essentials package is recommended for most users. Both the + Essentials and Complete distributions are available as an .msi + file for use with the Windows Installer. The Noinstall + distribution is packaged as Zip archives. To use Zip archives, you + must have a tool that can unpack .zip files. + + When using the MSI installers you can automate the installation + process. For more information, see Section 2.5.3.2, "Installing + MySQL Automatically using MSI." To automate the creation of a + MySQL instance, see Section 2.5.4.13, "Creating an Instance from + the Command Line." - 8. Install the distribution: -shell> make install - You might need to run this command as root. - If you want to set up an option file, use one of those present - in the support-files directory as a template. For example: -shell> cp support-files/my-medium.cnf /etc/my.cnf - You might need to run this command as root. - If you want to configure support for InnoDB tables, you should - edit the /etc/my.cnf file, remove the # character before the - option lines that start with innodb_..., and modify the option - values to be what you want. See Section 4.2.3.3, "Using Option - Files," and Section 13.6.2, "InnoDB Configuration." + Your choice of install package affects the installation process + you must follow. If you choose to install either the Essentials or + Complete install packages, see Section 2.5.3, "Installing MySQL + with the MSI Package." If you choose to install MySQL from the + Noinstall archive, see Section 2.5.5, "Installing MySQL from a + noinstall Zip Archive." - 9. Change location into the installation directory: -shell> cd /usr/local/mysql - 10. If you ran the make install command as root, the installed - files will be owned by root. Ensure that the installation is - accessible to mysql by executing the following commands as - root in the installation directory: -shell> chown -R mysql . -shell> chgrp -R mysql . - The first command changes the owner attribute of the files to - the mysql user. The second changes the group attribute to the - mysql group. - 11. If you have not installed MySQL before, you must create the - MySQL data directory and initialize the grant tables: -shell> bin/mysql_install_db --user=mysql - If you run the command as root, include the --user option as - shown. If you run the command while logged in as mysql, you - can omit the --user option. - The command should create the data directory and its contents - with mysql as the owner. - After using mysql_install_db to create the grant tables for - MySQL, you must restart the server manually. The mysqld_safe - command to do this is shown in a later step. - 12. Most of the MySQL installation can be owned by root if you - like. The exception is that the data directory must be owned - by mysql. To accomplish this, run the following commands as - root in the installation directory: -shell> chown -R root . -shell> chown -R mysql var - 13. If you want MySQL to start automatically when you boot your - machine, you can copy support-files/mysql.server to the - location where your system has its startup files. More - information can be found in the support-files/mysql.server - script itself; see also Section 2.11.2.2, "Starting and - Stopping MySQL Automatically." - 14. You can set up new accounts using the bin/mysql_setpermission - script if you install the DBI and DBD::mysql Perl modules. See - Section 4.6.14, "mysql_setpermission --- Interactively Set - Permissions in Grant Tables." For Perl module installation - instructions, see Section 2.15, "Perl Installation Notes." +2.5.3. Installing MySQL with the MSI Package + + The MSI package are designed to install and configure MySQL in + such a way that you can immediately get started using MySQL. + + The MySQL Installation Wizard and MySQL Config Wizard are + available in the Essentials and Complete install packages. They + are recommended for most standard MySQL installations. Exceptions + include users who need to install multiple instances of MySQL on a + single server host and advanced users who want complete control of + server configuration. - After everything has been installed, you should test your - distribution. To start the MySQL server, use the following - command: -shell> /usr/local/mysql/bin/mysqld_safe --user=mysql & + * For information on installing using the GUI MSI installer + process, see Section 2.5.3, "Installing MySQL with the MSI + Package." - If you run the command as root, you should use the --user option - as shown. The value of the option is the name of the login account - that you created in the first step to use for running the server. - If you run the command while logged in as that user, you can omit - the --user option. + * For information on installing using the command line using the + MSI package, see Section 2.5.3.2, "Installing MySQL + Automatically using MSI." + + * If you have previously installed MySQL using the MSI package + and want to remove MySQL, see Section 2.5.3.3, "Removing MySQL + Installed from the MSI Package." - If the command fails immediately and prints mysqld ended, you can - find some information in the host_name.err file in the data - directory. + The workflow sequence for using the installer is shown in the + figure below: - More information about mysqld_safe is given in Section 4.3.2, - "mysqld_safe --- MySQL Server Startup Script." + Figure 2.3. Installation Workflow for Windows using MSI Installer + Installation Workflow for Windows using MSI Installer Note - The accounts that are listed in the MySQL grant tables initially - have no passwords. After starting the server, you should set up - passwords for them using the instructions in Section 2.11, - "Post-Installation Setup and Testing." + Microsoft Windows XP and later include a firewall which + specifically blocks ports. If you plan on using MySQL through a + network port then you should open and create an exception for this + port before performing the installation. To check and if necessary + add an exception to the firewall settings: -2.10.2. Typical configure Options + 1. First ensure that you are logged in as an Administrator or a + user with Administrator privileges. - The configure script gives you a great deal of control over how - you configure a MySQL source distribution. Typically you do this - using options on the configure command line. You can also affect - configure using certain environment variables. See Section 2.14, - "Environment Variables." For a full list of options supported by - configure, run this command: -shell> ./configure --help + 2. Go to the Control Panel, and double click the Windows Firewall + icon. - A list of the available configure options is provided in the table - below. + 3. Choose the Allow a program through Windows Firewall option and + click the Add port button. - Table 2.1. Build (configure) Reference - Formats Description Default Introduced Removed - --bindir=DIR User executables EPREFIX/bin - --build=BUILD Configure for building on BUILD guessed - --cache-file=FILE Cache test results in FILE disabled - -C Alias for `--cache-file=config.cache' - --config-cache - --datadir=DIR Read-only architecture-independent data PREFIX/share + 4. Enter MySQL into the Name text box and 3306 (or the port of + your choice) into the Port number text box. - --disable-FEATURE Do not include FEATURE - --disable-dependency-tracking Disable dependency tracking - --disable-grant-options Disable GRANT options - --disable-largefile Omit support for large files - --disable-libtool-lock Disable libtool lock - --disable-thread-safe-client Compile the client without threads - 5.1.7 - --enable-FEATURE Enable FEATURE - --enable-assembler Use assembler versions of some string functions - if available - --enable-dependency-tracking Do not reject slow dependency - extractors - --enable-fast-install Optimize for fast installation yes - --enable-local-infile Enable LOAD DATA LOCAL INFILE disabled - --enable-shared Build shared libraries yes - --enable-static Build static libraries yes - --enable-thread-safe-client Compile the client with threads - --exec-prefix=EPREFIX Install architecture-dependent files in - EPREFIX - -h Display this help and exit - --help - --help=short Display options specific to this package - --help=recursive Display the short help of all the included - packages - --host=HOST Cross-compile to build programs to run on HOST - --includedir=DIR C header files PREFIX/include - --infodir=DIR Info documentation PREFIX/info - --libdir=DIR Object code libraries EPREFIX/lib - --libexecdir=DIR Program executables EPREFIX/libexec - --localstatedir=DIR Modifiable single-machine data PREFIX/var - --mandir=DIR man documentation PREFIX/man - -n Do not create output files - --no-create - --oldincludedir=DIR C header files for non-gcc /usr/include - --prefix=PREFIX Install architecture-independent files in PREFIX + 5. Also ensure that the TCP protocol radio button is selected. - --program-prefix=PREFIX Prepend PREFIX to installed program names + 6. If you wish, you can also limit access to the MySQL server by + choosing the Change scope button. - --program-suffix=SUFFIX Append SUFFIX to installed program names + 7. Confirm your choices by clicking the OK button. - --program-transform-name=PROGRAM run sed PROGRAM on installed - program names - -q Do not print `checking...' messages - --quiet - --sbindir=DIR System admin executables EPREFIX/sbin - --sharedstatedir=DIR Modifiable architecture-independent data - PREFIX/com - --srcdir=DIR Find the sources in DIR configure directory or .. - --sysconfdir=DIR Read-only single-machine data PREFIX/etc - --target=TARGET Configure for building compilers for TARGET - -V Display version information and exit - --version - --with-PACKAGE Use PACKAGE - --with-archive-storage-engine Enable the Archive Storage Engine no + Additionally, when running the MySQL Installation Wizard on + Windows Vista, ensure that you are logged in as a user with + administrative rights. - --with-atomic-ops Implement atomic operations using pthread - rwlocks or atomic CPU instructions for multi-processor 5.1.12 - --with-berkeley-db Use BerkeleyDB located in DIR no - --with-berkeley-db-includes Find Berkeley DB headers in DIR - --with-berkeley-db-libs Find Berkeley DB libraries in DIR - --with-big-tables Support tables with more than 4 G rows even on - 32 bit platforms - --with-blackhole-storage-engine Enable the Blackhole Storage - Engine no - --with-charset Default character set - --with-client-ldflags Extra linking arguments for clients - --with-collation Default collation - --with-comment Comment about compilation environment - --with-csv-storage-engine Enable the CSV Storage Engine yes - --with-darwin-mwcc Use Metrowerks CodeWarrior wrappers on OS - X/Darwin - --with-debug Add debug code 5.1.7 - --with-debug=full Add debug code (adds memory checker, very slow) +Note - --with-embedded-privilege-control Build parts to check user's - privileges (only affects embedded library) - --with-embedded-server Build the embedded server - --with-error-inject Enable error injection in MySQL Server - 5.1.11 - --with-example-storage-engine Enable the Example Storage Engine no + When using Windows Vista, you may want to disable User Account + Control (UAC) before performing the installation. If you do not do + so, then MySQL may be identified as a security risk, which will + mean that you need to enable MySQL. You can disable the security + checking by following these instructions: - --with-extra-charsets Use charsets in addition to default - --with-fast-mutexes Compile with fast mutexes enabled 5.1.5 - --with-federated-storage-engine Enable federated storage engine no - 5.1.3 5.1.9 - --with-gnu-ld Assume the C compiler uses GNU ld no - --with-innodb Enable innobase storage engine no 5.1.3 5.1.9 - --with-lib-ccflags Extra CC options for libraries - --with-libwrap=DIR Compile in libwrap (tcp_wrappers) support - --with-low-memory Try to use less memory to compile to avoid - memory limitations - --with-machine-type Set the machine type, like "powerpc" - --with-max-indexes=N Sets the maximum number of indexes per table - 64 - --with-mysqld-ldflags Extra linking arguments for mysqld - --with-mysqld-libs Extra libraries to link with for mysqld - --with-mysqld-user What user the mysqld daemon shall be run as + 1. Open Control Panel. - --with-mysqlmanager Build the mysqlmanager binary Build if server - is built - --with-named-curses-libs Use specified curses libraries - --with-named-thread-libs Use specified thread libraries - --with-ndb-ccflags Extra CC options for ndb compile - --with-ndb-docs Include the NDB Cluster ndbapi and mgmapi - documentation - --with-ndb-port Port for NDB Cluster management server - --with-ndb-port-base Port for NDB Cluster management server - --with-ndb-sci=DIR Provide MySQL with a custom location of sci - library - --with-ndb-test Include the NDB Cluster ndbapi test programs - --with-ndbcluster Include the NDB Cluster table handler no - --with-openssl=DIR Include the OpenSSL support - --with-openssl-includes Find OpenSSL headers in DIR - --with-openssl-libs Find OpenSSL libraries in DIR - --with-other-libc=DIR Link against libc and other standard - libraries installed in the specified nonstandard location - --with-pic Try to use only PIC/non-PIC objects Use both - --with-plugin-PLUGIN Forces the named plugin to be linked into - mysqld statically 5.1.11 - --with-plugins Plugins to include in mysqld none 5.1.11 - --with-pstack Use the pstack backtrace library - --with-pthread Force use of pthread library - --with-row-based-replication Include row-based replication 5.1.5 - 5.1.6 - --with-server-suffix Append value to the version string - --with-ssl=DIR Include SSL support 5.1.11 - --with-system-type Set the system type, like "sun-solaris10" - --with-tags Include additional configurations automatic - --with-tcp-port Which port to use for MySQL services 3306 - --with-unix-socket-path Where to put the unix-domain socket - --with-yassl Include the yaSSL support - --with-zlib-dir=no|bundled|DIR Provide MySQL with a custom - location of compression library - --without-PACKAGE Do not use PACKAGE - --without-bench Skip building of the benchmark suite - --without-debug Build a production version without debugging code + 2. Under the User Accounts and Family Safety, select Add or + remove user accounts. - --without-docs Skip building of the documentation - --without-extra-tools Skip building utilities in the tools - directory - --without-geometry Do not build geometry-related parts - --without-libedit Use system libedit instead of bundled copy - --without-man Skip building of the man pages - --without-ndb-binlog Disable ndb binlog 5.1.6 - --without-ndb-debug Disable special ndb debug features - --without-plugin-PLUGIN Exclude PLUGIN 5.1.11 - --without-query-cache Do not build query cache - --without-readline Use system readline instead of bundled copy + 3. Click on the Got to the main User Accounts page link. - --without-row-based-replication Don't include row-based - replication 5.1.7 5.1.14 - --without-server Only build the client - --without-uca Skip building of the national Unicode collations + 4. Click on Turn User Account Control on or off. You may be + prompted to provide permission to change this setting. Click + Continue. - Some of the configure options available are described here. For - options that may be of use if you have difficulties building - MySQL, see Section 2.10.4, "Dealing with Problems Compiling - MySQL." + 5. Deselect or unceck the checkbox next to Use User Account + Control (UAC) to help protect your computer. Click OK to save + the setting. - * To compile just the MySQL client libraries and client programs - and not the server, use the --without-server option: -shell> ./configure --without-server - If you have no C++ compiler, some client programs such as - mysql cannot be compiled because they require C++.. In this - case, you can remove the code in configure that tests for the - C++ compiler and then run ./configure with the - --without-server option. The compile step should still try to - build all clients, but you can ignore any warnings about files - such as mysql.cc. (If make stops, try make -k to tell it to - continue with the rest of the build even if errors occur.) + You will need to restart to complete the process. Click Restart + Now to reboot the machine and apply the changes. You can then + follow the instructions below for installing Windows. - * If you want to build the embedded MySQL library (libmysqld.a), - use the --with-embedded-server option. +2.5.3.1. Using the MySQL Installation Wizard - * If you don't want your log files and database directories - located under /usr/local/var, use a configure command - something like one of these: -shell> ./configure --prefix=/usr/local/mysql -shell> ./configure --prefix=/usr/local \ - --localstatedir=/usr/local/mysql/data - The first command changes the installation prefix so that - everything is installed under /usr/local/mysql rather than the - default of /usr/local. The second command preserves the - default installation prefix, but overrides the default - location for database directories (normally /usr/local/var) - and changes it to /usr/local/mysql/data. - You can also specify the installation directory and data - directory locations at server startup time by using the - --basedir and --datadir options. These can be given on the - command line or in an MySQL option file, although it is more - common to use an option file. See Section 4.2.3.3, "Using - Option Files." + MySQL Installation Wizard is an installer for the MySQL server + that uses the latest installer technologies for Microsoft Windows. + The MySQL Installation Wizard, in combination with the MySQL + Config Wizard, allows a user to install and configure a MySQL + server that is ready for use immediately after installation. - * If you are using Unix and you want the MySQL socket file - location to be somewhere other than the default location - (normally in the directory /tmp or /var/run), use a configure - command like this: -shell> ./configure \ - --with-unix-socket-path=/usr/local/mysql/tmp/mysql.sock - The socket file name must be an absolute path name. You can - also change the location of mysql.sock at server startup by - using a MySQL option file. See Section B.1.4.5, "How to - Protect or Change the MySQL Unix Socket File." + The MySQL Installation Wizard uses the standard Microsoft + Installer Engine (MSI) system is the standard installer for all + MySQL server distributions, version 4.1.5 and higher. Users of + previous versions of MySQL need to shut down and remove their + existing MySQL installations manually before installing MySQL with + the MySQL Installation Wizard. See Section 2.5.3.1.6, "Upgrading + MySQL with the Installation Wizard," for more information on + upgrading from a previous version. - * If you want to compile statically linked programs (for - example, to make a binary distribution, to get better - performance, or to work around problems with some Red Hat - Linux distributions), run configure like this: -shell> ./configure --with-client-ldflags=-all-static \ - --with-mysqld-ldflags=-all-static + If you are upgrading an installation from MySQL 5.1.31 or earlier + to MySQL 5.1.32 or later, read the notes provided in Section + 2.5.3.1.6, "Upgrading MySQL with the Installation Wizard." - * If you are using gcc and don't have libg++ or libstdc++ - installed, you can tell configure to use gcc as your C++ - compiler: -shell> CC=gcc CXX=gcc ./configure - When you use gcc as your C++ compiler, it does not attempt to - link in libg++ or libstdc++. This may be a good thing to do - even if you have those libraries installed. Some versions of - them have caused strange problems for MySQL users in the past. - The following list indicates some compilers and environment - variable settings that are commonly used with each one. + The Microsoft Windows Installer Engine was updated with the + release of Windows XP; those using a previous version of Windows + can reference this Microsoft Knowledge Base article + (http://support.microsoft.com/default.aspx?scid=kb;EN-US;292539) + for information on upgrading to the latest version of the Windows + Installer Engine. - + gcc 2.7.2: -CC=gcc CXX=gcc CXXFLAGS="-O3 -felide-constructors" + In addition, Microsoft has introduced the WiX (Windows Installer + XML) toolkit. This is the first highly acknowledged Open Source + project from Microsoft. We have switched to WiX because it is an + Open Source project and it allows us to handle the complete + Windows installation process in a flexible manner using scripts. - + gcc 2.95.2: -CFLAGS="-O3 -mpentiumpro" CXX=gcc CXXFLAGS="-O3 -mpentiumpro \ --felide-constructors -fno-exceptions -fno-rtti" + Improving the MySQL Installation Wizard depends on the support and + feedback of users like you. If you find that the MySQL + Installation Wizard is lacking some feature important to you, or + if you discover a bug, please report it in our bugs database using + the instructions given in Section 1.6, "How to Report Bugs or + Problems." - + pgcc 2.90.29 or newer: -CFLAGS="-O3 -mpentiumpro -mstack-align-double" CXX=gcc \ -CXXFLAGS="-O3 -mpentiumpro -mstack-align-double \ --felide-constructors -fno-exceptions -fno-rtti" - In most cases, you can get a reasonably optimized MySQL binary - by using the options from the preceding list and adding the - following options to the configure line: ---prefix=/usr/local/mysql --enable-assembler \ ---with-mysqld-ldflags=-all-static - The full configure line would, in other words, be something - like the following for all recent gcc versions: -CFLAGS="-O3 -mpentiumpro" CXX=gcc CXXFLAGS="-O3 -mpentiumpro \ --felide-constructors -fno-exceptions -fno-rtti" ./configure \ ---prefix=/usr/local/mysql --enable-assembler \ ---with-mysqld-ldflags=-all-static - The binaries we provide on the MySQL Web site at - http://dev.mysql.com/downloads/ are all compiled with full - optimization and should be perfect for most users. See Section - 2.1.2.4, "MySQL Binaries Compiled by Sun Microsystems, Inc.." - There are some configuration settings you can tweak to build - an even faster binary, but these are only for advanced users. - See Section 7.5.1, "How Compiling and Linking Affects the - Speed of MySQL." - If the build fails and produces errors about your compiler or - linker not being able to create the shared library - libmysqlclient.so.N (where N is a version number), you can - work around this problem by giving the --disable-shared option - to configure. In this case, configure does not build a shared - libmysqlclient.so.N library. +2.5.3.1.1. Downloading and Starting the MySQL Installation Wizard - * By default, MySQL uses the latin1 (cp1252 West European) - character set. To change the default set, use the - --with-charset option: -shell> ./configure --with-charset=CHARSET - CHARSET may be one of binary, armscii8, ascii, big5, cp1250, - cp1251, cp1256, cp1257, cp850, cp852, cp866, cp932, dec8, - eucjpms, euckr, gb2312, gbk, geostd8, greek, hebrew, hp8, - keybcs2, koi8r, koi8u, latin1, latin2, latin5, latin7, macce, - macroman, sjis, swe7, tis620, ucs2, ujis, utf8. See Section - 9.2, "The Character Set Used for Data and Sorting." - (Additional character sets might be available. Check the - output from ./configure --help for the current list.) - The default collation may also be specified. MySQL uses the - latin1_swedish_ci collation by default. To change this, use - the --with-collation option: -shell> ./configure --with-collation=COLLATION - To change both the character set and the collation, use both - the --with-charset and --with-collation options. The collation - must be a legal collation for the character set. (Use the SHOW - COLLATION statement to determine which collations are - available for each character set.) - With the configure option --with-extra-charsets=LIST, you can - define which additional character sets should be compiled into - the server. LIST is one of the following: + The MySQL installation packages can be downloaded from + http://dev.mysql.com/downloads/. If the package you download is + contained within a Zip archive, you need to extract the archive + first. - + A list of character set names separated by spaces + The process for starting the wizard depends on the contents of the + installation package you download. If there is a setup.exe file + present, double-click it to start the installation process. If + there is an .msi file present, double-click it to start the + installation process. - + complex to include all character sets that can't be - dynamically loaded +2.5.3.1.2. Choosing an Install Type - + all to include all character sets into the binaries - Clients that want to convert characters between the server and - the client should use the SET NAMES statement. See Section - 5.1.5, "Session System Variables," and Section 9.1.4, - "Connection Character Sets and Collations." + There are three installation types available: Typical, Complete, + and Custom. - * To configure MySQL with debugging code, use the --with-debug - option: -shell> ./configure --with-debug - This causes a safe memory allocator to be included that can - find some errors and that provides output about what is - happening. See MySQL Internals: Porting - (http://forge.mysql.com/wiki/MySQL_Internals_Porting) - As of MySQL 5.1.12, using --with-debug to configure MySQL with - debugging support enables you to use the - --debug="d,parser_debug" option when you start the server. - This causes the Bison parser that is used to process SQL - statements to dump a parser trace to the server's standard - error output. Typically, this output is written to the error - log. + The Typical installation type installs the MySQL server, the mysql + command-line client, and the command-line utilities. The + command-line clients and utilities include mysqldump, myisamchk, + and several other tools to help you manage the MySQL server. - * If your client programs are using threads, you must compile a - thread-safe version of the MySQL client library with the - --enable-thread-safe-client configure option. This creates a - libmysqlclient_r library with which you should link your - threaded applications. See Section 21.10.17, "How to Make a - Threaded Client." + The Complete installation type installs all components included in + the installation package. The full installation package includes + components such as the embedded server library, the benchmark + suite, support scripts, and documentation. - * Some features require that the server be built with - compression library support, such as the COMPRESS() and - UNCOMPRESS() functions, and compression of the client/server - protocol. The --with-zlib-dir=no|bundled|DIR option provides - control for compression library support. The value no - explicitly disables compression support. bundled causes the - zlib library bundled in the MySQL sources to be used. A DIR - path name specifies where to find the compression library - sources. + The Custom installation type gives you complete control over which + packages you wish to install and the installation path that is + used. See Section 2.5.3.1.3, "The Custom Install Dialog," for more + information on performing a custom install. - * It is possible to build MySQL with large table support using - the --with-big-tables option. - This option causes the variables that store table row counts - to be declared as unsigned long long rather than unsigned - long. This enables tables to hold up to approximately - 1.844E+19 ((2^32)^2) rows rather than 2^32 (~4.295E+09) rows. - Previously it was necessary to pass -DBIG_TABLES to the - compiler manually in order to enable this feature. + If you choose the Typical or Complete installation types and click + the Next button, you advance to the confirmation screen to verify + your choices and begin the installation. If you choose the Custom + installation type and click the Next button, you advance to the + custom installation dialog, described in Section 2.5.3.1.3, "The + Custom Install Dialog." - * Run configure with the --disable-grant-options option to cause - the --bootstrap, --skip-grant-tables, and --init-file options - for mysqld to be disabled. For Windows, the configure.js - script recognizes the DISABLE_GRANT_OPTIONS flag, which has - the same effect. The capability is available as of MySQL - 5.1.15. +2.5.3.1.3. The Custom Install Dialog - * This option allows MySQL Community Server features to be - enabled. Additional options may be required for individual - features, such as --enable-profiling to enable statement - profiling. This option was added in MySQL 5.1.24. It is - enabled by default as of MySQL 5.1.28; to disable it, use - --disable-community-features. + If you wish to change the installation path or the specific + components that are installed by the MySQL Installation Wizard, + choose the Custom installation type. - * When given with --enable-community-features, the - --enable-profiling option enables the statement profiling - capability exposed by the SHOW PROFILE and SHOW PROFILES - statements. (See Section 12.5.5.33, "SHOW PROFILES Syntax.") - This option was added in MySQL 5.1.24. It is enabled by - default as of MySQL 5.1.28; to disable it, use - --disable-profiling. + A tree view on the left side of the custom install dialog lists + all available components. Components that are not installed have a + red X icon; components that are installed have a gray icon. To + change whether a component is installed, click on that component's + icon and choose a new option from the drop-down list that appears. - * See Section 2.13, "Operating System-Specific Notes," for - options that pertain to particular operating systems. + You can change the default installation path by clicking the + Change... button to the right of the displayed installation path. - * See Section 5.5.7.2, "Using SSL Connections," for options that - pertain to configuring MySQL to support secure (encrypted) - connections. + After choosing your installation components and installation path, + click the Next button to advance to the confirmation dialog. - * Several configure options apply to plugin selection and - building: ---with-plugins=PLUGIN[,PLUGIN]... ---with-plugins=GROUP ---with-plugin-PLUGIN ---without-plugin-PLUGIN - PLUGIN is an individual plugin name such as csv or archive. - As shorthand, GROUP is a configuration group name such as none - (select no plugins) or all (select all plugins). - You can build a plugin as static (compiled into the server) or - dynamic (built as a dynamic library that must be installed - using the INSTALL PLUGIN statement before it can be used). - Some plugins might not support static or dynamic build. - configure --help shows the following information pertaining to - plugins: +2.5.3.1.4. The Confirmation Dialog - + The plugin-related options + Once you choose an installation type and optionally choose your + installation components, you advance to the confirmation dialog. + Your installation type and installation path are displayed for you + to review. - + The names of all available plugins + To install MySQL if you are satisfied with your settings, click + the Install button. To change your settings, click the Back + button. To exit the MySQL Installation Wizard without installing + MySQL, click the Cancel button. - + For each plugin, a description of its purpose, which - build types it supports (static or dynamic), and which - plugin groups it is a part of. - --with-plugins can take a list of one or more plugin names - separated by commas, or a plugin group name. The named plugins - are configured to be built as static plugins. - --with-plugin-PLUGIN configures the given plugin to be built - as a static plugin. - --without-plugin-PLUGIN disables the given plugin from being - built. - If a plugin is named both with a --with and --without option, - the result is undefined. - For any plugin that is not explicitly selected or disabled, it - is selected to be built dynamically if it supports dynamic - build, and not built if it does not support dynamic build. - (Thus, in the case that no plugin options are given, all - plugins that support dynamic build are selected to be built as - dynamic plugins. Plugins that do not support dynamic build are - not built.) + After installation is complete, you have the option of registering + with the MySQL web site. Registration gives you access to post in + the MySQL forums at forums.mysql.com (http://forums.mysql.com) + along with the ability to report bugs at bugs.mysql.com + (http://bugs.mysql.com) and to subscribe to our newsletter. The + final screen of the installer provides a summary of the + installation and gives you the option to launch the MySQL Config + Wizard, which you can use to create a configuration file, install + the MySQL service, and configure security settings. -2.10.3. Installing from the Development Source Tree +2.5.3.1.5. Changes Made by MySQL Installation Wizard -Caution + Once you click the Install button, the MySQL Installation Wizard + begins the installation process and makes certain changes to your + system which are described in the sections that follow. - You should read this section only if you are interested in helping - us test our new code. If you just want to get MySQL up and running - on your system, you should use a standard release distribution - (either a binary or source distribution). + Changes to the Registry - To obtain the most recent development source tree, you first need - to download and install Bazaar. You can obtain Bazaar from the - Bazaar VCS Website (http://bazaar-vcs.org) Bazaar is supported by - any platform that supports Python, and is therefore compatible - with any Linux, Unix, Windows or Mac OS X host. Instructions for - downloading and installing Bazaar on the different platforms are - available on the Bazaar website. + The MySQL Installation Wizard creates one Windows registry key in + a typical install situation, located in + HKEY_LOCAL_MACHINE\SOFTWARE\MySQL AB. - All MySQL projects are hosted on Launchpad - (http://launchpad.net/) MySQL projects, including MySQL server, - MySQL Workbench and others are available from the Sun/MySQL - Engineering (http://launchpad.net/~mysql) page. For the - repositories related only to MySQL server, see the MySQL Server - (http://launchpad.net/mysql-server) page. + The MySQL Installation Wizard creates a key named after the major + version of the server that is being installed, such as MySQL + Server 5.1. It contains two string values, Location and Version. + The Location string contains the path to the installation + directory. In a default installation it contains C:\Program + Files\MySQL\MySQL Server 5.1\. The Version string contains the + release number. For example, for an installation of MySQL Server + 5.1.41, the key contains a value of 5.1.41. - To build under Unix/Linux, you must have the following tools - installed: + These registry keys are used to help external tools identify the + installed location of the MySQL server, preventing a complete scan + of the hard-disk to determine the installation path of the MySQL + server. The registry keys are not required to run the server, and + if you install MySQL using the noinstall Zip archive, the registry + keys are not created. + + Changes to the Start Menu + + The MySQL Installation Wizard creates a new entry in the Windows + Start menu under a common MySQL menu heading named after the major + version of MySQL that you have installed. For example, if you + install MySQL 5.1, the MySQL Installation Wizard creates a MySQL + Server 5.1 section in the Start menu. - * GNU make, available from http://www.gnu.org/software/make/. - Although some platforms come with their own make - implementations, it is highly recommended that you use GNU - make. It may already be available on your system as gmake. + The following entries are created within the new Start menu + section: - * autoconf 2.58 (or newer), available from - http://www.gnu.org/software/autoconf/. + * MySQL Command Line Client: This is a shortcut to the mysql + command-line client and is configured to connect as the root + user. The shortcut prompts for a root user password when you + connect. - * automake 1.8.1, available from - http://www.gnu.org/software/automake/. + * MySQL Server Instance Config Wizard: This is a shortcut to the + MySQL Config Wizard. Use this shortcut to configure a newly + installed server, or to reconfigure an existing server. - * libtool 1.5, available from - http://www.gnu.org/software/libtool/. + * MySQL Documentation: This is a link to the MySQL server + documentation that is stored locally in the MySQL server + installation directory. This option is not available when the + MySQL server is installed using the Essentials installation + package. - * m4, available from http://www.gnu.org/software/m4/. + Changes to the File System - * bison, available from http://www.gnu.org/software/bison/. You - should use the latest version of bison where possible. Version - 1.75 and version 2.1 are known to work. There have been - reported problems with bison 1.875. If you experience - problems, upgrade to a later, rather than earlier, version. - Versions of bison older than 1.75 may report this error: -sql_yacc.yy:#####: fatal error: maximum table size (32767) exceeded - The maximum table size is not actually exceeded; the error is - caused by bugs in older versions of bison. + The MySQL Installation Wizard by default installs the MySQL 5.1 + server to C:\Program Files\MySQL\MySQL Server 5.1, where Program + Files is the default location for applications in your system, and + 5.1 is the major version of your MySQL server. This is the + recommended location for the MySQL server, replacing the former + default location C:\mysql. - To build under Windows you will need a copy of Microsoft Visual - C++ 2005 Express Edition, Visual Studio .Net 2003 (7.1), or Visual - Studio 2005 (8.0) compiler system. + By default, all MySQL applications are stored in a common + directory at C:\Program Files\MySQL, where Program Files is the + default location for applications in your Windows installation. A + typical MySQL installation on a developer machine might look like + this: +C:\Program Files\MySQL\MySQL Server 5.1 +C:\Program Files\MySQL\MySQL Workbench 5.1 OSS - Once you have the necessary tools installed, you first need to - create a local branch of the MySQL source code on your machine: + This approach makes it easier to manage and maintain all MySQL + applications installed on a particular system. - 1. To obtain a copy of the MySQL source code, you must create a - new Bazaar branch. If you do not already have a Bazaar - repository directory set up, you need to initialize a new - directory: -shell> mkdir mysql-server -shell> bzr init-repo --trees mysql-server + In MySQL 5.1.23 and earlier, the default location for the data + files used by MySQL is located within the corresponding MySQL + Server installation directory. For MySQL 5.1.24 and later, the + default location of the data directory is the AppData directory + configured for the user that installed the MySQL application. - 2. Once you have an initialized directory, you can branch from - the public MySQL server repositories. To create a branch of a - specific version: -shell> cd mysql-server -shell> bzr branch lp:mysql-server/5.1 mysql-5.1 +2.5.3.1.6. Upgrading MySQL with the Installation Wizard - 3. The initial download will take some time to complete, - depending on the speed of your connection. Please be patient. - Once you have downloaded the first tree, additional trees - should take significantly less time to download. + The MySQL Installation Wizard can perform server upgrades + automatically using the upgrade capabilities of MSI. That means + you do not need to remove a previous installation manually before + installing a new release. The installer automatically shuts down + and removes the previous MySQL service before installing the new + version. - 4. When building from the Bazaar branch, you may want to create a - copy of your active branch so that you can make configuration - and other changes without affecting the original branch - contents. You can achieve this by branching from the original - branch: -shell> bzr branch mysql-5.1 mysql-5.1-build + Automatic upgrades are available only when upgrading between + installations that have the same major and minor version numbers. + For example, you can upgrade automatically from MySQL 5.1.34 to + MySQL 5.1.37, but not from MySQL 5.0 to MySQL 5.1. - Once you have the local branch, you can start to build MySQL - server from the source code. On Windows, the build process is - different from Unix/Linux. To continue building MySQL on Windows, - see Section 2.10.6, "Installing MySQL from Source on Windows." - - On Unix/Linux you need to use the autoconf system to create the - configure script so that you can configure the build environment - before building. - - 1. The following example shows the typical commands required to - configure a source tree. The first cd command changes location - into the top-level directory of the tree; replace mysql-5.1 - with the appropriate directory name. + In MySQL 5.1.32 and later, the EXE version of the MSI installer + packages were removed. When upgrading an existing MySQL + installation from the old EXE based installer to the MSI based + installer, please keep the following notes in mind: + + * The MSI installer will not identify an existing installation + that was installed using the old EXE installer. This means + that the installer will not stop the existing server, or + detect that the existing password is required before + installing the new version. To work around this: + + 1. Stop the current server manually using net stop or + mysqladmin shutdown. + + 2. Remove the existing installation manually by using the + Add/Remove Programs control panel. This will keep the + existing configuration and data files, as these are not + removed automatically. + + 3. Install the new version of MySQL using the MSI installer. + When running the installation, skip updating the security + by deselecting the checkbox on the security screen. + + 4. Complete the installation, and then start the server + again. You should be able to login with your existing + user and password credentials. + + * You can only upgrade the version and release using the MSI + installer. For example, you can upgrade an open source + installation with an open source installer. You cannot upgrade + an open source installation using the enterprise installer. + + See Section 2.5.7, "Upgrading MySQL on Windows." + +2.5.3.2. Installing MySQL Automatically using MSI + + The Microsoft Installer (MSI) supports a both a quiet and a + passive mode that can be used to install MySQL automatically + without requireing intervention. You can use this either in + scripts to automatically install MySQL or through a terminal + connection such as Telnet where you do not have access to the + standard Windows user interface. The MSI packages can also be used + in combination with Microsoft's Group Policy system (part of + Windows Server 2003 and Windows Server 2008) to install MySQL + across multiple machines. + + To install MySQL from one of the MSI packages automatically from + the command line (or within a script), you need to use the + msiexec.exe tool. For example, to perform a quiet installation + (which shows no dialog boxes or progress): +shell> msiexec /i /quiet mysql-5.1.39.msi + + The /i indicates that you want to perform an installation. The + /quiet option indicates that you want no interactive elements. + + To provide a dialog box showing the progress during installation, + and the dialog boxes providing information on the installation and + registration of MySQL, use /passive mode instead of /quiet: +shell> msiexec /i /passive mysql-5.1.39.msi + + Regardless of the mode of the installation, installing the package + in this manner performs a 'Typical' installation, and installs the + default components into the standard location. + + You can also use this method to uninstall MySQL by using the + /uninstall or /x options: +shell> msiexec /x /quiet mysql-5.1.39.msi + + To install MySQL and configure a MySQL instance from the command + line, see Section 2.5.4.13, "Creating an Instance from the Command + Line." + + For information on using MSI packages to install software + automatically using Group Policy, see How to use Group Policy to + remotely install software in Windows Server 2003 + (http://support.microsoft.com/kb/816102) -Note - For MySQL 5.1.12 and earlier, you must separately configure - the INNODB storage engine. You can do this by running the - following command from the main source directory: -shell> (cd storage/innobase; autoreconf --force --install) -shell> cd mysql-5.1 -shell> autoreconf --force --install -shell> ./configure # Add your favorite options here -shell> make - Or you can use BUILD/autorun.sh as a shortcut for the - following sequence of commands: -shell> aclocal; autoheader -shell> libtoolize --automake --force -shell> automake --force --add-missing; autoconf - The command line that changes directory into the - storage/innobase directory is used to configure the InnoDB - storage engine. You can omit this lines if you do not require - InnoDB support. +2.5.3.3. Removing MySQL Installed from the MSI Package -Note - Beginning with MySQL 5.1, code specific to storage engines has - been moved under a storage directory. For example, InnoDB code - is now found in storage/innobase and NDBCLUSTER code is in - storage/ndb. - If you get some strange errors during this stage, verify that - you have the correct version of the libtool installed. - A collection of our standard configuration scripts is located - in the BUILD/ subdirectory. For example, you may find it more - convenient to use the BUILD/compile-pentium-debug script than - the preceding set of shell commands. To compile on a different - architecture, modify the script by removing flags that are - Pentium-specific, or use another script that may be more - appropriate. These scripts are provided on an "as-is" basis. - They are not officially maintained and their contents may - change from release to release. + To uninstall a MySQL where you have used the MSI packages, you + must use the Add/Remove Programs tool within Control Panel. To do + this: - 2. When the build is done, run make install. Be careful with this - on a production machine; the command may overwrite your live - release installation. If you have another installation of - MySQL, run ./configure with different values for the --prefix, - --with-tcp-port, and --with-unix-socket-path options than - those used for your production server. + 1. Right click on the start menu and choose Control Panel. - 3. Play hard with your new installation and try to make the new - features crash. Start by running make test. See Section - 22.1.2, "MySQL Test Suite." + 2. If the Control Panel is set to category mode (you will see + Pick a category at the top of the Control Panel window), + double click on Add or Remove Programs. If the Control is set + to classic mode, doubgle click on the Add or Remove Programs + icon. + + 3. Find MySQL in the list of installed software. MySQL Server is + installed against major version numbers (MySQL 5.0, MySQL 5.1, + etc.). Select the version that you want to remove and click + Remove. + + 4. You will be prompted to confirm the removal. Click Yes to + remove MySQL. + + When MySQL is removed using this method, only the installed + components are removed. Any database information (including the + tables and data), import or export files, log files, and binary + logs produced during execution are kept in their configured + location. - 4. If you have gotten to the make stage, but the distribution - does not compile, please enter the problem into our bugs - database using the instructions given in Section 1.6, "How to - Report Bugs or Problems." If you have installed the latest - versions of the required GNU tools, and they crash trying to - process our configuration files, please report that also. - However, if you execute aclocal and get a command not found - error or a similar problem, do not report it. Instead, make - sure that all the necessary tools are installed and that your - PATH variable is set correctly so that your shell can find - them. +2.5.4. MySQL Server Instance Config Wizard - 5. After initially copying the repository with bzr to obtain the - source tree, you should use pull option to periodically update - your local copy. To do this any time after you have set up the - repository, use this command: -shell> bzr pull + The MySQL Server Instance Config Wizard helps automate the process + of configuring your server. It creates a custom MySQL + configuration file (my.ini or my.cnf) by asking you a series of + questions and then applying your responses to a template to + generate the configuration file that is tuned to your + installation. - 6. You can examine the changeset comments for the tree by using - the log option to bzr: -shell> bzr log - You can also browse changesets, comments, and source code - online. To browse this information for MySQL 5.1, go to - http://launchpad.net/mysql-server/. - If you see diffs or code that you have a question about, do - not hesitate to send email to the MySQL internals mailing - list. See Section 1.5.1, "MySQL Mailing Lists." Also, if you - think you have a better idea on how to do something, send an - email message to the list with a patch. + The complete and essential MSI installation packages include the + MySQL Server Instance Config Wizard in the MySQL 5.1 server. The + MySQL Server Instance Config Wizard is only available for Windows. + + The workflow sequence for using the MySQL Server Instance Config + Wizard is shown in the figure below: + + Figure 2.4. MySQL Server Instance Config Wizard Workflow + MySQL Server Instance Config Wizard Workflow + +2.5.4.1. Starting the MySQL Server Instance Config Wizard + + The MySQL Server Instance Config Wizard is normally started as + part of the installation process. You should only need to run the + MySQL Server Instance Config Wizard again when you need to change + the configuration parameters of your server. -2.10.4. Dealing with Problems Compiling MySQL + If you chose not to open a port prior to installing MySQL on + Windows Vista, you can choose to use the MySQL Server Instance + Config Wizard after installation. However, you must open a port in + the Windows Firewall. To do this see the instructions given in + Section 2.5.3.1.1, "Downloading and Starting the MySQL + Installation Wizard." Rather than opening a port, you also have + the option of adding MySQL as a program that bypasses the Windows + Firewall. One or the other option is sufficient --- you need not + do both. Additionally, when running the MySQL Server Config Wizard + on Windows Vista ensure that you are logged in as a user with + administrative rights. + MySQL Server Instance Config Wizard - All MySQL programs compile cleanly for us with no warnings on - Solaris or Linux using gcc. On other systems, warnings may occur - due to differences in system include files. See Section 2.10.5, - "MIT-pthreads Notes," for warnings that may occur when using - MIT-pthreads. For other problems, check the following list. + You can launch the MySQL Config Wizard by clicking the MySQL + Server Instance Config Wizard entry in the MySQL section of the + Windows Start menu. - The solution to many problems involves reconfiguring. If you do - need to reconfigure, take note of the following: + Alternatively, you can navigate to the bin directory of your MySQL + installation and launch the MySQLInstanceConfig.exe file directly. - * If configure is run after it has previously been run, it may - use information that was gathered during its previous - invocation. This information is stored in config.cache. When - configure starts up, it looks for that file and reads its - contents if it exists, on the assumption that the information - is still correct. That assumption is invalid when you - reconfigure. + The MySQL Server Instance Config Wizard places the my.ini file in + the installation directory for the MySQL server. This helps + associate configuration files with particular server instances. - * Each time you run configure, you must run make again to - recompile. However, you may want to remove old object files - from previous builds first because they were compiled using - different configuration options. + To ensure that the MySQL server knows where to look for the my.ini + file, an argument similar to this is passed to the MySQL server as + part of the service installation: +--defaults-file="C:\Program Files\MySQL\MySQL Server 5.1\my.ini" - To prevent old configuration information or object files from - being used, run these commands before re-running configure: -shell> rm config.cache -shell> make clean + Here, C:\Program Files\MySQL\MySQL Server 5.1 is replaced with the + installation path to the MySQL Server. The --defaults-file option + instructs the MySQL server to read the specified file for + configuration options when it starts. - Alternatively, you can run make distclean. + Apart from making changes to the my.ini file by running the MySQL + Server Instance Config Wizard again, you can modify it by opening + it with a text editor and making any necessary changes. You can + also modify the server configuration with the MySQL Administrator + (http://www.mysql.com/products/administrator/) utility. For more + information about server configuration, see Section 5.1.2, "Server + Command Options." - The following list describes some of the problems when compiling - MySQL that have been found to occur most often: + MySQL clients and utilities such as the mysql and mysqldump + command-line clients are not able to locate the my.ini file + located in the server installation directory. To configure the + client and utility applications, create a new my.ini file in the + Windows installation directory (for example, C:\WINDOWS). - * If you get errors such as the ones shown here when compiling - sql_yacc.cc, you probably have run out of memory or swap - space: -Internal compiler error: program cc1plus got fatal signal 11 -Out of virtual memory -Virtual memory exhausted - The problem is that gcc requires a huge amount of memory to - compile sql_yacc.cc with inline functions. Try running - configure with the --with-low-memory option: -shell> ./configure --with-low-memory - This option causes -fno-inline to be added to the compile line - if you are using gcc and -O0 if you are using something else. - You should try the --with-low-memory option even if you have - so much memory and swap space that you think you can't - possibly have run out. This problem has been observed to occur - even on systems with generous hardware configurations, and the - --with-low-memory option usually fixes it. + Under Windows Server 2003, Windows Server 2000, Windows XP, and + Windows Vista MySQL Server Instance Config Wizard will configure + MySQL to work as a Windows service. To start and stop MySQL you + use the Services application that is supplied as part of the + Windows Administrator Tools. + +2.5.4.2. Choosing a Maintenance Option + + If the MySQL Server Instance Config Wizard detects an existing + configuration file, you have the option of either reconfiguring + your existing server, or removing the server instance by deleting + the configuration file and stopping and removing the MySQL + service. - * By default, configure picks c++ as the compiler name and GNU - c++ links with -lg++. If you are using gcc, that behavior can - cause problems during configuration such as this: -configure: error: installation or configuration problem: -C++ compiler cannot create executables. - You might also observe problems during compilation related to - g++, libg++, or libstdc++. - One cause of these problems is that you may not have g++, or - you may have g++ but not libg++, or libstdc++. Take a look at - the config.log file. It should contain the exact reason why - your C++ compiler didn't work. To work around these problems, - you can use gcc as your C++ compiler. Try setting the - environment variable CXX to "gcc -O3". For example: -shell> CXX="gcc -O3" ./configure - This works because gcc compiles C++ source files as well as - g++ does, but does not link in libg++ or libstdc++ by default. - Another way to fix these problems is to install g++, libg++, - and libstdc++. However, do not use libg++ or libstdc++ with - MySQL because this only increases the binary size of mysqld - without providing any benefits. Some versions of these - libraries have also caused strange problems for MySQL users in - the past. + To reconfigure an existing server, choose the Re-configure + Instance option and click the Next button. Any existing + configuration file is not overwritten, but renamed (within the + same directory) using a timestamp (Windows) or sequential number + (Linux). To remove the existing server instance, choose the Remove + Instance option and click the Next button. - * If your compile fails with errors such as any of the - following, you must upgrade your version of make to GNU make: -making all in mit-pthreads -make: Fatal error in reader: Makefile, line 18: -Badly formed macro assignment - Or: -make: file `Makefile' line 18: Must be a separator (: - Or: -pthread.h: No such file or directory - Solaris and FreeBSD are known to have troublesome make - programs. - GNU make 3.75 is known to work. + If you choose the Remove Instance option, you advance to a + confirmation window. Click the Execute button. The MySQL Server + Config Wizard stops and removes the MySQL service, and then + deletes the configuration file. The server installation and its + data folder are not removed. - * If you want to define flags to be used by your C or C++ - compilers, do so by adding the flags to the CFLAGS and - CXXFLAGS environment variables. You can also specify the - compiler names this way using CC and CXX. For example: -shell> CC=gcc -shell> CFLAGS=-O3 -shell> CXX=gcc -shell> CXXFLAGS=-O3 -shell> export CC CFLAGS CXX CXXFLAGS - See Section 2.1.2.4, "MySQL Binaries Compiled by Sun - Microsystems, Inc.," for a list of flag definitions that have - been found to be useful on various systems. + If you choose the Re-configure Instance option, you advance to the + Configuration Type dialog where you can choose the type of + installation that you wish to configure. - * If you get errors such as those shown here when compiling - mysqld, configure did not correctly detect the type of the - last argument to accept(), getsockname(), or getpeername(): -cxx: Error: mysqld.cc, line 645: In this statement, the referenced - type of the pointer value ''length'' is ''unsigned long'', - which is not compatible with ''int''. -new_sock = accept(sock, (struct sockaddr *)&cAddr, &length); - To fix this, edit the config.h file (which is generated by - configure). Look for these lines: -/* Define as the base type of the last arg to accept */ -#define SOCKET_SIZE_TYPE XXX - Change XXX to size_t or int, depending on your operating - system. (You must do this each time you run configure because - configure regenerates config.h.) +2.5.4.3. Choosing a Configuration Type - * The sql_yacc.cc file is generated from sql_yacc.yy. Normally, - the build process does not need to create sql_yacc.cc because - MySQL comes with a pre-generated copy. However, if you do need - to re-create it, you might encounter this error: -"sql_yacc.yy", line xxx fatal: default action causes potential... - This is a sign that your version of yacc is deficient. You - probably need to install bison (the GNU version of yacc) and - use that instead. + When you start the MySQL Server Instance Config Wizard for a new + MySQL installation, or choose the Re-configure Instance option for + an existing installation, you advance to the Configuration Type + dialog. + MySQL Server Instance Config Wizard: Configuration Type - * On Debian Linux 3.0, you need to install gawk instead of the - default mawk. + There are two configuration types available: Detailed + Configuration and Standard Configuration. The Standard + Configuration option is intended for new users who want to get + started with MySQL quickly without having to make many decisions + about server configuration. The Detailed Configuration option is + intended for advanced users who want more fine-grained control + over server configuration. - * If you need to debug mysqld or a MySQL client, run configure - with the --with-debug option, and then recompile and link your - clients with the new client library. See MySQL Internals: - Porting (http://forge.mysql.com/wiki/MySQL_Internals_Porting) + If you are new to MySQL and need a server configured as a + single-user developer machine, the Standard Configuration should + suit your needs. Choosing the Standard Configuration option causes + the MySQL Config Wizard to set all configuration options + automatically with the exception of Service Options and Security + Options. - * If you get a compilation error on Linux (for example, SuSE - Linux 8.1 or Red Hat Linux 7.3) similar to the following one, - you probably do not have g++ installed: -libmysql.c:1329: warning: passing arg 5 of `gethostbyname_r' from -incompatible pointer type -libmysql.c:1329: too few arguments to function `gethostbyname_r' -libmysql.c:1329: warning: assignment makes pointer from integer -without a cast -make[2]: *** [libmysql.lo] Error 1 - By default, the configure script attempts to determine the - correct number of arguments by using g++ (the GNU C++ - compiler). This test yields incorrect results if g++ is not - installed. There are two ways to work around this problem: + The Standard Configuration sets options that may be incompatible + with systems where there are existing MySQL installations. If you + have an existing MySQL installation on your system in addition to + the installation you wish to configure, the Detailed Configuration + option is recommended. - + Make sure that the GNU C++ g++ is installed. On some - Linux distributions, the required package is called gpp; - on others, it is named gcc-c++. + To complete the Standard Configuration, please refer to the + sections on Service Options and Security Options in Section + 2.5.4.10, "The Service Options Dialog," and Section 2.5.4.11, "The + Security Options Dialog," respectively. - + Use gcc as your C++ compiler by setting the CXX - environment variable to gcc: -export CXX="gcc" - You must run configure again after making either of those - changes. +2.5.4.4. The Server Type Dialog -2.10.5. MIT-pthreads Notes + There are three different server types available to choose from. + The server type that you choose affects the decisions that the + MySQL Server Instance Config Wizard makes with regard to memory, + disk, and processor usage. + MySQL Server Instance Config Wizard: Server Type - This section describes some of the issues involved in using - MIT-pthreads. + * Developer Machine: Choose this option for a typical desktop + workstation where MySQL is intended only for personal use. It + is assumed that many other desktop applications are running. + The MySQL server is configured to use minimal system + resources. - On Linux, you should not use MIT-pthreads. Use the installed - LinuxThreads implementation instead. See Section 2.13.1, "Linux - Notes." + * Server Machine: Choose this option for a server machine where + the MySQL server is running alongside other server + applications such as FTP, email, and Web servers. The MySQL + server is configured to use a moderate portion of the system + resources. - If your system does not provide native thread support, you should - build MySQL using the MIT-pthreads package. This includes older - FreeBSD systems, SunOS 4.x, Solaris 2.4 and earlier, and some - others. See Section 2.1.1, "Operating Systems Supported by MySQL - Community Server." + * Dedicated MySQL Server Machine: Choose this option for a + server machine that is intended to run only the MySQL server. + It is assumed that no other applications are running. The + MySQL server is configured to use all available system + resources. - MIT-pthreads is not part of the MySQL 5.1 source distribution. If - you require this package, you need to download it separately from - http://dev.mysql.com/Downloads/Contrib/pthreads-1_60_beta6-mysql.t - ar.gz +Note - After downloading, extract this source archive into the top level - of the MySQL source directory. It creates a new subdirectory named - mit-pthreads. + By selecting one of the preconfigured configurations, the values + and settings of various options in your my.cnf or my.ini will be + altered accordingly. The default values and options as described + in the reference manual may therefore be different to the options + and values that were created during the execution of the Config + Wizard. - * On most systems, you can force MIT-pthreads to be used by - running configure with the --with-mit-threads option: -shell> ./configure --with-mit-threads - Building in a nonsource directory is not supported when using - MIT-pthreads because we want to minimize our changes to this - code. +2.5.4.5. The Database Usage Dialog - * The checks that determine whether to use MIT-pthreads occur - only during the part of the configuration process that deals - with the server code. If you have configured the distribution - using --without-server to build only the client code, clients - do not know whether MIT-pthreads is being used and use Unix - socket file connections by default. Because Unix socket files - do not work under MIT-pthreads on some platforms, this means - you need to use -h or --host with a value other than localhost - when you run client programs. + The Database Usage dialog allows you to indicate the storage + engines that you expect to use when creating MySQL tables. The + option you choose determines whether the InnoDB storage engine is + available and what percentage of the server resources are + available to InnoDB. + MySQL Server Instance Config Wizard: Usage Dialog - * When MySQL is compiled using MIT-pthreads, system locking is - disabled by default for performance reasons. You can tell the - server to use system locking with the --external-locking - option. This is needed only if you want to be able to run two - MySQL servers against the same data files, but that is not - recommended, anyway. + * Multifunctional Database: This option enables both the InnoDB + and MyISAM storage engines and divides resources evenly + between the two. This option is recommended for users who use + both storage engines on a regular basis. - * Sometimes the pthread bind() command fails to bind to a socket - without any error message (at least on Solaris). The result is - that all connections to the server fail. For example: -shell> mysqladmin version -mysqladmin: connect to server at '' failed; -error: 'Can't connect to mysql server on localhost (146)' - The solution to this problem is to kill the mysqld server and - restart it. This has happened to us only when we have forcibly - stopped the server and restarted it immediately. + * Transactional Database Only: This option enables both the + InnoDB and MyISAM storage engines, but dedicates most server + resources to the InnoDB storage engine. This option is + recommended for users who use InnoDB almost exclusively and + make only minimal use of MyISAM. - * With MIT-pthreads, the sleep() system call isn't interruptible - with SIGINT (break). This is noticeable only when you run - mysqladmin --sleep. You must wait for the sleep() call to - terminate before the interrupt is served and the process - stops. + * Non-Transactional Database Only: This option disables the + InnoDB storage engine completely and dedicates all server + resources to the MyISAM storage engine. This option is + recommended for users who do not use InnoDB. - * When linking, you might receive warning messages like these - (at least on Solaris); they can be ignored: -ld: warning: symbol `_iob' has differing sizes: - (file /my/local/pthreads/lib/libpthread.a(findfp.o) value=0x4; -file /usr/lib/libc.so value=0x140); - /my/local/pthreads/lib/libpthread.a(findfp.o) definition taken -ld: warning: symbol `__iob' has differing sizes: - (file /my/local/pthreads/lib/libpthread.a(findfp.o) value=0x4; -file /usr/lib/libc.so value=0x140); - /my/local/pthreads/lib/libpthread.a(findfp.o) definition taken + The Config Wizard uses a template to generate the server + configuration file. The Database Usage dialog sets one of the + following option strings: +Multifunctional Database: MIXED +Transactional Database Only: INNODB +Non-Transactional Database Only: MYISAM - * Some other warnings also can be ignored: -implicit declaration of function `int strtoll(...)' -implicit declaration of function `int strtoul(...)' + When these options are processed through the default template + (my-template.ini) the result is: +Multifunctional Database: +default-storage-engine=InnoDB +_myisam_pct=50 - * We have not been able to make readline work with MIT-pthreads. - (This is not necessary, but may be of interest to some.) +Transactional Database Only: +default-storage-engine=InnoDB +_myisam_pct=5 -2.10.6. Installing MySQL from Source on Windows +Non-Transactional Database Only: +default-storage-engine=MyISAM +_myisam_pct=100 +skip-innodb - These instructions describe how to build binaries from source for - MySQL 5.1 on Windows. Instructions are provided for building - binaries from a standard source distribution or from the Bazaar - tree that contains the latest development source. + The _myisam_pct value is used to calculate the percentage of + resources dedicated to MyISAM. The remaining resources are + allocated to InnoDB. -Note +2.5.4.6. The InnoDB Tablespace Dialog - The instructions here are strictly for users who want to test - MySQL on Microsoft Windows from the latest source distribution or - from the Bazaar tree. For production use, we do not advise using a - MySQL server built by yourself from source. Normally, it is best - to use precompiled binary distributions of MySQL that are built - specifically for optimal performance on Windows by Sun - Microsystems, Inc. Instructions for installing binary - distributions are available in Section 2.3, "Installing MySQL on - Windows." + Some users may want to locate the InnoDB tablespace files in a + different location than the MySQL server data directory. Placing + the tablespace files in a separate location can be desirable if + your system has a higher capacity or higher performance storage + device available, such as a RAID storage system. + MySQL Server Instance Config Wizard: InnoDB Data Tablespace - To build MySQL on Windows from source, you must satisfy the - following system, compiler, and resource requirements: + To change the default location for the InnoDB tablespace files, + choose a new drive from the drop-down list of drive letters and + choose a new path from the drop-down list of paths. To create a + custom path, click the ... button. - * Windows 2000, Windows XP, or newer version. - Windows Vista is supported when using Visual Studio 2005 - provided you have installed the following updates: + If you are modifying the configuration of an existing server, you + must click the Modify button before you change the path. In this + situation you must move the existing tablespace files to the new + location manually before starting the server. - + Microsoft Visual Studio 2005 Professional Edition - ENU - Service Pack 1 (KB926601) - (http://support.microsoft.com/?kbid=926601) +2.5.4.7. The Concurrent Connections Dialog - + Security Update for Microsoft Visual Studio 2005 - Professional Edition - ENU (KB937061) - (http://support.microsoft.com/?kbid=937061) + To prevent the server from running out of resources, it is + important to limit the number of concurrent connections to the + MySQL server that can be established. The Concurrent Connections + dialog allows you to choose the expected usage of your server, and + sets the limit for concurrent connections accordingly. It is also + possible to set the concurrent connection limit manually. + MySQL Server Instance Config Wizard: Connections - + Update for Microsoft Visual Studio 2005 Professional - Edition - ENU (KB932232) - (http://support.microsoft.com/?kbid=932232) + * Decision Support (DSS)/OLAP: Choose this option if your server + does not require a large number of concurrent connections. The + maximum number of connections is set at 100, with an average + of 20 concurrent connections assumed. - * CMake, which can be downloaded from http://www.cmake.org. - After installing, modify your path to include the cmake - binary. + * Online Transaction Processing (OLTP): Choose this option if + your server requires a large number of concurrent connections. + The maximum number of connections is set at 500. - * Microsoft Visual C++ 2005 Express Edition, Visual Studio .Net - 2003 (7.1), or Visual Studio 2005 (8.0) compiler system. + * Manual Setting: Choose this option to set the maximum number + of concurrent connections to the server manually. Choose the + number of concurrent connections from the drop-down box + provided, or enter the maximum number of connections into the + drop-down box if the number you desire is not listed. - * If you are using Visual C++ 2005 Express Edition, you must - also install an appropriate Platform SDK. More information and - links to downloads for various Windows platforms is available - from - http://www.microsoft.com/downloads/details.aspx?familyid=0baf2 - b35-c656-4969-ace8-e4c0c0716adb. +2.5.4.8. The Networking and Strict Mode Options Dialog - * If you are compiling from a Bazaar tree or making changes to - the parser, you need bison for Windows, which can be - downloaded from - http://gnuwin32.sourceforge.net/packages/bison.htm. Download - the package labeled "Complete package, excluding sources". - After installing the package, modify your path to include the - bison binary and ensure that this binary is accessible from - Visual Studio. + Use the Networking Options dialog to enable or disable TCP/IP + networking and to configure the port number that is used to + connect to the MySQL server. + MySQL Server Instance Config Wizard: Network Configuration - * Cygwin might be necessary if you want to run the test script - or package the compiled binaries and support files into a Zip - archive. (Cygwin is needed only to test or package the - distribution, not to build it.) Cygwin is available from - http://cygwin.com. + TCP/IP networking is enabled by default. To disable TCP/IP + networking, uncheck the box next to the Enable TCP/IP Networking + option. - * 3GB to 5GB of disk space. + Port 3306 is used by default. To change the port used to access + MySQL, choose a new port number from the drop-down box or type a + new port number directly into the drop-down box. If the port + number you choose is in use, you are prompted to confirm your + choice of port number. - The exact system requirements can be found here: - http://msdn.microsoft.com/vstudio/Previous/2003/sysreqs/default.as - px and - http://msdn.microsoft.com/vstudio/products/sysreqs/default.aspx + Set the Server SQL Mode to either enable or disable strict mode. + Enabling strict mode (default) makes MySQL behave more like other + database management systems. If you run applications that rely on + MySQL's old "forgiving" behavior, make sure to either adapt those + applications or to disable strict mode. For more information about + strict mode, see Section 5.1.8, "Server SQL Modes." - You also need a MySQL source distribution for Windows, which can - be obtained two ways: +2.5.4.9. The Character Set Dialog - * Obtain a source distribution packaged by Sun Microsystems, - Inc. These are available from http://dev.mysql.com/downloads/. + The MySQL server supports multiple character sets and it is + possible to set a default server character set that is applied to + all tables, columns, and databases unless overridden. Use the + Character Set dialog to change the default character set of the + MySQL server. + MySQL Server Instance Config Wizard: Character Set - * Package a source distribution yourself from the latest Bazaar - developer source tree. For instructions on pulling the latest - source files, see Section 2.10.3, "Installing from the - Development Source Tree." + * Standard Character Set: Choose this option if you want to use + latin1 as the default server character set. latin1 is used for + English and many Western European languages. - If you find something not working as expected, or you have - suggestions about ways to improve the current build process on - Windows, please send a message to the win32 mailing list. See - Section 1.5.1, "MySQL Mailing Lists." + * Best Support For Multilingualism: Choose this option if you + want to use utf8 as the default server character set. This is + a Unicode character set that can store characters from many + different languages. -2.10.6.1. Building MySQL from Source Using CMake and Visual Studio + * Manual Selected Default Character Set / Collation: Choose this + option if you want to pick the server's default character set + manually. Choose the desired character set from the provided + drop-down list. - You can build MySQL on Windows by using a combination of cmake and - Microsoft Visual Studio .NET 2003 (7.1), Microsoft Visual Studio - 2005 (8.0) or Microsoft Visual C++ 2005 Express Edition. You must - have the appropriate Microsoft Platform SDK installed. +2.5.4.10. The Service Options Dialog -Note + On Windows platforms, the MySQL server can be installed as a + Windows service. When installed this way, the MySQL server can be + started automatically during system startup, and even restarted + automatically by Windows in the event of a service failure. - To compile from the source code on Windows you must use the - standard source distribution (for example, mysql-5.0.45.tar.gz). - You build from the same distribution as used to build MySQL on - Unix, Linux and other platforms. Do not use the Windows Source - distributions as they do not contain the necessary configuration - script and other files. + The MySQL Server Instance Config Wizard installs the MySQL server + as a service by default, using the service name MySQL. If you do + not wish to install the service, uncheck the box next to the + Install As Windows Service option. You can change the service name + by picking a new service name from the drop-down box provided or + by entering a new service name into the drop-down box. - Follow this procedure to build MySQL: +Note - 1. If you are installing from a packaged source distribution, - create a work directory (for example, C:\workdir), and unpack - the source distribution there using WinZip or another Windows - tool that can read .zip files. This directory is the work - directory in the following instructions. + Service names can include any legal character except forward (/) + or backward (\) slashes, and must be less than 256 characters + long. - 2. Using a command shell, navigate to the work directory and run - the following command: -C:\workdir>win\configure.js options - If you have associated the .js file extension with an - application such as a text editor, then you may need to use - the following command to force configure.js to be executed as - a script: -C:\workdir>cscript win\configure.js options - These options are available for configure.js: +Warning - + WITH_INNOBASE_STORAGE_ENGINE: Enable the InnoDB storage - engine. + If you are installing multiple versions of MySQL onto the same + machine, you must choose a different service name for each version + that you install. If you do not choose a different service for + each installed version then the service manager information will + be inconsistent and this will cause problems when you try to + uninstall a previous version. - + WITH_PARTITION_STORAGE_ENGINE: Enable user-defined - partitioning. + If you have already installed multiple versions using the same + service name, you must manually edit the contents of the + HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services parameters + within the Windows registry to update the association of the + service name with the correct server version. - + WITH_ARCHIVE_STORAGE_ENGINE: Enable the ARCHIVE storage - engine. + Typically, when installing multiple versions you create a service + name based on the version information. For example, you might + install MySQL 5.x as mysql5, or specific versions such as MySQL + 5.1.30 as mysql50130. - + WITH_BLACKHOLE_STORAGE_ENGINE: Enable the BLACKHOLE - storage engine. + To install the MySQL server as a service but not have it started + automatically at startup, uncheck the box next to the Launch the + MySQL Server Automatically option. - + WITH_EXAMPLE_STORAGE_ENGINE: Enable the EXAMPLE storage - engine. +2.5.4.11. The Security Options Dialog - + WITH_FEDERATED_STORAGE_ENGINE: Enable the FEDERATED - storage engine. + It is strongly recommended that you set a root password for your + MySQL server, and the MySQL Server Instance Config Wizard requires + by default that you do so. If you do not wish to set a root + password, uncheck the box next to the Modify Security Settings + option. + MySQL Server Instance Config Wizard: Security - + WITH_NDBCLUSTER_STORAGE_ENGINE (experimental): Enable the - NDBCLUSTER storage engine in the MySQL server; cause - binaries for the MySQL Cluster management and data node, - management client, and other programs to be built. - This option is supported only in MySQL Cluster NDB 7.0 - (NDBCLUSTER storage engine versions 6.4.0 and later) - using the MySQL Cluster sources. It cannot be used to - enable clustering support in other MySQL source trees or - distributions. + To set the root password, enter the desired password into both the + New root password and Confirm boxes. If you are reconfiguring an + existing server, you need to enter the existing root password into + the Current root password box. - + MYSQL_SERVER_SUFFIX=suffix: Server suffix, default none. + To allow root logins from across the network, check the box next + to the Enable root access from remote machines option. This + decreases the security of your root account. - + COMPILATION_COMMENT=comment: Server comment, default - "Source distribution". + To create an anonymous user account, check the box next to the + Create An Anonymous Account option. Creating an anonymous account + can decrease server security and cause login and permission + difficulties. For this reason, it is not recommended. - + MYSQL_TCP_PORT=port: Server port, default 3306. +2.5.4.12. The Confirmation Dialog - + DISABLE_GRANT_OPTIONS: Disables the --bootstrap, - --skip-grant-tables, and --init-file options for mysqld. - This option is available as of MySQL 5.1.15. - For example (type the command on one line): -C:\workdir>win\configure.js WITH_INNOBASE_STORAGE_ENGINE - WITH_PARTITION_STORAGE_ENGINE MYSQL_SERVER_SUFFIX=-pro + The final dialog in the MySQL Server Instance Config Wizard is the + Confirmation Dialog. To start the configuration process, click the + Execute button. To return to a previous dialog, click the Back + button. To exit the MySQL Server Instance Config Wizard without + configuring the server, click the Cancel button. + MySQL Server Instance Config Wizard: Confirmation - 3. From the work directory, execute the win\build-vs8.bat or - win\build-vs71.bat file, depending on the version of Visual - Studio you have installed. The script invokes CMake, which - generates the mysql.sln solution file. - You can also use win\build-vs8_x64.bat to build the 64-bit - version of MySQL. However, you cannot build the 64-bit version - with Visual Studio Express Edition. You must use Visual Studio - 2005 (8.0) or higher. + After you click the Execute button, the MySQL Server Instance + Config Wizard performs a series of tasks and displays the progress + onscreen as the tasks are performed. - 4. From the work directory, open the generated mysql.sln file - with Visual Studio and select the proper configuration using - the Configuration menu. The menu provides Debug, Release, - RelwithDebInfo, MinRelInfo options. Then select Solution > - Build to build the solution. - Remember the configuration that you use in this step. It is - important later when you run the test script because that - script needs to know which configuration you used. + The MySQL Server Instance Config Wizard first determines + configuration file options based on your choices using a template + prepared by MySQL developers and engineers. This template is named + my-template.ini and is located in your server installation + directory. - 5. Test the server. The server built using the preceding - instructions expects that the MySQL base directory and data - directory are C:\mysql and C:\mysql\data by default. If you - want to test your server using the source tree root directory - and its data directory as the base directory and data - directory, you need to tell the server their path names. You - can either do this on the command line with the --basedir and - --datadir options, or by placing appropriate options in an - option file. (See Section 4.2.3.3, "Using Option Files.") If - you have an existing data directory elsewhere that you want to - use, you can specify its path name instead. - When the server is running in standalone fashion or as a - service based on your configuration, try to connect to it from - the mysql interactive command-line utility. - You can also run the standard test script, mysql-test-run.pl. - This script is written in Perl, so you'll need either Cygwin - or ActiveState Perl to run it. You may also need to install - the modules required by the script. To run the test script, - change location into the mysql-test directory under the work - directory, set the MTR_VS_CONFIG environment variable to the - configuration you selected earlier (or use the --vs-config - option), and invoke mysql-test-run.pl. For example (using - Cygwin and the bash shell): -shell> cd mysql-test -shell> export MTR_VS_CONFIG=debug -shell> ./mysql-test-run.pl --force --timer -shell> ./mysql-test-run.pl --force --timer --ps-protocol + The MySQL Config Wizard then writes these options to the + corresponding configuration file. - When you are satisfied that the programs you have built are - working correctly, stop the server. Now you can install the - distribution. One way to do this is to use the make_win_bin_dist - script in the scripts directory of the MySQL source distribution - (see Section 4.4.2, "make_win_bin_dist --- Package MySQL - Distribution as ZIP Archive"). This is a shell script, so you must - have Cygwin installed if you want to use it. It creates a Zip - archive of the built executables and support files that you can - unpack in the location at which you want to install MySQL. + If you chose to create a service for the MySQL server, the MySQL + Server Instance Config Wizard creates and starts the service. If + you are reconfiguring an existing service, the MySQL Server + Instance Config Wizard restarts the service to apply your + configuration changes. + + If you chose to set a root password, the MySQL Config Wizard + connects to the server, sets your new root password, and applies + any other security settings you may have selected. + + After the MySQL Server Instance Config Wizard has completed its + tasks, it displays a summary. Click the Finish button to exit the + MySQL Server Config Wizard. + +2.5.4.13. Creating an Instance from the Command Line + + In addition to using the GUI interface to the MySQL Server + Instance Config Wizard, you can also create instances + automatically from the command line. + + To use the MySQL Server Instance Config Wizard on the command + line, you need to use the MySQLInstanceConfig.exe command that is + installed with MySQL in the bin directory within the installation + directory. MySQLInstanceConfig.exe takes a number of command-line + arguments the set the properties that would normally be selected + through the GUI interface, and then creates a new configuration + file (my.ini) by combining these selections with a template + configuration file to produce the working configuration file. + + The main command line options are provided in the table below. + Some of the options are required, while some options are optional. + + Table 2.4. MySQL Server Instance Config Wizard Command Line + Options + Option Description + Required Parameters + -nPRODUCTNAME The name of the instance when installed + -pPATH Path of the base directory for installation. This is + equivalent to the directory when using the basedir configuration + parameter + -vVERSION The version tag to use for this installation + Action to Perform + -i Install an instance + -r Remove an instance + -s Stop an existing instance + -q Perform the operation quietly + -lFILENAME Sae the installation progress in a logfile + Config File to Use + -tFILENAME Path to the template config file that will be used to + generate the installed configuration file + -cFILENAME Path to a config file to be generated + + The -t and -c options work together to set the configuration + parameters for a new instance. The -t option specifies the + template configuration file to use as the basic configuration, + which are then merged with the configuration parameters generated + by the MySQL Server Instance Config Wizard into the configuration + file specified by the -c option. + + A sample template file, my-template.ini is provided in the + toplevel MySQL installation directory. The file contains elements + are replaced automatically by the MySQL Server Instance Config + Wizard during configuration. + + If you specify a configuration file that already exists, the + existing configuration file will be saved in the file with the + original, with the date and time added. For example, the mysql.ini + will be copied to mysql 2009-10-27 1646.ini.bak. + + The parameters that you can specify on the command line are listed + in the table below. + + Table 2.5. MySQL Server Instance Config Wizard Parameters + Parameter Description + ServiceName=$ Specify the name of the service to be created + AddBinToPath={yes | no} Specifies whether to add the binary + directory of MySQL to the standard PATH environment variable + ServerType={DEVELOPMENT | SERVER | DEDICATED} Specify the server + type. For more information, see Section 2.5.4.4, "The Server Type + Dialog" + DatabaseType={MIXED | INNODB | MYISAM} Specify the default + database type. For more information, see Section 2.5.4.5, "The + Database Usage Dialog" + ConnectionUsage={DSS | OLTP} Specify the type of connection + support, this automates the setting for the number of concurrent + connections (see the ConnectionCount parameter). For more + information, see Section 2.5.4.7, "The Concurrent Connections + Dialog" + ConnectionCount=# Specify the number of concurrent connections to + support. For more information, see Section 2.5.4.4, "The Server + Type Dialog" + SkipNetworking={yes | no} Specify whether network support should + be supported. Specifying yes disables network access altogether + Port=# Specify the network port number to use for network + connections. For more information, see Section 2.5.4.8, "The + Networking and Strict Mode Options Dialog" + StrictMode={yes | no} Specify whether to use the strict SQL mode. + For more information, see Section 2.5.4.8, "The Networking and + Strict Mode Options Dialog" + Charset=$ Specify the default character set. For more information, + see Section 2.5.4.9, "The Character Set Dialog" + RootPassword=$ Specify the root password + RootCurrentPassword=$ Specify the current root password then + stopping and/or reconfiguring an existing service - It is also possible to install MySQL by copying directories and - files directly: +Note - 1. Create the directories where you want to install MySQL. For - example, to install into C:\mysql, use these commands: -C:\> mkdir C:\mysql -C:\> mkdir C:\mysql\bin -C:\> mkdir C:\mysql\data -C:\> mkdir C:\mysql\share -C:\> mkdir C:\mysql\scripts - If you want to compile other clients and link them to MySQL, - you should also create several additional directories: -C:\> mkdir C:\mysql\include -C:\> mkdir C:\mysql\lib -C:\> mkdir C:\mysql\lib\debug -C:\> mkdir C:\mysql\lib\opt - If you want to benchmark MySQL, create this directory: -C:\> mkdir C:\mysql\sql-bench - Benchmarking requires Perl support. See Section 2.15, "Perl - Installation Notes." + When specifying options on the command line, you can enclose the + entire command-line option and the value you are specifying using + double quotes. This enables you to use spaces in the options. For + example, "-cC:\mysql.ini". + + The following command installs a MySQL Server 5.1 instance from + the directory C:\Program Files\MySQL\MySQL Server 5.1 using the + service name MySQL51 and setting the root password to 1234. +shell> MySQLInstanceConfig.exe -i -q "-lC:\mysql_install_log.txt" » + "-nMySQL Server 5.1" "-pC:\Program Files\MySQL\MySQL Server 5.1" - +v5.1.39 » + "-tmy-template.ini" "-cC:\mytest.ini" ServerType=DEVELOPMENT Datab +aseType=MIXED » + ConnectionUsage=DSS Port=3311 ServiceName=MySQL51 RootPassword=123 +4 + + In the above example, a log file will be generated in + mysql_install_log.txt containing the information about the + instance creation process. The log file generated by the above + example is shown below: +Welcome to the MySQL Server Instance Configuration Wizard 1.0.16.0 +Date: 2009-10-27 17:07:21 + +Installing service ... + +Product Name: MySQL Server 5.1 +Version: 5.1.39 +Installation Path: C:\Program Files\MySQL\MySQL Server 5.1\ + +Creating configuration file C:\mytest.ini using template my-template. +ini. +Options: +DEVELOPMENT +MIXED +DSS +STRICTMODE + +Variables: +port: 3311 +default-character-set: latin1 +basedir: "C:/Program Files/MySQL/MySQL Server 5.1/" +datadir: "C:/Program Files/MySQL/MySQL Server 5.1/Data/" + + +Creating Windows service entry. +Service name: "MySQL51" +Parameters: "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" -- +defaults-file="C:\mytest.ini" MySQL51. +Windows service MySQL51 installed. + + When using the command-line, the return values in the following + table indicate an error performing the specified option. + + Table 2.6. Return Value from MySQL Server Instance Config Wizard + Value Description + 2 Configuration template file cannot be found + 3 The Windows service entry cannot be created + 4 Could not connect to the Service Control Manager + 5 The MySQL service cannot be started + 6 The MySQL service cannot be stopped + 7 The security settings cannot be applied + 8 The configuration file cannot be written + 9 The Windows service entry cannot be removed + + You can perform an installation of MySQL automatically using the + MSI packe. For more information, see Section 2.5.3.2, "Installing + MySQL Automatically using MSI." - 2. From the work directory, copy into the C:\mysql directory the - following directories: -C:\> cd \workdir -C:\workdir> copy client_release\*.exe C:\mysql\bin -C:\workdir> copy client_debug\mysqld.exe C:\mysql\bin\mysqld-debug.ex -e -C:\workdir> xcopy scripts\*.* C:\mysql\scripts /E -C:\workdir> xcopy share\*.* C:\mysql\share /E - If you want to compile other clients and link them to MySQL, - you should also copy several libraries and header files: -C:\workdir> copy lib_debug\mysqlclient.lib C:\mysql\lib\debug -C:\workdir> copy lib_debug\libmysql.* C:\mysql\lib\debug -C:\workdir> copy lib_debug\zlib.* C:\mysql\lib\debug -C:\workdir> copy lib_release\mysqlclient.lib C:\mysql\lib\opt -C:\workdir> copy lib_release\libmysql.* C:\mysql\lib\opt -C:\workdir> copy lib_release\zlib.* C:\mysql\lib\opt -C:\workdir> copy include\*.h C:\mysql\include -C:\workdir> copy libmysql\libmysql.def C:\mysql\include - If you want to benchmark MySQL, you should also do this: -C:\workdir> xcopy sql-bench\*.* C:\mysql\bench /E +2.5.5. Installing MySQL from a noinstall Zip Archive - After installation, set up and start the server in the same way as - for binary Windows distributions. See Section 2.3, "Installing - MySQL on Windows." + Users who are installing from the noinstall package can use the + instructions in this section to manually install MySQL. The + process for installing MySQL from a Zip archive is as follows: -2.10.7. Compiling MySQL Clients on Windows + 1. Extract the archive to the desired install directory - In your source files, you should include my_global.h before - mysql.h: -#include <my_global.h> -#include <mysql.h> + 2. Create an option file + + 3. Choose a MySQL server type - my_global.h includes any other files needed for Windows - compatibility (such as windows.h) if you compile your program on - Windows. + 4. Start the MySQL server - You can either link your code with the dynamic libmysql.lib - library, which is just a wrapper to load in libmysql.dll on - demand, or link with the static mysqlclient.lib library. + 5. Secure the default user accounts - The MySQL client libraries are compiled as threaded libraries, so - you should also compile your code to be multi-threaded. + This process is described in the sections that follow. -2.11. Post-Installation Setup and Testing +2.5.5.1. Extracting the Install Archive - After installing MySQL, there are some issues that you should - address. For example, on Unix, you should initialize the data - directory and create the MySQL grant tables. On all platforms, an - important security concern is that the initial accounts in the - grant tables have no passwords. You should assign passwords to - prevent unauthorized access to the MySQL server. Optionally, you - can create time zone tables to enable recognition of named time - zones. + To install MySQL manually, do the following: - The following sections include post-installation procedures that - are specific to Windows systems and to Unix systems. Another - section, Section 2.11.2.3, "Starting and Troubleshooting the MySQL - Server," applies to all platforms; it describes what to do if you - have trouble getting the server to start. Section 2.11.3, - "Securing the Initial MySQL Accounts," also applies to all - platforms. You should follow its instructions to make sure that - you have properly protected your MySQL accounts by assigning - passwords to them. + 1. If you are upgrading from a previous version please refer to + Section 2.5.7, "Upgrading MySQL on Windows," before beginning + the upgrade process. - When you are ready to create additional user accounts, you can - find information on the MySQL access control system and account - management in Section 5.4, "The MySQL Access Privilege System," - and Section 5.5, "MySQL User Account Management." + 2. Make sure that you are logged in as a user with administrator + privileges. -2.11.1. Windows Post-Installation Procedures + 3. Choose an installation location. Traditionally, the MySQL + server is installed in C:\mysql. The MySQL Installation Wizard + installs MySQL under C:\Program Files\MySQL. If you do not + install MySQL at C:\mysql, you must specify the path to the + install directory during startup or in an option file. See + Section 2.5.5.2, "Creating an Option File." - On Windows, the data directory and the grant tables do not have to - be created. MySQL Windows distributions include the grant tables - with a set of preinitialized accounts in the mysql database under - the data directory. It is unnecessary to run the mysql_install_db - script that is used on Unix. Regarding passwords, if you installed - MySQL using the Windows Installation Wizard, you may have already - assigned passwords to the accounts. (See Section 2.3.3, "Using the - MySQL Installation Wizard.") Otherwise, use the - password-assignment procedure given in Section 2.11.3, "Securing - the Initial MySQL Accounts." + 4. Extract the install archive to the chosen installation + location using your preferred Zip archive tool. Some tools may + extract the archive to a folder within your chosen + installation location. If this occurs, you can move the + contents of the subfolder into the chosen installation + location. - Before setting up passwords, you might want to try running some - client programs to make sure that you can connect to the server - and that it is operating properly. Make sure that the server is - running (see Section 2.3.9, "Starting the Server for the First - Time"), and then issue the following commands to verify that you - can retrieve information from the server. The output should be - similar to what is shown here: -C:\> C:\mysql\bin\mysqlshow -+--------------------+ -| Databases | -+--------------------+ -| information_schema | -| mysql | -| test | -+--------------------+ +2.5.5.2. Creating an Option File -C:\> C:\mysql\bin\mysqlshow mysql -Database: mysql -+---------------------------+ -| Tables | -+---------------------------+ -| columns_priv | -| db | -| event | -| func | -| general_log | -| help_category | -| help_keyword | -| help_relation | -| help_topic | -| host | -| plugin | -| proc | -| procs_priv | -| servers | -| slow_log | -| tables_priv | -| time_zone | -| time_zone_leap_second | -| time_zone_name | -| time_zone_transition | -| time_zone_transition_type | -| user | -+---------------------------+ + If you need to specify startup options when you run the server, + you can indicate them on the command line or place them in an + option file. For options that are used every time the server + starts, you may find it most convenient to use an option file to + specify your MySQL configuration. This is particularly true under + the following circumstances: + * The installation or data directory locations are different + from the default locations (C:\Program Files\MySQL\MySQL + Server 5.1 and C:\Program Files\MySQL\MySQL Server 5.1\data). -C:\> C:\mysql\bin\mysql -e "SELECT Host,Db,User FROM db" mysql -+------+-------+------+ -| host | db | user | -+------+-------+------+ -| % | test% | | -+------+-------+------+ + * You need to tune the server settings, such as memory, cache, + or InnoDB configuration information. - You may need to specify a different directory from the one shown; - if you used the Windows Installation Wizard, then the default - directory is C:\Program Files\MySQL\MySQL Server 5.1, and the - mysql and mysqlshow client programs are in C:\Program - Files\MySQL\MySQL Server 5.1\bin. See Section 2.3.3, "Using the - MySQL Installation Wizard," for more information. + When the MySQL server starts on Windows, it looks for option files + in several locations, such as the Windows directory, C:\, and the + MySQL installation directory (for the full list of locations, see + Section 4.2.3.3, "Using Option Files"). The Windows directory + typically is named something like C:\WINDOWS. You can determine + its exact location from the value of the WINDIR environment + variable using the following command: +C:\> echo %WINDIR% - If you have already secured the initial MySQL accounts, you may - need to use the -u and -p options to supply a user name and - password to the mysqlshow and mysql client programs; otherwise the - programs may fail with an error, or you may not be able to view - all databases. For example, if you have assigned the password - "secretpass" to the MySQL root account, then you can invoke - mysqlshow and mysql as shown here: -C:\> C:\mysql\bin\mysqlshow -uroot -psecretpass -+--------------------+ -| Databases | -+--------------------+ -| information_schema | -| mysql | -| test | -+--------------------+ + MySQL looks for options in each location first in the my.ini file, + and then in the my.cnf file. However, to avoid confusion, it is + best if you use only one file. If your PC uses a boot loader where + C: is not the boot drive, your only option is to use the my.ini + file. Whichever option file you use, it must be a plain text file. -C:\> C:\mysql\bin\mysqlshow -uroot -psecretpass mysql -Database: mysql -+---------------------------+ -| Tables | -+---------------------------+ -| columns_priv | -| db | -| event | -| func | -| general_log | -| help_category | -| help_keyword | -| help_relation | -| help_topic | -| host | -| plugin | -| proc | -| procs_priv | -| servers | -| slow_log | -| tables_priv | -| time_zone | -| time_zone_leap_second | -| time_zone_name | -| time_zone_transition | -| time_zone_transition_type | -| user | -+---------------------------+ + You can also make use of the example option files included with + your MySQL distribution; see Section 4.2.3.3.2, "Preconfigured + Option Files." + An option file can be created and modified with any text editor, + such as Notepad. For example, if MySQL is installed in E:\mysql + and the data directory is in E:\mydata\data, you can create an + option file containing a [mysqld] section to specify values for + the basedir and datadir options: +[mysqld] +# set basedir to your installation path +basedir=E:/mysql +# set datadir to the location of your data directory +datadir=E:/mydata/data -C:\> C:\mysql\bin\mysql -uroot -psecretpass -e "SELECT Host,Db,User F -ROM db" mysql -+------+-------+------+ -| host | db | user | -+------+-------+------+ -| % | test% | | -+------+-------+------+ + Note that Windows path names are specified in option files using + (forward) slashes rather than backslashes. If you do use + backslashes, double them: +[mysqld] +# set basedir to your installation path +basedir=E:\\mysql +# set datadir to the location of your data directory +datadir=E:\\mydata\\data - For more information about these programs, see Section 4.5.6, - "mysqlshow --- Display Database, Table, and Column Information," - and Section 4.5.1, "mysql --- The MySQL Command-Line Tool." + The rules for use of backslash in option file values are given in + Section 4.2.3.3, "Using Option Files." - If you are running a version of Windows that supports services and - you want the MySQL server to run automatically when Windows - starts, see Section 2.3.11, "Starting MySQL as a Windows Service." + MySQL Enterprise For expert advice on the start-up options + appropriate to your circumstances, subscribe to the MySQL + Enterprise Monitor. For more information, see + http://www.mysql.com/products/enterprise/advisors.html. -2.11.2. Unix Post-Installation Procedures + In MySQL 5.1.23 and earlier, the MySQL installer places the data + directory directly under the directory where you install MySQL. On + MySQL 5.1.24 and later, the data directory is located within the + AppData directory for the user running MySQL. - After installing MySQL on Unix, you need to initialize the grant - tables, start the server, and make sure that the server works - satisfactorily. You may also wish to arrange for the server to be - started and stopped automatically when your system starts and - stops. You should also assign passwords to the accounts in the - grant tables. + If you would like to use a data directory in a different location, + you should copy the entire contents of the data directory to the + new location. For example, if you want to use E:\mydata as the + data directory instead, you must do two things: - On Unix, the grant tables are set up by the mysql_install_db - program. For some installation methods, this program is run for - you automatically: + 1. Move the entire data directory and all of its contents from + the default location (for example C:\Program Files\MySQL\MySQL + Server 5.1\data) to E:\mydata. - * If you install MySQL on Linux using RPM distributions, the - server RPM runs mysql_install_db. + 2. Use a --datadir option to specify the new data directory + location each time you start the server. - * If you install MySQL on Mac OS X using a PKG distribution, the - installer runs mysql_install_db. +2.5.5.3. Selecting a MySQL Server Type - Otherwise, you'll need to run mysql_install_db yourself. + The following table shows the available servers for Windows in + MySQL 5.1.20 and earlier. + Binary Description + mysqld-nt Optimized binary with named-pipe support + mysqld Optimized binary without named-pipe support + mysqld-debug Like mysqld-nt, but compiled with full debugging and + automatic memory allocation checking - The following procedure describes how to initialize the grant - tables (if that has not previously been done) and then start the - server. It also suggests some commands that you can use to test - whether the server is accessible and working properly. For - information about starting and stopping the server automatically, - see Section 2.11.2.2, "Starting and Stopping MySQL Automatically." + The following table shows the available servers for Windows in + MySQL 5.1.21 and later. + Binary Description + mysqld Optimized binary with named-pipe support + mysqld-debug Like mysqld, but compiled with full debugging and + automatic memory allocation checking - After you complete the procedure and have the server running, you - should assign passwords to the accounts created by - mysql_install_db. Instructions for doing so are given in Section - 2.11.3, "Securing the Initial MySQL Accounts." + All of the preceding binaries are optimized for modern Intel + processors, but should work on any Intel i386-class or higher + processor. - In the examples shown here, the server runs under the user ID of - the mysql login account. This assumes that such an account exists. - Either create the account if it does not exist, or substitute the - name of a different existing login account that you plan to use - for running the server. + Each of the servers in a distribution support the same set of + storage engines. The SHOW ENGINES statement displays which engines + a given server supports. - 1. Change location into the top-level directory of your MySQL - installation, represented here by BASEDIR: -shell> cd BASEDIR - BASEDIR is likely to be something like /usr/local/mysql or - /usr/local. The following steps assume that you are located in - this directory. + All Windows MySQL 5.1 servers have support for symbolic linking of + database directories. - 2. If necessary, run the mysql_install_db program to set up the - initial MySQL grant tables containing the privileges that - determine how users are allowed to connect to the server. - You'll need to do this if you used a distribution type for - which the installation procedure doesn't run the program for - you. - Typically, mysql_install_db needs to be run only the first - time you install MySQL, so you can skip this step if you are - upgrading an existing installation, However, mysql_install_db - does not overwrite any existing privilege tables, so it should - be safe to run in any circumstances. - To initialize the grant tables, use one of the following - commands, depending on whether mysql_install_db is located in - the bin or scripts directory: -shell> bin/mysql_install_db --user=mysql -shell> scripts/mysql_install_db --user=mysql - It might be necessary to specify other options such as - --basedir or --datadir if mysql_install_db does not use the - correct locations for the installation directory or data - directory. For example: -shell> bin/mysql_install_db --user=mysql \ - --basedir=/opt/mysql/mysql \ - --datadir=/opt/mysql/mysql/data - The mysql_install_db script creates the server's data - directory. Under the data directory, it creates directories - for the mysql database that holds all database privileges and - the test database that you can use to test MySQL. The script - also creates privilege table entries for root and - anonymous-user accounts. The accounts have no passwords - initially. A description of their initial privileges is given - in Section 2.11.3, "Securing the Initial MySQL Accounts." - Briefly, these privileges allow the MySQL root user to do - anything, and allow anybody to create or use databases with a - name of test or starting with test_. - It is important to make sure that the database directories and - files are owned by the mysql login account so that the server - has read and write access to them when you run it later. To - ensure this, the --user option should be used as shown if you - run mysql_install_db as root. Otherwise, you should execute - the script while logged in as mysql, in which case you can - omit the --user option from the command. - mysql_install_db creates several tables in the mysql database, - including user, db, host, tables_priv, columns_priv, func, and - others. See Section 5.4, "The MySQL Access Privilege System," - for a complete listing and description of these tables. - If you don't want to have the test database, you can remove it - with mysqladmin -u root drop test after starting the server. - If you have trouble with mysql_install_db at this point, see - Section 2.11.2.1, "Problems Running mysql_install_db." + MySQL supports TCP/IP on all Windows platforms. MySQL servers on + Windows support named pipes as indicated in the following list. + However, the default is to use TCP/IP regardless of platform. + (Named pipes are slower than TCP/IP in many Windows + configurations.) - 3. Start the MySQL server: -shell> bin/mysqld_safe --user=mysql & - It is important that the MySQL server be run using an - unprivileged (non-root) login account. To ensure this, the - --user option should be used as shown if you run mysqld_safe - as system root. Otherwise, you should execute the script while - logged in to the system as mysql, in which case you can omit - the --user option from the command. - Further instructions for running MySQL as an unprivileged user - are given in Section 5.3.5, "How to Run MySQL as a Normal - User." - If you neglected to create the grant tables before proceeding - to this step, the following message appears in the error log - file when you start the server: -mysqld: Can't find file: 'host.frm' - If you have other problems starting the server, see Section - 2.11.2.3, "Starting and Troubleshooting the MySQL Server." + Use of named pipes is subject to these conditions: - 4. Use mysqladmin to verify that the server is running. The - following commands provide simple tests to check whether the - server is up and responding to connections: -shell> bin/mysqladmin version -shell> bin/mysqladmin variables - The output from mysqladmin version varies slightly depending - on your platform and version of MySQL, but should be similar - to that shown here: -shell> bin/mysqladmin version -mysqladmin Ver 14.12 Distrib 5.1.39, for pc-linux-gnu on i686 -... + * Named pipes are enabled only if you start the server with the + --enable-named-pipe option. It is necessary to use this option + explicitly because some users have experienced problems with + shutting down the MySQL server when named pipes were used. -Server version 5.1.39 -Protocol version 10 -Connection Localhost via UNIX socket -UNIX socket /var/lib/mysql/mysql.sock -Uptime: 14 days 5 hours 5 min 21 sec + * For MySQL 5.1.20 and earlier, named-pipe connections are + allowed only by the mysqld-nt and mysqld-debug servers. For + MySQL 5.1.21 and later, the mysqld and mysqld-debug servers + both contain support for named-pipe connections. -Threads: 1 Questions: 366 Slow queries: 0 -Opens: 0 Flush tables: 1 Open tables: 19 -Queries per second avg: 0.000 - To see what else you can do with mysqladmin, invoke it with - the --help option. +Note - 5. Verify that you can shut down the server: -shell> bin/mysqladmin -u root shutdown + Most of the examples in this manual use mysqld as the server name. + If you choose to use a different server, such as mysqld-nt or + mysqld-debug, make the appropriate substitutions in the commands + that are shown in the examples. - 6. Verify that you can start the server again. Do this by using - mysqld_safe or by invoking mysqld directly. For example: -shell> bin/mysqld_safe --user=mysql --log & - If mysqld_safe fails, see Section 2.11.2.3, "Starting and - Troubleshooting the MySQL Server." +2.5.5.4. Starting the Server for the First Time - 7. Run some simple tests to verify that you can retrieve - information from the server. The output should be similar to - what is shown here: -shell> bin/mysqlshow -+-----------+ -| Databases | -+-----------+ -| mysql | -| test | -+-----------+ + This section gives a general overview of starting the MySQL + server. The following sections provide more specific information + for starting the MySQL server from the command line or as a + Windows service. -shell> bin/mysqlshow mysql -Database: mysql -+---------------------------+ -| Tables | -+---------------------------+ -| columns_priv | -| db | -| func | -| help_category | -| help_keyword | -| help_relation | -| help_topic | -| host | -| proc | -| procs_priv | -| tables_priv | -| time_zone | -| time_zone_leap_second | -| time_zone_name | -| time_zone_transition | -| time_zone_transition_type | -| user | -+---------------------------+ + The information here applies primarily if you installed MySQL + using the Noinstall version, or if you wish to configure and test + MySQL manually rather than with the GUI tools. -shell> bin/mysql -e "SELECT Host,Db,User FROM db" mysql -+------+--------+------+ -| host | db | user | -+------+--------+------+ -| % | test | | -| % | test_% | | -+------+--------+------+ + The examples in these sections assume that MySQL is installed + under the default location of C:\Program Files\MySQL\MySQL Server + 5.1. Adjust the path names shown in the examples if you have MySQL + installed in a different location. - 8. There is a benchmark suite in the sql-bench directory (under - the MySQL installation directory) that you can use to compare - how MySQL performs on different platforms. The benchmark suite - is written in Perl. It requires the Perl DBI module that - provides a database-independent interface to the various - databases, and some other additional Perl modules: -DBI -DBD::mysql -Data::Dumper -Data::ShowTable - These modules can be obtained from CPAN - (http://www.cpan.org/) See also Section 2.15.1, "Installing - Perl on Unix." - The sql-bench/Results directory contains the results from many - runs against different databases and platforms. To run all - tests, execute these commands: -shell> cd sql-bench -shell> perl run-all-tests - If you don't have the sql-bench directory, you probably - installed MySQL using RPM files other than the source RPM. - (The source RPM includes the sql-bench benchmark directory.) - In this case, you must first install the benchmark suite - before you can use it. There are separate benchmark RPM files - named mysql-bench-VERSION.i386.rpm that contain benchmark code - and data. - If you have a source distribution, there are also tests in its - tests subdirectory that you can run. For example, to run - auto_increment.tst, execute this command from the top-level - directory of your source distribution: -shell> mysql -vvf test < ./tests/auto_increment.tst - The expected result of the test can be found in the - ./tests/auto_increment.res file. + Clients have two options. They can use TCP/IP, or they can use a + named pipe if the server supports named-pipe connections. - 9. At this point, you should have the server running. However, - none of the initial MySQL accounts have a password, so you - should assign passwords using the instructions found in - Section 2.11.3, "Securing the Initial MySQL Accounts." + MySQL for Windows also supports shared-memory connections if the + server is started with the --shared-memory option. Clients can + connect through shared memory by using the --protocol=MEMORY + option. - The MySQL 5.1 installation procedure creates time zone tables in - the mysql database. However, you must populate the tables manually - using the instructions in Section 9.7, "MySQL Server Time Zone - Support." + For information about which server binary to run, see Section + 2.5.5.3, "Selecting a MySQL Server Type." -2.11.2.1. Problems Running mysql_install_db + Testing is best done from a command prompt in a console window (or + "DOS window"). In this way you can have the server display status + messages in the window where they are easy to see. If something is + wrong with your configuration, these messages make it easier for + you to identify and fix any problems. - The purpose of the mysql_install_db script is to generate new - MySQL privilege tables. It does not overwrite existing MySQL - privilege tables, and it does not affect any other data. + To start the server, enter this command: +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" --console - If you want to re-create your privilege tables, first stop the - mysqld server if it is running. Then rename the mysql directory - under the data directory to save it, and then run - mysql_install_db. Suppose that your current directory is the MySQL - installation directory and that mysql_install_db is located in the - bin directory and the data directory is named data. To rename the - mysql database and re-run mysql_install_db, use these commands. -shell> mv data/mysql data/mysql.old -shell> bin/mysql_install_db --user=mysql + For a server that includes InnoDB support, you should see the + messages similar to those following as it starts (the path names + and sizes may differ): +InnoDB: The first specified datafile c:\ibdata\ibdata1 did not exist: +InnoDB: a new database to be created! +InnoDB: Setting file c:\ibdata\ibdata1 size to 209715200 +InnoDB: Database physically writes the file full: wait... +InnoDB: Log file c:\iblogs\ib_logfile0 did not exist: new to be creat +ed +InnoDB: Setting log file c:\iblogs\ib_logfile0 size to 31457280 +InnoDB: Log file c:\iblogs\ib_logfile1 did not exist: new to be creat +ed +InnoDB: Setting log file c:\iblogs\ib_logfile1 size to 31457280 +InnoDB: Log file c:\iblogs\ib_logfile2 did not exist: new to be creat +ed +InnoDB: Setting log file c:\iblogs\ib_logfile2 size to 31457280 +InnoDB: Doublewrite buffer not found: creating new +InnoDB: Doublewrite buffer created +InnoDB: creating foreign key constraint system tables +InnoDB: foreign key constraint system tables created +011024 10:58:25 InnoDB: Started - When you run mysql_install_db, you might encounter the following - problems: + When the server finishes its startup sequence, you should see + something like this, which indicates that the server is ready to + service client connections: +mysqld: ready for connections +Version: '5.1.41' socket: '' port: 3306 - * mysql_install_db fails to install the grant tables - You may find that mysql_install_db fails to install the grant - tables and terminates after displaying the following messages: -Starting mysqld daemon with databases from XXXXXX -mysqld ended - In this case, you should examine the error log file very - carefully. The log should be located in the directory XXXXXX - named by the error message and should indicate why mysqld - didn't start. If you do not understand what happened, include - the log when you post a bug report. See Section 1.6, "How to - Report Bugs or Problems." + The server continues to write to the console any further + diagnostic output it produces. You can open a new console window + in which to run client programs. - * There is a mysqld process running - This indicates that the server is running, in which case the - grant tables have probably been created already. If so, there - is no need to run mysql_install_db at all because it needs to - be run only once (when you install MySQL the first time). + If you omit the --console option, the server writes diagnostic + output to the error log in the data directory (C:\Program + Files\MySQL\MySQL Server 5.1\data by default). The error log is + the file with the .err extension. - * Installing a second mysqld server does not work when one - server is running - This can happen when you have an existing MySQL installation, - but want to put a new installation in a different location. - For example, you might have a production installation, but you - want to create a second installation for testing purposes. - Generally the problem that occurs when you try to run a second - server is that it tries to use a network interface that is in - use by the first server. In this case, you should see one of - the following error messages: -Can't start server: Bind on TCP/IP port: -Address already in use -Can't start server: Bind on unix socket... - For instructions on setting up multiple servers, see Section - 5.6, "Running Multiple MySQL Servers on the Same Machine." +Note - * You do not have write access to the /tmp directory - If you do not have write access to create temporary files or a - Unix socket file in the default location (the /tmp directory), - an error occurs when you run mysql_install_db or the mysqld - server. - You can specify different locations for the temporary - directory and Unix socket file by executing these commands - prior to starting mysql_install_db or mysqld, where - some_tmp_dir is the full path name to some directory for which - you have write permission: -shell> TMPDIR=/some_tmp_dir/ -shell> MYSQL_UNIX_PORT=/some_tmp_dir/mysql.sock -shell> export TMPDIR MYSQL_UNIX_PORT - Then you should be able to run mysql_install_db and start the - server with these commands: -shell> bin/mysql_install_db --user=mysql -shell> bin/mysqld_safe --user=mysql & - If mysql_install_db is located in the scripts directory, - modify the first command to scripts/mysql_install_db. - See Section B.1.4.5, "How to Protect or Change the MySQL Unix - Socket File," and Section 2.14, "Environment Variables." + The accounts that are listed in the MySQL grant tables initially + have no passwords. After starting the server, you should set up + passwords for them using the instructions in Section 2.13, + "Post-Installation Setup and Testing." - There are some alternatives to running the mysql_install_db script - provided in the MySQL distribution: +2.5.5.5. Starting MySQL from the Windows Command Line - * If you want the initial privileges to be different from the - standard defaults, you can modify mysql_install_db before you - run it. However, it is preferable to use GRANT and REVOKE to - change the privileges after the grant tables have been set up. - In other words, you can run mysql_install_db, and then use - mysql -u root mysql to connect to the server as the MySQL root - user so that you can issue the necessary GRANT and REVOKE - statements. - If you want to install MySQL on several machines with the same - privileges, you can put the GRANT and REVOKE statements in a - file and execute the file as a script using mysql after - running mysql_install_db. For example: -shell> bin/mysql_install_db --user=mysql -shell> bin/mysql -u root < your_script_file - By doing this, you can avoid having to issue the statements - manually on each machine. + The MySQL server can be started manually from the command line. + This can be done on any version of Windows. - * It is possible to re-create the grant tables completely after - they have previously been created. You might want to do this - if you're just learning how to use GRANT and REVOKE and have - made so many modifications after running mysql_install_db that - you want to wipe out the tables and start over. - To re-create the grant tables, remove all the .frm, .MYI, and - .MYD files in the mysql database directory. Then run the - mysql_install_db script again. + To start the mysqld server from the command line, you should start + a console window (or "DOS window") and enter this command: +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" - * You can start mysqld manually using the --skip-grant-tables - option and add the privilege information yourself using mysql: -shell> bin/mysqld_safe --user=mysql --skip-grant-tables & -shell> bin/mysql mysql - From mysql, manually execute the SQL commands contained in - mysql_install_db. Make sure that you run mysqladmin - flush-privileges or mysqladmin reload afterward to tell the - server to reload the grant tables. - Note that by not using mysql_install_db, you not only have to - populate the grant tables manually, you also have to create - them first. + The path to mysqld may vary depending on the install location of + MySQL on your system. -2.11.2.2. Starting and Stopping MySQL Automatically + You can stop the MySQL server by executing this command: +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqladmin" -u root + shutdown - Generally, you start the mysqld server in one of these ways: +Note - * Invoke mysqld directly. This works on any platform. + If the MySQL root user account has a password, you need to invoke + mysqladmin with the -p option and supply the password when + prompted. - * Run the MySQL server as a Windows service. The service can be - set to start the server automatically when Windows starts, or - as a manual service that you start on request. For - instructions, see Section 2.3.11, "Starting MySQL as a Windows - Service." + This command invokes the MySQL administrative utility mysqladmin + to connect to the server and tell it to shut down. The command + connects as the MySQL root user, which is the default + administrative account in the MySQL grant system. Note that users + in the MySQL grant system are wholly independent from any login + users under Windows. - * Invoke mysqld_safe, which tries to determine the proper - options for mysqld and then runs it with those options. This - script is used on Unix and Unix-like systems. See Section - 4.3.2, "mysqld_safe --- MySQL Server Startup Script." + If mysqld doesn't start, check the error log to see whether the + server wrote any messages there to indicate the cause of the + problem. The error log is located in the C:\Program + Files\MySQL\MySQL Server 5.1\data directory. It is the file with a + suffix of .err. You can also try to start the server as mysqld + --console; in this case, you may get some useful information on + the screen that may help solve the problem. - * Invoke mysql.server. This script is used primarily at system - startup and shutdown on systems that use System V-style run - directories, where it usually is installed under the name - mysql. The mysql.server script starts the server by invoking - mysqld_safe. See Section 4.3.3, "mysql.server --- MySQL Server - Startup Script." + The last option is to start mysqld with the --standalone and + --debug options. In this case, mysqld writes a log file + C:\mysqld.trace that should contain the reason why mysqld doesn't + start. See MySQL Internals: Porting + (http://forge.mysql.com/wiki/MySQL_Internals_Porting) - * On Mac OS X, install a separate MySQL Startup Item package to - enable the automatic startup of MySQL on system startup. The - Startup Item starts the server by invoking mysql.server. See - Section 2.5, "Installing MySQL on Mac OS X," for details. + Use mysqld --verbose --help to display all the options that mysqld + supports. - The mysqld_safe and mysql.server scripts and the Mac OS X Startup - Item can be used to start the server manually, or automatically at - system startup time. mysql.server and the Startup Item also can be - used to stop the server. +2.5.5.6. Starting MySQL as a Windows Service - To start or stop the server manually using the mysql.server - script, invoke it with start or stop arguments: -shell> mysql.server start -shell> mysql.server stop + On Windows, the recommended way to run MySQL is to install it as a + Windows service, whereby MySQL starts and stops automatically when + Windows starts and stops. A MySQL server installed as a service + can also be controlled from the command line using NET commands, + or with the graphical Services utility. Generally, to install + MySQL as a Windows service you should be logged in using an + account that has administrator rights. - Before mysql.server starts the server, it changes location to the - MySQL installation directory, and then invokes mysqld_safe. If you - want the server to run as some specific user, add an appropriate - user option to the [mysqld] group of the /etc/my.cnf option file, - as shown later in this section. (It is possible that you will need - to edit mysql.server if you've installed a binary distribution of - MySQL in a nonstandard location. Modify it to change location into - the proper directory before it runs mysqld_safe. If you do this, - your modified version of mysql.server may be overwritten if you - upgrade MySQL in the future, so you should make a copy of your - edited version that you can reinstall.) + The Services utility (the Windows Service Control Manager) can be + found in the Windows Control Panel (under Administrative Tools on + Windows 2000, XP, Vista and Server 2003). To avoid conflicts, it + is advisable to close the Services utility while performing server + installation or removal operations from the command line. - mysql.server stop stops the server by sending a signal to it. You - can also stop the server manually by executing mysqladmin - shutdown. + Before installing MySQL as a Windows service, you should first + stop the current server if it is running by using the following + command: +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqladmin" + -u root shutdown - To start and stop MySQL automatically on your server, you need to - add start and stop commands to the appropriate places in your - /etc/rc* files. +Note - If you use the Linux server RPM package - (MySQL-server-VERSION.rpm), the mysql.server script is installed - in the /etc/init.d directory with the name mysql. You need not - install it manually. See Section 2.4, "Installing MySQL from RPM - Packages on Linux," for more information on the Linux RPM - packages. + If the MySQL root user account has a password, you need to invoke + mysqladmin with the -p option and supply the password when + prompted. - Some vendors provide RPM packages that install a startup script - under a different name such as mysqld. + This command invokes the MySQL administrative utility mysqladmin + to connect to the server and tell it to shut down. The command + connects as the MySQL root user, which is the default + administrative account in the MySQL grant system. Note that users + in the MySQL grant system are wholly independent from any login + users under Windows. - If you install MySQL from a source distribution or using a binary - distribution format that does not install mysql.server - automatically, you can install it manually. The script can be - found in the support-files directory under the MySQL installation - directory or in a MySQL source tree. + Install the server as a service using this command: +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" --install - To install mysql.server manually, copy it to the /etc/init.d - directory with the name mysql, and then make it executable. Do - this by changing location into the appropriate directory where - mysql.server is located and executing these commands: -shell> cp mysql.server /etc/init.d/mysql -shell> chmod +x /etc/init.d/mysql + The service-installation command does not start the server. + Instructions for that are given later in this section. - Older Red Hat systems use the /etc/rc.d/init.d directory rather - than /etc/init.d. Adjust the preceding commands accordingly. - Alternatively, first create /etc/init.d as a symbolic link that - points to /etc/rc.d/init.d: -shell> cd /etc -shell> ln -s rc.d/init.d . + To make it easier to invoke MySQL programs, you can add the path + name of the MySQL bin directory to your Windows system PATH + environment variable: - After installing the script, the commands needed to activate it to - run at system startup depend on your operating system. On Linux, - you can use chkconfig: -shell> chkconfig --add mysql + * On the Windows desktop, right-click on the My Computer icon, + and select Properties. - On some Linux systems, the following command also seems to be - necessary to fully enable the mysql script: -shell> chkconfig --level 345 mysql on + * Next select the Advanced tab from the System Properties menu + that appears, and click the Environment Variables button. - On FreeBSD, startup scripts generally should go in - /usr/local/etc/rc.d/. The rc(8) manual page states that scripts in - this directory are executed only if their basename matches the - *.sh shell file name pattern. Any other files or directories - present within the directory are silently ignored. In other words, - on FreeBSD, you should install the mysql.server script as - /usr/local/etc/rc.d/mysql.server.sh to enable automatic startup. + * Under System Variables, select Path, and then click the Edit + button. The Edit System Variable dialogue should appear. - As an alternative to the preceding setup, some operating systems - also use /etc/rc.local or /etc/init.d/boot.local to start - additional services on startup. To start up MySQL using this - method, you could append a command like the one following to the - appropriate startup file: -/bin/sh -c 'cd /usr/local/mysql; ./bin/mysqld_safe --user=mysql &' + * Place your cursor at the end of the text appearing in the + space marked Variable Value. (Use the End key to ensure that + your cursor is positioned at the very end of the text in this + space.) Then enter the complete path name of your MySQL bin + directory (for example, C:\Program Files\MySQL\MySQL Server + 5.1\bin), Note that there should be a semicolon separating + this path from any values present in this field. Dismiss this + dialogue, and each dialogue in turn, by clicking OK until all + of the dialogues that were opened have been dismissed. You + should now be able to invoke any MySQL executable program by + typing its name at the DOS prompt from any directory on the + system, without having to supply the path. This includes the + servers, the mysql client, and all MySQL command-line + utilities such as mysqladmin and mysqldump. + You should not add the MySQL bin directory to your Windows + PATH if you are running multiple MySQL servers on the same + machine. - For other systems, consult your operating system documentation to - see how to install startup scripts. +Warning - You can add options for mysql.server in a global /etc/my.cnf file. - A typical /etc/my.cnf file might look like this: -[mysqld] -datadir=/usr/local/mysql/var -socket=/var/tmp/mysql.sock -port=3306 -user=mysql + You must exercise great care when editing your system PATH by + hand; accidental deletion or modification of any portion of the + existing PATH value can leave you with a malfunctioning or even + unusable system. -[mysql.server] -basedir=/usr/local/mysql + The following additional arguments can be used in MySQL 5.1 when + installing the service: - The mysql.server script supports the following options: basedir, - datadir, and pid-file. If specified, they must be placed in an - option file, not on the command line. mysql.server supports only - start and stop as command-line arguments. + * You can specify a service name immediately following the + --install option. The default service name is MySQL. - The following table shows which option groups the server and each - startup script read from option files. - Script Option Groups - mysqld [mysqld], [server], [mysqld-major_version] - mysqld_safe [mysqld], [server], [mysqld_safe] - mysql.server [mysqld], [mysql.server], [server] + * If a service name is given, it can be followed by a single + option. By convention, this should be + --defaults-file=file_name to specify the name of an option + file from which the server should read options when it starts. + The use of a single option other than --defaults-file is + possible but discouraged. --defaults-file is more flexible + because it enables you to specify multiple startup options for + the server by placing them in the named option file. - [mysqld-major_version] means that groups with names like - [mysqld-5.0] and [mysqld-5.1] are read by servers having versions - 5.0.x, 5.1.x, and so forth. This feature can be used to specify - options that can be read only by servers within a given release - series. + * You can also specify a --local-service option following the + service name. This causes the server to run using the + LocalService Windows account that has limited system + privileges. This account is available only for Windows XP or + newer. If both --defaults-file and --local-service are given + following the service name, they can be in any order. - For backward compatibility, mysql.server also reads the - [mysql_server] group and mysqld_safe also reads the [safe_mysqld] - group. However, you should update your option files to use the - [mysql.server] and [mysqld_safe] groups instead when using MySQL - 5.1. + For a MySQL server that is installed as a Windows service, the + following rules determine the service name and option files that + the server uses: - See Section 4.2.3.3, "Using Option Files." + * If the service-installation command specifies no service name + or the default service name (MySQL) following the --install + option, the server uses the a service name of MySQL and reads + options from the [mysqld] group in the standard option files. -2.11.2.3. Starting and Troubleshooting the MySQL Server + * If the service-installation command specifies a service name + other than MySQL following the --install option, the server + uses that service name. It reads options from the [mysqld] + group and the group that has the same name as the service in + the standard option files. This allows you to use the [mysqld] + group for options that should be used by all MySQL services, + and an option group with the service name for use by the + server installed with that service name. - This section provides troubleshooting suggestions for problems - starting the server on Unix. If you are using Windows, see Section - 2.3.13, "Troubleshooting a MySQL Installation Under Windows." + * If the service-installation command specifies a + --defaults-file option after the service name, the server + reads options only from the [mysqld] group of the named file + and ignores the standard option files. - If you have problems starting the server, here are some things to - try: + As a more complex example, consider the following command: +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" + --install MySQL --defaults-file=C:\my-opts.cnf - * Check the error log to see why the server does not start. + Here, the default service name (MySQL) is given after the + --install option. If no --defaults-file option had been given, + this command would have the effect of causing the server to read + the [mysqld] group from the standard option files. However, + because the --defaults-file option is present, the server reads + options from the [mysqld] option group, and only from the named + file. - * Specify any special options needed by the storage engines you - are using. + You can also specify options as Start parameters in the Windows + Services utility before you start the MySQL service. - * Make sure that the server knows where to find the data - directory. + Once a MySQL server has been installed as a service, Windows + starts the service automatically whenever Windows starts. The + service also can be started immediately from the Services utility, + or by using a NET START MySQL command. The NET command is not case + sensitive. - * Make sure that the server can access the data directory. The - ownership and permissions of the data directory and its - contents must be set such that the server can read and modify - them. + When run as a service, mysqld has no access to a console window, + so no messages can be seen there. If mysqld does not start, check + the error log to see whether the server wrote any messages there + to indicate the cause of the problem. The error log is located in + the MySQL data directory (for example, C:\Program + Files\MySQL\MySQL Server 5.1\data). It is the file with a suffix + of .err. - * Verify that the network interfaces the server wants to use are - available. + When a MySQL server has been installed as a service, and the + service is running, Windows stops the service automatically when + Windows shuts down. The server also can be stopped manually by + using the Services utility, the NET STOP MySQL command, or the + mysqladmin shutdown command. - Some storage engines have options that control their behavior. You - can create a my.cnf file and specify startup options for the - engines that you plan to use. If you are going to use storage - engines that support transactional tables (InnoDB, NDB), be sure - that you have them configured the way you want before starting the - server: + You also have the choice of installing the server as a manual + service if you do not wish for the service to be started + automatically during the boot process. To do this, use the + --install-manual option rather than the --install option: +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" --install-m +anual - * If you are using InnoDB tables, see Section 13.6.2, "InnoDB - Configuration." + To remove a server that is installed as a service, first stop it + if it is running by executing NET STOP MySQL. Then use the + --remove option to remove it: +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" --remove - * If you are using MySQL Cluster, see Section 17.3, "MySQL - Cluster Configuration." + If mysqld is not running as a service, you can start it from the + command line. For instructions, see Section 2.5.5.5, "Starting + MySQL from the Windows Command Line." - MySQL Enterprise For expert advice on start-up options appropriate - to your circumstances, subscribe to The MySQL Enterprise Monitor. - For more information, see - http://www.mysql.com/products/enterprise/advisors.html. + Please see Section 2.5.6, "Troubleshooting a MySQL Installation + Under Windows," if you encounter difficulties during installation. - Storage engines will use default option values if you specify - none, but it is recommended that you review the available options - and specify explicit values for those for which the defaults are - not appropriate for your installation. +2.5.5.7. Testing The MySQL Installation - When the mysqld server starts, it changes location to the data - directory. This is where it expects to find databases and where it - expects to write log files. The server also writes the pid - (process ID) file in the data directory. + You can test whether the MySQL server is working by executing any + of the following commands: +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqlshow" +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqlshow" -u root +mysql +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqladmin" version + status proc +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysql" test - The data directory location is hardwired in when the server is - compiled. This is where the server looks for the data directory by - default. If the data directory is located somewhere else on your - system, the server will not work properly. You can determine what - the default path settings are by invoking mysqld with the - --verbose and --help options. +Note - If the default locations don't match the MySQL installation layout - on your system, you can override them by specifying options to - mysqld or mysqld_safe on the command line or in an option file. + By default, mysqlshow will try to connect using the ODBC user. + This user is not created by default. You should specify a valid + user, or root with the right password to check the operation of + the server. - To specify the location of the data directory explicitly, use the - --datadir option. However, normally you can tell mysqld the - location of the base directory under which MySQL is installed and - it looks for the data directory there. You can do this with the - --basedir option. + If mysqld is slow to respond to TCP/IP connections from client + programs, there is probably a problem with your DNS. In this case, + start mysqld with the --skip-name-resolve option and use only + localhost and IP numbers in the Host column of the MySQL grant + tables. - To check the effect of specifying path options, invoke mysqld with - those options followed by the --verbose and --help options. For - example, if you change location into the directory where mysqld is - installed and then run the following command, it shows the effect - of starting the server with a base directory of /usr/local: -shell> ./mysqld --basedir=/usr/local --verbose --help + You can force a MySQL client to use a named-pipe connection rather + than TCP/IP by specifying the --pipe or --protocol=PIPE option, or + by specifying . (period) as the host name. Use the --socket option + to specify the name of the pipe if you do not want to use the + default pipe name. - You can specify other options such as --datadir as well, but - --verbose and --help must be the last options. + Note that if you have set a password for the root account, deleted + the anonymous account, or created a new user account, then you + must use the appropriate -u and -p options with the commands shown + above in order to connect with the MySQL Server. See Section + 4.2.2, "Connecting to the MySQL Server." - Once you determine the path settings you want, start the server - without --verbose and --help. + For more information about mysqlshow, see Section 4.5.6, + "mysqlshow --- Display Database, Table, and Column Information." - If mysqld is currently running, you can find out what path - settings it is using by executing this command: -shell> mysqladmin variables +2.5.6. Troubleshooting a MySQL Installation Under Windows - Or: -shell> mysqladmin -h host_name variables + When installing and running MySQL for the first time, you may + encounter certain errors that prevent the MySQL server from + starting. The purpose of this section is to help you diagnose and + correct some of these errors. - host_name is the name of the MySQL server host. + Your first resource when troubleshooting server issues is the + error log. The MySQL server uses the error log to record + information relevant to the error that prevents the server from + starting. The error log is located in the data directory specified + in your my.ini file. The default data directory location is + C:\Program Files\MySQL\MySQL Server 5.1\data. See Section 5.2.2, + "The Error Log." - If you get Errcode 13 (which means Permission denied) when - starting mysqld, this means that the privileges of the data - directory or its contents do not allow the server access. In this - case, you change the permissions for the involved files and - directories so that the server has the right to use them. You can - also start the server as root, but this raises security issues and - should be avoided. + Another source of information regarding possible errors is the + console messages displayed when the MySQL service is starting. Use + the NET START MySQL command from the command line after installing + mysqld as a service to see any error messages regarding the + starting of the MySQL server as a service. See Section 2.5.5.6, + "Starting MySQL as a Windows Service." + + The following examples show other common error messages you may + encounter when installing MySQL and starting the server for the + first time: - On Unix, change location into the data directory and check the - ownership of the data directory and its contents to make sure the - server has access. For example, if the data directory is - /usr/local/mysql/var, use this command: -shell> ls -la /usr/local/mysql/var + * If the MySQL server cannot find the mysql privileges database + or other critical files, you may see these messages: +System error 1067 has occurred. +Fatal error: Can't open privilege tables: Table 'mysql.host' doesn't +exist + These messages often occur when the MySQL base or data + directories are installed in different locations than the + default locations (C:\Program Files\MySQL\MySQL Server 5.1 and + C:\Program Files\MySQL\MySQL Server 5.1\data, respectively). + This situation may occur when MySQL is upgraded and installed + to a new location, but the configuration file is not updated + to reflect the new location. In addition, there may be old and + new configuration files that conflict. Be sure to delete or + rename any old configuration files when upgrading MySQL. + If you have installed MySQL to a directory other than + C:\Program Files\MySQL\MySQL Server 5.1, you need to ensure + that the MySQL server is aware of this through the use of a + configuration (my.ini) file. The my.ini file needs to be + located in your Windows directory, typically C:\WINDOWS. You + can determine its exact location from the value of the WINDIR + environment variable by issuing the following command from the + command prompt: +C:\> echo %WINDIR% + An option file can be created and modified with any text + editor, such as Notepad. For example, if MySQL is installed in + E:\mysql and the data directory is D:\MySQLdata, you can + create the option file and set up a [mysqld] section to + specify values for the basedir and datadir options: +[mysqld] +# set basedir to your installation path +basedir=E:/mysql +# set datadir to the location of your data directory +datadir=D:/MySQLdata + Note that Windows path names are specified in option files + using (forward) slashes rather than backslashes. If you do use + backslashes, double them: +[mysqld] +# set basedir to your installation path +basedir=C:\\Program Files\\MySQL\\MySQL Server 5.1 +# set datadir to the location of your data directory +datadir=D:\\MySQLdata + The rules for use of backslash in option file values are given + in Section 4.2.3.3, "Using Option Files." + If you change the datadir value in your MySQL configuration + file, you must move the contents of the existing MySQL data + directory before restarting the MySQL server. + See Section 2.5.5.2, "Creating an Option File." - If the data directory or its files or subdirectories are not owned - by the login account that you use for running the server, change - their ownership to that account. If the account is named mysql, - use these commands: -shell> chown -R mysql /usr/local/mysql/var -shell> chgrp -R mysql /usr/local/mysql/var + * If you reinstall or upgrade MySQL without first stopping and + removing the existing MySQL service and install MySQL using + the MySQL Config Wizard, you may see this error: +Error: Cannot create Windows service for MySql. Error: 0 + This occurs when the Config Wizard tries to install the + service and finds an existing service with the same name. + One solution to this problem is to choose a service name other + than mysql when using the configuration wizard. This allows + the new service to be installed correctly, but leaves the + outdated service in place. Although this is harmless, it is + best to remove old services that are no longer in use. + To permanently remove the old mysql service, execute the + following command as a user with administrative privileges, on + the command-line: +C:\> sc delete mysql +[SC] DeleteService SUCCESS + If the sc utility is not available for your version of + Windows, download the delsrv utility from + http://www.microsoft.com/windows2000/techinfo/reskit/tools/exi + sting/delsrv-o.asp and use the delsrv mysql syntax. - If the server fails to start up correctly, check the error log. - Log files are located in the data directory (typically C:\Program - Files\MySQL\MySQL Server 5.1\data on Windows, - /usr/local/mysql/data for a Unix binary distribution, and - /usr/local/var for a Unix source distribution). Look in the data - directory for files with names of the form host_name.err and - host_name.log, where host_name is the name of your server host. - Then examine the last few lines of these files. On Unix, you can - use tail to display them: -shell> tail host_name.err -shell> tail host_name.log +2.5.7. Upgrading MySQL on Windows - The error log should contain information that indicates why the - server couldn't start. + This section lists some of the steps you should take when + upgrading MySQL on Windows. - If either of the following errors occur, it means that some other - program (perhaps another mysqld server) is using the TCP/IP port - or Unix socket file that mysqld is trying to use: -Can't start server: Bind on TCP/IP port: Address already in use -Can't start server: Bind on unix socket... + 1. Review Section 2.4.1, "Upgrading MySQL," for additional + information on upgrading MySQL that is not specific to + Windows. - Use ps to determine whether you have another mysqld server - running. If so, shut down the server before starting mysqld again. - (If another server is running, and you really want to run multiple - servers, you can find information about how to do so in Section - 5.6, "Running Multiple MySQL Servers on the Same Machine.") + 2. You should always back up your current MySQL installation + before performing an upgrade. See Section 6.1, "Database + Backup Methods." - If no other server is running, try to execute the command telnet - your_host_name tcp_ip_port_number. (The default MySQL port number - is 3306.) Then press Enter a couple of times. If you don't get an - error message like telnet: Unable to connect to remote host: - Connection refused, some other program is using the TCP/IP port - that mysqld is trying to use. You'll need to track down what - program this is and disable it, or else tell mysqld to listen to a - different port with the --port option. In this case, you'll also - need to specify the port number for client programs when - connecting to the server via TCP/IP. + 3. Download the latest Windows distribution of MySQL from + http://dev.mysql.com/downloads/. - Another reason the port might be inaccessible is that you have a - firewall running that blocks connections to it. If so, modify the - firewall settings to allow access to the port. + 4. Before upgrading MySQL, you must stop the server. If the + server is installed as a service, stop the service with the + following command from the command prompt: +C:\> NET STOP MySQL + If you are not running the MySQL server as a service, use the + following command to stop it: +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqladmin" -u root + shutdown - If the server starts but you can't connect to it, you should make - sure that you have an entry in /etc/hosts that looks like this: -127.0.0.1 localhost +Note + If the MySQL root user account has a password, you need to + invoke mysqladmin with the -p option and supply the password + when prompted. - This problem occurs only on systems that do not have a working - thread library and for which MySQL must be configured to use - MIT-pthreads. + 5. When upgrading to MySQL 5.1 from a version previous to 4.1.5, + or when upgrading from a version of MySQL installed from a Zip + archive to a version of MySQL installed with the MySQL + Installation Wizard, you must manually remove the previous + installation and MySQL service (if the server is installed as + a service). + To remove the MySQL service, use the following command: +C:\> C:\mysql\bin\mysqld --remove + If you do not remove the existing service, the MySQL + Installation Wizard may fail to properly install the new MySQL + service. - If you cannot get mysqld to start, you can try to make a trace - file to find the problem by using the --debug option. See MySQL - Internals: Porting - (http://forge.mysql.com/wiki/MySQL_Internals_Porting) + 6. When upgrading from MySQL 5.1.23 to MySQL 5.1.24, the change + in the default location of the data directory from a directory + within the MySQL installation to the AppData folder means that + you must manually copy the data files from your old + installation to the new location. -2.11.3. Securing the Initial MySQL Accounts + 7. If you are using the MySQL Installation Wizard, start the + wizard as described in Section 2.5.3.1, "Using the MySQL + Installation Wizard." - Part of the MySQL installation process is to set up the mysql - database that contains the grant tables: + 8. If you are installing MySQL from a Zip archive, extract the + archive. You may either overwrite your existing MySQL + installation (usually located at C:\mysql), or install it into + a different directory, such as C:\mysql5. Overwriting the + existing installation is recommended. - * Windows distributions contain preinitialized grant tables that - are installed automatically. + 9. If you were running MySQL as a Windows service and you had to + remove the service earlier in this procedure, reinstall the + service. (See Section 2.5.5.6, "Starting MySQL as a Windows + Service.") + 10. Restart the server. For example, use NET START MySQL if you + run MySQL as a service, or invoke mysqld directly otherwise. + 11. If you encounter errors, see Section 2.5.6, "Troubleshooting a + MySQL Installation Under Windows." - * On Unix, the grant tables are populated by the - mysql_install_db program. Some installation methods run this - program for you. Others require that you execute it manually. - For details, see Section 2.11.2, "Unix Post-Installation - Procedures." +2.5.8. Windows Post-Installation Procedures - The grant tables define the initial MySQL user accounts and their - access privileges. These accounts are set up as follows: + On Windows, the data directory and the grant tables do not have to + be created. MySQL Windows distributions include the grant tables + with a set of preinitialized accounts in the mysql database under + the data directory. It is unnecessary to run the mysql_install_db + script that is used on Unix. Regarding passwords, if you installed + MySQL using the Windows Installation Wizard, you may have already + assigned passwords to the accounts. (See Section 2.5.3.1, "Using + the MySQL Installation Wizard.") Otherwise, use the + password-assignment procedure given in Section 2.13.2, "Securing + the Initial MySQL Accounts." - * Accounts with the user name root are created. These are - superuser accounts that can do anything. The initial root - account passwords are empty, so anyone can connect to the - MySQL server as root --- without a password --- and be granted - all privileges. + Before setting up passwords, you might want to try running some + client programs to make sure that you can connect to the server + and that it is operating properly. Make sure that the server is + running (see Section 2.5.5.4, "Starting the Server for the First + Time"), and then issue the following commands to verify that you + can retrieve information from the server. The output should be + similar to what is shown here: +C:\> C:\mysql\bin\mysqlshow ++--------------------+ +| Databases | ++--------------------+ +| information_schema | +| mysql | +| test | ++--------------------+ - + On Windows, one root account is created; this account - allows connecting from the local host only. The Windows - installer will optionally create an account allowing for - connections from any host only if the user selects the - Enable root access from remote machines option during - installation. +Note - + On Unix, both root accounts are for connections from the - local host. Connections must be made from the local host - by specifying a host name of localhost for one of the - accounts, or the actual host name or IP number for the - other. + The above may not work if the correct user does not exist. If you + installed using the MSI packages and used the MySQL Server + Instance Config Wizard, then the root will haqve been created + automatically with the password you supplied. In this case, you + should use the -u and -p options where you will be prompted for + the password. - * Two anonymous-user accounts are created, each with an empty - user name. The anonymous accounts have no password, so anyone - can use them to connect to the MySQL server. +Note - + On Windows, one anonymous account is for connections from - the local host. It has no global privileges. (Before - MySQL 5.1.16, it has all global privileges, just like the - root accounts.) The other is for connections from any - host and has all privileges for the test database and for - other databases with names that start with test. + The list of installed databases may vary, but will always include + the minimum of mysql and information_schema. In most cases, the + test database will also be installed automatically. + + If you specify the name of the database, then a list of the tables + within a given database will be displayed: +C:\> C:\mysql\bin\mysqlshow mysql +Database: mysql ++---------------------------+ +| Tables | ++---------------------------+ +| columns_priv | +| db | +| event | +| func | +| general_log | +| help_category | +| help_keyword | +| help_relation | +| help_topic | +| host | +| plugin | +| proc | +| procs_priv | +| servers | +| slow_log | +| tables_priv | +| time_zone | +| time_zone_leap_second | +| time_zone_name | +| time_zone_transition | +| time_zone_transition_type | +| user | ++---------------------------+ - + On Unix, both anonymous accounts are for connections from - the local host. Connections must be made from the local - host by specifying a host name of localhost for one of - the accounts, or the actual host name or IP number for - the other. These accounts have all privileges for the - test database and for other databases with names that - start with test_. - As noted, none of the initial accounts have passwords. This means - that your MySQL installation is unprotected until you do something - about it: +C:\> C:\mysql\bin\mysql -e "SELECT Host,Db,User FROM db" mysql ++------+-------+------+ +| host | db | user | ++------+-------+------+ +| % | test% | | ++------+-------+------+ - * If you want to prevent clients from connecting as anonymous - users without a password, you should either assign a password - to each anonymous account or else remove the accounts. + You may need to specify a different directory from the one shown; + if you used the Windows Installation Wizard, then the default + directory is C:\Program Files\MySQL\MySQL Server 5.1, and the + mysql and mysqlshow client programs are in C:\Program + Files\MySQL\MySQL Server 5.1\bin. See Section 2.5.3.1, "Using the + MySQL Installation Wizard," for more information. - * You should assign a password to each MySQL root account. + If you have already secured the initial MySQL accounts, you may + need to use the -u and -p options to supply a user name and + password to the mysqlshow and mysql client programs; otherwise the + programs may fail with an error, or you may not be able to view + all databases. For example, if you have assigned the password + "secretpass" to the MySQL root account, then you can invoke + mysqlshow and mysql as shown here: +C:\> C:\mysql\bin\mysqlshow -uroot -psecretpass ++--------------------+ +| Databases | ++--------------------+ +| information_schema | +| mysql | +| test | ++--------------------+ - The following instructions describe how to set up passwords for - the initial MySQL accounts, first for the anonymous accounts and - then for the root accounts. Replace "newpwd" in the examples with - the actual password that you want to use. The instructions also - cover how to remove the anonymous accounts, should you prefer not - to allow anonymous access at all. +C:\> C:\mysql\bin\mysqlshow -uroot -psecretpass mysql +Database: mysql ++---------------------------+ +| Tables | ++---------------------------+ +| columns_priv | +| db | +| event | +| func | +| general_log | +| help_category | +| help_keyword | +| help_relation | +| help_topic | +| host | +| plugin | +| proc | +| procs_priv | +| servers | +| slow_log | +| tables_priv | +| time_zone | +| time_zone_leap_second | +| time_zone_name | +| time_zone_transition | +| time_zone_transition_type | +| user | ++---------------------------+ - You might want to defer setting the passwords until later, so that - you don't need to specify them while you perform additional setup - or testing. However, be sure to set them before using your - installation for production purposes. - Anonymous Account Password Assignment +C:\> C:\mysql\bin\mysql -uroot -psecretpass -e "SELECT Host,Db,User F +ROM db" mysql ++------+-------+------+ +| host | db | user | ++------+-------+------+ +| % | test% | | ++------+-------+------+ - To assign passwords to the anonymous accounts, connect to the - server as root and then use either SET PASSWORD or UPDATE. In - either case, be sure to encrypt the password using the PASSWORD() - function. + For more information about these programs, see Section 4.5.6, + "mysqlshow --- Display Database, Table, and Column Information," + and Section 4.5.1, "mysql --- The MySQL Command-Line Tool." - To use SET PASSWORD on Windows, do this: -shell> mysql -u root -mysql> SET PASSWORD FOR ''@'localhost' = PASSWORD('newpwd'); -mysql> SET PASSWORD FOR ''@'%' = PASSWORD('newpwd'); + If you are running a version of Windows that supports services and + you want the MySQL server to run automatically when Windows + starts, see Section 2.5.5.6, "Starting MySQL as a Windows + Service." - To use SET PASSWORD on Unix, do this: -shell> mysql -u root -mysql> SET PASSWORD FOR ''@'localhost' = PASSWORD('newpwd'); -mysql> SET PASSWORD FOR ''@'host_name' = PASSWORD('newpwd'); +2.5.9. MySQL on Windows Compared to MySQL on Unix - In the second SET PASSWORD statement, replace host_name with the - name of the server host. This is the name that is specified in the - Host column of the non-localhost record for root in the user - table. If you don't know what host name this is, issue the - following statement before using SET PASSWORD: -mysql> SELECT Host, User FROM mysql.user; + MySQL for Windows has proven itself to be very stable. The Windows + version of MySQL has the same features as the corresponding Unix + version, with the following exceptions: - Look for the record that has root in the User column and something - other than localhost in the Host column. Then use that Host value - in the second SET PASSWORD statement. + * Limited number of ports + Windows systems have about 4,000 ports available for client + connections, and after a connection on a port closes, it takes + two to four minutes before the port can be reused. In + situations where clients connect to and disconnect from the + server at a high rate, it is possible for all available ports + to be used up before closed ports become available again. If + this happens, the MySQL server appears to be unresponsive even + though it is running. Note that ports may be used by other + applications running on the machine as well, in which case the + number of ports available to MySQL is lower. + For more information about this problem, see + http://support.microsoft.com/default.aspx?scid=kb;en-us;196271 + . - Anonymous Account Removal + * Concurrent reads + MySQL depends on the pread() and pwrite() system calls to be + able to mix INSERT and SELECT. Currently, we use mutexes to + emulate pread() and pwrite(). We intend to replace the file + level interface with a virtual interface in the future so that + we can use the readfile()/writefile() interface to get more + speed. The current implementation limits the number of open + files that MySQL 5.1 can use to 2,048, which means that you + cannot run as many concurrent threads on Windows as on Unix. - If you prefer to remove the anonymous accounts instead, do so as - follows: -shell> mysql -u root -mysql> DROP USER ''; + * Blocking read + MySQL uses a blocking read for each connection. That has the + following implications if named-pipe connections are enabled: - The DROP statement applies both to Windows and to Unix. On - Windows, if you want to remove only the anonymous account that has - the same privileges as root, do this instead: -shell> mysql -u root -mysql> DROP USER ''@'localhost'; + + A connection is not disconnected automatically after + eight hours, as happens with the Unix version of MySQL. - That account allows anonymous access but has full privileges, so - removing it improves security. + + If a connection hangs, it is not possible to break it + without killing MySQL. - root Account Password Assignment + + mysqladmin kill does not work on a sleeping connection. - You can assign passwords to the root accounts in several ways. The - following discussion demonstrates three methods: + + mysqladmin shutdown cannot abort as long as there are + sleeping connections. + We plan to fix this problem in the future. - * Use the SET PASSWORD statement + * ALTER TABLE + While you are executing an ALTER TABLE statement, the table is + locked from being used by other threads. This has to do with + the fact that on Windows, you can't delete a file that is in + use by another thread. In the future, we may find some way to + work around this problem. - * Use the mysqladmin command-line client program + * DATA DIRECTORY and INDEX DIRECTORY + The DATA DIRECTORY and INDEX DIRECTORY options for CREATE + TABLE are ignored on Windows, because Windows doesn't support + symbolic links. These options also are ignored on systems that + have a nonfunctional realpath() call. - * Use the UPDATE statement + * DROP DATABASE + You cannot drop a database that is in use by another thread. - To assign passwords using SET PASSWORD, connect to the server as - root and issue SET PASSWORD statements. Be sure to encrypt the - password using the PASSWORD() function. + * Case-insensitive names + File names are not case sensitive on Windows, so MySQL + database and table names are also not case sensitive on + Windows. The only restriction is that database and table names + must be specified using the same case throughout a given + statement. See Section 8.2.2, "Identifier Case Sensitivity." - For Windows, do this: -shell> mysql -u root -mysql> SET PASSWORD FOR 'root'@'localhost' = PASSWORD('newpwd'); -mysql> SET PASSWORD FOR 'root'@'%' = PASSWORD('newpwd'); + * Directory and file names + On Windows, MySQL Server supports only directory and file + names that are compatible with the current ANSI code pages. + For example, the following Japanese directory name will not + work in the Western locale (code page 1252): +datadir="C:/维基百科关于中文维基百科" + The same limitation applies to directory and file names + referred to in SQL statements, such as the data file path name + in LOAD DATA INFILE. - For Unix, do this: -shell> mysql -u root -mysql> SET PASSWORD FOR 'root'@'localhost' = PASSWORD('newpwd'); -mysql> SET PASSWORD FOR 'root'@'host_name' = PASSWORD('newpwd'); + * The "\" path name separator character + Path name components in Windows are separated by the "\" + character, which is also the escape character in MySQL. If you + are using LOAD DATA INFILE or SELECT ... INTO OUTFILE, use + Unix-style file names with "/" characters: +mysql> LOAD DATA INFILE 'C:/tmp/skr.txt' INTO TABLE skr; +mysql> SELECT * INTO OUTFILE 'C:/tmp/skr.txt' FROM skr; + Alternatively, you must double the "\" character: +mysql> LOAD DATA INFILE 'C:\\tmp\\skr.txt' INTO TABLE skr; +mysql> SELECT * INTO OUTFILE 'C:\\tmp\\skr.txt' FROM skr; - In the second SET PASSWORD statement, replace host_name with the - name of the server host. This is the same host name that you used - when you assigned the anonymous account passwords. + * Problems with pipes + Pipes do not work reliably from the Windows command-line + prompt. If the pipe includes the character ^Z / CHAR(24), + Windows thinks that it has encountered end-of-file and aborts + the program. + This is mainly a problem when you try to apply a binary log as + follows: +C:\> mysqlbinlog binary_log_file | mysql --user=root + If you have a problem applying the log and suspect that it is + because of a ^Z / CHAR(24) character, you can use the + following workaround: +C:\> mysqlbinlog binary_log_file --result-file=/tmp/bin.sql +C:\> mysql --user=root --execute "source /tmp/bin.sql" + The latter command also can be used to reliably read in any + SQL file that may contain binary data. - If the user table contains an account with User and Host values of - 'root' and '127.0.0.1', use an additional SET PASSWORD statement - to set that account's password: -mysql> SET PASSWORD FOR 'root'@'127.0.0.1' = PASSWORD('newpwd'); + * Access denied for user error + If MySQL cannot resolve your host name properly, you may get + the following error when you attempt to run a MySQL client + program to connect to a server running on the same machine: +Access denied for user 'some_user'@'unknown' +to database 'mysql' + To fix this problem, you should create a file named + \windows\hosts containing the following information: +127.0.0.1 localhost - To assign passwords to the root accounts using mysqladmin, execute - the following commands: -shell> mysqladmin -u root password "newpwd" -shell> mysqladmin -u root -h host_name password "newpwd" + Here are some open issues for anyone who might want to help us + improve MySQL on Windows: + + * Add macros to use the faster thread-safe increment/decrement + methods provided by Windows. + +2.5.10. Installing MySQL from Source on Windows - These commands apply both to Windows and to Unix. In the second - command, replace host_name with the name of the server host. The - double quotes around the password are not always necessary, but - you should use them if the password contains spaces or other - characters that are special to your command interpreter. + These instructions describe how to build binaries from source for + MySQL 5.1 on Windows. Instructions are provided for building + binaries from a standard source distribution or from the Bazaar + tree that contains the latest development source. - The mysqladmin method of setting the root account passwords does - not set the password for the 'root'@'127.0.0.1' account. To do so, - use SET PASSWORD as shown earlier. +Note - You can also use UPDATE to modify the user table directly. The - following UPDATE statement assigns a password to all root - accounts: -shell> mysql -u root -mysql> UPDATE mysql.user SET Password = PASSWORD('newpwd') - -> WHERE User = 'root'; -mysql> FLUSH PRIVILEGES; + The instructions here are strictly for users who want to test + MySQL on Microsoft Windows from the latest source distribution or + from the Bazaar tree. For production use, we do not advise using a + MySQL server built by yourself from source. Normally, it is best + to use precompiled binary distributions of MySQL that are built + specifically for optimal performance on Windows by Sun + Microsystems, Inc. Instructions for installing binary + distributions are available in Section 2.5, "Installing MySQL on + Windows." - The UPDATE statement applies both to Windows and to Unix. + To build MySQL on Windows from source, you must satisfy the + following system, compiler, and resource requirements: - After the passwords have been set, you must supply the appropriate - password whenever you connect to the server. For example, if you - want to use mysqladmin to shut down the server, you can do so - using this command: -shell> mysqladmin -u root -p shutdown -Enter password: (enter root password here) + * Windows 2000, Windows XP, or newer version. + Windows Vista is supported when using Visual Studio 2005 + provided you have installed the following updates: -Note + + Microsoft Visual Studio 2005 Professional Edition - ENU + Service Pack 1 (KB926601) + (http://support.microsoft.com/?kbid=926601) - If you forget your root password after setting it up, Section - B.1.4.1, "How to Reset the Root Password," covers the procedure - for resetting it. + + Security Update for Microsoft Visual Studio 2005 + Professional Edition - ENU (KB937061) + (http://support.microsoft.com/?kbid=937061) - To set up additional accounts, you can use the GRANT statement. - For instructions, see Section 5.5.2, "Adding User Accounts." + + Update for Microsoft Visual Studio 2005 Professional + Edition - ENU (KB932232) + (http://support.microsoft.com/?kbid=932232) -2.12. Upgrading or Downgrading MySQL + * CMake, which can be downloaded from http://www.cmake.org. + After installing, modify your path to include the cmake + binary. -2.12.1. Upgrading MySQL + * Microsoft Visual C++ 2005 Express Edition, Visual Studio .Net + 2003 (7.1), or Visual Studio 2005 (8.0) compiler system. - As a general rule, to upgrade from one release series to another, - you should go to the next series rather than skipping a series. To - upgrade from a release series previous to MySQL 5.0, upgrade to - each successive release series in turn until you have reached - MySQL 5.0, and then proceed with the upgrade to MySQL 5.1. For - example, if you currently are running MySQL 4.0 and wish to - upgrade to a newer series, upgrade to MySQL 4.1 first before - upgrading to 5.0, and so forth. For information on upgrading to - MySQL 5.0, see the MySQL 5.0 Reference Manual; for earlier - releases, see the MySQL 3.23, 4.0, 4.1 Reference Manual. + * If you are using Visual C++ 2005 Express Edition, you must + also install an appropriate Platform SDK. More information and + links to downloads for various Windows platforms is available + from + http://www.microsoft.com/downloads/details.aspx?familyid=0baf2 + b35-c656-4969-ace8-e4c0c0716adb. - To upgrade from MySQL 5.0 to 5.1, use the items in the following - checklist as a guide: + * If you are compiling from a Bazaar tree or making changes to + the parser, you need bison for Windows, which can be + downloaded from + http://gnuwin32.sourceforge.net/packages/bison.htm. Download + the package labeled "Complete package, excluding sources". + After installing the package, modify your path to include the + bison binary and ensure that this binary is accessible from + Visual Studio. - * Before any upgrade, back up your databases, including the - mysql database that contains the grant tables. See Section - 6.1, "Database Backups." + * Cygwin might be necessary if you want to run the test script + or package the compiled binaries and support files into a Zip + archive. (Cygwin is needed only to test or package the + distribution, not to build it.) Cygwin is available from + http://cygwin.com. - * Read all the notes in Section 2.12.1.1, "Upgrading from MySQL - 5.0 to 5.1." These notes enable you to identify upgrade issues - that apply to your current MySQL installation. Some - incompatibilities discussed in that section require your - attention before upgrading. Others should be dealt with after - upgrading. + * 3GB to 5GB of disk space. - * Read Appendix C, "MySQL Change History" as well, which - provides information about features that are new in MySQL 5.1 - or differ from those found in MySQL 5.0. + The exact system requirements for Visual Studio can be found here: + http://msdn.microsoft.com/vstudio/Previous/2003/sysreqs/default.as + px and + http://msdn.microsoft.com/vstudio/products/sysreqs/default.aspx - * After you upgrade to a new version of MySQL, run mysql_upgrade - (see Section 4.4.8, "mysql_upgrade --- Check Tables for MySQL - Upgrade"). This program checks your tables, and attempts to - repair them if necessary. It also updates your grant tables to - make sure that they have the current structure so that you can - take advantage of any new capabilities. (Some releases of - MySQL introduce changes to the structure of the grant tables - to add new privileges or features.) + You also need a MySQL source distribution for Windows, which can + be obtained two ways: - * If you are running MySQL Server on Windows, see Section - 2.3.14, "Upgrading MySQL on Windows." + * Obtain a source distribution packaged by Sun Microsystems, + Inc. These are available from http://dev.mysql.com/downloads/. - * If you are using replication, see Section 16.3.3, "Upgrading a - Replication Setup," for information on upgrading your - replication setup. + * Package a source distribution yourself from the latest Bazaar + developer source tree. For instructions on pulling the latest + source files, see Section 2.3.3, "Installing from the + Development Source Tree." - * If you are upgrading an installation originally produced by - installing multiple RPM packages, it is best to upgrade all - the packages, not just some. For example, if you previously - installed the server and client RPMs, do not upgrade just the - server RPM. + If you find something not working as expected, or you have + suggestions about ways to improve the current build process on + Windows, please send a message to the win32 mailing list. See + Section 1.5.1, "MySQL Mailing Lists." - * As of MySQL 5.1.9, the mysqld-max server is included in binary - distributions. There is no separate MySQL-Max distribution. As - of MySQL 5.1.12, there is no mysqld-max server at all in - binary distributions. They contain a server that includes the - features previously included in mysqld-max. +2.5.10.1. Building MySQL from Source Using CMake and Visual Studio - * If you have created a user-defined function (UDF) with a given - name and upgrade MySQL to a version that implements a new - built-in function with the same name, the UDF becomes - inaccessible. To correct this, use DROP FUNCTION to drop the - UDF, and then use CREATE FUNCTION to re-create the UDF with a - different nonconflicting name. The same is true if the new - version of MySQL implements a built-in function with the same - name as an existing stored function. See Section 8.2.4, - "Function Name Parsing and Resolution," for the rules - describing how the server interprets references to different - kinds of functions. + You can build MySQL on Windows by using a combination of cmake and + Microsoft Visual Studio .NET 2003 (7.1), Microsoft Visual Studio + 2005 (8.0) or Microsoft Visual C++ 2005 Express Edition. You must + have the appropriate Microsoft Platform SDK installed. - You can always move the MySQL format files and data files between - different versions on systems with the same architecture as long - as you stay within versions for the same release series of MySQL. +Note - If you are cautious about using new versions, you can always - rename your old mysqld before installing a newer one. For example, - if you are using MySQL 5.0.13 and want to upgrade to 5.1.10, - rename your current server from mysqld to mysqld-5.0.13. If your - new mysqld then does something unexpected, you can simply shut it - down and restart with your old mysqld. + To compile from the source code on Windows you must use the + standard source distribution (for example, mysql-5.1.41.tar.gz). + You build from the same distribution as used to build MySQL on + Unix, Linux and other platforms. Do not use the Windows Source + distributions as they do not contain the necessary configuration + script and other files. - If, after an upgrade, you experience problems with recompiled - client programs, such as Commands out of sync or unexpected core - dumps, you probably have used old header or library files when - compiling your programs. In this case, you should check the date - for your mysql.h file and libmysqlclient.a library to verify that - they are from the new MySQL distribution. If not, recompile your - programs with the new headers and libraries. + Follow this procedure to build MySQL: - If problems occur, such as that the new mysqld server does not - start or that you cannot connect without a password, verify that - you do not have an old my.cnf file from your previous - installation. You can check this with the --print-defaults option - (for example, mysqld --print-defaults). If this command displays - anything other than the program name, you have an active my.cnf - file that affects server or client operation. + 1. If you are installing from a packaged source distribution, + create a work directory (for example, C:\workdir), and unpack + the source distribution there using WinZip or another Windows + tool that can read .zip files. This directory is the work + directory in the following instructions. - If your MySQL installation contains a large amount of data that - might take a long time to convert after an in-place upgrade, you - might find it useful to create a "dummy" database instance for - assessing what conversions might be needed and the work involved - to perform them. Make a copy of your MySQL instance that contains - a full copy of the mysql database, plus all other databases - without data. Run your upgrade procedure on this dummy instance to - see what actions might be needed so that you can better evaluate - the work involved when performing actual data conversion on your - original database instance. + 2. Using a command shell, navigate to the work directory and run + the following command: +C:\workdir>win\configure.js options + If you have associated the .js file extension with an + application such as a text editor, then you may need to use + the following command to force configure.js to be executed as + a script: +C:\workdir>cscript win\configure.js options + These options are available for configure.js: - It is a good idea to rebuild and reinstall the Perl DBD::mysql - module whenever you install a new release of MySQL. The same - applies to other MySQL interfaces as well, such as PHP mysql - extensions and the Python MySQLdb module. + + WITH_INNOBASE_STORAGE_ENGINE: Enable the InnoDB storage + engine. -2.12.1.1. Upgrading from MySQL 5.0 to 5.1 + + WITH_PARTITION_STORAGE_ENGINE: Enable user-defined + partitioning. - After upgrading a 5.0 installation to 5.0.10 or above, it is - necessary to upgrade your grant tables. Otherwise, creating stored - procedures and functions might not work. To perform this upgrade, - run mysql_upgrade. + + WITH_ARCHIVE_STORAGE_ENGINE: Enable the ARCHIVE storage + engine. -Note + + WITH_BLACKHOLE_STORAGE_ENGINE: Enable the BLACKHOLE + storage engine. - It is good practice to back up your data before installing any new - version of software. Although MySQL works very hard to ensure a - high level of quality, you should protect your data by making a - backup. + + WITH_EXAMPLE_STORAGE_ENGINE: Enable the EXAMPLE storage + engine. - To upgrade to 5.1 from any previous version, MySQL recommends that - you dump your tables with mysqldump before upgrading and reload - the dump file after upgrading. + + WITH_FEDERATED_STORAGE_ENGINE: Enable the FEDERATED + storage engine. - In general, you should do the following when upgrading from MySQL - 5.0 to 5.1: + + WITH_NDBCLUSTER_STORAGE_ENGINE (experimental): Enable the + NDBCLUSTER storage engine in the MySQL server; cause + binaries for the MySQL Cluster management and data node, + management client, and other programs to be built. + This option is supported only in MySQL Cluster NDB 7.0 + (NDBCLUSTER storage engine versions 6.4.0 and later) + using the MySQL Cluster sources. It cannot be used to + enable clustering support in other MySQL source trees or + distributions. - * Read all the items in the following sections to see whether - any of them might affect your applications: + + MYSQL_SERVER_SUFFIX=suffix: Server suffix, default none. - + Section 2.12.1, "Upgrading MySQL," has general update - information. + + COMPILATION_COMMENT=comment: Server comment, default + "Source distribution". - + The items in the change lists found later in this section - enable you to identify upgrade issues that apply to your - current MySQL installation. + + MYSQL_TCP_PORT=port: Server port, default 3306. + + + DISABLE_GRANT_OPTIONS: Disables the --bootstrap, + --skip-grant-tables, and --init-file options for mysqld. + This option is available as of MySQL 5.1.15. + For example (type the command on one line): +C:\workdir>win\configure.js WITH_INNOBASE_STORAGE_ENGINE + WITH_PARTITION_STORAGE_ENGINE MYSQL_SERVER_SUFFIX=-pro - + The MySQL 5.1 change history describes significant new - features you can use in 5.1 or that differ from those - found in MySQL 5.0. Some of these changes may result in - incompatibilities. See Section C.1, "Changes in Release - 5.1.x (Production)." + 3. From the work directory, execute the win\build-vs8.bat or + win\build-vs71.bat file, depending on the version of Visual + Studio you have installed. The script invokes CMake, which + generates the mysql.sln solution file. + You can also use win\build-vs8_x64.bat to build the 64-bit + version of MySQL. However, you cannot build the 64-bit version + with Visual Studio Express Edition. You must use Visual Studio + 2005 (8.0) or higher. - * Note particularly any changes that are marked Known issue or - Incompatible change. These incompatibilities with earlier - versions of MySQL may require your attention before you - upgrade. - Our aim is to avoid these changes, but occasionally they are - necessary to correct problems that would be worse than an - incompatibility between releases. If any upgrade issue - applicable to your installation involves an incompatibility - that requires special handling, follow the instructions given - in the incompatibility description. Often this will involve a - dump and reload, or use of a statement such as CHECK TABLE or - REPAIR TABLE. - For dump and reload instructions, see Section 2.12.4, - "Rebuilding or Repairing Tables or Indexes." Any procedure - that involves REPAIR TABLE with the USE_FRM option must be - done before upgrading. Use of this statement with a version of - MySQL different from the one used to create the table (that - is, using it after upgrading) may damage the table. See - Section 12.5.2.6, "REPAIR TABLE Syntax." + 4. From the work directory, open the generated mysql.sln file + with Visual Studio and select the proper configuration using + the Configuration menu. The menu provides Debug, Release, + RelwithDebInfo, MinRelInfo options. Then select Solution > + Build to build the solution. + Remember the configuration that you use in this step. It is + important later when you run the test script because that + script needs to know which configuration you used. - * After you upgrade to a new version of MySQL, run mysql_upgrade - (see Section 4.4.8, "mysql_upgrade --- Check Tables for MySQL - Upgrade"). This program checks your tables, and attempts to - repair them if necessary. It also updates your grant tables to - make sure that they have the current structure so that you can - take advantage of any new capabilities. (Some releases of - MySQL introduce changes to the structure of the grant tables - to add new privileges or features.) + 5. Test the server. The server built using the preceding + instructions expects that the MySQL base directory and data + directory are C:\mysql and C:\mysql\data by default. If you + want to test your server using the source tree root directory + and its data directory as the base directory and data + directory, you need to tell the server their path names. You + can either do this on the command line with the --basedir and + --datadir options, or by placing appropriate options in an + option file. (See Section 4.2.3.3, "Using Option Files.") If + you have an existing data directory elsewhere that you want to + use, you can specify its path name instead. + When the server is running in standalone fashion or as a + service based on your configuration, try to connect to it from + the mysql interactive command-line utility. + You can also run the standard test script, mysql-test-run.pl. + This script is written in Perl, so you'll need either Cygwin + or ActiveState Perl to run it. You may also need to install + the modules required by the script. To run the test script, + change location into the mysql-test directory under the work + directory, set the MTR_VS_CONFIG environment variable to the + configuration you selected earlier (or use the --vs-config + option), and invoke mysql-test-run.pl. For example (using + Cygwin and the bash shell): +shell> cd mysql-test +shell> export MTR_VS_CONFIG=debug +shell> ./mysql-test-run.pl --force --timer +shell> ./mysql-test-run.pl --force --timer --ps-protocol - * Check Section 2.12.3, "Checking Whether Table Indexes Must Be - Rebuilt," to see whether changes to character sets or - collations were made that affect your table indexes. If so, - you will need to rebuild the affected indexes using the - instructions in Section 2.12.4, "Rebuilding or Repairing - Tables or Indexes." + When you are satisfied that the programs you have built are + working correctly, stop the server. Now you can install the + distribution. One way to do this is to use the make_win_bin_dist + script in the scripts directory of the MySQL source distribution + (see Section 4.4.2, "make_win_bin_dist --- Package MySQL + Distribution as ZIP Archive"). This is a shell script, so you must + have Cygwin installed if you want to use it. It creates a Zip + archive of the built executables and support files that you can + unpack in the location at which you want to install MySQL. - * If you are running MySQL Server on Windows, see Section - 2.3.14, "Upgrading MySQL on Windows." + It is also possible to install MySQL by copying directories and + files directly: - * If you are using replication, see Section 16.3.3, "Upgrading a - Replication Setup," for information on upgrading your - replication setup. + 1. Create the directories where you want to install MySQL. For + example, to install into C:\mysql, use these commands: +C:\> mkdir C:\mysql +C:\> mkdir C:\mysql\bin +C:\> mkdir C:\mysql\data +C:\> mkdir C:\mysql\share +C:\> mkdir C:\mysql\scripts + If you want to compile other clients and link them to MySQL, + you should also create several additional directories: +C:\> mkdir C:\mysql\include +C:\> mkdir C:\mysql\lib +C:\> mkdir C:\mysql\lib\debug +C:\> mkdir C:\mysql\lib\opt + If you want to benchmark MySQL, create this directory: +C:\> mkdir C:\mysql\sql-bench + Benchmarking requires Perl support. See Section 2.15, "Perl + Installation Notes." - If your MySQL installation contains a large amount of data that - might take a long time to convert after an in-place upgrade, you - might find it useful to create a "dummy" database instance for - assessing what conversions might be needed and the work involved - to perform them. Make a copy of your MySQL instance that contains - a full copy of the mysql database, plus all other databases - without data. Run your upgrade procedure on this dummy instance to - see what actions might be needed so that you can better evaluate - the work involved when performing actual data conversion on your - original database instance. + 2. From the work directory, copy into the C:\mysql directory the + following directories: +C:\> cd \workdir +C:\workdir> copy client_release\*.exe C:\mysql\bin +C:\workdir> copy client_debug\mysqld.exe C:\mysql\bin\mysqld-debug.ex +e +C:\workdir> xcopy scripts\*.* C:\mysql\scripts /E +C:\workdir> xcopy share\*.* C:\mysql\share /E + If you want to compile other clients and link them to MySQL, + you should also copy several libraries and header files: +C:\workdir> copy lib_debug\mysqlclient.lib C:\mysql\lib\debug +C:\workdir> copy lib_debug\libmysql.* C:\mysql\lib\debug +C:\workdir> copy lib_debug\zlib.* C:\mysql\lib\debug +C:\workdir> copy lib_release\mysqlclient.lib C:\mysql\lib\opt +C:\workdir> copy lib_release\libmysql.* C:\mysql\lib\opt +C:\workdir> copy lib_release\zlib.* C:\mysql\lib\opt +C:\workdir> copy include\*.h C:\mysql\include +C:\workdir> copy libmysql\libmysql.def C:\mysql\include + If you want to benchmark MySQL, you should also do this: +C:\workdir> xcopy sql-bench\*.* C:\mysql\bench /E - MySQL Enterprise MySQL Enterprise subscribers will find more - information about upgrading in the Knowledge Base articles found - at Upgrading - (https://kb.mysql.com/search.php?cat=search&category=41) Access - to the MySQL Knowledge Base collection of articles is one of the - advantages of subscribing to MySQL Enterprise. For more - information, see - http://www.mysql.com/products/enterprise/advisors.html. + After installation, set up and start the server in the same way as + for binary Windows distributions. See Section 2.5, "Installing + MySQL on Windows." - The following lists describe changes that may affect applications - and that you should watch out for when upgrading to MySQL 5.1. +2.5.11. Compiling MySQL Clients on Windows - Configuration Changes: + In your source files, you should include my_global.h before + mysql.h: +#include <my_global.h> +#include <mysql.h> - * Before MySQL 5.1.11, to build MySQL from source with SSL - support enabled, you would invoke configure with either the - --with-openssl or --with-yassl option. In MySQL 5.1.11, those - options both have been replaced by the --with-ssl option. By - default, --with-ssl causes the bundled yaSSL library to be - used. To select OpenSSL instead, give the option as - --with-ssl=path, where path is the directory where the OpenSSL - header files and libraries are located. + my_global.h includes any other files needed for Windows + compatibility (such as windows.h) if you compile your program on + Windows. - Server Changes: + You can either link your code with the dynamic libmysql.lib + library, which is just a wrapper to load in libmysql.dll on + demand, or link with the static mysqlclient.lib library. - * Known issue: Dumps performed by using mysqldump to generate a - dump file before the upgrade and reloading the file after - upgrading are subject to the following problem: - Before MySQL 5.0.40, mysqldump displays SPATIAL index - definitions using prefix lengths for the indexed columns. - These prefix lengths are accepted in MySQL 5.0, but not as of - MySQL 5.1. If you use mysqldump from versions of MySQL older - than 5.0.40, any table containing SPATIAL indexes will cause - an error when the dump file is reloaded into MySQL 5.1 or - higher. - For example, a table definition might look like this when - dumped in MySQL 5.0: -CREATE TABLE `t` ( - `g` geometry NOT NULL, - SPATIAL KEY `g` (`g`(32)) -) ENGINE=MyISAM DEFAULT CHARSET=latin1 - The SPATIAL index definition will not be accepted in MySQL - 5.1. To work around this, edit the dump file to remove the - prefix: -CREATE TABLE `t` ( - `g` geometry NOT NULL, - SPATIAL KEY `g` (`g`) -) ENGINE=MyISAM DEFAULT CHARSET=latin1 - Dump files can be large, so it may be preferable to dump table - definitions and data separately to make it easier to edit the - definitions: -shell> mysqldump --no-data other_args > definitions.sql -shell> mysqldump --no-create-info other_args > data.sql - Then edit definitions.sql before reloading definitions.sql and - data.sql, in that order. - If you upgrade to a version of MySQL 5.0 higher than 5.0.40 - before upgrading to MySQL 5.1, this problem does not occur. + The MySQL client libraries are compiled as threaded libraries, so + you should also compile your code to be multi-threaded. - * Known issue: Before MySQL 5.1.30, the CHECK TABLE ... FOR - UPGRADE statement did not check for incompatible collation - changes made in MySQL 5.1.24. (This also affects mysqlcheck - and mysql_upgrade, which cause that statement to be executed.) - Prior to the fix made in 5.1.30, a binary upgrade (performed - without dumping tables with mysqldump before the upgrade and - reloading the dump file after the upgrade) would corrupt - tables. After the fix, CHECK TABLE ... FOR UPGRADE properly - detects the problem and warns about tables that need repair. - However, the fix is not backward compatible and can result in - a downgrading problem under these circumstances: +2.6. Installing MySQL on Linux - 1. Perform a binary upgrade to a version of MySQL that - includes the fix. + The following sections covers the installation of Linux using + RPMs. For information on using a generic binary package using tar, + see Section 2.2, "Installing MySQL from Generic Binaries on + Unix/Linux." For information on installing from source, see + Section 2.3, "MySQL Installation Using a Source Distribution." - 2. Run CHECK TABLE ... FOR UPGRADE (or mysqlcheck or - mysql_upgrade) to upgrade tables. + mysql.server can be found in the support-files directory under the + MySQL installation directory or in a MySQL source tree. You can + install it as /etc/init.d/mysql for automatic MySQL startup and + shutdown. See Section 2.13.1.2, "Starting and Stopping MySQL + Automatically." - 3. Perform a binary downgrade to a version of MySQL that - does not include the fix. - The solution is to dump tables with mysqldump before the - downgrade and reload the dump file after the downgrade. - Alternatively, drop and recreate affected indexes. +2.6.1. Installing MySQL from RPM Packages on Linux - * Known issue: MySQL introduces encoding for table names that - have non-ASCII characters (see Section 8.2.3, "Mapping of - Identifiers to File Names"). After a binary upgrade from MySQL - 5.0 to 5.1 or higher, the server recognizes names that have - non-ASCII characters and adds a #mysql50# prefix to them. - As of MySQL 5.1.31, mysql_upgrade encodes these names by - executing the following command: -mysqlcheck --all-databases --check-upgrade --fix-db-names --fix-table --names - Prior to MySQL 5.1.31, mysql_upgrade does not execute this - command, so you should execute it manually if you have - database or table names that contain nonalphanumeric - characters. - Prior to MySQL 5.1.23, the mysqlcheck command does not perform - the name encoding for views. To work around this problem, drop - each affected view and recreate it. - mysqlcheck cannot fix names that contain literal instances of - the @ character that is used for encoding special characters. - If you have databases or tables that contain this character, - use mysqldump to dump them before upgrading to MySQL 5.1, and - then reload the dump file after upgrading. + The recommended way to install MySQL on RPM-based Linux + distributions is by using the RPM packages. The RPMs that we + provide to the community should work on all versions of Linux that + support RPM packages and use glibc 2.3. To obtain RPM packages, + see Section 2.1.3, "How to Get MySQL." - * Known issue: When upgrading from MySQL 5.0 to versions of 5.1 - prior to 5.1.23, running mysqlcheck (or mysql_upgrade, which - runs mysqlcheck) to upgrade tables fails for names that must - be written as quoted identifiers. To work around this problem, - rename each affected table to a name that does not require - quoting: -RENAME TABLE `tab``le_a` TO table_a; -RENAME TABLE `table b` TO table_b; - After renaming the tables, run the mysql_upgrade program. Then - rename the tables back to their original names: -RENAME TABLE table_a TO `tab``le_a`; -RENAME TABLE table_b TO `table b`; + For non-RPM Linux distributions, you can install MySQL using a + .tar.gz package. See Section 2.2, "Installing MySQL from Generic + Binaries on Unix/Linux." - * Known issue: In connection with view creation, the server - created arc directories inside database directories and - maintained useless copies of .frm files there. Creation and - renaming procedures of those copies as well as creation of arc - directories has been discontinued in MySQL 5.1.29. - This change does cause a problem when downgrading to older - server versions which manifests itself under these - circumstances: + We do provide some platform-specific RPMs; the difference between + a platform-specific RPM and a generic RPM is that a + platform-specific RPM is built on the targeted platform and is + linked dynamically whereas a generic RPM is linked statically with + LinuxThreads. - 1. Create a view v_orig in MySQL 5.1.29 or higher. +Note - 2. Rename the view to v_new and then back to v_orig. + RPM distributions of MySQL often are provided by other vendors. Be + aware that they may differ in features and capabilities from those + built by us, and that the instructions in this manual do not + necessarily apply to installing them. The vendor's instructions + should be consulted instead. - 3. Downgrade to an older 5.1.x server and run mysql_upgrade. + In most cases, you need to install only the MySQL-server and + MySQL-client packages to get a functional MySQL installation. The + other packages are not required for a standard installation. - 4. Try to rename v_orig to v_new again. This operation - fails. - As a workaround to avoid this problem, use either of these - approaches: + RPMs for MySQL Cluster. Beginning with MySQL 5.1.24, standard + MySQL server RPMs built by MySQL no longer provide support for the + NDBCLUSTER storage engine. MySQL Cluster users wanting to upgrade + MySQL 5.1.23 or earlier installations from RPMs built by MySQL + should upgrade to MySQL Cluster NDB 6.2 or MySQL Cluster NDB 6.3; + RPMs that should work with most Linux distributions are available + for both of these release series. - + Dump your data using mysqldump before downgrading and - reload the dump file after downgrading. +Important - + Instead of renaming a view after the downgrade, drop it - and recreate it. + When upgrading a MySQL Cluster RPM installation, you must upgrade + all installed RPMs, including the Server and Client RPMs. - * Incompatible change: Character set or collation changes were - made in MySQL 5.1.21, 5.1.23, and 5.1.24 that may require - table indexes to be rebuilt. For details, see Section 2.12.3, - "Checking Whether Table Indexes Must Be Rebuilt." + For more information about installing MySQL Cluster from RPMs, see + Section 17.2.1, "MySQL Cluster Multi-Computer Installation." - * Incompatible change: In MySQL 5.1.36, options for loading - plugins such as pluggable storage engines were changed from - boolean to tristate format. The implementations overlap, but - if you previously used options of the form --plugin_name=0 or - --plugin_name=1, you should instead use --plugin_name=OFF or - --plugin_name=ON, respectively. For details, see Section - 5.1.3, "Server Options for Loading Plugins." + For upgrades, if your installation was originally produced by + installing multiple RPM packages, it is best to upgrade all the + packages, not just some. For example, if you previously installed + the server and client RPMs, do not upgrade just the server RPM. - * Incompatible change: From MySQL 5.1.24 to 5.1.31, the UPDATE - statement was changed such that assigning NULL to a NOT NULL - column caused an error even when strict SQL mode was not - enabled. The original behavior before MySQL 5.1.24 was that - such assignments caused an error only in strict SQL mode, and - otherwise set the column to the implicit default value for the - column data type and generated a warning. (For information - about implicit default values, see Section 10.1.4, "Data Type - Default Values.") - The change caused compatibility problems for applications that - relied on the original behavior. It also caused replication - problems between servers that had the original behavior and - those that did not, for applications that assigned NULL to NOT - NULL columns in UPDATE statements without strict SQL mode - enabled. The change was reverted in MySQL 5.1.32 so that - UPDATE again had the original behavior. Problems can still - occur if you replicate between servers that have the modified - UPDATE behavior and those that do not. + The RPM packages shown in the following list are available. The + names shown here use a suffix of .glibc23.i386.rpm, but particular + packages can have different suffixes, described later. - * Incompatible change: As of MySQL 5.1.29, the default binary - logging mode has been changed from MIXED to STATEMENT for - compatibility with MySQL 5.0. + * MySQL-server-VERSION.glibc23.i386.rpm + The MySQL server. You need this unless you only want to + connect to a MySQL server running on another machine. - * Incompatible change: In MySQL 5.1.25, a change was made to the - way that the server handles prepared statements. This affects - prepared statements processed at the SQL level (using the - PREPARE statement) and those processed using the binary - client-server protocol (using the mysql_stmt_prepare() C API - function). - Previously, changes to metadata of tables or views referred to - in a prepared statement could cause a server crash when the - statement was next executed, or perhaps an error at execute - time with a crash occurring later. For example, this could - happen after dropping a table and recreating it with a - different definition. - Now metadata changes to tables or views referred to by - prepared statements are detected and cause automatic - repreparation of the statement when it is next executed. - Metadata changes occur for DDL statements such as those that - create, drop, alter, rename, or truncate tables, or that - analyze, optimize, or repair tables. Repreparation also occurs - after referenced tables or views are flushed from the table - definition cache, either implicitly to make room for new - entries in the cache, or explicitly due to FLUSH TABLES. - Repreparation is automatic, but to the extent that it occurs, - performance of prepared statements is diminished. - Table content changes (for example, with INSERT or UPDATE) do - not cause repreparation, nor do SELECT statements. - An incompatibility with previous versions of MySQL is that a - prepared statement may now return a different set of columns - or different column types from one execution to the next. For - example, if the prepared statement is SELECT * FROM t1, - altering t1 to contain a different number of columns causes - the next execution to return a number of columns different - from the previous execution. - Older versions of the client library cannot handle this change - in behavior. For applications that use prepared statements - with the new server, an upgrade to the new client library is - strongly recommended. - Along with this change to statement repreparation, the default - value of the table_definition_cache system variable has been - increased from 128 to 256. The purpose of this increase is to - lessen the chance that prepared statements will need - repreparation due to referred-to tables/views having been - flushed from the cache to make room for new entries. - A new status variable, Com_stmt_reprepare, has been introduced - to track the number of repreparations. + * MySQL-client-VERSION.glibc23.i386.rpm + The standard MySQL client programs. You probably always want + to install this package. - * Incompatible change: As of MySQL 5.1.23, within a stored - routine, it is no longer allowable to declare a cursor for a - SHOW or DESCRIBE statement. This happened to work in some - instances, but is no longer supported. In many cases, a - workaround for this change is to use the cursor with a SELECT - query to read from an INFORMATION_SCHEMA table that produces - the same information as the SHOW statement. + * MySQL-devel-VERSION.glibc23.i386.rpm + The libraries and include files that are needed if you want to + compile other MySQL clients, such as the Perl modules. - * Incompatible change: SHOW CREATE VIEW displays view - definitions using an AS alias_name clause for each column. If - a column is created from an expression, the default alias is - the expression text, which can be quite long. As of MySQL - 5.1.23, aliases for column names in CREATE VIEW statements are - checked against the maximum column length of 64 characters - (not the maximum alias length of 256 characters). As a result, - views created from the output of SHOW CREATE VIEW fail if any - column alias exceeds 64 characters. This can cause problems - for replication or loading dump files. For additional - information and workarounds, see Section D.4, "Restrictions on - Views." + * MySQL-debuginfo-VERSION.glibc23.i386.rpm + This package contains debugging information. debuginfo RPMs + are never needed to use MySQL software; this is true both for + the server and for client programs. However, they contain + additional information that might be needed by a debugger to + analyze a crash. - * Incompatible change: MySQL 5.1 implements support for a plugin - API that allows the loading and unloading of components at - runtime, without restarting the server. Section 22.2, "The - MySQL Plugin Interface." The plugin API requires the - mysql.plugin table. After upgrading from an older version of - MySQL, you should run the mysql_upgrade command to create this - table. See Section 4.4.8, "mysql_upgrade --- Check Tables for - MySQL Upgrade." - Plugins are installed in the directory named by the plugin_dir - system variable. This variable also controls the location from - which the server loads user-defined functions (UDFs), which is - a change from earlier versions of MySQL. That is, all UDF - library files now must be installed in the plugin directory. - When upgrading from an older version of MySQL, you must - migrate your UDF files to the plugin directory. + * MySQL-shared-VERSION.glibc23.i386.rpm + This package contains the shared libraries + (libmysqlclient.so*) that certain languages and applications + need to dynamically load and use MySQL. It contains + single-threaded and thread-safe libraries. If you install this + package, do not install the MySQL-shared-compat package. - * Incompatible change: The table_cache system variable has been - renamed to table_open_cache. Any scripts that refer to - table_cache must be updated to use the new name. + * MySQL-shared-compat-VERSION.glibc23.i386.rpm + This package includes the shared libraries for MySQL 3.23, + 4.0, and so on, up to the current release. It contains + single-threaded and thread-safe libraries. Install this + package instead of MySQL-shared if you have applications + installed that are dynamically linked against older versions + of MySQL but you want to upgrade to the current version + without breaking the library dependencies. - * Incompatible change: Several issues were identified for stored - programs (stored procedures and functions, triggers, and - events) and views containing non-ASCII symbols. These issues - involved conversion errors due to incomplete character set - information when translating these objects to and from stored - format. - To address these problems, the representation for these - objects was changed in MySQL 5.1.21. However, the fixes affect - all stored programs and views. (For example, you will see - warnings about "no creation context.") To avoid warnings from - the server about the use of old definitions from any release - prior to 5.1.21, you should dump stored programs and views - with mysqldump after upgrading to 5.1.21 or higher, and then - reload them to recreate them with new definitions. Invoke - mysqldump with a --default-character-set option that names the - non-ASCII character set that was used for the definitions when - the objects were originally defined. + * MySQL-shared-compat-advanced-gpl-VERSION.glibc23.i386.rpm, + MySQL-shared-compat-advanced-VERSION.glibc23.i386.rpm + These are like the MySQL-shared-compat package, but are for + the "MySQL Enterprise Server - Advanced Edition" products. + Install these packages rather than the normal + MySQL-shared-compat package if you want to included shared + client libraries for older MySQL versions. - * Incompatible change: As of MySQL 5.1.20, mysqld_safe supports - error logging to syslog on systems that support the logger - command. The new --syslog and --skip-syslog options can be - used instead of the --log-error option to control logging - behavior, as described in Section 4.3.2, "mysqld_safe --- - MySQL Server Startup Script." - In 5.1.21 and up, the default is --skip-syslog, which is - compatible with the default behavior of writing an error log - file for releases prior to 5.1.20. - In 5.1.20 only, the following conditions apply: 1) The default - is to use syslog, which is not compatible with releases prior - to 5.1.20. 2) Logging to syslog may fail to operate correctly - in some cases. For these reasons, avoid using MySQL 5.1.20. + * MySQL-embedded-VERSION.glibc23.i386.rpm + The embedded MySQL server library. - * Incompatible change: As of MySQL 5.1.15, InnoDB rolls back - only the last statement on a transaction timeout. A new - option, --innodb_rollback_on_timeout, causes InnoDB to abort - and roll back the entire transaction if a transaction timeout - occurs (the same behavior as in MySQL 4.1). + * MySQL-ndb-management-VERSION.glibc23.i386.rpm, + MySQL-ndb-storage-VERSION.glibc23.i386.rpm, + MySQL-ndb-tools-VERSION.glibc23.i386.rpm, + MySQL-ndb-extra-VERSION.glibc23.i386.rpm + Packages that contain additional files for MySQL Cluster + installations. + +Note + The MySQL-ndb-tools RPM requires a working installation of + perl. Prior to MySQL 5.1.18, the DBI and HTML::Template + packages were also required. See Section 2.15, "Perl + Installation Notes," and Section 17.4.21, "ndb_size.pl --- + NDBCLUSTER Size Requirement Estimator," for more information. + + * MySQL-test-VERSION.glibc23.i386.rpm + This package includes the MySQL test suite. + + * MySQL-VERSION.src.rpm + This contains the source code for all of the previous + packages. It can also be used to rebuild the RPMs on other + architectures (for example, Alpha or SPARC). - * Incompatible change: As of MySQL 5.1.15, the following - conditions apply to enabling the read_only system variable: + The suffix of RPM package names (following the VERSION value) has + the following syntax: +.PLATFORM.CPU.rpm - + If you attempt to enable read_only while you have any - explicit locks (acquired with LOCK TABLES or have a - pending transaction, an error will occur. + The PLATFORM and CPU values indicate the type of system for which + the package is built. PLATFORM indicates the platform and CPU + indicates the processor type or family. - + If other clients hold explicit table locks or have - pending transactions, the attempt to enable read_only - blocks until the locks are released and the transactions - end. While the attempt to enable read_only is pending, - requests by other clients for table locks or to begin - transactions also block until read_only has been set. + All packages are dynamically linked against glibc 2.3. The + PLATFORM value indicates whether the package is platform + independent or intended for a specific platform, as shown in the + following table. + glibc23 Platform independent, should run on any Linux distribution + that supports glibc 2.3 + rhel3, rhel4 Red Hat Enterprise Linux 3 or 4 + sles9, sles10 SuSE Linux Enterprise Server 9 or 10 - + read_only can be enabled while you hold a global read - lock (acquired with FLUSH TABLES WITH READ LOCK) because - that does not involve table locks. - Previously, the attempt to enable read_only would return - immediately even if explicit locks or transactions were - pending, so some data changes could occur for statements - executing in the server at the same time. + In MySQL 5.1, only glibc23 packages are available currently. - * Incompatible change: The number of function names affected by - IGNORE_SPACE was reduced significantly in MySQL 5.1.13, from - about 200 to about 30. (For details about IGNORE_SPACE, see - Section 8.2.4, "Function Name Parsing and Resolution.") This - change improves the consistency of parser operation. However, - it also introduces the possibility of incompatibility for old - SQL code that relies on the following conditions: + The CPU value indicates the processor type or family for which the + package is built. + i386 x86 processor, 386 and up + i586 x86 processor, Pentium and up + x86_64 64-bit x86 processor + ia64 Itanium (IA-64) processor - + IGNORE_SPACE is disabled. + To see all files in an RPM package (for example, a MySQL-server + RPM), run a command like this: +shell> rpm -qpl MySQL-server-VERSION.glibc23.i386.rpm - + The presence or absence of whitespace following a - function name is used to distinguish between a built-in - function and stored function that have the same name (for - example, PI() versus PI ()). - For functions that are no longer affected by IGNORE_SPACE as - of MySQL 5.1.13, that strategy no longer works. Either of the - following approaches can be used if you have code that is - subject to the preceding incompatibility: + To perform a standard minimal installation, install the server and + client RPMs: +shell> rpm -i MySQL-server-VERSION.glibc23.i386.rpm +shell> rpm -i MySQL-client-VERSION.glibc23.i386.rpm - + If a stored function has a name that conflicts with a - built-in function, refer to the stored function with a - schema name qualifier, regardless of whether whitespace - is present. For example, write schema_name.PI() or - schema_name.PI (). + To install only the client programs, install just the client RPM: +shell> rpm -i MySQL-client-VERSION.glibc23.i386.rpm - + Alternatively, rename the stored function to use a - nonconflicting name and change invocations of the - function to use the new name. + RPM provides a feature to verify the integrity and authenticity of + packages before installing them. If you would like to learn more + about this feature, see Section 2.1.4, "Verifying Package + Integrity Using MD5 Checksums or GnuPG." - * Incompatible change: For utf8 columns, the full-text parser - incorrectly considered several nonword punctuation and - whitespace characters as word characters, causing some - searches to return incorrect results. The fix involves a - change to the full-text parser in MySQL 5.1.12, so as of - 5.1.12, any tables that have FULLTEXT indexes on utf8 columns - must be repaired with REPAIR TABLE: -REPAIR TABLE tbl_name QUICK; + The server RPM places data under the /var/lib/mysql directory. The + RPM also creates a login account for a user named mysql (if one + does not exist) to use for running the MySQL server, and creates + the appropriate entries in /etc/init.d/ to start the server + automatically at boot time. (This means that if you have performed + a previous installation and have made changes to its startup + script, you may want to make a copy of the script so that you + don't lose it when you install a newer RPM.) See Section 2.13.1.2, + "Starting and Stopping MySQL Automatically," for more information + on how MySQL can be started automatically on system startup. - * Incompatible change: Storage engines can be pluggable at - runtime, so the distinction between disabled and invalid - storage engines no longer applies. As of MySQL 5.1.12, this - affects the NO_ENGINE_SUBSTITUTION SQL mode, as described in - Section 5.1.8, "Server SQL Modes." + If you want to install the MySQL RPM on older Linux distributions + that do not support initialization scripts in /etc/init.d + (directly or via a symlink), you should create a symbolic link + that points to the location where your initialization scripts + actually are installed. For example, if that location is + /etc/rc.d/init.d, use these commands before installing the RPM to + create /etc/init.d as a symbolic link that points there: +shell> cd /etc +shell> ln -s rc.d/init.d . - * Incompatible change: The structure of FULLTEXT indexes has - been changed in MySQL 5.1.6. After upgrading to MySQL 5.1.6 or - greater, any tables that have FULLTEXT indexes must be - repaired with REPAIR TABLE: -REPAIR TABLE tbl_name QUICK; + However, all current major Linux distributions should support the + new directory layout that uses /etc/init.d, because it is required + for LSB (Linux Standard Base) compliance. - * Incompatible change: In MySQL 5.1.6, when log tables were - implemented, the default log destination for the general query - and slow query log was TABLE. As of MySQL 5.1.21, this default - has been changed to FILE, which is compatible with MySQL 5.0, - but incompatible with earlier releases of MySQL 5.1. If you - are upgrading from MySQL 5.0 to 5.1.21 or higher, no logging - option changes should be necessary. However, if you are - upgrading from 5.1.6 through 5.1.20 to 5.1.21 or higher and - were using TABLE logging, use the --log-output=TABLE option - explicitly to preserve your server's table-logging behavior. + If the RPM files that you install include MySQL-server, the mysqld + server should be up and running after installation. You should be + able to start using MySQL. - * Incompatible change: For ENUM columns that had enumeration - values containing commas, the commas were mapped to 0xff - internally. However, this rendered the commas - indistinguishable from true 0xff characters in the values. - This no longer occurs. However, the fix requires that you dump - and reload any tables that have ENUM columns containing true - 0xff in their values: Dump the tables using mysqldump with the - current server before upgrading from a version of MySQL 5.1 - older than 5.1.15 to version 5.1.15 or newer. + If something goes wrong, you can find more information in the + binary installation section. See Section 2.2, "Installing MySQL + from Generic Binaries on Unix/Linux." - * As of MySQL 5.1.12, the lc_time_names system variable - specifies the locale that controls the language used to - display day and month names and abbreviations. This variable - affects the output from the DATE_FORMAT(), DAYNAME() and - MONTHNAME() functions. See Section 9.8, "MySQL Server Locale - Support." +Note - * As of MySQL 5.1.9, mysqld_safe no longer implicitly invokes - mysqld-max if it exists. Instead, it invokes mysqld unless a - --mysqld or --mysqld-version option is given to specify - another server explicitly. If you previously relied on the - implicit invocation of mysqld-max, you should use an - appropriate option now. As of MySQL 5.1.12, there is no longer - any separate mysqld-max server, so no change should be - necessary. + The accounts that are listed in the MySQL grant tables initially + have no passwords. After starting the server, you should set up + passwords for them using the instructions in Section 2.13, + "Post-Installation Setup and Testing." - SQL Changes: + During RPM installation, a user named mysql and a group named + mysql are created on the system. This is done using the useradd, + groupadd, and usermod commands. Those commands require appropriate + administrative privileges, which is ensured for locally managed + users and groups (as listed in the /etc/passwd and /etc/group + files) by the RPM installation process being run by root. - * Known issue: Prior to MySQL 5.1.17, the parser accepted - invalid code in SQL condition handlers, leading to server - crashes or unexpected execution behavior in stored programs. - Specifically, the parser allowed a condition handler to refer - to labels for blocks that enclose the handler declaration. - This was incorrect because block label scope does not include - the code for handlers declared within the labeled block. - As of 5.1.17, the parser rejects this invalid construct, but - if you perform a binary upgrade (without dumping and reloading - your databases), existing handlers that contain the construct - still are invalid and should be rewritten even if they appear - to function as you expect. - To find affected handlers, use mysqldump to dump all stored - procedures and functions, triggers, and events. Then attempt - to reload them into an upgraded server. Handlers that contain - illegal label references will be rejected. - For more information about condition handlers and writing them - to avoid invalid jumps, see Section 12.8.4.2, "DECLARE for - Handlers." + For nonlocal user management (LDAP, NIS, and so forth), the + administrative tools may require additional authentication (such + as a password), and will fail if the installing user does not + provide this authentication. Even if they fail, the RPM + installation will not abort but succeed, and this is intentional. + If they failed, some of the intended transfer of ownership may be + missing, and it is recommended that the system administrator then + manually ensures some appropriate user andgroup exists and + manually transfers ownership following the actions in the RPM spec + file. - * Incompatible change: The parser accepted statements that - contained /* ... */ that were not properly closed with */, - such as SELECT 1 /* + 2. As of MySQL 5.1.23, statements that - contain unclosed /*-comments now are rejected with a syntax - error. - This fix has the potential to cause incompatibilities. Because - of Bug#26302: http://bugs.mysql.com/26302, which caused the - trailing */ to be truncated from comments in views, stored - routines, triggers, and events, it is possible that objects of - those types may have been stored with definitions that now - will be rejected as syntactically invalid. Such objects should - be dropped and re-created so that their definitions do not - contain truncated comments. +2.7. Installing MySQL on Mac OS X - * Incompatible change: Multiple-table DELETE statements - containing ambiguous aliases could have unintended side - effects such as deleting rows from the wrong table. Example: -DELETE FROM t1 AS a2 USING t1 AS a1 INNER JOIN t2 AS a2; - As of MySQL 5.1.23, alias declarations can be declared only in - the table_references part. Elsewhere in the statement, alias - references are allowed but not alias declarations. Statements - containing aliases that are no longer allowed must be - rewritten. + MySQL for Mac OS X is available in a number of different forms: - * Incompatible change: As of MySQL 5.1.8, TYPE = engine_name is - still accepted as a synonym for the ENGINE = engine_name table - option but generates a warning. You should note that this - option is not available in MySQL 5.1.7, and is removed - altogether as of MySQL 6.0 and produces a syntax error. - TYPE has been deprecated since MySQL 4.0. + * Native Package Installer format, which uses the native Mac OS + X installer to walk you through the installation of MySQL. For + more information, see Section 2.7.1, "Installing MySQL Using + the Installation Package." You can use the package installer + with Mac OS X 10.3 and later, and available for both PowerPC + and Intel architectures, and both 32-bit and 64-bit + architectures. There is no Universal Binary available using + the package installation method. The user you use to perform + the installation must have administrator privileges. + + * Tar package format, which uses a file packaged using the Unix + tar and gzip commands. To use this method, you will need to + open a Terminal window. You do not need administrator + privileges using this method, as you can install the MySQL + server anywhere using this method. For more information on + using this method, you can use the generic instructions for + using a tarball, Section 2.2, "Installing MySQL from Generic + Binaries on Unix/Linux."You can use the package installer with + Mac OS X 10.3 and later, and available for both PowerPC and + Intel architectures, and both 32-bit and 64-bit architectures. + A Universal Binary, incorporating both Power PC and Intel + architectures and 32-bit and 64-bit binaries is available. + In addition to the core installation, the Package Installer + also includes Section 2.7.2, "Installing the MySQL Startup + Item" and Section 2.7.3, "Installing and Using the MySQL + Preference Pane," both of which simplify the management of + your installation. + + * Mac OS X server includes a version of MySQL as standard. If + you want to use a more recent version than that supplied with + the Mac OS X server release, you can make use of the package + or tar formats. For more information on using the MySQL + bundled with Mac OS X, see Section 2.7.4, "Using MySQL on Mac + OS X Server." - * Incompatible change: The namespace for triggers changed in - MySQL 5.0.10. Previously, trigger names had to be unique per - table. Now they must be unique within the schema (database). - An implication of this change is that DROP TRIGGER syntax now - uses a schema name instead of a table name (schema name is - optional and, if omitted, the current schema will be used). - When upgrading from a version of MySQL 5 older than 5.0.10 to - MySQL 5.0.10 or newer, you must drop all triggers and - re-create them or DROP TRIGGER will not work after the - upgrade. Here is a suggested procedure for doing this: + For additional information on using MySQL on Mac OS X, see Section + 2.7.5, "MySQL Installation on Mac OS X Notes." - 1. Upgrade to MySQL 5.0.10 or later to be able to access - trigger information in the INFORMATION_SCHEMA.TRIGGERS - table. (This should work even for pre-5.0.10 triggers.) +2.7.1. Installing MySQL Using the Installation Package - 2. Dump all trigger definitions using the following SELECT - statement: -SELECT CONCAT('CREATE TRIGGER ', t.TRIGGER_SCHEMA, '.', t.TRIGGER_NAM -E, - ' ', t.ACTION_TIMING, ' ', t.EVENT_MANIPULATION, ' ON ' -, - t.EVENT_OBJECT_SCHEMA, '.', t.EVENT_OBJECT_TABLE, - ' FOR EACH ROW ', t.ACTION_STATEMENT, '//' ) -INTO OUTFILE '/tmp/triggers.sql' -FROM INFORMATION_SCHEMA.TRIGGERS AS t; - The statement uses INTO OUTFILE, so you must have the - FILE privilege. The file will be created on the server - host. Use a different file name if you like. To be 100% - safe, inspect the trigger definitions in the triggers.sql - file, and perhaps make a backup of the file. + You can install MySQL on Mac OS X 10.3.x ("Panther") or newer + using a Mac OS X binary package in PKG format instead of the + binary tarball distribution. Please note that older versions of + Mac OS X (for example, 10.1.x or 10.2.x) are not supported by this + package. - 3. Stop the server and drop all triggers by removing all - .TRG files in your database directories. Change location - to your data directory and issue this command: -shell> rm */*.TRG + The package is located inside a disk image (.dmg) file that you + first need to mount by double-clicking its icon in the Finder. It + should then mount the image and display its contents. - 4. Start the server and re-create all triggers using the - triggers.sql file: -mysql> delimiter // ; -mysql> source /tmp/triggers.sql // +Note - 5. Check that all triggers were successfully created using - the SHOW TRIGGERS statement. + Before proceeding with the installation, be sure to shut down all + running MySQL server instances by either using the MySQL Manager + Application (on Mac OS X Server) or via mysqladmin shutdown on the + command line. - * Incompatible change: MySQL 5.1.6 introduces the TRIGGER - privilege. Previously, the SUPER privilege was needed to - create or drop triggers. Now those operations require the - TRIGGER privilege. This is a security improvement because you - no longer need to grant users the SUPER privilege to enable - them to create triggers. However, the requirement that the - account named in a trigger's DEFINER clause must have the - SUPER privilege has changed to a requirement for the TRIGGER - privilege. When upgrading from a previous version of MySQL 5.0 - or 5.1 to MySQL 5.1.6 or newer, be sure to update your grant - tables by running mysql_upgrade. This will assign the TRIGGER - privilege to all accounts that had the SUPER privilege. If you - fail to update the grant tables, triggers may fail when - activated. After updating the grant tables, you can revoke the - SUPER privilege from those accounts that no longer otherwise - require it. + When installing from the package version, you should also install + the MySQL Preference Pane, which will allow you to control the + startup and execution of your MySQL server from System + Preferences. For more information, see Section 2.7.3, "Installing + and Using the MySQL Preference Pane." + + When installing using the package installer, the files are + installed into a directory within /usr/local matching the name of + the installation version and platform. For example, the installer + file mysql-5.1.39-osx10.5-x86_64.pkg installs MySQL into + /usr/local/mysql-5.1.39-osx10.5-x86_64 . The installation layout + of the directory is as shown in the following table: + Directory Contents of Directory + bin Client programs and the mysqld server + data Log files, databases + docs Manual in Info format + include Include (header) files + lib Libraries + man Unix manual pages + mysql-test MySQL test suite + scripts Contains the mysql_install_db script + share/mysql Error message files + sql-bench Benchmarks + support-files Scripts and sample configuration files + /tmp/mysql.sock The location of the MySQL Unix socket + + During the package installer process, a symbolic link from + /usr/local/mysql to the version/platform specific directory + created during installation will be created automatically. + + 1. Download and open the MySQL package installer, which is + provided on a disk image (.dmg). Double-click to open the disk + image, which includes the main MySQL installation package, the + MySQLStartupItem.pkg installation package, and the + MySQL.prefPane. + + 2. Double-click on the MySQL installer package. It will be named + according to the version of MySQL you have downloaded. For + example, if you have downloaded MySQL 5.1.39, double-click + mysql-5.1.39-osx10.5-x86.pkg. + + 3. You will be presented with the openin installer dialog. Click + Continue to begihn installation. + MySQL Package Installer: Step 1 + + 4. A copy of the installation instructions and other important + information relevant to this installation are display. Click + Continue . + + 5. If you have downloaded the community version of MySQL, you + will be shown a copy of the relevent GNU General Public + License. Click Continue . + + 6. Select the drive you want to use to install the MySQL Startup + Item. The drive must have a valid, bootable, Mac OS X + operating system installed. Click Continue. + MySQL Package Installer: Step 4 + + 7. You will be asked to confirm the details of the installation, + including the space required for the installation. To change + the drive on which the startup item is installed you can click + either Go Back or Change Install Location.... To install the + startup item, click Install. - * Some keywords are reserved in MySQL 5.1 that were not reserved - in MySQL 5.0. See Section 8.3, "Reserved Words." + 8. Once the installation has been completed successfully, you + will be given an Install Succeeded message. - * The LOAD DATA FROM MASTER and LOAD TABLE FROM MASTER - statements are deprecated. See Section 12.6.2.2, "LOAD DATA - FROM MASTER Syntax," for recommended alternatives. + Once you have completed the basic installation, you must complete + the post-installation steps as specifed in Section 2.13, + "Post-Installation Setup and Testing." - * The INSTALL PLUGIN and UNINSTALL PLUGIN statements that are - used for the plugin API are new. So is the WITH PARSER clause - for FULLTEXT index creation that associates a parser plugin - with a full-text index. Section 22.2, "The MySQL Plugin - Interface." + For convenience, you may also want to install the Section 2.7.2, + "Installing the MySQL Startup Item" and Section 2.7.3, "Installing + and Using the MySQL Preference Pane." + +2.7.2. Installing the MySQL Startup Item + + The MySQL Installation Package includes a startup item that can be + used to automatically startup and shutdown MySQL during boot. + + To install the MySQL Startup Item: + + 1. Download and open the MySQL package installer, which is + provided on a disk image (.dmg). Double-click to open the disk + image, which includes the main MySQL installation package, the + MySQLStartupItem.pkg installation package, and the + MySQL.prefPane. + + 2. Double-click on the MySQLStartItem.pkg file to start the + installation process. + + 3. You will be presented with the Install MySQL Startup Item + dialog. + MySQL Startup Item Installer: Step 1 + Click Continue to continue the installation process. + + 4. A copy of the installation instructions and other important + information relevant to this installation are display. Click + Continue . + + 5. Select the drive you want to use to install the MySQL Startup + Item. The drive must have a valid, bootable, Mac OS X + operating system installed. Click Continue. + MySQL Startup Item Installer: Step 3 + + 6. You will be asked to confirm the details of the installation. + To change the drive on which the startup item is installed you + can click either Go Back or Change Install Location.... To + install the startup item, click Install. + + 7. Once the installation has been completed successfully, you + will be given an Install Succeeded message. + MySQL Startup Item Installer: Step 5 - C API Changes: + The Startup Item for MySQL is installed into + /Library/StartupItems/MySQLCOM. The Startup Item installation adds + a variable MYSQLCOM=-YES- to the system configuration file + /etc/hostconfig. If you want to disable the automatic startup of + MySQL, simply change this variable to MYSQLCOM=-NO-. - * Incompatible change: As of MySQL 5.1.7, the - mysql_stmt_attr_get() C API function returns a boolean rather - than an unsigned int for STMT_ATTR_UPDATE_MAX_LENGTH. - (Bug#16144: http://bugs.mysql.com/16144) + After the installation, you can start up MySQL by running the + following commands in a terminal window. You must have + administrator privileges to perform this task. -2.12.2. Downgrading MySQL + If you have installed the Startup Item, use this command to start + the server: +shell> sudo /Library/StartupItems/MySQLCOM/MySQLCOM start - This section describes what you should do to downgrade to an older - MySQL version in the unlikely case that the previous version - worked better than the new one. + You may be prompted for your password to complete the startup. - If you are downgrading within the same release series (for - example, from 5.0.13 to 5.0.12) the general rule is that you just - have to install the new binaries on top of the old ones. There is - no need to do anything with the databases. As always, however, it - is always a good idea to make a backup. + If you have installed the Startup Item, use this command to stop + the server: +shell> sudo /Library/StartupItems/MySQLCOM/MySQLCOM stop + + You may be prompted for your password to complete the shutdown. + +2.7.3. Installing and Using the MySQL Preference Pane + + The MySQL Package installer disk image also includes a custom + MySQL Preference Pane that enables you to start, stop and control + automated startup during boot of your MySQL installation. + + To install the MySQL Preference Pane: + + 1. Download and open the MySQL package installer package, which + is provided on a disk image (.dmg). Double-click to open the + disk image, which includes the main MySQL installation + package, the MySQLStartupItem.pkg installation package, and + the MySQL.prefPane. + + 2. Double click on MySQL.prefPane. The MySQL System Preferences + will open. + + 3. If this is the first time you have installed the preference + pane, you will be asked to confirm installation and whether + you want to install the preference pane for all users, or only + the current user. To install the preference pane for all users + you will need administrator privileges. If necessary, you will + be prompted for the username and password for a user with + administrator privileges. + + 4. If you already have the MySQL Preference Pane installed, you + will be asked to confirm whether you want to overwrite the + existing MySQL Preference Pane. - The following items form a checklist of things you should do - whenever you perform a downgrade: +Note - * Read the upgrading section for the release series from which - you are downgrading to be sure that it does not have any - features you really need. See Section 2.12.1, "Upgrading - MySQL." + The MySQL Preference Pane only starts and stops MySQL installation + installed from the MySQL package installation that have been + installed in the default location. + + Once the MySQL Preference Pane has been installed, you can control + your MySQL server instance using the preference pane. To use the + preference pane, open the System Preferences... from the Apple + menu. Select the MySQL preference pane by clicking on the MySQL + logo within the Other section of the preference panes list. + MySQL Preference Pane + + The MySQL Preference Pane shows the current status of the MySQL + server, showing stopped (in red) if the server is not running and + running (in green) if the server has already been started. The + preference pane will also show the current setting for whether the + MySQL server has been set to start up automatically. + + * To start MySQL using the preference pane: + Click Start MySQL Server. You may be prompted for the username + and password of a user with administrator privileges to start + the MySQL server. + + * To stop MySQL using the preference pane: + Click Stop MySQL Server. You may be prompted for the username + and password of a user with administrator privileges to + shutdown the MySQL server. + + * To automatically start the MySQL server when the system boots: + Check the checkbox next to Automatically Start MySQL Server on + Startup. + + * To disable the automatic starting of the MySQL server when the + system boots: + Uncheck the checkbox next to Automatically Start MySQL Server + on Startup. - * If there is a downgrading section for that version, you should - read that as well. + You can close the System Preferences... once you have completed + your settings. - * To see which new features were added between the version to - which you are downgrading and your current version, see the - change logs (Appendix C, "MySQL Change History"). +2.7.4. Using MySQL on Mac OS X Server - * Check Section 2.12.3, "Checking Whether Table Indexes Must Be - Rebuilt," to see whether changes to character sets or - collations were made between your current version of MySQL and - the version to which you are downgrading. If so and these - changes affect your table indexes, you will need to rebuild - the affected indexes using the instructions in Section 2.12.4, - "Rebuilding or Repairing Tables or Indexes." + If you are running Mac OS X Server, a version of MySQL should + already be installed. The following table shows the versions of + MySQL that ship with Mac OS X Server versions. + Mac OS X Server Version MySQL Version + 10.2-10.2.2 3.23.51 + 10.2.3-10.2.6 3.23.53 + 10.3 4.0.14 + 10.3.2 4.0.16 + 10.4.0 4.1.10a + 10.5.0 5.0.45 + 10.6.0 5.0.82 - In most cases, you can move the MySQL format files and data files - between different versions on the same architecture as long as you - stay within versions for the same release series of MySQL. + The installation layout of MySQL on Mac OS X Server is as shown in + the table below: + Directory Contents of Directory + /usr/bin Client programs + /var/mysql Log files, databases + /usr/libexec The mysqld server + /usr/share/man Unix manual pages + /usr/share/mysql/mysql-test MySQL test suite + /usr/share/mysql Contains the mysql_install_db script + /var/mysql/mysql.sock The location of the MySQL Unix socket - If you downgrade from one release series to another, there may be - incompatibilities in table storage formats. In this case, use - mysqldump to dump your tables before downgrading. After - downgrading, reload the dump file using mysql or mysqlimport to - re-create your tables. For examples, see Section 2.12.5, "Copying - MySQL Databases to Another Machine." +Note - A typical symptom of a downward-incompatible table format change - when you downgrade is that you cannot open tables. In that case, - use the following procedure: + The MySQL server bundled with Mac OS X Server does not include the + MySQL client libraries and header files required if you want to + access and use MySQL from a third-party driver, such as Perl DBI + or PHP. For more information on obtaining and installing MySQL + libraries, see Mac OS X Server version 10.5: MySQL libraries + available for download (http://support.apple.com/kb/TA25017) + Alternatively, you can ignore the bundled MySQL server and install + MySQL from the package or tarball installation. + + For more information on managing the bundled MySQL instance in Mac + OS X Server 10.5, see Mac OS X Server: Web Technologies + Administration For Version 10.5 Leopard + (http://images.apple.com/server/macosx/docs/Web_Technologies_Admin + _v10.5.pdf). For more information on managing the bundled MySQL + instance in Mac OS X Server 10.6, see Mac OS X Server: Web + Technologies Administration Version 10.6 Snow Leopard + (http://manuals.info.apple.com/en_US/WebTech_v10.6.pdf) + +2.7.5. MySQL Installation on Mac OS X Notes + + You should keep the following issues and notes in mind: + + * The default location for the MySQL Unix socket is different on + Mac OS X and Mac OS X Server depending on the installation + type you chose. The default locations by installation are as + follows: - 1. Stop the older MySQL server that you are downgrading to. + Package Installer from MySQL /tmp/mysql.sock + Tarball from MySQL /tmp/mysql.sock + MySQL Bundled with Mac OS X Server /var/mysql/mysql.sock + To prevent issues, you should either change the configuration + of the socket used within your application (for example, + changing php.ini), or you should configure the socket location + using a MySQL configuration file and the socket option. For + more information, see Section 5.1.2, "Server Command Options." + + * You may need (or want) to create a specific mysql user to own + the MySQL directory and data. On Mac OS X 10.4 and lower you + can do this by using the Netinfo Manager application, located + within the Utilities folder within the Applications folder. On + Mac OS X 10.5 and later you can do this through the Directory + Utility. From Mac OS X 10.5 and later (including Mac OS X + Server 10.5) the mysql should already exist. For use in single + user mode, an entry for _mysql (note the underscore prefix) + should already exist within the system /etc/passwd file. - 2. Restart the newer MySQL server you are downgrading from. + * Due to a bug in the Mac OS X package installer, you may see + this error message in the destination disk selection dialog: +You cannot install this software on this disk. (null) + If this error occurs, simply click the Go Back button once to + return to the previous screen. Then click Continue to advance + to the destination disk selection again, and you should be + able to choose the destination disk correctly. We have + reported this bug to Apple and it is investigating this + problem. + + * Because the MySQL package installer installs the MySQL + contents into a version and platform specific directory, you + can use this to upgrade and migrate your database between + versions. You will need to either copy the data directory from + the old version to the new version, or alternatively specify + an alternative datadir value to set location of the data + directory. - 3. Dump any tables that were inaccessible to the older server by - using mysqldump to create a dump file. + * You might want to add aliases to your shell's resource file to + make it easier to access commonly used programs such as mysql + and mysqladmin from the command line. The syntax for bash is: +alias mysql=/usr/local/mysql/bin/mysql +alias mysqladmin=/usr/local/mysql/bin/mysqladmin + For tcsh, use: +alias mysql /usr/local/mysql/bin/mysql +alias mysqladmin /usr/local/mysql/bin/mysqladmin + Even better, add /usr/local/mysql/bin to your PATH environment + variable. You can do this by modifying the appropriate startup + file for your shell. For more information, see Section 4.2.1, + "Invoking MySQL Programs." + + * After you have copied over the MySQL database files from the + previous installation and have successfully started the new + server, you should consider removing the old installation + files to save disk space. Additionally, you should also remove + older versions of the Package Receipt directories located in + /Library/Receipts/mysql-VERSION.pkg. - 4. Stop the newer MySQL server and restart the older one. +2.8. Installing MySQL on Solaris - 5. Reload the dump file into the older server. Your tables should - be accessible. + To obtain a binary MySQL distribution for Solaris in tarball or + PKG format, http://dev.mysql.com/downloads/mysql/5.1.html. - It might also be the case that the structure of the system tables - in the mysql database has changed and that downgrading introduces - some loss of functionality or requires some adjustments. Here are - some examples: + If you install MySQL using a binary tarball distribution on + Solaris, you may run into trouble even before you get the MySQL + distribution unpacked, as the Solaris tar cannot handle long file + names. This means that you may see errors when you try to unpack + MySQL. - * Trigger creation requires the TRIGGER privilege as of MySQL - 5.1. In MySQL 5.0, there is no TRIGGER privilege and SUPER is - required instead. If you downgrade from MySQL 5.1 to 5.0, you - will need to give the SUPER privilege to those accounts that - had the TRIGGER privilege in 5.1. + If this occurs, you must use GNU tar (gtar) to unpack the + distribution. - * Triggers were added in MySQL 5.0, so if you downgrade from 5.0 - to 4.1, you cannot use triggers at all. + You can install MySQL on Solaris using a binary package in PKG + format instead of the binary tarball distribution. Before + installing using the binary PKG format, you should create the + mysql user and group, for example: +groupadd mysql +useradd -g mysql mysql -2.12.2.1. Downgrading to MySQL 5.0 + Some basic PKG-handling commands follow: - When downgrading to MySQL 5.0 from MySQL 5.1 or a later version, - you should keep in mind the following issues relating to features - found in MySQL 5.1 and later, but not in MySQL 5.0: + * To add a package: +pkgadd -d package_name.pkg - * Partitioning. MySQL 5.0 does not support user-defined - partitioning. If a table was created as a partitioned table in - 5.1 (or if an table created in a previous version of MySQL was - altered to include partitions after an upgrade to 5.1), the - table is accessible after downgrade only if you do one of the - following: + * To remove a package: +pkgrm package_name - + Export the table using mysqldump and then drop it in - MySQL 5.1; import the table again following the downgrade - to MySQL 5.0. + * To get a full list of installed packages: +pkginfo - + Prior to the downgrade, remove the table's partitioning - using ALTER TABLE table_name REMOVE PARTITIONING. + * To get detailed information for a package: +pkginfo -l package_name - * Event Scheduler. MySQL 5.0 does not support scheduled events. - If your databases contain scheduled event definitions, you - should prevent them from being dumped when you use mysqldump - by using the --skip-events option. (See Section 4.5.4, - "mysqldump --- A Database Backup Program.") + * To list the files belonging to a package: +pkgchk -v package_name - * Stored routines. MySQL 5.1.21 added a number of new columns - to the mysql.proc table in which stored routine definitions - are stored. If you are downgrading from MySQL 5.1.21 or later - to MySQL 5.0, you cannot import the MySQL 5.1 routine - definitions into MySQL 5.0.46 or earlier using the dump of - mysql.proc created by mysqldump (such as when using the - --all-databases option). Instead, you should run mysqldump - --routines prior to performing the downgrade and run the - stored routines DDL statements following the downgrade. - See Bug#11986: http://bugs.mysql.com/11986, - Bug#30029: http://bugs.mysql.com/30029, and - Bug#30660: http://bugs.mysql.com/30660, for more information. + * To get packaging information for an arbitrary file: +pkgchk -l -p file_name - * Triggers. Trigger creation requires the TRIGGER privilege as - of MySQL 5.1. In MySQL 5.0, there is no TRIGGER privilege and - SUPER is required instead. If you downgrade from MySQL 5.1 to - 5.0, you will need to give the SUPER privilege to those - accounts that had the TRIGGER privilege in 5.1. +2.8.1. Solaris Notes -2.12.3. Checking Whether Table Indexes Must Be Rebuilt + For information about installing MySQL on Solaris using PKG + distributions, see Section 2.8, "Installing MySQL on Solaris." - A binary upgrade or downgrade is one that installs one version of - MySQL "in place" over an existing version, without dumping and - reloading tables: + On Solaris, you may run into trouble even before you get the MySQL + distribution unpacked, as the Solaris tar cannot handle long file + names. This means that you may see errors when you try to unpack + MySQL. - 1. Stop the server for the existing version if it is running. + If this occurs, you must use GNU tar (gtar) to unpack the + distribution. - 2. Install a different version of MySQL. This is an upgrade if - the new version is higher than the original version, a - downgrade if the version is lower. + If you have an UltraSPARC system, you can get 4% better + performance by adding -mcpu=v8 -Wa,-xarch=v8plusa to the CFLAGS + and CXXFLAGS environment variables. - 3. Start the server for the new version. + If you have Sun's Forte 5.0 (or newer) compiler, you can run + configure like this: +CC=cc CFLAGS="-Xa -fast -native -xstrconst -mt" \ +CXX=CC CXXFLAGS="-noex -mt" \ +./configure --prefix=/usr/local/mysql --enable-assembler - In many cases, the tables from the previous version of MySQL can - be used without change by the new version. However, sometimes - modifications are made to the handling of character sets or - collations that change the character sort order, which causes the - ordering of entries in any index that uses an affected character - set or collation to be incorrect. Such changes result in several - possible problems: + To create a 64-bit binary with Sun's Forte compiler, use the + following configuration options: +CC=cc CFLAGS="-Xa -fast -native -xstrconst -mt -xarch=v9" \ +CXX=CC CXXFLAGS="-noex -mt -xarch=v9" ASFLAGS="-xarch=v9" \ +./configure --prefix=/usr/local/mysql --enable-assembler - * Comparison results that differ from previous results + To create a 64-bit Solaris binary using gcc, add -m64 to CFLAGS + and CXXFLAGS and remove --enable-assembler from the configure + line. - * Inability to find some index values due to misordered index - entries + In the MySQL benchmarks, we obtained a 4% speed increase on + UltraSPARC when using Forte 5.0 in 32-bit mode, as compared to + using gcc 3.2 with the -mcpu flag. - * Misordered ORDER BY results + If you create a 64-bit mysqld binary, it is 4% slower than the + 32-bit binary, but can handle more threads and memory. - * Tables that CHECK TABLE reports as being in need of repair + When using Solaris 10 for x86_64, you should mount any file + systems on which you intend to store InnoDB files with the + forcedirectio option. (By default mounting is done without this + option.) Failing to do so will cause a significant drop in + performance when using the InnoDB storage engine on this platform. - The solution to these problems is to rebuild any indexes that use - an affected character set or collation, either by dropping and - re-creating the indexes, or by dumping and reloading the entire - table. For information about rebuilding indexes, see Section - 2.12.4, "Rebuilding or Repairing Tables or Indexes." + If you get a problem with fdatasync or sched_yield, you can fix + this by adding LIBS=-lrt to the configure line - To check whether a table has indexes that must be rebuilt, consult - the following list. It indicates which versions of MySQL - introduced character set or collation changes that require indexes - to be rebuilt. Each entry indicates the version in which the - change occurred and the character sets or collations that the - change affects. If the change is associated with a particular bug - report, the bug number is given. + Solaris does not provide static versions of all system libraries + (libpthreads and libdl), so you cannot compile MySQL with + --static. If you try to do so, you get one of the following + errors: +ld: fatal: library -ldl: not found +undefined reference to `dlopen' +cannot find -lrt + + If you link your own MySQL client programs, you may see the + following error at runtime: +ld.so.1: fatal: libmysqlclient.so.#: +open failed: No such file or directory + + This problem can be avoided by one of the following methods: + + * Link clients with the -Wl,r/full/path/to/libmysqlclient.so + flag rather than with -Lpath). - The list applies both for binary upgrades and downgrades. For - example, Bug#29461: http://bugs.mysql.com/29461 was fixed in MySQL - 5.0.48, so it applies to upgrades from versions older than 5.0.48 - to 5.0.48 or newer, and also to downgrades from 5.0.48 or newer to - versions older than 5.0.48. + * Copy libmysqclient.so to /usr/lib. - If you have tables with indexes that are affected, rebuild the - indexes using the instructions given in Section 2.12.4, - "Rebuilding or Repairing Tables or Indexes." + * Add the path name of the directory where libmysqlclient.so is + located to the LD_RUN_PATH environment variable before running + your client. - In many cases, you can use CHECK TABLE ... FOR UPGRADE to identify - tables for which index rebuilding is required. (It will report: - Table upgrade required. Please do "REPAIR TABLE `tbl_name`" or - dump/reload to fix it!) In these cases, you can also use - mysqlcheck --check-upgrade or mysql_upgrade, which execute CHECK - TABLE. However, the use of CHECK TABLE applies only after - upgrades, not downgrades. Also, CHECK TABLE is not applicable to - all storage engines. For details about which storage engines CHECK - TABLE supports, see Section 12.5.2.3, "CHECK TABLE Syntax." + If you have problems with configure trying to link with -lz when + you don't have zlib installed, you have two options: - Changes that cause index rebuilding to be necessary: + * If you want to be able to use the compressed communication + protocol, you need to get and install zlib from ftp.gnu.org. - * MySQL 5.0.48 (Bug#29461: http://bugs.mysql.com/29461) - Affects indexes for columns that use any of these character - sets: eucjpms, euc_kr, gb2312, latin7, macce, ujis - Affected tables can be detected by CHECK TABLE ... FOR UPGRADE - as of MySQL 5.1.29, 6.0.8 (see - Bug#39585: http://bugs.mysql.com/39585) + * Run configure with the --with-named-z-libs=no option when + building MySQL. - * MySQL 5.0.48 (Bug#27562: http://bugs.mysql.com/27562) - Affects indexes that use the ascii_general_ci collation for - columns that contain any of these characters: '`' GRAVE - ACCENT, '[' LEFT SQUARE BRACKET, '\' REVERSE SOLIDUS, ']' - RIGHT SQUARE BRACKET, '~' TILDE - Affected tables can be detected by CHECK TABLE ... FOR UPGRADE - as of MySQL 5.1.29, 6.0.8 (see - Bug#39585: http://bugs.mysql.com/39585) + If you are using gcc and have problems with loading user-defined + functions (UDFs) into MySQL, try adding -lgcc to the link line for + the UDF. - * MySQL 5.1.21 (Bug#29461: http://bugs.mysql.com/29461) - Affects indexes for columns that use any of these character - sets: eucjpms, euc_kr, gb2312, latin7, macce, ujis - Affected tables can be detected by CHECK TABLE ... FOR UPGRADE - as of MySQL 5.1.29, 6.0.8 (see - Bug#39585: http://bugs.mysql.com/39585) + If you would like MySQL to start automatically, you can copy + support-files/mysql.server to /etc/init.d and create a symbolic + link to it named /etc/rc3.d/S99mysql.server. - * MySQL 5.1.23 (Bug#27562: http://bugs.mysql.com/27562) - Affects indexes that use the ascii_general_ci collation for - columns that contain any of these characters: '`' GRAVE - ACCENT, '[' LEFT SQUARE BRACKET, '\' REVERSE SOLIDUS, ']' - RIGHT SQUARE BRACKET, '~' TILDE - Affected tables can be detected by CHECK TABLE ... FOR UPGRADE - as of MySQL 5.1.29, 6.0.8 (see - Bug#39585: http://bugs.mysql.com/39585) + If too many processes try to connect very rapidly to mysqld, you + should see this error in the MySQL log: +Error in accept: Protocol error - * MySQL 5.1.24 (Bug#27877: http://bugs.mysql.com/27877) - Affects indexes that use the utf8_general_ci or - ucs2_general_ci collation for columns that contain 'ß' LATIN - SMALL LETTER SHARP S (German). - Affected tables can be detected by CHECK TABLE ... FOR UPGRADE - as of MySQL 5.1.30, 6.0.8 (see - Bug#40053: http://bugs.mysql.com/40053) + You might try starting the server with the --back_log=50 option as + a workaround for this. (Use -O back_log=50 before MySQL 4.) - * * MySQL 6.0.1 (WL#3664) - Affects indexes that use the latin2_czech_cs collation. - Affected tables can be detected by CHECK TABLE ... FOR UPGRADE - as of MySQL 5.4.4, 6.0.9 (see - Bug#40054: http://bugs.mysql.com/40054) - MySQL 6.0.5 (Bug#33452: http://bugs.mysql.com/33452) - Affects indexes that use the latin2_czech_cs collation. - Affected tables can be detected by CHECK TABLE ... FOR UPGRADE - as of MySQL 5.4.4, 6.0.9 (see - Bug#40054: http://bugs.mysql.com/40054) + To configure the generation of core files on Solaris you should + use the coreadm command. Because of the security implications of + generating a core on a setuid() application, by default, Solaris + does not support core files on setuid() programs. However, you can + modify this behavior using coreadm. If you enable setuid() core + files for the current user, they will be generated using the mode + 600 and owned by the superuser. - * MySQL 6.0.5 (Bug#27877: http://bugs.mysql.com/27877) - Affects indexes that use the utf8_general_ci or - ucs2_general_ci collation for columns that contain 'ß' LATIN - SMALL LETTER SHARP S (German). - Affected tables can be detected by CHECK TABLE ... FOR UPGRADE - as of MySQL 6.0.8 (see - Bug#40053: http://bugs.mysql.com/40053) +2.9. Installing MySQL on i5/OS - * MySQL 6.0.6 (Bug#25420: http://bugs.mysql.com/25420) - Affects indexes for columns that use the following collations, - if the columns contain the indicated characters: - big5_chinese_ci: '~' TILDE or '`' GRAVE ACCENT; - cp866_general_ci: j LATIN SMALL LETTER J; gb2312_chinese_ci: - '~' TILDE; gbk_chinese_ci: '~' TILDE - Affected tables can be detected by CHECK TABLE ... FOR UPGRADE - as of MySQL 5.4.4, 6.0.9 (see - Bug#40054: http://bugs.mysql.com/40054) + The i5/OS POWER MySQL package was created in cooperation with IBM. + MySQL works within the Portable Application Solution Environment + (PASE) on the System i series of hardware and will also provide + database services for the Zend Core for i5/OS. -2.12.4. Rebuilding or Repairing Tables or Indexes + MySQL for i5/OS is provided both as a tar file and as a save file + (.savf) package that can be downloaded and installed directly + without any additional installation steps required. To install + MySQL using the tar file, see Section 2.2, "Installing MySQL from + Generic Binaries on Unix/Linux." - This section describes how to rebuild a table. This can be - necessitated by changes to MySQL such as how data types are - handled or changes to character set handling. For example, an - error in a collation might have been corrected, necessitating a - table rebuild to rebuild the indexes for character columns that - use the collation. It might also be that a table repair or upgrade - should be done as indicated by a table check operation such as - that performed by CHECK TABLE, mysqlcheck, or mysql_upgrade. + MySQL is only supported on i5/OS V5R4 or later releases. The i5/OS + PASE must be installed for MySQL to operate. You must be able to + login as a user in *SECOFR class. - Methods for rebuilding a table include dumping and reloading it, - or using ALTER TABLE or REPAIR TABLE. + You should the installation notes and tips for i5/OS before + starting installation. See i5/OS Installation Notes. + + Before Installation: Note - If you are rebuilding tables because a different version of MySQL - will not handle them after a binary upgrade or downgrade, you must - use the dump-and-reload method. Dump the tables before upgrading - or downgrading (using your original version of MySQL), and reload - the tables after upgrading or downgrading (after installing the - new version). + The installation package will use an existing configuration if you + have previously installed MySQL (which is identified by looking + for the file /etc/my.cnf). The values for the data directory + (DATADIR) and owner of the MySQL files (USRPRF) specified during + the installation will be ignored, and the values determined from + the /etc/my.cnf will be used instead. - If you use the dump-and-reload method of rebuilding tables only - for the purpose of rebuilding indexes, you can perform the dump - either before or after upgrading or downgrading. Reloading still - must be done afterward. + If you want to change these parameters during a new install, you + should temporarily rename /etc/my.cnf, install MySQL using the new + parameters you want to use, and then merge your previous + /etc/my.cnf configuration settings with the new /etc/my.cnf file + that is created during installation. - To re-create a table by dumping and reloading it, use mysqldump to - create a dump file and mysql to reload the file: -shell> mysqldump db_name t1 > dump.sql -shell> mysql db_name < dump.sql + * You must have a user profile with PASE with suitable + privileges. The user should be within the *SECOFR class, such + as the QSECOFR user ID. You can use the WRKUSRPRF command to + check your user profile. + + * For network connections to MySQL, you must have TCP/IP + enabled. You should also check the following: + + + Ensure that a name has defined for the system. Run the + Configure TCP/IP (CFGTCP) command and select option 12 + (Change TCP/IP domain information) to display this + setting. Make sure that a value is listed in the Host + name field. - To recreate all the tables in a single database, specify the - database name without any following table name: -shell> mysqldump db_name > dump.sql -shell> mysql db_name < dump.sql + + Make sure that the system has a loopback entry which + represents the localhost or 127.0.0.1. - To recreate all tables in all databases, use the --all-databases - option: -shell> mysqldump --all-databases > dump.sql -shell> mysql < dump.sql + + Ensure that the IP address of the IBM i machine is mapped + correctly to the host name. - To rebuild a table with ALTER TABLE, use a statement that - "changes" the table to use the storage engine that it already has. - For example, if t1 is a MyISAM table, use this statement: -mysql> ALTER TABLE t1 ENGINE = MyISAM; + To install MySQL on i5/OS, follow these steps: - If you are not sure which storage engine to specify in the ALTER - TABLE statement, use SHOW CREATE TABLE to display the table - definition. + 1. On the System i machine, create a save file that will be used + to receive the downloaded installation save file. The file + should be located within the General Purpose Library (QGPL): +CRTSAVF FILE(QGPL/MYSQLINST) TESXT('MySQL Save file') - If you must rebuild a table because a table checking operation - indicates that the table is corrupt or needs an upgrade, you can - use REPAIR TABLE if that statement supports the table's storage - engine. For example, to repair a MyISAM table, use this statement: -mysql> REPAIR TABLE t1; + 2. Download the MySQL installation save file in 32-bit + (mysql-5.1.39-i5os-power-32bit.savf) or 64-bit + (mysql-5.1.39-i5os-power-64bit.savf) from MySQL Downloads + (http://dev.mysql.com/downloads) - For storage engines such as InnoDB that REPAIR TABLE does not - support, use mysqldump to create a dump file and mysql to reload - the file, as described earlier. + 3. You need to FTP the downloaded .savf file directly into the + QGPL/MYSQLINST file on the System i server. You can do this + through FTP using the following steps after logging in to the + System i machine: +ftp> bin +ftp> cd qgpl +ftp> put mysql-5.1.39-i5os-power.savf mysqlinst - For specifics about which storage engines REPAIR TABLE supports, - see Section 12.5.2.6, "REPAIR TABLE Syntax." + 4. Log into the System i server using a user in the *SECOFR + class, such as the QSECOFR user ID. -2.12.5. Copying MySQL Databases to Another Machine + 5. You need to restore the installation library stored in the + .savf save file: +RSTLIB MYSQLINST DEV(*SAVF) SAVF(QGPL/MYSQLINST) MBROPT(*ALL) ALWOBJD +IF(*ALL) - You can copy the .frm, .MYI, and .MYD files for MyISAM tables - between different architectures that support the same - floating-point format. (MySQL takes care of any byte-swapping - issues.) See Section 13.5, "The MyISAM Storage Engine." +Note + You can ignore the security changes-type message at the bottom + of the installation panel. - In cases where you need to transfer databases between different - architectures, you can use mysqldump to create a file containing - SQL statements. You can then transfer the file to the other - machine and feed it as input to the mysql client. + 6. Once you have finished restoring the MYSQLINST library, check + that all the necessary objects for installation are on the + system by using the Display Library (DSPLIB) command: +DSPLIB LIB(MYSQLINST) - Use mysqldump --help to see what options are available. + 7. You need to execute the installation command, + MYSQLINST/INSMYSQL. You can specify three parameter settings + during installation: - The easiest (although not the fastest) way to move a database - between two machines is to run the following commands on the - machine on which the database is located: -shell> mysqladmin -h 'other_hostname' create db_name -shell> mysqldump db_name | mysql -h 'other_hostname' db_name + + DIR('/QOpenSys/usr/local/mysql') sets the installation + location for the MySQL files. The directory will be + created if it does not already exist. + + + DATADIR('/QOpenSys/usr/local/mysql/data') sets the + location of the directory that will be used to store the + database files and binary logs. The default setting is + /QOpenSys/usr/local/mysql/data. Note that if the + installer detects an existing installation (due to the + existence of /etc/my.cnf), then the existing setting will + be used instead of the default. - If you want to copy a database from a remote machine over a slow - network, you can use these commands: -shell> mysqladmin create db_name -shell> mysqldump -h 'other_hostname' --compress db_name | mysql db_na -me + + USRPRF(MYSQL) sets the user profile that will own the + files that are installed. The profile will be created if + it does not already exist. - You can also store the dump in a file, transfer the file to the - target machine, and then load the file into the database there. - For example, you can dump a database to a compressed file on the - source machine like this: -shell> mysqldump --quick db_name | gzip > db_name.gz +Note + You should choose an appropriate user for using the MySQL + server installation. The user will be used whenever you + need to do any administration on the MySQL server. + Once you have set the appropriate parameters, you can begin + the installation. + The installation copies all the necessary files into a + directory matching the DIR configuration value; sets the + ownership on those files, sets up the MySQL environment and + creates the MySQL configuration file (in /etc/my.cnf) + completing all the steps in a typical binary installation + process automatically. If this is a new installation of MySQL, + or if the installer detects that this is a new version + (because the /etc/my.cnf file does not exist), then the + initial core MySQL databases will also be created during + installation. + Once the installation has been completed, you will get a + notice advising you to set the password for the root user. For + more information, Section 2.13, "Post-Installation Setup and + Testing." - Transfer the file containing the database contents to the target - machine and run these commands there: -shell> mysqladmin create db_name -shell> gunzip < db_name.gz | mysql db_name + 8. Once the installation has completed, you can delete the + installation file: +DLTLIB LIB(MYSQLINST) - You can also use mysqldump and mysqlimport to transfer the - database. For large tables, this is much faster than simply using - mysqldump. In the following commands, DUMPDIR represents the full - path name of the directory you use to store the output from - mysqldump. + Upgrading an existing MySQL instance - First, create the directory for the output files and dump the - database: -shell> mkdir DUMPDIR -shell> mysqldump --tab=DUMPDIR db_name + You need to execute the upgrade command, MYSQLINST/UPGMYSQL. You + must specify 6 parameters to perform an upgrade: - Then transfer the files in the DUMPDIR directory to some - corresponding directory on the target machine and load the files - into MySQL there: -shell> mysqladmin create db_name # create database -shell> cat DUMPDIR/*.sql | mysql db_name # create tables in databas -e -shell> mysqlimport db_name DUMPDIR/*.txt # load data into tables + * DIR('/QOpenSys/usr/local/') --- sets the installation location + for the MySQL files. The directory will be created if it does + not already exist. This is the directory that the MySQL server + will be installed into, inside a directory with a name + matching the version and release. For example if installing + MySQL 5.1.39 with the DIR set to /QOpenSys/usr/local/ would + result in /QOpenSys/usr/local/mysql-5.1.39-i5os-power64 and a + symbolic link to this directory will be created in + /QOpenSys/usr/local/mysql. + + * DATADIR('/QOpenSys/mysql/data') --- sets the location of the + directory that will be upgraded. + + * USRPRF('MYSQL') --- sets the user profile that will own the + files that are installed. The profile will be created if it + does not already exist; if it is created as part of the + upgrade process, it will be disabled initially. You may wish + to enable this user profile so that it can be used to start + the MySQL server later. It is best practice to use the one + previously created during the first installation. + + * MYSQLUSR('root user') --- any user account in the current + MySQL server with SUPER privileges. + + * PASSWORD('root user password') --- the password for the above + account. This is necessary as the upgrade starts the MySQL + server to upgrade the tables and the password is need to be + able to shutdown the MySQL server. + + * CURINST('path to previous install') --- the full path to the + installation that is being upgraded. For example an + installation in /QOpenSys/usr/local/ will be + /QOpenSys/usr/local/msyql-5.1.30-i5os-power64. Failure to + specify this option may result in corruption of your existing + data files. + + For example: +MYSQLINST/UPGMYSQL DIR('/QOpenSys/usr/local/') DATADIR('/QOpenSys/mys +ql/data') » + USERPRF(MYSQL) MYSQLUSR('root') PASSWORD('root') CURINST('/QOpen +Sys/usr/local/mysql-5.1.30-i5os-power64') + + You should receive a Program Message indicating UPGRADE + SUCCESSFUL! upon completion or an error message if there is a + problem.You can view the upgrade programs progression and the + error in the text file upgrade.log in the installation directory. - Do not forget to copy the mysql database because that is where the - grant tables are stored. You might have to run commands as the - MySQL root user on the new machine until you have the mysql - database in place. + To start MySQL: - After you import the mysql database on the new machine, execute - mysqladmin flush-privileges so that the server reloads the grant - table information. + 1. Log into the System i server using the user profile create or + specified during installation. By default, this is MYSQL. -2.13. Operating System-Specific Notes +Note + You should start mysqld_safe using a user that in the PASE + environment has the id=0 (the equivalent of the standard Unix + root user). If you do not use a user with this ID then the + system will be unable to change the user when executing mysqld + as set using --user option. If this happens, mysqld may be + unable to read the files located within the MySQL data + directory and the execution will fail. -2.13.1. Linux Notes + 2. Enter the PASE environment using call qp2term. - This section discusses issues that have been found to occur on - Linux. The first few subsections describe general operating - system-related issues, problems that can occur when using binary - or source distributions, and post-installation issues. The - remaining subsections discuss problems that occur with Linux on - specific platforms. + 3. Start the MySQL server by changing to the installation + directory and running mysqld_safe, specifying the user name + used to install the server. The installer conveniently + installs a symbolic link to the installation directory + (mysql-5.0.42-i5os-power-32bit) as /opt/mysql/mysql: +> cd /opt/mysql/mysql +> bin/mysqld_safe --user=mysql & + You should see a message similar to the following: +Starting mysqld daemon with databases » + from /opt/mysql/mysql-enterprise-5.0.42-i5os-power-32bit/data - Note that most of these problems occur on older versions of Linux. - If you are running a recent version, you may see none of them. + If you are having problems starting MySQL server, see Section + 2.13.1.3, "Starting and Troubleshooting the MySQL Server." -2.13.1.1. Linux Operating System Notes + To stop MySQL: - MySQL needs at least Linux version 2.0. + 1. Log into the System i server using the user profile create or + specified during installation. By default, this is MYSQL. -Warning + 2. Enter the PASE environment using call qp2term. - We have seen some strange problems with Linux 2.2.14 and MySQL on - SMP systems. We also have reports from some MySQL users that they - have encountered serious stability problems using MySQL with - kernel 2.2.14. If you are using this kernel, you should upgrade to - 2.2.19 (or newer) or to a 2.4 kernel. If you have a multiple-CPU - box, you should seriously consider using 2.4 because it gives you - a significant speed boost. Your system should be more stable. - - When using LinuxThreads, you should see a minimum of three mysqld - processes running. These are in fact threads. There is one thread - for the LinuxThreads manager, one thread to handle connections, - and one thread to handle alarms and signals. - -2.13.1.2. Linux Binary Distribution Notes - - The Linux-Intel binary and RPM releases of MySQL are configured - for the highest possible speed. We are always trying to use the - fastest stable compiler available. - - The binary release is linked with -static, which means you do not - normally need to worry about which version of the system libraries - you have. You need not install LinuxThreads, either. A program - linked with -static is slightly larger than a dynamically linked - program, but also slightly faster (3-5%). However, one problem - with a statically linked program is that you can't use - user-defined functions (UDFs). If you are going to write or use - UDFs (this is something for C or C++ programmers only), you must - compile MySQL yourself using dynamic linking. - - A known issue with binary distributions is that on older Linux - systems that use libc (such as Red Hat 4.x or Slackware), you get - some (nonfatal) issues with host name resolution. If your system - uses libc rather than glibc2, you probably will encounter some - difficulties with host name resolution and getpwnam(). This - happens because glibc (unfortunately) depends on some external - libraries to implement host name resolution and getpwent(), even - when compiled with -static. These problems manifest themselves in - two ways: - - * You may see the following error message when you run - mysql_install_db: -Sorry, the host 'xxxx' could not be looked up - You can deal with this by executing mysql_install_db --force, - which does not execute the resolveip test in mysql_install_db. - The downside is that you cannot use host names in the grant - tables: except for localhost, you must use IP numbers instead. - If you are using an old version of MySQL that does not support - --force, you must manually remove the resolveip test in - mysql_install_db using a text editor. - - * You also may see the following error when you try to run - mysqld with the --user option: -getpwnam: No such file or directory - To work around this problem, start mysqld by using the su - command rather than by specifying the --user option. This - causes the system itself to change the user ID of the mysqld - process so that mysqld need not do so. - - Another solution, which solves both problems, is not to use a - binary distribution. Obtain a MySQL source distribution (in RPM or - tar.gz format) and install that instead. - - On some Linux 2.2 versions, you may get the error Resource - temporarily unavailable when clients make a great many new - connections to a mysqld server over TCP/IP. The problem is that - Linux has a delay between the time that you close a TCP/IP socket - and the time that the system actually frees it. There is room for - only a finite number of TCP/IP slots, so you encounter the - resource-unavailable error if clients attempt too many new TCP/IP - connections over a short period of time. For example, you may see - the error when you run the MySQL test-connect benchmark over - TCP/IP. - - We have inquired about this problem a few times on different Linux - mailing lists but have never been able to find a suitable - resolution. The only known "fix" is for clients to use persistent - connections, or, if you are running the database server and - clients on the same machine, to use Unix socket file connections - rather than TCP/IP connections. - -2.13.1.3. Linux Source Distribution Notes - - The following notes regarding glibc apply only to the situation - when you build MySQL yourself. If you are running Linux on an x86 - machine, in most cases it is much better for you to use our - binary. We link our binaries against the best patched version of - glibc we can find and with the best compiler options, in an - attempt to make it suitable for a high-load server. For a typical - user, even for setups with a lot of concurrent connections or - tables exceeding the 2GB limit, our binary is the best choice in - most cases. After reading the following text, if you are in doubt - about what to do, try our binary first to determine whether it - meets your needs. If you discover that it is not good enough, you - may want to try your own build. In that case, we would appreciate - a note about it so that we can build a better binary next time. - - MySQL uses LinuxThreads on Linux. If you are using an old Linux - version that doesn't have glibc2, you must install LinuxThreads - before trying to compile MySQL. You can obtain LinuxThreads from - http://dev.mysql.com/downloads/os-linux.html. - - Note that glibc versions before and including version 2.1.1 have a - fatal bug in pthread_mutex_timedwait() handling, which is used - when INSERT DELAYED statements are issued. Do not use INSERT - DELAYED before upgrading glibc. - - Note that Linux kernel and the LinuxThread library can by default - handle a maximum of 1,024 threads. If you plan to have more than - 1,000 concurrent connections, you need to make some changes to - LinuxThreads, as follows: - - * Increase PTHREAD_THREADS_MAX in - sysdeps/unix/sysv/linux/bits/local_lim.h to 4096 and decrease - STACK_SIZE in linuxthreads/internals.h to 256KB. The paths are - relative to the root of glibc. (Note that MySQL is not stable - with 600-1000 connections if STACK_SIZE is the default of - 2MB.) - - * Recompile LinuxThreads to produce a new libpthread.a library, - and relink MySQL against it. - - There is another issue that greatly hurts MySQL performance, - especially on SMP systems. The mutex implementation in - LinuxThreads in glibc 2.1 is very poor for programs with many - threads that hold the mutex only for a short time. This produces a - paradoxical result: If you link MySQL against an unmodified - LinuxThreads, removing processors from an SMP actually improves - MySQL performance in many cases. We have made a patch available - for glibc 2.1.3 to correct this behavior - (http://dev.mysql.com/Downloads/Linux/linuxthreads-2.1-patch) - - With glibc 2.2.2, MySQL uses the adaptive mutex, which is much - better than even the patched one in glibc 2.1.3. Be warned, - however, that under some conditions, the current mutex code in - glibc 2.2.2 overspins, which hurts MySQL performance. The - likelihood that this condition occurs can be reduced by re-nicing - the mysqld process to the highest priority. We have also been able - to correct the overspin behavior with a patch, available at - http://dev.mysql.com/Downloads/Linux/linuxthreads-2.2.2.patch. It - combines the correction of overspin, maximum number of threads, - and stack spacing all in one. You need to apply it in the - linuxthreads directory with patch -p0 - </tmp/linuxthreads-2.2.2.patch. We hope it is included in some - form in future releases of glibc 2.2. In any case, if you link - against glibc 2.2.2, you still need to correct STACK_SIZE and - PTHREAD_THREADS_MAX. We hope that the defaults is corrected to - some more acceptable values for high-load MySQL setup in the - future, so that the commands needed to produce your own build can - be reduced to ./configure; make; make install. - - If you use these patches to build a special static version of - libpthread.a, use it only for statically linking against MySQL. We - know that these patches are safe for MySQL and significantly - improve its performance, but we cannot say anything about their - effects on other applications. If you link other applications that - require LinuxThreads against the patched static version of the - library, or build a patched shared version and install it on your - system, you do so at your own risk. - - If you experience any strange problems during the installation of - MySQL, or with some common utilities hanging, it is very likely - that they are either library or compiler related. If this is the - case, using our binary resolves them. + 3. Stop the MySQL server by changing into the installation + directory and running mysqladmin, specifying the user name + used to install the server: +> cd /opt/mysql/mysql +> bin/mysqladmin -u root shutdown + If the session that you started and stopped MySQL are the + same, you may get the log output from mysqld: + STOPPING server from pid file » + /opt/mysql/mysql-enterprise-5.0.42-i5os-power-32bit/data/I5DBX.R +CHLAND.IBM.COM.pid + 070718 10:34:20 mysqld ended + If the sessions used to start and stop MySQL are different, + you will not receive any confirmation of the shutdown. - If you link your own MySQL client programs, you may see the - following error at runtime: -ld.so.1: fatal: libmysqlclient.so.#: -open failed: No such file or directory + Note and tips - This problem can be avoided by one of the following methods: + * A problem has been identified with the installation process on + DBCS systems. If you are having problems install MySQL on a + DBCS system, you need to change your job's coded character set + identifier (CSSID) to 37 (EBCDIC) before executing the install + command, INSMYSQL. To do this, determine your existing CSSID + (using DSPJOB and selecting option 2), execute CHGJOB + CSSID(37), run INSMYSQL to install MySQL and then execute + CHGJOB again with your original CSSID. - * Link clients with the -Wl,r/full/path/to/libmysqlclient.so - flag rather than with -Lpath). + * If you want to use the Perl scripts that are included with + MySQL, you need to download the iSeries Tools for Developers + (5799-PTL). See + http://www-03.ibm.com/servers/enable/site/porting/tools/. - * Copy libmysqclient.so to /usr/lib. +2.10. Installing MySQL on FreeBSD - * Add the path name of the directory where libmysqlclient.so is - located to the LD_RUN_PATH environment variable before running - your client. + This section provides information about using MySQL on variants of + FreeBSD Unix. - If you are using the Fujitsu compiler (fcc/FCC), you may have some - problems compiling MySQL because the Linux header files are very - gcc oriented. The following configure line should work with - fcc/FCC: -CC=fcc CFLAGS="-O -K fast -K lib -K omitfp -Kpreex -D_GNU_SOURCE \ - -DCONST=const -DNO_STRTOLL_PROTO" \ -CXX=FCC CXXFLAGS="-O -K fast -K lib \ - -K omitfp -K preex --no_exceptions --no_rtti -D_GNU_SOURCE \ - -DCONST=const -Dalloca=__builtin_alloca -DNO_STRTOLL_PROTO \ - '-D_EXTERN_INLINE=static __inline'" \ -./configure \ - --prefix=/usr/local/mysql --enable-assembler \ - --with-mysqld-ldflags=-all-static --disable-shared \ - --with-low-memory + The easiest (and preferred) way to install MySQL is to use the + mysql-server and mysql-client ports available at + http://www.freebsd.org/. Using these ports gives you the following + benefits: -2.13.1.4. Linux Post-Installation Notes + * A working MySQL with all optimizations enabled that are known + to work on your version of FreeBSD. - mysql.server can be found in the support-files directory under the - MySQL installation directory or in a MySQL source tree. You can - install it as /etc/init.d/mysql for automatic MySQL startup and - shutdown. See Section 2.11.2.2, "Starting and Stopping MySQL - Automatically." + * Automatic configuration and build. - If MySQL cannot open enough files or connections, it may be that - you have not configured Linux to handle enough files. + * Startup scripts installed in /usr/local/etc/rc.d. - In Linux 2.2 and onward, you can check the number of allocated - file handles as follows: -shell> cat /proc/sys/fs/file-max -shell> cat /proc/sys/fs/dquot-max -shell> cat /proc/sys/fs/super-max - - If you have more than 16MB of memory, you should add something - like the following to your init scripts (for example, - /etc/init.d/boot.local on SuSE Linux): -echo 65536 > /proc/sys/fs/file-max -echo 8192 > /proc/sys/fs/dquot-max -echo 1024 > /proc/sys/fs/super-max - - You can also run the echo commands from the command line as root, - but these settings are lost the next time your computer restarts. - - Alternatively, you can set these parameters on startup by using - the sysctl tool, which is used by many Linux distributions - (including SuSE Linux 8.0 and later). Put the following values - into a file named /etc/sysctl.conf: -# Increase some values for MySQL -fs.file-max = 65536 -fs.dquot-max = 8192 -fs.super-max = 1024 - - You should also add the following to /etc/my.cnf: -[mysqld_safe] -open-files-limit=8192 - - This should allow the server a limit of 8,192 for the combined - number of connections and open files. - - The STACK_SIZE constant in LinuxThreads controls the spacing of - thread stacks in the address space. It needs to be large enough so - that there is plenty of room for each individual thread stack, but - small enough to keep the stack of some threads from running into - the global mysqld data. Unfortunately, as we have experimentally - discovered, the Linux implementation of mmap() successfully unmaps - a mapped region if you ask it to map out an address currently in - use, zeroing out the data on the entire page instead of returning - an error. So, the safety of mysqld or any other threaded - application depends on the "gentlemanly" behavior of the code that - creates threads. The user must take measures to make sure that the - number of running threads at any given time is sufficiently low - for thread stacks to stay away from the global heap. With mysqld, - you should enforce this behavior by setting a reasonable value for - the max_connections variable. - - If you build MySQL yourself, you can patch LinuxThreads for better - stack use. See Section 2.13.1.3, "Linux Source Distribution - Notes." If you do not want to patch LinuxThreads, you should set - max_connections to a value no higher than 500. It should be even - less if you have a large key buffer, large heap tables, or some - other things that make mysqld allocate a lot of memory, or if you - are running a 2.2 kernel with a 2GB patch. If you are using our - binary or RPM version, you can safely set max_connections at 1500, - assuming no large key buffer or heap tables with lots of data. The - more you reduce STACK_SIZE in LinuxThreads the more threads you - can safely create. Values between 128KB and 256KB are recommended. - - If you use a lot of concurrent connections, you may suffer from a - "feature" in the 2.2 kernel that attempts to prevent fork bomb - attacks by penalizing a process for forking or cloning a child. - This causes MySQL not to scale well as you increase the number of - concurrent clients. On single-CPU systems, we have seen this - manifest as very slow thread creation; it may take a long time to - connect to MySQL (as long as one minute), and it may take just as - long to shut it down. On multiple-CPU systems, we have observed a - gradual drop in query speed as the number of clients increases. In - the process of trying to find a solution, we have received a - kernel patch from one of our users who claimed it helped for his - site. This patch is available at - http://dev.mysql.com/Downloads/Patches/linux-fork.patch. We have - done rather extensive testing of this patch on both development - and production systems. It has significantly improved MySQL - performance without causing any problems and is recommended for - users who still run high-load servers on 2.2 kernels. - - This issue has been fixed in the 2.4 kernel, so if you are not - satisfied with the current performance of your system, rather than - patching your 2.2 kernel, it might be easier to upgrade to 2.4. On - SMP systems, upgrading also gives you a nice SMP boost in addition - to fixing the fairness bug. - - We have tested MySQL on the 2.4 kernel on a two-CPU machine and - found MySQL scales much better. There was virtually no slowdown on - query throughput all the way up to 1,000 clients, and the MySQL - scaling factor (computed as the ratio of maximum throughput to the - throughput for one client) was 180%. We have observed similar - results on a four-CPU system: Virtually no slowdown as the number - of clients was increased up to 1,000, and a 300% scaling factor. - Based on these results, for a high-load SMP server using a 2.2 - kernel, it is definitely recommended to upgrade to the 2.4 kernel - at this point. - - We have discovered that it is essential to run the mysqld process - with the highest possible priority on the 2.4 kernel to achieve - maximum performance. This can be done by adding a renice -20 $$ - command to mysqld_safe. In our testing on a four-CPU machine, - increasing the priority resulted in a 60% throughput increase with - 400 clients. - - We are currently also trying to collect more information on how - well MySQL performs with a 2.4 kernel on four-way and eight-way - systems. If you have access such a system and have done some - benchmarks, please send an email message to benchmarks(a)mysql.com - with the results. We will review them for inclusion in the manual. - - If you see a dead mysqld server process with ps, this usually - means that you have found a bug in MySQL or you have a corrupted - table. See Section B.1.4.2, "What to Do If MySQL Keeps Crashing." - - To get a core dump on Linux if mysqld dies with a SIGSEGV signal, - you can start mysqld with the --core-file option. Note that you - also probably need to raise the core file size by adding ulimit -c - 1000000 to mysqld_safe or starting mysqld_safe with - --core-file-size=1000000. See Section 4.3.2, "mysqld_safe --- - MySQL Server Startup Script." - -2.13.1.5. Linux x86 Notes - - MySQL requires libc 5.4.12 or newer. It is known to work with libc - 5.4.46. glibc 2.0.6 and later should also work. There have been - some problems with the glibc RPMs from Red Hat, so if you have - problems, check whether there are any updates. The glibc 2.0.7-19 - and 2.0.7-29 RPMs are known to work. - - If you are using Red Hat 8.0 or a new glibc 2.2.x library, you may - see mysqld die in gethostbyaddr(). This happens because the new - glibc library requires a stack size greater than 128KB for this - call. To fix the problem, start mysqld with the - --thread-stack=192K option. (Use -O thread_stack=192K before MySQL - 4.) This stack size is the default on MySQL 4.0.10 and above, so - you should not see the problem. - - If you are using gcc 3.0 and above to compile MySQL, you must - install the libstdc++v3 library before compiling MySQL; if you - don't do this, you get an error about a missing __cxa_pure_virtual - symbol during linking. - - On some older Linux distributions, configure may produce an error - like this: -Syntax error in sched.h. Change _P to __P in the -/usr/include/sched.h file. -See the Installation chapter in the Reference Manual. - - Just do what the error message says. Add an extra underscore to - the _P macro name that has only one underscore, and then try - again. - - You may get some warnings when compiling. Those shown here can be - ignored: -mysqld.cc -o objs-thread/mysqld.o -mysqld.cc: In function `void init_signals()': -mysqld.cc:315: warning: assignment of negative value `-1' to -`long unsigned int' -mysqld.cc: In function `void * signal_hand(void *)': -mysqld.cc:346: warning: assignment of negative value `-1' to -`long unsigned int' - - If mysqld always dumps core when it starts, the problem may be - that you have an old /lib/libc.a. Try renaming it, and then remove - sql/mysqld and do a new make install and try again. This problem - has been reported on some Slackware installations. - - If you get the following error when linking mysqld, it means that - your libg++.a is not installed correctly: -/usr/lib/libc.a(putc.o): In function `_IO_putc': -putc.o(.text+0x0): multiple definition of `_IO_putc' - - You can avoid using libg++.a by running configure like this: -shell> CXX=gcc ./configure - -2.13.1.6. Linux SPARC Notes - - In some implementations, readdir_r() is broken. The symptom is - that the SHOW DATABASES statement always returns an empty set. - This can be fixed by removing HAVE_READDIR_R from config.h after - configuring and before compiling. - -2.13.1.7. Linux Alpha Notes - - We have tested MySQL 5.1 on Alpha with our benchmarks and test - suite, and it appears to work well. - - We currently build the MySQL binary packages on SuSE Linux 7.0 for - AXP, kernel 2.4.4-SMP, Compaq C compiler (V6.2-505) and Compaq C++ - compiler (V6.3-006) on a Compaq DS20 machine with an Alpha EV6 - processor. + * The ability to use pkg_info -L to see which files are + installed. - You can find the preceding compilers at - http://www.support.compaq.com/alpha-tools/. By using these - compilers rather than gcc, we get about 9-14% better MySQL - performance. - - For MySQL on Alpha, we use the -arch generic flag to our compile - options, which ensures that the binary runs on all Alpha - processors. We also compile statically to avoid library problems. - The configure command looks like this: -CC=ccc CFLAGS="-fast -arch generic" CXX=cxx \ -CXXFLAGS="-fast -arch generic -noexceptions -nortti" \ -./configure --prefix=/usr/local/mysql --disable-shared \ - --with-extra-charsets=complex --enable-thread-safe-client \ - --with-mysqld-ldflags=-non_shared --with-client-ldflags=-non_shar -ed + * The ability to use pkg_delete to remove MySQL if you no longer + want it on your machine. - Some known problems when running MySQL on Linux-Alpha: + The MySQL build process requires GNU make (gmake) to work. If GNU + make is not available, you must install it first before compiling + MySQL. - * Debugging threaded applications like MySQL does not work with - gdb 4.18. You should use gdb 5.1 instead. + The recommended way to compile and install MySQL on FreeBSD with + gcc (2.95.2 and up) is: +CC=gcc CFLAGS="-O2 -fno-strength-reduce" \ + CXX=gcc CXXFLAGS="-O2 -fno-rtti -fno-exceptions \ + -felide-constructors -fno-strength-reduce" \ + ./configure --prefix=/usr/local/mysql --enable-assembler +gmake +gmake install +cd /usr/local/mysql +bin/mysql_install_db --user=mysql +bin/mysqld_safe & - * If you try linking mysqld statically when using gcc, the - resulting image dumps core at startup time. In other words, do - not use --with-mysqld-ldflags=-all-static with gcc. - -2.13.1.8. Linux PowerPC Notes - - MySQL should work on MkLinux with the newest glibc package (tested - with glibc 2.0.7). - -2.13.1.9. Linux MIPS Notes - - To get MySQL to work on Qube2 (Linux Mips), you need the newest - glibc libraries. glibc-2.0.7-29C2 is known to work. You must also - use gcc 2.95.2 or newer). - -2.13.1.10. Linux IA-64 Notes - - To get MySQL to compile on Linux IA-64, we use the following - configure command for building with gcc 2.96: -CC=gcc \ -CFLAGS="-O3 -fno-omit-frame-pointer" \ -CXX=gcc \ -CXXFLAGS="-O3 -fno-omit-frame-pointer -felide-constructors \ - -fno-exceptions -fno-rtti" \ - ./configure --prefix=/usr/local/mysql \ - "--with-comment=Official MySQL binary" \ - --with-extra-charsets=complex - - On IA-64, the MySQL client binaries use shared libraries. This - means that if you install our binary distribution at a location - other than /usr/local/mysql, you need to add the path of the - directory where you have libmysqlclient.so installed either to the - /etc/ld.so.conf file or to the value of your LD_LIBRARY_PATH - environment variable. - - See Section B.1.3.1, "Problems Linking to the MySQL Client - Library." - -2.13.1.11. SELinux Notes - - RHEL4 comes with SELinux, which supports tighter access control - for processes. If SELinux is enabled (SELINUX in - /etc/selinux/config is set to enforcing, SELINUXTYPE is set to - either targeted or strict), you might encounter problems - installing Sun Microsystems, Inc. RPM packages. - - Red Hat has an update that solves this. It involves an update of - the "security policy" specification to handle the install - structure of the RPMs provided by Sun Microsystems, Inc. For - further information, see - https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=167551 and - http://rhn.redhat.com/errata/RHBA-2006-0049.html. - - The preceding discussion applies only to RHEL4. The patch is - unnecessary for RHEL5. - -2.13.2. Mac OS X Notes - - On Mac OS X, tar cannot handle long file names. If you need to - unpack a .tar.gz distribution, use gnutar instead. - -2.13.2.1. Mac OS X 10.x (Darwin) - - MySQL should work without major problems on Mac OS X 10.x - (Darwin). - - Known issues: - - * If you have problems with performance under heavy load, try - using the --skip-thread-priority option to mysqld. This runs - all threads with the same priority. On Mac OS X, this gives - better performance, at least until Apple fixes its thread - scheduler. - - * The connection times (wait_timeout, interactive_timeout and - net_read_timeout) values are not honored. - This is probably a signal handling problem in the thread - library where the signal doesn't break a pending read and we - hope that a future update to the thread libraries will fix - this. - - Our binary for Mac OS X is compiled on Darwin 6.3 with the - following configure line: -CC=gcc CFLAGS="-O3 -fno-omit-frame-pointer" CXX=gcc \ -CXXFLAGS="-O3 -fno-omit-frame-pointer -felide-constructors \ - -fno-exceptions -fno-rtti" \ - ./configure --prefix=/usr/local/mysql \ - --with-extra-charsets=complex --enable-thread-safe-client \ - --enable-local-infile --disable-shared - - See Section 2.5, "Installing MySQL on Mac OS X." - -2.13.2.2. Mac OS X Server 1.2 (Rhapsody) - - For current versions of Mac OS X Server, no operating system - changes are necessary before compiling MySQL. Compiling for the - Server platform is the same as for the client version of Mac OS X. + FreeBSD is known to have a very low default file handle limit. See + Section B.5.2.18, "'File' Not Found and Similar Errors." Start the + server by using the --open-files-limit option for mysqld_safe, or + raise the limits for the mysqld user in /etc/login.conf and + rebuild it with cap_mkdb /etc/login.conf. Also be sure that you + set the appropriate class for this user in the password file if + you are not using the default (use chpass mysqld-user-name). See + Section 4.3.2, "mysqld_safe --- MySQL Server Startup Script." - For older versions (Mac OS X Server 1.2, a.k.a. Rhapsody), you - must first install a pthread package before trying to configure - MySQL. + In current versions of FreeBSD (at least 4.x and greater), you may + increase the limit on the amount of memory available for a process + by adding the following entries to the /boot/loader.conf file and + rebooting the machine (these are not settings that can be changed + at run time with the sysctl command): +kern.maxdsiz="1073741824" # 1GB +kern.dfldsiz="1073741824" # 1GB +kern.maxssiz="134217728" # 128MB - See Section 2.5, "Installing MySQL on Mac OS X." + For older versions of FreeBSD, you must recompile your kernel to + change the maximum data segment size for a process. In this case, + you should look at the MAXDSIZ option in the LINT config file for + more information. -2.13.3. Solaris Notes + If you get problems with the current date in MySQL, setting the TZ + variable should help. See Section 2.14, "Environment Variables." - For information about installing MySQL on Solaris using PKG - distributions, see Section 2.6, "Installing MySQL on Solaris." +2.11. Installing MySQL on HP-UX - On Solaris, you may run into trouble even before you get the MySQL - distribution unpacked, as the Solaris tar cannot handle long file + If you install MySQL using a binary tarball distribution on HP-UX, + you may run into trouble even before you get the MySQL + distribution unpacked, as the HP-UX tar cannot handle long file names. This means that you may see errors when you try to unpack MySQL. If this occurs, you must use GNU tar (gtar) to unpack the distribution. - Sun native threads work only on Solaris 2.5 and higher. For - Solaris 2.4 and earlier, MySQL automatically uses MIT-pthreads. - See Section 2.10.5, "MIT-pthreads Notes." - - If you get the following error from configure, it means that you - have something wrong with your compiler installation: -checking for restartable system calls... configure: error can not -run test programs while cross compiling - - In this case, you should upgrade your compiler to a newer version. - You may also be able to solve this problem by inserting the - following row into the config.cache file: -ac_cv_sys_restartable_syscalls=${ac_cv_sys_restartable_syscalls='no'} - - If you are using Solaris on a SPARC, the recommended compiler is - gcc 2.95.2 or 3.2. You can find this at http://gcc.gnu.org/. Note - that gcc 2.8.1 does not work reliably on SPARC. - - The recommended configure line when using gcc 2.95.2 is: -CC=gcc CFLAGS="-O3" \ -CXX=gcc CXXFLAGS="-O3 -felide-constructors -fno-exceptions -fno-rtti" - \ -./configure --prefix=/usr/local/mysql --with-low-memory \ - --enable-assembler + Because of some critical bugs in the standard HP-UX libraries, you + should install the following patches before trying to run MySQL on + HP-UX 11.0: +PHKL_22840 Streams cumulative +PHNE_22397 ARPA cumulative - If you have an UltraSPARC system, you can get 4% better - performance by adding -mcpu=v8 -Wa,-xarch=v8plusa to the CFLAGS - and CXXFLAGS environment variables. + This solves the problem of getting EWOULDBLOCK from recv() and + EBADF from accept() in threaded applications. - If you have Sun's Forte 5.0 (or newer) compiler, you can run - configure like this: -CC=cc CFLAGS="-Xa -fast -native -xstrconst -mt" \ -CXX=CC CXXFLAGS="-noex -mt" \ -./configure --prefix=/usr/local/mysql --enable-assembler + If you are using gcc 2.95.1 on an unpatched HP-UX 11.x system, you + may get the following error: +In file included from /usr/include/unistd.h:11, + from ../include/global.h:125, + from mysql_priv.h:15, + from item.cc:19: +/usr/include/sys/unistd.h:184: declaration of C function ... +/usr/include/sys/pthread.h:440: previous declaration ... +In file included from item.h:306, + from mysql_priv.h:158, + from item.cc:19: - To create a 64-bit binary with Sun's Forte compiler, use the - following configuration options: -CC=cc CFLAGS="-Xa -fast -native -xstrconst -mt -xarch=v9" \ -CXX=CC CXXFLAGS="-noex -mt -xarch=v9" ASFLAGS="-xarch=v9" \ -./configure --prefix=/usr/local/mysql --enable-assembler + The problem is that HP-UX does not define pthreads_atfork() + consistently. It has conflicting prototypes in + /usr/include/sys/unistd.h:184 and /usr/include/sys/pthread.h:440. - To create a 64-bit Solaris binary using gcc, add -m64 to CFLAGS - and CXXFLAGS and remove --enable-assembler from the configure - line. + One solution is to copy /usr/include/sys/unistd.h into + mysql/include and edit unistd.h and change it to match the + definition in pthread.h. Look for this line: +extern int pthread_atfork(void (*prepare)(), void (*parent)(), + void (*child)()); - In the MySQL benchmarks, we obtained a 4% speed increase on - UltraSPARC when using Forte 5.0 in 32-bit mode, as compared to - using gcc 3.2 with the -mcpu flag. + Change it to look like this: +extern int pthread_atfork(void (*prepare)(void), void (*parent)(void) +, + void (*child)(void)); - If you create a 64-bit mysqld binary, it is 4% slower than the - 32-bit binary, but can handle more threads and memory. + After making the change, the following configure line should work: +CFLAGS="-fomit-frame-pointer -O3 -fpic" CXX=gcc \ +CXXFLAGS="-felide-constructors -fno-exceptions -fno-rtti -O3" \ +./configure --prefix=/usr/local/mysql --disable-shared - When using Solaris 10 for x86_64, you should mount any file - systems on which you intend to store InnoDB files with the - forcedirectio option. (By default mounting is done without this - option.) Failing to do so will cause a significant drop in - performance when using the InnoDB storage engine on this platform. + If you are using HP-UX compiler, you can use the following command + (which has been tested with cc B.11.11.04): +CC=cc CXX=aCC CFLAGS=+DD64 CXXFLAGS=+DD64 ./configure \ + --with-extra-character-set=complex - If you get a problem with fdatasync or sched_yield, you can fix - this by adding LIBS=-lrt to the configure line + You can ignore any errors of the following type: +aCC: warning 901: unknown option: `-3': use +help for online +documentation - For compilers older than WorkShop 5.3, you might have to edit the - configure script. Change this line: -#if !defined(__STDC__) || __STDC__ != 1 - - To this: -#if !defined(__STDC__) - - If you turn on __STDC__ with the -Xc option, the Sun compiler - can't compile with the Solaris pthread.h header file. This is a - Sun bug (broken compiler or broken include file). - - If mysqld issues the following error message when you run it, you - have tried to compile MySQL with the Sun compiler without enabling - the -mt multi-thread option: -libc internal error: _rmutex_unlock: rmutex not held - - Add -mt to CFLAGS and CXXFLAGS and recompile. - - If you are using the SFW version of gcc (which comes with Solaris - 8), you must add /opt/sfw/lib to the environment variable - LD_LIBRARY_PATH before running configure. - - If you are using the gcc available from sunfreeware.com, you may - have many problems. To avoid this, you should recompile gcc and - GNU binutils on the machine where you are running them. - - If you get the following error when compiling MySQL with gcc, it - means that your gcc is not configured for your version of Solaris: -shell> gcc -O3 -g -O2 -DDBUG_OFF -o thr_alarm ... -./thr_alarm.c: In function `signal_hand': -./thr_alarm.c:556: too many arguments to function `sigwait' - - The proper thing to do in this case is to get the newest version - of gcc and compile it with your current gcc compiler. At least for - Solaris 2.5, almost all binary versions of gcc have old, unusable - include files that break all programs that use threads, and - possibly other programs as well. + If you get the following error from configure, verify that you + don't have the path to the K&R compiler before the path to the + HP-UX C and C++ compiler: +checking for cc option to accept ANSI C... no +configure: error: MySQL requires an ANSI C compiler (and a C++ compil +er). +Try gcc. See the Installation chapter in the Reference Manual. - Solaris does not provide static versions of all system libraries - (libpthreads and libdl), so you cannot compile MySQL with - --static. If you try to do so, you get one of the following - errors: -ld: fatal: library -ldl: not found -undefined reference to `dlopen' -cannot find -lrt + Another reason for not being able to compile is that you didn't + define the +DD64 flags as just described. - If you link your own MySQL client programs, you may see the - following error at runtime: -ld.so.1: fatal: libmysqlclient.so.#: -open failed: No such file or directory + Another possibility for HP-UX 11 is to use the MySQL binaries + provided at http://dev.mysql.com/downloads/, which we have built + and tested ourselves. We have also received reports that the HP-UX + 10.20 binaries supplied by MySQL can be run successfully on HP-UX + 11. If you encounter problems, you should be sure to check your + HP-UX patch level. - This problem can be avoided by one of the following methods: +2.12. Installing MySQL on AIX - * Link clients with the -Wl,r/full/path/to/libmysqlclient.so - flag rather than with -Lpath). + Automatic detection of xlC is missing from Autoconf, so a number + of variables need to be set before running configure. The + following example uses the IBM compiler: +export CC="xlc_r -ma -O3 -qstrict -qoptimize=3 -qmaxmem=8192 " +export CXX="xlC_r -ma -O3 -qstrict -qoptimize=3 -qmaxmem=8192" +export CFLAGS="-I /usr/local/include" +export LDFLAGS="-L /usr/local/lib" +export CPPFLAGS=$CFLAGS +export CXXFLAGS=$CFLAGS - * Copy libmysqclient.so to /usr/lib. +./configure --prefix=/usr/local \ + --localstatedir=/var/mysql \ + --sbindir='/usr/local/bin' \ + --libexecdir='/usr/local/bin' \ + --enable-thread-safe-client \ + --enable-large-files - * Add the path name of the directory where libmysqlclient.so is - located to the LD_RUN_PATH environment variable before running - your client. + The preceding options are used to compile the MySQL distribution + that can be found at http://www-frec.bull.com/. - If you have problems with configure trying to link with -lz when - you don't have zlib installed, you have two options: + If you change the -O3 to -O2 in the preceding configure line, you + must also remove the -qstrict option. This is a limitation in the + IBM C compiler. - * If you want to be able to use the compressed communication - protocol, you need to get and install zlib from ftp.gnu.org. + If you are using gcc to compile MySQL, you must use the + -fno-exceptions flag, because the exception handling in gcc is not + thread-safe! There are also some known problems with IBM's + assembler that may cause it to generate bad code when used with + gcc. - * Run configure with the --with-named-z-libs=no option when - building MySQL. + Use the following configure line with gcc 2.95 on AIX: +CC="gcc -pipe -mcpu=power -Wa,-many" \ +CXX="gcc -pipe -mcpu=power -Wa,-many" \ +CXXFLAGS="-felide-constructors -fno-exceptions -fno-rtti" \ +./configure --prefix=/usr/local/mysql --with-low-memory - If you are using gcc and have problems with loading user-defined - functions (UDFs) into MySQL, try adding -lgcc to the link line for - the UDF. + The -Wa,-many option is necessary for the compile to be + successful. IBM is aware of this problem but is in no hurry to fix + it because of the workaround that is available. We don't know if + the -fno-exceptions is required with gcc 2.95, but because MySQL + doesn't use exceptions and the option generates faster code, you + should always use it with gcc. - If you would like MySQL to start automatically, you can copy - support-files/mysql.server to /etc/init.d and create a symbolic - link to it named /etc/rc3.d/S99mysql.server. + If you get a problem with assembler code, try changing the + -mcpu=xxx option to match your CPU. Typically power2, power, or + powerpc may need to be used. Alternatively, you might need to use + 604 or 604e. We are not positive but suspect that power would + likely be safe most of the time, even on a power2 machine. - If too many processes try to connect very rapidly to mysqld, you - should see this error in the MySQL log: -Error in accept: Protocol error + If you don't know what your CPU is, execute a uname -m command. It + produces a string that looks like 000514676700, with a format of + xxyyyyyymmss where xx and ss are always 00, yyyyyy is a unique + system ID and mm is the ID of the CPU Planar. A chart of these + values can be found at + http://www16.boulder.ibm.com/pseries/en_US/cmds/aixcmds5/uname.htm + . - You might try starting the server with the --back_log=50 option as - a workaround for this. (Use -O back_log=50 before MySQL 4.) + This gives you a machine type and a machine model you can use to + determine what type of CPU you have. - Solaris doesn't support core files for setuid() applications, so - you can't get a core file from mysqld if you are using the --user - option. + If you have problems with threads on AIX 5.3, you should upgrade + AIX 5.3 to technology level 7 (5300-07). + + If you have problems with signals (MySQL dies unexpectedly under + high load), you may have found an OS bug with threads and signals. + In this case, you can tell MySQL not to use signals by configuring + as follows: +CFLAGS=-DDONT_USE_THR_ALARM CXX=gcc \ +CXXFLAGS="-felide-constructors -fno-exceptions -fno-rtti \ +-DDONT_USE_THR_ALARM" \ +./configure --prefix=/usr/local/mysql --with-debug \ + --with-low-memory + + This doesn't affect the performance of MySQL, but has the side + effect that you can't kill clients that are "sleeping" on a + connection with mysqladmin kill or mysqladmin shutdown. Instead, + the client dies when it issues its next command. + + On some versions of AIX, linking with libbind.a makes + getservbyname() dump core. This is an AIX bug and should be + reported to IBM. + + For AIX 4.2.1 and gcc, you have to make the following changes. -2.13.3.1. Solaris 2.7/2.8 Notes + After configuring, edit config.h and include/my_config.h and + change the line that says this: +#define HAVE_SNPRINTF 1 - Normally, you can use a Solaris 2.6 binary on Solaris 2.7 and 2.8. - Most of the Solaris 2.6 issues also apply for Solaris 2.7 and 2.8. + to this: +#undef HAVE_SNPRINTF - MySQL should be able to detect new versions of Solaris - automatically and enable workarounds for the following problems. + And finally, in mysqld.cc, you need to add a prototype for + initgroups(). +#ifdef _AIX41 +extern "C" int initgroups(const char *,int); +#endif - Solaris 2.7 / 2.8 has some bugs in the include files. You may see - the following error when you use gcc: -/usr/include/widec.h:42: warning: `getwc' redefined -/usr/include/wchar.h:326: warning: this is the location of the previo -us -definition + For 32-bit binaries, if you need to allocate a lot of memory to + the mysqld process, it is not enough to just use ulimit -d + unlimited. You may also have to modify mysqld_safe to add a line + something like this: +export LDR_CNTRL='MAXDATA=0x80000000' - If this occurs, you can fix the problem by copying - /usr/include/widec.h to .../lib/gcc-lib/os/gcc-version/include and - changing line 41 from this: -#if !defined(lint) && !defined(__lint) + You can find more information about using a lot of memory at + http://publib16.boulder.ibm.com/pseries/en_US/aixprggd/genprogc/lr + g_prg_support.htm. - To this: -#if !defined(lint) && !defined(__lint) && !defined(getwc) + Users of AIX 4.3 should use gmake instead of the make utility + included with AIX. - Alternatively, you can edit /usr/include/widec.h directly. Either - way, after you make the fix, you should remove config.cache and - run configure again. + As of AIX 4.1, the C compiler has been unbundled from AIX as a + separate product. gcc 3.3.2 can be obtained here: + ftp://ftp.software.ibm.com/aix/freeSoftware/aixtoolbox/RPMS/ppc/gc + c/ - If you get the following errors when you run make, it is because - configure didn't detect the curses.h file (probably because of the - error in /usr/include/widec.h): -In file included from mysql.cc:50: -/usr/include/term.h:1060: syntax error before `,' -/usr/include/term.h:1081: syntax error before `;' + The steps for compiling MySQL on AIX with gcc 3.3.2 are similar to + those for using gcc 2.95 (in particular, the need to edit config.h + and my_config.h after running configure). However, before running + configure, you should also patch the curses.h file as follows: +/opt/freeware/lib/gcc-lib/powerpc-ibm-aix5.2.0.0/3.3.2/include/curses +.h.ORIG + Mon Dec 26 02:17:28 2005 +--- /opt/freeware/lib/gcc-lib/powerpc-ibm-aix5.2.0.0/3.3.2/include/cu +rses.h +Mon Dec 26 02:40:13 2005 +*************** +*** 2023,2029 **** - The solution to this problem is to do one of the following: - 1. Configure with CFLAGS=-DHAVE_CURSES_H CXXFLAGS=-DHAVE_CURSES_H - ./configure. + #endif /* _AIX32_CURSES */ +! #if defined(__USE_FIXED_PROTOTYPES__) || defined(__cplusplus) || de +fined +(__STRICT_ANSI__) + extern int delwin (WINDOW *); + extern int endwin (void); + extern int getcurx (WINDOW *); +--- 2023,2029 ---- - 2. Edit /usr/include/widec.h as indicated in the preceding - discussion and re-run configure. - 3. Remove the #define HAVE_TERM line from the config.h file and - run make again. + #endif /* _AIX32_CURSES */ +! #if 0 && (defined(__USE_FIXED_PROTOTYPES__) || defined(__cplusplus) +|| defined +(__STRICT_ANSI__)) + extern int delwin (WINDOW *); + extern int endwin (void); + extern int getcurx (WINDOW *); - If your linker cannot find -lz when linking client programs, the - problem is probably that your libz.so file is installed in - /usr/local/lib. You can fix this problem by one of the following - methods: +2.13. Post-Installation Setup and Testing - * Add /usr/local/lib to LD_LIBRARY_PATH. + After installing MySQL, there are some issues that you should + address. For example, on Unix, you should initialize the data + directory and create the MySQL grant tables. On all platforms, an + important security concern is that the initial accounts in the + grant tables have no passwords. You should assign passwords to + prevent unauthorized access to the MySQL server. Optionally, you + can create time zone tables to enable recognition of named time + zones. - * Add a link to libz.so from /lib. + The following sections include post-installation procedures that + are specific to Windows systems and to Unix systems. Another + section, Section 2.13.1.3, "Starting and Troubleshooting the MySQL + Server," applies to all platforms; it describes what to do if you + have trouble getting the server to start. Section 2.13.2, + "Securing the Initial MySQL Accounts," also applies to all + platforms. You should follow its instructions to make sure that + you have properly protected your MySQL accounts by assigning + passwords to them. - * If you are using Solaris 8, you can install the optional zlib - from your Solaris 8 CD distribution. + When you are ready to create additional user accounts, you can + find information on the MySQL access control system and account + management in Section 5.4, "The MySQL Access Privilege System," + and Section 5.5, "MySQL User Account Management." - * Run configure with the --with-named-z-libs=no option when - building MySQL. +2.13.1. Unix Post-Installation Procedures -2.13.3.2. Solaris x86 Notes + After installing MySQL on Unix, you need to initialize the grant + tables, start the server, and make sure that the server works + satisfactorily. You may also wish to arrange for the server to be + started and stopped automatically when your system starts and + stops. You should also assign passwords to the accounts in the + grant tables. - On Solaris 8 on x86, mysqld dumps core if you remove the debug - symbols using strip. + On Unix, the grant tables are set up by the mysql_install_db + program. For some installation methods, this program is run for + you automatically: - If you are using gcc on Solaris x86 and you experience problems - with core dumps under load, you should use the following configure - command: -CC=gcc CFLAGS="-O3 -fomit-frame-pointer -DHAVE_CURSES_H" \ -CXX=gcc \ -CXXFLAGS="-O3 -fomit-frame-pointer -felide-constructors \ - -fno-exceptions -fno-rtti -DHAVE_CURSES_H" \ -./configure --prefix=/usr/local/mysql + * If you install MySQL on Linux using RPM distributions, the + server RPM runs mysql_install_db. - This avoids problems with the libstdc++ library and with C++ - exceptions. + * If you install MySQL on Mac OS X using a PKG distribution, the + installer runs mysql_install_db. - If this doesn't help, you should compile a debug version and run - it with a trace file or under gdb. See MySQL Internals: Porting - (http://forge.mysql.com/wiki/MySQL_Internals_Porting) + Otherwise, you'll need to run mysql_install_db yourself. -2.13.4. BSD Notes + The following procedure describes how to initialize the grant + tables (if that has not previously been done) and then start the + server. It also suggests some commands that you can use to test + whether the server is accessible and working properly. For + information about starting and stopping the server automatically, + see Section 2.13.1.2, "Starting and Stopping MySQL Automatically." - This section provides information about using MySQL on variants of - BSD Unix. + After you complete the procedure and have the server running, you + should assign passwords to the accounts created by + mysql_install_db. Instructions for doing so are given in Section + 2.13.2, "Securing the Initial MySQL Accounts." -2.13.4.1. FreeBSD Notes + In the examples shown here, the server runs under the user ID of + the mysql login account. This assumes that such an account exists. + Either create the account if it does not exist, or substitute the + name of a different existing login account that you plan to use + for running the server. - FreeBSD 4.x or newer is recommended for running MySQL, because the - thread package is much more integrated. To get a secure and stable - system, you should use only FreeBSD kernels that are marked - -RELEASE. + 1. Change location into the top-level directory of your MySQL + installation, represented here by BASEDIR: +shell> cd BASEDIR + BASEDIR is likely to be something like /usr/local/mysql or + /usr/local. The following steps assume that you are located in + this directory. - The easiest (and preferred) way to install MySQL is to use the - mysql-server and mysql-client ports available at - http://www.freebsd.org/. Using these ports gives you the following - benefits: + 2. If necessary, run the mysql_install_db program to set up the + initial MySQL grant tables containing the privileges that + determine how users are allowed to connect to the server. + You'll need to do this if you used a distribution type for + which the installation procedure doesn't run the program for + you. + Typically, mysql_install_db needs to be run only the first + time you install MySQL, so you can skip this step if you are + upgrading an existing installation, However, mysql_install_db + does not overwrite any existing privilege tables, so it should + be safe to run in any circumstances. + To initialize the grant tables, use one of the following + commands, depending on whether mysql_install_db is located in + the bin or scripts directory: +shell> bin/mysql_install_db --user=mysql +shell> scripts/mysql_install_db --user=mysql + It might be necessary to specify other options such as + --basedir or --datadir if mysql_install_db does not use the + correct locations for the installation directory or data + directory. For example: +shell> bin/mysql_install_db --user=mysql \ + --basedir=/opt/mysql/mysql \ + --datadir=/opt/mysql/mysql/data + The mysql_install_db script creates the server's data + directory. Under the data directory, it creates directories + for the mysql database that holds all database privileges and + the test database that you can use to test MySQL. The script + also creates privilege table entries for root and + anonymous-user accounts. The accounts have no passwords + initially. A description of their initial privileges is given + in Section 2.13.2, "Securing the Initial MySQL Accounts." + Briefly, these privileges allow the MySQL root user to do + anything, and allow anybody to create or use databases with a + name of test or starting with test_. + It is important to make sure that the database directories and + files are owned by the mysql login account so that the server + has read and write access to them when you run it later. To + ensure this, the --user option should be used as shown if you + run mysql_install_db as root. Otherwise, you should execute + the script while logged in as mysql, in which case you can + omit the --user option from the command. + mysql_install_db creates several tables in the mysql database, + including user, db, host, tables_priv, columns_priv, func, and + others. See Section 5.4, "The MySQL Access Privilege System," + for a complete listing and description of these tables. + If you don't want to have the test database, you can remove it + with mysqladmin -u root drop test after starting the server. + If you have trouble with mysql_install_db at this point, see + Section 2.13.1.1, "Problems Running mysql_install_db." - * A working MySQL with all optimizations enabled that are known - to work on your version of FreeBSD. + 3. Start the MySQL server: +shell> bin/mysqld_safe --user=mysql & + It is important that the MySQL server be run using an + unprivileged (non-root) login account. To ensure this, the + --user option should be used as shown if you run mysqld_safe + as system root. Otherwise, you should execute the script while + logged in to the system as mysql, in which case you can omit + the --user option from the command. + Further instructions for running MySQL as an unprivileged user + are given in Section 5.3.5, "How to Run MySQL as a Normal + User." + If you neglected to create the grant tables before proceeding + to this step, the following message appears in the error log + file when you start the server: +mysqld: Can't find file: 'host.frm' + If you have other problems starting the server, see Section + 2.13.1.3, "Starting and Troubleshooting the MySQL Server." - * Automatic configuration and build. + 4. Use mysqladmin to verify that the server is running. The + following commands provide simple tests to check whether the + server is up and responding to connections: +shell> bin/mysqladmin version +shell> bin/mysqladmin variables + The output from mysqladmin version varies slightly depending + on your platform and version of MySQL, but should be similar + to that shown here: +shell> bin/mysqladmin version +mysqladmin Ver 14.12 Distrib 5.1.41, for pc-linux-gnu on i686 +... - * Startup scripts installed in /usr/local/etc/rc.d. +Server version 5.1.41 +Protocol version 10 +Connection Localhost via UNIX socket +UNIX socket /var/lib/mysql/mysql.sock +Uptime: 14 days 5 hours 5 min 21 sec - * The ability to use pkg_info -L to see which files are - installed. +Threads: 1 Questions: 366 Slow queries: 0 +Opens: 0 Flush tables: 1 Open tables: 19 +Queries per second avg: 0.000 + To see what else you can do with mysqladmin, invoke it with + the --help option. - * The ability to use pkg_delete to remove MySQL if you no longer - want it on your machine. + 5. Verify that you can shut down the server: +shell> bin/mysqladmin -u root shutdown - It is recommended you use MIT-pthreads on FreeBSD 2.x, and native - threads on FreeBSD 3 and up. It is possible to run with native - threads on some late 2.2.x versions, but you may encounter - problems shutting down mysqld. - - Unfortunately, certain function calls on FreeBSD are not yet fully - thread-safe. Most notably, this includes the gethostbyname() - function, which is used by MySQL to convert host names into IP - addresses. Under certain circumstances, the mysqld process - suddenly causes 100% CPU load and is unresponsive. If you - encounter this problem, try to start MySQL using the - --skip-name-resolve option. - - Alternatively, you can link MySQL on FreeBSD 4.x against the - LinuxThreads library, which avoids a few of the problems that the - native FreeBSD thread implementation has. For a very good - comparison of LinuxThreads versus native threads, see Jeremy - Zawodny's article FreeBSD or Linux for your MySQL Server? at - http://jeremy.zawodny.com/blog/archives/000697.html. - - Known problem when using LinuxThreads on FreeBSD is: - - * The connection times (wait_timeout, interactive_timeout and - net_read_timeout) values are not honored. The symptom is that - persistent connections can hang for a very long time without - getting closed down and that a 'kill' for a thread will not - take affect until the thread does it a new command - This is probably a signal handling problem in the thread - library where the signal doesn't break a pending read. This is - supposed to be fixed in FreeBSD 5.0 + 6. Verify that you can start the server again. Do this by using + mysqld_safe or by invoking mysqld directly. For example: +shell> bin/mysqld_safe --user=mysql --log & + If mysqld_safe fails, see Section 2.13.1.3, "Starting and + Troubleshooting the MySQL Server." - The MySQL build process requires GNU make (gmake) to work. If GNU - make is not available, you must install it first before compiling - MySQL. + 7. Run some simple tests to verify that you can retrieve + information from the server. The output should be similar to + what is shown here: +shell> bin/mysqlshow ++-----------+ +| Databases | ++-----------+ +| mysql | +| test | ++-----------+ - The recommended way to compile and install MySQL on FreeBSD with - gcc (2.95.2 and up) is: -CC=gcc CFLAGS="-O2 -fno-strength-reduce" \ - CXX=gcc CXXFLAGS="-O2 -fno-rtti -fno-exceptions \ - -felide-constructors -fno-strength-reduce" \ - ./configure --prefix=/usr/local/mysql --enable-assembler -gmake -gmake install -cd /usr/local/mysql -bin/mysql_install_db --user=mysql -bin/mysqld_safe & +shell> bin/mysqlshow mysql +Database: mysql ++---------------------------+ +| Tables | ++---------------------------+ +| columns_priv | +| db | +| func | +| help_category | +| help_keyword | +| help_relation | +| help_topic | +| host | +| proc | +| procs_priv | +| tables_priv | +| time_zone | +| time_zone_leap_second | +| time_zone_name | +| time_zone_transition | +| time_zone_transition_type | +| user | ++---------------------------+ - If you notice that configure uses MIT-pthreads, you should read - the MIT-pthreads notes. See Section 2.10.5, "MIT-pthreads Notes." +shell> bin/mysql -e "SELECT Host,Db,User FROM db" mysql ++------+--------+------+ +| host | db | user | ++------+--------+------+ +| % | test | | +| % | test_% | | ++------+--------+------+ - If you get an error from make install that it can't find - /usr/include/pthreads, configure didn't detect that you need - MIT-pthreads. To fix this problem, remove config.cache, and then - re-run configure with the --with-mit-threads option. - - Be sure that your name resolver setup is correct. Otherwise, you - may experience resolver delays or failures when connecting to - mysqld. Also make sure that the localhost entry in the /etc/hosts - file is correct. The file should start with a line similar to - this: -127.0.0.1 localhost localhost.your.domain + 8. There is a benchmark suite in the sql-bench directory (under + the MySQL installation directory) that you can use to compare + how MySQL performs on different platforms. The benchmark suite + is written in Perl. It requires the Perl DBI module that + provides a database-independent interface to the various + databases, and some other additional Perl modules: +DBI +DBD::mysql +Data::Dumper +Data::ShowTable + These modules can be obtained from CPAN + (http://www.cpan.org/) See also Section 2.15.1, "Installing + Perl on Unix." + The sql-bench/Results directory contains the results from many + runs against different databases and platforms. To run all + tests, execute these commands: +shell> cd sql-bench +shell> perl run-all-tests + If you don't have the sql-bench directory, you probably + installed MySQL using RPM files other than the source RPM. + (The source RPM includes the sql-bench benchmark directory.) + In this case, you must first install the benchmark suite + before you can use it. There are separate benchmark RPM files + named mysql-bench-VERSION.i386.rpm that contain benchmark code + and data. + If you have a source distribution, there are also tests in its + tests subdirectory that you can run. For example, to run + auto_increment.tst, execute this command from the top-level + directory of your source distribution: +shell> mysql -vvf test < ./tests/auto_increment.tst + The expected result of the test can be found in the + ./tests/auto_increment.res file. - FreeBSD is known to have a very low default file handle limit. See - Section B.1.2.18, "'File' Not Found and Similar Errors." Start the - server by using the --open-files-limit option for mysqld_safe, or - raise the limits for the mysqld user in /etc/login.conf and - rebuild it with cap_mkdb /etc/login.conf. Also be sure that you - set the appropriate class for this user in the password file if - you are not using the default (use chpass mysqld-user-name). See - Section 4.3.2, "mysqld_safe --- MySQL Server Startup Script." + 9. At this point, you should have the server running. However, + none of the initial MySQL accounts have a password, so you + should assign passwords using the instructions found in + Section 2.13.2, "Securing the Initial MySQL Accounts." - FreeBSD limits the size of a process to 512MB, even if you have - much more RAM available on the system. So you may get an error - such as this: -Out of memory (Needed 16391 bytes) + The MySQL 5.1 installation procedure creates time zone tables in + the mysql database. However, you must populate the tables manually + using the instructions in Section 9.7, "MySQL Server Time Zone + Support." - In current versions of FreeBSD (at least 4.x and greater), you may - increase this limit by adding the following entries to the - /boot/loader.conf file and rebooting the machine (these are not - settings that can be changed at run time with the sysctl command): -kern.maxdsiz="1073741824" # 1GB -kern.dfldsiz="1073741824" # 1GB -kern.maxssiz="134217728" # 128MB +2.13.1.1. Problems Running mysql_install_db - For older versions of FreeBSD, you must recompile your kernel to - change the maximum data segment size for a process. In this case, - you should look at the MAXDSIZ option in the LINT config file for - more information. + The purpose of the mysql_install_db script is to generate new + MySQL privilege tables. It does not overwrite existing MySQL + privilege tables, and it does not affect any other data. - If you get problems with the current date in MySQL, setting the TZ - variable should help. See Section 2.14, "Environment Variables." + If you want to re-create your privilege tables, first stop the + mysqld server if it is running. Then rename the mysql directory + under the data directory to save it, and then run + mysql_install_db. Suppose that your current directory is the MySQL + installation directory and that mysql_install_db is located in the + bin directory and the data directory is named data. To rename the + mysql database and re-run mysql_install_db, use these commands. +shell> mv data/mysql data/mysql.old +shell> bin/mysql_install_db --user=mysql -2.13.4.2. NetBSD Notes + When you run mysql_install_db, you might encounter the following + problems: - To compile on NetBSD, you need GNU make. Otherwise, the build - process fails when make tries to run lint on C++ files. + * mysql_install_db fails to install the grant tables + You may find that mysql_install_db fails to install the grant + tables and terminates after displaying the following messages: +Starting mysqld daemon with databases from XXXXXX +mysqld ended + In this case, you should examine the error log file very + carefully. The log should be located in the directory XXXXXX + named by the error message and should indicate why mysqld + didn't start. If you do not understand what happened, include + the log when you post a bug report. See Section 1.6, "How to + Report Bugs or Problems." -2.13.4.3. OpenBSD 2.5 Notes + * There is a mysqld process running + This indicates that the server is running, in which case the + grant tables have probably been created already. If so, there + is no need to run mysql_install_db at all because it needs to + be run only once (when you install MySQL the first time). - On OpenBSD 2.5, you can compile MySQL with native threads with the - following options: -CFLAGS=-pthread CXXFLAGS=-pthread ./configure --with-mit-threads=no + * Installing a second mysqld server does not work when one + server is running + This can happen when you have an existing MySQL installation, + but want to put a new installation in a different location. + For example, you might have a production installation, but you + want to create a second installation for testing purposes. + Generally the problem that occurs when you try to run a second + server is that it tries to use a network interface that is in + use by the first server. In this case, you should see one of + the following error messages: +Can't start server: Bind on TCP/IP port: +Address already in use +Can't start server: Bind on unix socket... + For instructions on setting up multiple servers, see Section + 5.6, "Running Multiple MySQL Servers on the Same Machine." -2.13.4.4. BSD/OS Version 2.x Notes + * You do not have write access to the /tmp directory + If you do not have write access to create temporary files or a + Unix socket file in the default location (the /tmp directory), + an error occurs when you run mysql_install_db or the mysqld + server. + You can specify different locations for the temporary + directory and Unix socket file by executing these commands + prior to starting mysql_install_db or mysqld, where + some_tmp_dir is the full path name to some directory for which + you have write permission: +shell> TMPDIR=/some_tmp_dir/ +shell> MYSQL_UNIX_PORT=/some_tmp_dir/mysql.sock +shell> export TMPDIR MYSQL_UNIX_PORT + Then you should be able to run mysql_install_db and start the + server with these commands: +shell> bin/mysql_install_db --user=mysql +shell> bin/mysqld_safe --user=mysql & + If mysql_install_db is located in the scripts directory, + modify the first command to scripts/mysql_install_db. + See Section B.5.4.5, "How to Protect or Change the MySQL Unix + Socket File," and Section 2.14, "Environment Variables." - If you get the following error when compiling MySQL, your ulimit - value for virtual memory is too low: -item_func.h: In method -`Item_func_ge::Item_func_ge(const Item_func_ge &)': -item_func.h:28: virtual memory exhausted -make[2]: *** [item_func.o] Error 1 + There are some alternatives to running the mysql_install_db script + provided in the MySQL distribution: - Try using ulimit -v 80000 and run make again. If this doesn't work - and you are using bash, try switching to csh or sh; some BSDI - users have reported problems with bash and ulimit. + * If you want the initial privileges to be different from the + standard defaults, you can modify mysql_install_db before you + run it. However, it is preferable to use GRANT and REVOKE to + change the privileges after the grant tables have been set up. + In other words, you can run mysql_install_db, and then use + mysql -u root mysql to connect to the server as the MySQL root + user so that you can issue the necessary GRANT and REVOKE + statements. + If you want to install MySQL on several machines with the same + privileges, you can put the GRANT and REVOKE statements in a + file and execute the file as a script using mysql after + running mysql_install_db. For example: +shell> bin/mysql_install_db --user=mysql +shell> bin/mysql -u root < your_script_file + By doing this, you can avoid having to issue the statements + manually on each machine. - If you are using gcc, you may also use have to use the - --with-low-memory flag for configure to be able to compile - sql_yacc.cc. + * It is possible to re-create the grant tables completely after + they have previously been created. You might want to do this + if you're just learning how to use GRANT and REVOKE and have + made so many modifications after running mysql_install_db that + you want to wipe out the tables and start over. + To re-create the grant tables, remove all the .frm, .MYI, and + .MYD files in the mysql database directory. Then run the + mysql_install_db script again. - If you get problems with the current date in MySQL, setting the TZ - variable should help. See Section 2.14, "Environment Variables." + * You can start mysqld manually using the --skip-grant-tables + option and add the privilege information yourself using mysql: +shell> bin/mysqld_safe --user=mysql --skip-grant-tables & +shell> bin/mysql mysql + From mysql, manually execute the SQL commands contained in + mysql_install_db. Make sure that you run mysqladmin + flush-privileges or mysqladmin reload afterward to tell the + server to reload the grant tables. + Note that by not using mysql_install_db, you not only have to + populate the grant tables manually, you also have to create + them first. -2.13.4.5. BSD/OS Version 3.x Notes +2.13.1.2. Starting and Stopping MySQL Automatically - Upgrade to BSD/OS 3.1. If that is not possible, install BSDIpatch - M300-038. + Generally, you start the mysqld server in one of these ways: - Use the following command when configuring MySQL: -env CXX=shlicc++ CC=shlicc2 \ -./configure \ - --prefix=/usr/local/mysql \ - --localstatedir=/var/mysql \ - --without-perl \ - --with-unix-socket-path=/var/mysql/mysql.sock - - The following is also known to work: -env CC=gcc CXX=gcc CXXFLAGS=-O3 \ -./configure \ - --prefix=/usr/local/mysql \ - --with-unix-socket-path=/var/mysql/mysql.sock - - You can change the directory locations if you wish, or just use - the defaults by not specifying any locations. - - If you have problems with performance under heavy load, try using - the --skip-thread-priority option to mysqld. This runs all threads - with the same priority. On BSDI 3.1, this gives better - performance, at least until BSDI fixes its thread scheduler. - - If you get the error virtual memory exhausted while compiling, you - should try using ulimit -v 80000 and running make again. If this - doesn't work and you are using bash, try switching to csh or sh; - some BSDI users have reported problems with bash and ulimit. - -2.13.4.6. BSD/OS Version 4.x Notes - - BSDI 4.x has some thread-related bugs. If you want to use MySQL on - this, you should install all thread-related patches. At least - M400-023 should be installed. - - On some BSDI 4.x systems, you may get problems with shared - libraries. The symptom is that you can't execute any client - programs, for example, mysqladmin. In this case, you need to - reconfigure not to use shared libraries with the --disable-shared - option to configure. - - Some customers have had problems on BSDI 4.0.1 that the mysqld - binary after a while can't open tables. This occurs because some - library/system-related bug causes mysqld to change current - directory without having asked for that to happen. - - The fix is to either upgrade MySQL to at least version 3.23.34 or, - after running configure, remove the line #define HAVE_REALPATH - from config.h before running make. - - Note that this means that you can't symbolically link a database - directories to another database directory or symbolic link a table - to another database on BSDI. (Making a symbolic link to another - disk is okay). + * Invoke mysqld directly. This works on any platform. -2.13.5. Other Unix Notes + * Run the MySQL server as a Windows service. The service can be + set to start the server automatically when Windows starts, or + as a manual service that you start on request. For + instructions, see Section 2.5.5.6, "Starting MySQL as a + Windows Service." -2.13.5.1. HP-UX Version 10.20 Notes + * Invoke mysqld_safe, which tries to determine the proper + options for mysqld and then runs it with those options. This + script is used on Unix and Unix-like systems. See Section + 4.3.2, "mysqld_safe --- MySQL Server Startup Script." - If you install MySQL using a binary tarball distribution on HP-UX, - you may run into trouble even before you get the MySQL - distribution unpacked, as the HP-UX tar cannot handle long file - names. This means that you may see errors when you try to unpack - MySQL. + * Invoke mysql.server. This script is used primarily at system + startup and shutdown on systems that use System V-style run + directories, where it usually is installed under the name + mysql. The mysql.server script starts the server by invoking + mysqld_safe. See Section 4.3.3, "mysql.server --- MySQL Server + Startup Script." - If this occurs, you must use GNU tar (gtar) to unpack the - distribution. + * On Mac OS X, install a separate MySQL Startup Item package to + enable the automatic startup of MySQL on system startup. The + Startup Item starts the server by invoking mysql.server. See + Section 2.7, "Installing MySQL on Mac OS X," for details. - There are a couple of small problems when compiling MySQL on - HP-UX. Use gcc instead of the HP-UX native compiler, because gcc - produces better code. - - Use gcc 2.95 on HP-UX. Don't use high optimization flags (such as - -O6) because they may not be safe on HP-UX. - - The following configure line should work with gcc 2.95: -CFLAGS="-I/opt/dce/include -fpic" \ -CXXFLAGS="-I/opt/dce/include -felide-constructors -fno-exceptions \ --fno-rtti" \ -CXX=gcc \ -./configure --with-pthread \ - --with-named-thread-libs='-ldce' \ - --prefix=/usr/local/mysql --disable-shared - - The following configure line should work with gcc 3.1: -CFLAGS="-DHPUX -I/opt/dce/include -O3 -fPIC" CXX=gcc \ -CXXFLAGS="-DHPUX -I/opt/dce/include -felide-constructors \ - -fno-exceptions -fno-rtti -O3 -fPIC" \ -./configure --prefix=/usr/local/mysql \ - --with-extra-charsets=complex --enable-thread-safe-client \ - --enable-local-infile --with-pthread \ - --with-named-thread-libs=-ldce --with-lib-ccflags=-fPIC - --disable-shared + The mysqld_safe and mysql.server scripts and the Mac OS X Startup + Item can be used to start the server manually, or automatically at + system startup time. mysql.server and the Startup Item also can be + used to stop the server. -2.13.5.2. HP-UX Version 11.x Notes + To start or stop the server manually using the mysql.server + script, invoke it with start or stop arguments: +shell> mysql.server start +shell> mysql.server stop - If you install MySQL using a binary tarball distribution on HP-UX, - you may run into trouble even before you get the MySQL - distribution unpacked, as the HP-UX tar cannot handle long file - names. This means that you may see errors when you try to unpack - MySQL. + Before mysql.server starts the server, it changes location to the + MySQL installation directory, and then invokes mysqld_safe. If you + want the server to run as some specific user, add an appropriate + user option to the [mysqld] group of the /etc/my.cnf option file, + as shown later in this section. (It is possible that you will need + to edit mysql.server if you've installed a binary distribution of + MySQL in a nonstandard location. Modify it to change location into + the proper directory before it runs mysqld_safe. If you do this, + your modified version of mysql.server may be overwritten if you + upgrade MySQL in the future, so you should make a copy of your + edited version that you can reinstall.) - If this occurs, you must use GNU tar (gtar) to unpack the - distribution. + mysql.server stop stops the server by sending a signal to it. You + can also stop the server manually by executing mysqladmin + shutdown. - Because of some critical bugs in the standard HP-UX libraries, you - should install the following patches before trying to run MySQL on - HP-UX 11.0: -PHKL_22840 Streams cumulative -PHNE_22397 ARPA cumulative + To start and stop MySQL automatically on your server, you need to + add start and stop commands to the appropriate places in your + /etc/rc* files. - This solves the problem of getting EWOULDBLOCK from recv() and - EBADF from accept() in threaded applications. + If you use the Linux server RPM package + (MySQL-server-VERSION.rpm), the mysql.server script is installed + in the /etc/init.d directory with the name mysql. You need not + install it manually. See Section 2.6.1, "Installing MySQL from RPM + Packages on Linux," for more information on the Linux RPM + packages. - If you are using gcc 2.95.1 on an unpatched HP-UX 11.x system, you - may get the following error: -In file included from /usr/include/unistd.h:11, - from ../include/global.h:125, - from mysql_priv.h:15, - from item.cc:19: -/usr/include/sys/unistd.h:184: declaration of C function ... -/usr/include/sys/pthread.h:440: previous declaration ... -In file included from item.h:306, - from mysql_priv.h:158, - from item.cc:19: + Some vendors provide RPM packages that install a startup script + under a different name such as mysqld. - The problem is that HP-UX does not define pthreads_atfork() - consistently. It has conflicting prototypes in - /usr/include/sys/unistd.h:184 and /usr/include/sys/pthread.h:440. + If you install MySQL from a source distribution or using a binary + distribution format that does not install mysql.server + automatically, you can install it manually. The script can be + found in the support-files directory under the MySQL installation + directory or in a MySQL source tree. - One solution is to copy /usr/include/sys/unistd.h into - mysql/include and edit unistd.h and change it to match the - definition in pthread.h. Look for this line: -extern int pthread_atfork(void (*prepare)(), void (*parent)(), - void (*child)()); + To install mysql.server manually, copy it to the /etc/init.d + directory with the name mysql, and then make it executable. Do + this by changing location into the appropriate directory where + mysql.server is located and executing these commands: +shell> cp mysql.server /etc/init.d/mysql +shell> chmod +x /etc/init.d/mysql - Change it to look like this: -extern int pthread_atfork(void (*prepare)(void), void (*parent)(void) -, - void (*child)(void)); + Older Red Hat systems use the /etc/rc.d/init.d directory rather + than /etc/init.d. Adjust the preceding commands accordingly. + Alternatively, first create /etc/init.d as a symbolic link that + points to /etc/rc.d/init.d: +shell> cd /etc +shell> ln -s rc.d/init.d . - After making the change, the following configure line should work: -CFLAGS="-fomit-frame-pointer -O3 -fpic" CXX=gcc \ -CXXFLAGS="-felide-constructors -fno-exceptions -fno-rtti -O3" \ -./configure --prefix=/usr/local/mysql --disable-shared + After installing the script, the commands needed to activate it to + run at system startup depend on your operating system. On Linux, + you can use chkconfig: +shell> chkconfig --add mysql - If you are using HP-UX compiler, you can use the following command - (which has been tested with cc B.11.11.04): -CC=cc CXX=aCC CFLAGS=+DD64 CXXFLAGS=+DD64 ./configure \ - --with-extra-character-set=complex + On some Linux systems, the following command also seems to be + necessary to fully enable the mysql script: +shell> chkconfig --level 345 mysql on - You can ignore any errors of the following type: -aCC: warning 901: unknown option: `-3': use +help for online -documentation + On FreeBSD, startup scripts generally should go in + /usr/local/etc/rc.d/. The rc(8) manual page states that scripts in + this directory are executed only if their basename matches the + *.sh shell file name pattern. Any other files or directories + present within the directory are silently ignored. In other words, + on FreeBSD, you should install the mysql.server script as + /usr/local/etc/rc.d/mysql.server.sh to enable automatic startup. - If you get the following error from configure, verify that you - don't have the path to the K&R compiler before the path to the - HP-UX C and C++ compiler: -checking for cc option to accept ANSI C... no -configure: error: MySQL requires an ANSI C compiler (and a C++ compil -er). -Try gcc. See the Installation chapter in the Reference Manual. + As an alternative to the preceding setup, some operating systems + also use /etc/rc.local or /etc/init.d/boot.local to start + additional services on startup. To start up MySQL using this + method, you could append a command like the one following to the + appropriate startup file: +/bin/sh -c 'cd /usr/local/mysql; ./bin/mysqld_safe --user=mysql &' - Another reason for not being able to compile is that you didn't - define the +DD64 flags as just described. + For other systems, consult your operating system documentation to + see how to install startup scripts. - Another possibility for HP-UX 11 is to use the MySQL binaries - provided at http://dev.mysql.com/downloads/, which we have built - and tested ourselves. We have also received reports that the HP-UX - 10.20 binaries supplied by MySQL can be run successfully on HP-UX - 11. If you encounter problems, you should be sure to check your - HP-UX patch level. + You can add options for mysql.server in a global /etc/my.cnf file. + A typical /etc/my.cnf file might look like this: +[mysqld] +datadir=/usr/local/mysql/var +socket=/var/tmp/mysql.sock +port=3306 +user=mysql -2.13.5.3. IBM-AIX notes +[mysql.server] +basedir=/usr/local/mysql - Automatic detection of xlC is missing from Autoconf, so a number - of variables need to be set before running configure. The - following example uses the IBM compiler: -export CC="xlc_r -ma -O3 -qstrict -qoptimize=3 -qmaxmem=8192 " -export CXX="xlC_r -ma -O3 -qstrict -qoptimize=3 -qmaxmem=8192" -export CFLAGS="-I /usr/local/include" -export LDFLAGS="-L /usr/local/lib" -export CPPFLAGS=$CFLAGS -export CXXFLAGS=$CFLAGS + The mysql.server script supports the following options: basedir, + datadir, and pid-file. If specified, they must be placed in an + option file, not on the command line. mysql.server supports only + start and stop as command-line arguments. -./configure --prefix=/usr/local \ - --localstatedir=/var/mysql \ - --sbindir='/usr/local/bin' \ - --libexecdir='/usr/local/bin' \ - --enable-thread-safe-client \ - --enable-large-files + The following table shows which option groups the server and each + startup script read from option files. + Script Option Groups + mysqld [mysqld], [server], [mysqld-major_version] + mysqld_safe [mysqld], [server], [mysqld_safe] + mysql.server [mysqld], [mysql.server], [server] - The preceding options are used to compile the MySQL distribution - that can be found at http://www-frec.bull.com/. + [mysqld-major_version] means that groups with names like + [mysqld-5.0] and [mysqld-5.1] are read by servers having versions + 5.0.x, 5.1.x, and so forth. This feature can be used to specify + options that can be read only by servers within a given release + series. - If you change the -O3 to -O2 in the preceding configure line, you - must also remove the -qstrict option. This is a limitation in the - IBM C compiler. + For backward compatibility, mysql.server also reads the + [mysql_server] group and mysqld_safe also reads the [safe_mysqld] + group. However, you should update your option files to use the + [mysql.server] and [mysqld_safe] groups instead when using MySQL + 5.1. - If you are using gcc to compile MySQL, you must use the - -fno-exceptions flag, because the exception handling in gcc is not - thread-safe! There are also some known problems with IBM's - assembler that may cause it to generate bad code when used with - gcc. + See Section 4.2.3.3, "Using Option Files." - Use the following configure line with gcc 2.95 on AIX: -CC="gcc -pipe -mcpu=power -Wa,-many" \ -CXX="gcc -pipe -mcpu=power -Wa,-many" \ -CXXFLAGS="-felide-constructors -fno-exceptions -fno-rtti" \ -./configure --prefix=/usr/local/mysql --with-low-memory +2.13.1.3. Starting and Troubleshooting the MySQL Server - The -Wa,-many option is necessary for the compile to be - successful. IBM is aware of this problem but is in no hurry to fix - it because of the workaround that is available. We don't know if - the -fno-exceptions is required with gcc 2.95, but because MySQL - doesn't use exceptions and the option generates faster code, you - should always use it with gcc. + This section provides troubleshooting suggestions for problems + starting the server on Unix. If you are using Windows, see Section + 2.5.6, "Troubleshooting a MySQL Installation Under Windows." - If you get a problem with assembler code, try changing the - -mcpu=xxx option to match your CPU. Typically power2, power, or - powerpc may need to be used. Alternatively, you might need to use - 604 or 604e. We are not positive but suspect that power would - likely be safe most of the time, even on a power2 machine. + If you have problems starting the server, here are some things to + try: - If you don't know what your CPU is, execute a uname -m command. It - produces a string that looks like 000514676700, with a format of - xxyyyyyymmss where xx and ss are always 00, yyyyyy is a unique - system ID and mm is the ID of the CPU Planar. A chart of these - values can be found at - http://www16.boulder.ibm.com/pseries/en_US/cmds/aixcmds5/uname.htm - . + * Check the error log to see why the server does not start. - This gives you a machine type and a machine model you can use to - determine what type of CPU you have. + * Specify any special options needed by the storage engines you + are using. - If you have problems with signals (MySQL dies unexpectedly under - high load), you may have found an OS bug with threads and signals. - In this case, you can tell MySQL not to use signals by configuring - as follows: -CFLAGS=-DDONT_USE_THR_ALARM CXX=gcc \ -CXXFLAGS="-felide-constructors -fno-exceptions -fno-rtti \ --DDONT_USE_THR_ALARM" \ -./configure --prefix=/usr/local/mysql --with-debug \ - --with-low-memory + * Make sure that the server knows where to find the data + directory. - This doesn't affect the performance of MySQL, but has the side - effect that you can't kill clients that are "sleeping" on a - connection with mysqladmin kill or mysqladmin shutdown. Instead, - the client dies when it issues its next command. + * Make sure that the server can access the data directory. The + ownership and permissions of the data directory and its + contents must be set such that the server can read and modify + them. - On some versions of AIX, linking with libbind.a makes - getservbyname() dump core. This is an AIX bug and should be - reported to IBM. + * Verify that the network interfaces the server wants to use are + available. - For AIX 4.2.1 and gcc, you have to make the following changes. + Some storage engines have options that control their behavior. You + can create a my.cnf file and specify startup options for the + engines that you plan to use. If you are going to use storage + engines that support transactional tables (InnoDB, NDB), be sure + that you have them configured the way you want before starting the + server: - After configuring, edit config.h and include/my_config.h and - change the line that says this: -#define HAVE_SNPRINTF 1 + * If you are using InnoDB tables, see Section 13.6.2, "InnoDB + Configuration." - to this: -#undef HAVE_SNPRINTF + * If you are using MySQL Cluster, see Section 17.3, "MySQL + Cluster Configuration." - And finally, in mysqld.cc, you need to add a prototype for - initgroups(). -#ifdef _AIX41 -extern "C" int initgroups(const char *,int); -#endif + MySQL Enterprise For expert advice on start-up options appropriate + to your circumstances, subscribe to The MySQL Enterprise Monitor. + For more information, see + http://www.mysql.com/products/enterprise/advisors.html. - For 32-bit binaries, if you need to allocate a lot of memory to - the mysqld process, it is not enough to just use ulimit -d - unlimited. You may also have to modify mysqld_safe to add a line - something like this: -export LDR_CNTRL='MAXDATA=0x80000000' + Storage engines will use default option values if you specify + none, but it is recommended that you review the available options + and specify explicit values for those for which the defaults are + not appropriate for your installation. - You can find more information about using a lot of memory at - http://publib16.boulder.ibm.com/pseries/en_US/aixprggd/genprogc/lr - g_prg_support.htm. + When the mysqld server starts, it changes location to the data + directory. This is where it expects to find databases and where it + expects to write log files. The server also writes the pid + (process ID) file in the data directory. - Users of AIX 4.3 should use gmake instead of the make utility - included with AIX. + The data directory location is hardwired in when the server is + compiled. This is where the server looks for the data directory by + default. If the data directory is located somewhere else on your + system, the server will not work properly. You can determine what + the default path settings are by invoking mysqld with the + --verbose and --help options. - As of AIX 4.1, the C compiler has been unbundled from AIX as a - separate product. gcc 3.3.2 can be obtained here: - ftp://ftp.software.ibm.com/aix/freeSoftware/aixtoolbox/RPMS/ppc/gc - c/ + If the default locations don't match the MySQL installation layout + on your system, you can override them by specifying options to + mysqld or mysqld_safe on the command line or in an option file. - The steps for compiling MySQL on AIX with gcc 3.3.2 are similar to - those for using gcc 2.95 (in particular, the need to edit config.h - and my_config.h after running configure). However, before running - configure, you should also patch the curses.h file as follows: -/opt/freeware/lib/gcc-lib/powerpc-ibm-aix5.2.0.0/3.3.2/include/curses -.h.ORIG - Mon Dec 26 02:17:28 2005 ---- /opt/freeware/lib/gcc-lib/powerpc-ibm-aix5.2.0.0/3.3.2/include/cu -rses.h -Mon Dec 26 02:40:13 2005 -*************** -*** 2023,2029 **** + To specify the location of the data directory explicitly, use the + --datadir option. However, normally you can tell mysqld the + location of the base directory under which MySQL is installed and + it looks for the data directory there. You can do this with the + --basedir option. + + To check the effect of specifying path options, invoke mysqld with + those options followed by the --verbose and --help options. For + example, if you change location into the directory where mysqld is + installed and then run the following command, it shows the effect + of starting the server with a base directory of /usr/local: +shell> ./mysqld --basedir=/usr/local --verbose --help + You can specify other options such as --datadir as well, but + --verbose and --help must be the last options. - #endif /* _AIX32_CURSES */ -! #if defined(__USE_FIXED_PROTOTYPES__) || defined(__cplusplus) || de -fined -(__STRICT_ANSI__) - extern int delwin (WINDOW *); - extern int endwin (void); - extern int getcurx (WINDOW *); ---- 2023,2029 ---- + Once you determine the path settings you want, start the server + without --verbose and --help. + If mysqld is currently running, you can find out what path + settings it is using by executing this command: +shell> mysqladmin variables - #endif /* _AIX32_CURSES */ -! #if 0 && (defined(__USE_FIXED_PROTOTYPES__) || defined(__cplusplus) -|| defined -(__STRICT_ANSI__)) - extern int delwin (WINDOW *); - extern int endwin (void); - extern int getcurx (WINDOW *); + Or: +shell> mysqladmin -h host_name variables -2.13.5.4. SunOS 4 Notes + host_name is the name of the MySQL server host. - On SunOS 4, MIT-pthreads is needed to compile MySQL. This in turn - means you need GNU make. + If you get Errcode 13 (which means Permission denied) when + starting mysqld, this means that the privileges of the data + directory or its contents do not allow the server access. In this + case, you change the permissions for the involved files and + directories so that the server has the right to use them. You can + also start the server as root, but this raises security issues and + should be avoided. - Some SunOS 4 systems have problems with dynamic libraries and - libtool. You can use the following configure line to avoid this - problem: -./configure --disable-shared --with-mysqld-ldflags=-all-static - - When compiling readline, you may get warnings about duplicate - defines. These can be ignored. - - When compiling mysqld, there are some implicit declaration of - function warnings. These can be ignored. - -2.13.5.5. Alpha-DEC-UNIX Notes (Tru64) - - If you are using egcs 1.1.2 on Digital Unix, you should upgrade to - gcc 2.95.2, because egcs on DEC has some serious bugs! - - When compiling threaded programs under Digital Unix, the - documentation recommends using the -pthread option for cc and cxx - and the -lmach -lexc libraries (in addition to -lpthread). You - should run configure something like this: -CC="cc -pthread" CXX="cxx -pthread -O" \ -./configure --with-named-thread-libs="-lpthread -lmach -lexc -lc" - - When compiling mysqld, you may see a couple of warnings like this: -mysqld.cc: In function void handle_connections()': -mysqld.cc:626: passing long unsigned int *' as argument 3 of -accept(int,sockadddr *, int *)' - - You can safely ignore these warnings. They occur because configure - can detect only errors, not warnings. - - If you start the server directly from the command line, you may - have problems with it dying when you log out. (When you log out, - your outstanding processes receive a SIGHUP signal.) If so, try - starting the server like this: -nohup mysqld [options] & - - nohup causes the command following it to ignore any SIGHUP signal - sent from the terminal. Alternatively, start the server by running - mysqld_safe, which invokes mysqld using nohup for you. See Section - 4.3.2, "mysqld_safe --- MySQL Server Startup Script." - - If you get a problem when compiling mysys/get_opt.c, just remove - the #define _NO_PROTO line from the start of that file. - - If you are using Compaq's CC compiler, the following configure - line should work: -CC="cc -pthread" -CFLAGS="-O4 -ansi_alias -ansi_args -fast -inline speed \ - -speculate all -arch host" -CXX="cxx -pthread" -CXXFLAGS="-O4 -ansi_alias -ansi_args -fast -inline speed \ - -speculate all -arch host -noexceptions -nortti" -export CC CFLAGS CXX CXXFLAGS -./configure \ - --prefix=/usr/local/mysql \ - --with-low-memory \ - --enable-large-files \ - --enable-shared=yes \ - --with-named-thread-libs="-lpthread -lmach -lexc -lc" -gnumake - - If you get a problem with libtool when compiling with shared - libraries as just shown, when linking mysql, you should be able to - get around this by issuing these commands: -cd mysql -/bin/sh ../libtool --mode=link cxx -pthread -O3 -DDBUG_OFF \ - -O4 -ansi_alias -ansi_args -fast -inline speed \ - -speculate all \ -arch host -DUNDEF_HAVE_GETHOSTBYNAME_R \ - -o mysql mysql.o readline.o sql_string.o completion_hash.o \ - ../readline/libreadline.a -lcurses \ - ../libmysql/.libs/libmysqlclient.so -lm -cd .. -gnumake -gnumake install -scripts/mysql_install_db - -2.13.5.6. Alpha-DEC-OSF/1 Notes - - If you have problems compiling and have DEC CC and gcc installed, - try running configure like this: -CC=cc CFLAGS=-O CXX=gcc CXXFLAGS=-O3 \ -./configure --prefix=/usr/local/mysql - - If you get problems with the c_asm.h file, you can create and use - a 'dummy' c_asm.h file with: -touch include/c_asm.h -CC=gcc CFLAGS=-I./include \ -CXX=gcc CXXFLAGS=-O3 \ -./configure --prefix=/usr/local/mysql - - Note that the following problems with the ld program can be fixed - by downloading the latest DEC (Compaq) patch kit from: - http://ftp.support.compaq.com/public/unix/. - - On OSF/1 V4.0D and compiler "DEC C V5.6-071 on Digital Unix V4.0 - (Rev. 878)," the compiler had some strange behavior (undefined asm - symbols). /bin/ld also appears to be broken (problems with _exit - undefined errors occurring while linking mysqld). On this system, - we have managed to compile MySQL with the following configure - line, after replacing /bin/ld with the version from OSF 4.0C: -CC=gcc CXX=gcc CXXFLAGS=-O3 ./configure --prefix=/usr/local/mysql - - With the Digital compiler "C++ V6.1-029," the following should - work: -CC=cc -pthread -CFLAGS=-O4 -ansi_alias -ansi_args -fast -inline speed \ - -speculate all -arch host -CXX=cxx -pthread -CXXFLAGS=-O4 -ansi_alias -ansi_args -fast -inline speed \ - -speculate all -arch host -noexceptions -nortti -export CC CFLAGS CXX CXXFLAGS -./configure --prefix=/usr/mysql/mysql \ - --with-mysqld-ldflags=-all-static --disable-shared \ - --with-named-thread-libs="-lmach -lexc -lc" - - In some versions of OSF/1, the alloca() function is broken. Fix - this by removing the line in config.h that defines 'HAVE_ALLOCA'. - - The alloca() function also may have an incorrect prototype in - /usr/include/alloca.h. This warning resulting from this can be - ignored. + On Unix, change location into the data directory and check the + ownership of the data directory and its contents to make sure the + server has access. For example, if the data directory is + /usr/local/mysql/var, use this command: +shell> ls -la /usr/local/mysql/var - configure uses the following thread libraries automatically: - --with-named-thread-libs="-lpthread -lmach -lexc -lc". + If the data directory or its files or subdirectories are not owned + by the login account that you use for running the server, change + their ownership to that account. If the account is named mysql, + use these commands: +shell> chown -R mysql /usr/local/mysql/var +shell> chgrp -R mysql /usr/local/mysql/var - When using gcc, you can also try running configure like this: -CFLAGS=-D_PTHREAD_USE_D4 CXX=gcc CXXFLAGS=-O3 ./configure ... + If it possible that even with correct ownership, MySQL may fail to + start up if there is other security software running on your + system that manages application access to various parts of the + file system. In this case, you may need to reconfigure that + software to enable mysqld to access the directories it uses during + normal operation. - If you have problems with signals (MySQL dies unexpectedly under - high load), you may have found an OS bug with threads and signals. - In this case, you can tell MySQL not to use signals by configuring - with: -CFLAGS=-DDONT_USE_THR_ALARM \ -CXXFLAGS=-DDONT_USE_THR_ALARM \ -./configure ... + If the server fails to start up correctly, check the error log. + Log files are located in the data directory (typically C:\Program + Files\MySQL\MySQL Server 5.1\data on Windows, + /usr/local/mysql/data for a Unix binary distribution, and + /usr/local/var for a Unix source distribution). Look in the data + directory for files with names of the form host_name.err and + host_name.log, where host_name is the name of your server host. + Then examine the last few lines of these files. On Unix, you can + use tail to display them: +shell> tail host_name.err +shell> tail host_name.log - This does not affect the performance of MySQL, but has the side - effect that you can't kill clients that are "sleeping" on a - connection with mysqladmin kill or mysqladmin shutdown. Instead, - the client dies when it issues its next command. + The error log should contain information that indicates why the + server couldn't start. + + If either of the following errors occur, it means that some other + program (perhaps another mysqld server) is using the TCP/IP port + or Unix socket file that mysqld is trying to use: +Can't start server: Bind on TCP/IP port: Address already in use +Can't start server: Bind on unix socket... - With gcc 2.95.2, you may encounter the following compile error: -sql_acl.cc:1456: Internal compiler error in `scan_region', -at except.c:2566 -Please submit a full bug report. - - To fix this, you should change to the sql directory and do a - cut-and-paste of the last gcc line, but change -O3 to -O0 (or add - -O0 immediately after gcc if you don't have any -O option on your - compile line). After this is done, you can just change back to the - top-level directory and run make again. - -2.13.5.7. SGI Irix Notes - - As of MySQL 5.0, we don't provide binaries for Irix any more. - - If you are using Irix 6.5.3 or newer, mysqld is able to create - threads only if you run it as a user that has CAP_SCHED_MGT - privileges (such as root) or give the mysqld server this privilege - with the following shell command: -chcap "CAP_SCHED_MGT+epi" /opt/mysql/libexec/mysqld - - You may have to undefine some symbols in config.h after running - configure and before compiling. - - In some Irix implementations, the alloca() function is broken. If - the mysqld server dies on some SELECT statements, remove the lines - from config.h that define HAVE_ALLOC and HAVE_ALLOCA_H. If - mysqladmin create doesn't work, remove the line from config.h that - defines HAVE_READDIR_R. You may have to remove the HAVE_TERM_H - line as well. - - SGI recommends that you install all the patches on this page as a - set: - http://support.sgi.com/surfzone/patches/patchset/6.2_indigo.rps.ht - ml - - At the very minimum, you should install the latest kernel rollup, - the latest rld rollup, and the latest libc rollup. - - You definitely need all the POSIX patches on this page, for - pthreads support: - - http://support.sgi.com/surfzone/patches/patchset/6.2_posix.rps.htm - l - - If you get the something like the following error when compiling - mysql.cc: -"/usr/include/curses.h", line 82: error(1084): -invalid combination of type - - Type the following in the top-level directory of your MySQL source - tree: -extra/replace bool curses_bool < /usr/include/curses.h > include/curs -es.h -make - - There have also been reports of scheduling problems. If only one - thread is running, performance is slow. Avoid this by starting - another client. This may lead to a two-to-tenfold increase in - execution speed thereafter for the other thread. This is a poorly - understood problem with Irix threads; you may have to improvise to - find solutions until this can be fixed. + Use ps to determine whether you have another mysqld server + running. If so, shut down the server before starting mysqld again. + (If another server is running, and you really want to run multiple + servers, you can find information about how to do so in Section + 5.6, "Running Multiple MySQL Servers on the Same Machine.") - If you are compiling with gcc, you can use the following configure - command: -CC=gcc CXX=gcc CXXFLAGS=-O3 \ -./configure --prefix=/usr/local/mysql --enable-thread-safe-client \ - --with-named-thread-libs=-lpthread - - On Irix 6.5.11 with native Irix C and C++ compilers ver. 7.3.1.2, - the following is reported to work -CC=cc CXX=CC CFLAGS='-O3 -n32 -TARG:platform=IP22 -I/usr/local/includ -e \ --L/usr/local/lib' CXXFLAGS='-O3 -n32 -TARG:platform=IP22 \ --I/usr/local/include -L/usr/local/lib' \ -./configure --prefix=/usr/local/mysql --with-innodb \ - --with-libwrap=/usr/local \ - --with-named-curses-libs=/usr/local/lib/libncurses.a - -2.13.5.8. SCO UNIX and OpenServer 5.0.x Notes - - The current port is tested only on sco3.2v5.0.5, sco3.2v5.0.6, and - sco3.2v5.0.7 systems. There has also been progress on a port to - sco3.2v4.2. Open Server 5.0.8 (Legend) has native threads and - allows files greater than 2GB. The current maximum file size is - 2GB. - - We have been able to compile MySQL with the following configure - command on OpenServer with gcc 2.95.3. -CC=gcc CFLAGS="-D_FILE_OFFSET_BITS=64 -O3" \ -CXX=gcc CXXFLAGS="-D_FILE_OFFSET_BITS=64 -O3" \ -./configure --prefix=/usr/local/mysql \ - --enable-thread-safe-client --with-innodb \ - --with-openssl --with-vio --with-extra-charsets=complex - - gcc is available at - ftp://ftp.sco.com/pub/openserver5/opensrc/gnutools-5.0.7Kj. - - This development system requires the OpenServer Execution - Environment Supplement oss646B on OpenServer 5.0.6 and oss656B and - The OpenSource libraries found in gwxlibs. All OpenSource tools - are in the opensrc directory. They are available at - ftp://ftp.sco.com/pub/openserver5/opensrc/. - - Use the latest production release of MySQL. - - SCO provides operating system patches at - ftp://ftp.sco.com/pub/openserver5 for OpenServer 5.0.[0-6] and - ftp://ftp.sco.com/pub/openserverv5/507 for OpenServer 5.0.7. - - SCO provides information about security fixes at - ftp://ftp.sco.com/pub/security/OpenServer for OpenServer 5.0.x. - - The maximum file size on an OpenServer 5.0.x system is 2GB. - - The total memory which can be allocated for streams buffers, - clists, and lock records cannot exceed 60MB on OpenServer 5.0.x. - - Streams buffers are allocated in units of 4096 byte pages, clists - are 70 bytes each, and lock records are 64 bytes each, so: -(NSTRPAGES x 4096) + (NCLIST x 70) + (MAX_FLCKREC x 64) <= 62914560 - - Follow this procedure to configure the Database Services option. - If you are unsure whether an application requires this, see the - documentation provided with the application. - - 1. Log in as root. - - 2. Enable the SUDS driver by editing the /etc/conf/sdevice.d/suds - file. Change the N in the second field to a Y. - - 3. Use mkdev aio or the Hardware/Kernel Manager to enable support - for asynchronous I/O and relink the kernel. To allow users to - lock down memory for use with this type of I/O, update the - aiomemlock(F) file. This file should be updated to include the - names of users that can use AIO and the maximum amounts of - memory they can lock down. - - 4. Many applications use setuid binaries so that you need to - specify only a single user. See the documentation provided - with the application to determine whether this is the case for - your application. + If no other server is running, try to execute the command telnet + your_host_name tcp_ip_port_number. (The default MySQL port number + is 3306.) Then press Enter a couple of times. If you don't get an + error message like telnet: Unable to connect to remote host: + Connection refused, some other program is using the TCP/IP port + that mysqld is trying to use. You'll need to track down what + program this is and disable it, or else tell mysqld to listen to a + different port with the --port option. In this case, you'll also + need to specify the port number for client programs when + connecting to the server via TCP/IP. - After you complete this process, reboot the system to create a new - kernel incorporating these changes. + Another reason the port might be inaccessible is that you have a + firewall running that blocks connections to it. If so, modify the + firewall settings to allow access to the port. - By default, the entries in /etc/conf/cf.d/mtune are set as - follows: -Value Default Min Max ------ ------- --- --- -NBUF 0 24 450000 -NHBUF 0 32 524288 -NMPBUF 0 12 512 -MAX_INODE 0 100 64000 -MAX_FILE 0 100 64000 -CTBUFSIZE 128 0 256 -MAX_PROC 0 50 16000 -MAX_REGION 0 500 160000 -NCLIST 170 120 16640 -MAXUP 100 15 16000 -NOFILES 110 60 11000 -NHINODE 128 64 8192 -NAUTOUP 10 0 60 -NGROUPS 8 0 128 -BDFLUSHR 30 1 300 -MAX_FLCKREC 0 50 16000 -PUTBUFSZ 8000 2000 20000 -MAXSLICE 100 25 100 -ULIMIT 4194303 2048 4194303 -* Streams Parameters -NSTREAM 64 1 32768 -NSTRPUSH 9 9 9 -NMUXLINK 192 1 4096 -STRMSGSZ 16384 4096 524288 -STRCTLSZ 1024 1024 1024 -STRMAXBLK 524288 4096 524288 -NSTRPAGES 500 0 8000 -STRSPLITFRAC 80 50 100 -NLOG 3 3 3 -NUMSP 64 1 256 -NUMTIM 16 1 8192 -NUMTRW 16 1 8192 -* Semaphore Parameters -SEMMAP 10 10 8192 -SEMMNI 10 10 8192 -SEMMNS 60 60 8192 -SEMMNU 30 10 8192 -SEMMSL 25 25 150 -SEMOPM 10 10 1024 -SEMUME 10 10 25 -SEMVMX 32767 32767 32767 -SEMAEM 16384 16384 16384 -* Shared Memory Parameters -SHMMAX 524288 131072 2147483647 -SHMMIN 1 1 1 -SHMMNI 100 100 2000 -FILE 0 100 64000 -NMOUNT 0 4 256 -NPROC 0 50 16000 -NREGION 0 500 160000 - - Set these values as follows: - - * NOFILES should be 4096 or 2048. - - * MAXUP should be 2048. - - To make changes to the kernel, use the idtune name parameter - command. idtune modifies the /etc/conf/cf.d/stune file for you. - For example, to change SEMMS to 200, execute this command as root: -# /etc/conf/bin/idtune SEMMNS 200 - - Then rebuild and reboot the kernel by issuing this command: -# /etc/conf/bin/idbuild -B && init 6 - - To tune the system, the proper parameter values to use depend on - the number of users accessing the application or database and size - the of the database (that is, the used buffer pool). The following - kernel parameters can be set with idtune: - - * SHMMAX (recommended setting: 128MB) and SHMSEG (recommended - setting: 15). These parameters have an influence on the MySQL - database engine to create user buffer pools. - - * NOFILES and MAXUP should be set to at least 2048. - - * MAXPROC should be set to at least 3000/4000 (depends on number - of users) or more. - - * The following formulas are recommended to calculate values for - SEMMSL, SEMMNS, and SEMMNU: -SEMMSL = 13 - 13 is what has been found to be the best for both Progress and - MySQL. -SEMMNS = SEMMSL x number of db servers to be run on the system - Set SEMMNS to the value of SEMMSL multiplied by the number of - database servers (maximum) that you are running on the system - at one time. -SEMMNU = SEMMNS - Set the value of SEMMNU to equal the value of SEMMNS. You - could probably set this to 75% of SEMMNS, but this is a - conservative estimate. - - You need to at least install the SCO OpenServer Linker and - Application Development Libraries or the OpenServer Development - System to use gcc. You cannot use the GCC Dev system without - installing one of these. - - You should get the FSU Pthreads package and install it first. This - can be found at - http://moss.csc.ncsu.edu/~mueller/ftp/pub/PART/pthreads.tar.gz. - You can also get a precompiled package from - ftp://ftp.zenez.com/pub/zenez/prgms/FSU-threads-3.14.tar.gz. - - FSU Pthreads can be compiled with SCO Unix 4.2 with tcpip, or - using OpenServer 3.0 or Open Desktop 3.0 (OS 3.0 ODT 3.0) with the - SCO Development System installed using a good port of GCC 2.5.x. - For ODT or OS 3.0, you need a good port of GCC 2.5.x. There are a - lot of problems without a good port. The port for this product - requires the SCO Unix Development system. Without it, you are - missing the libraries and the linker that is needed. You also need - SCO-3.2v4.2-includes.tar.gz. This file contains the changes to the - SCO Development include files that are needed to get MySQL to - build. You need to replace the existing system include files with - these modified header files. They can be obtained from - ftp://ftp.zenez.com/pub/zenez/prgms/SCO-3.2v4.2-includes.tar.gz. - - To build FSU Pthreads on your system, all you should need to do is - run GNU make. The Makefile in FSU-threads-3.14.tar.gz is set up to - make FSU-threads. - - You can run ./configure in the threads/src directory and select - the SCO OpenServer option. This command copies Makefile.SCO5 to - Makefile. Then run make. + If the server starts but you can't connect to it, you should make + sure that you have an entry in /etc/hosts that looks like this: +127.0.0.1 localhost - To install in the default /usr/include directory, log in as root, - and then cd to the thread/src directory and run make install. + This problem occurs only on systems that do not have a working + thread library and for which MySQL must be configured to use + MIT-pthreads. - Remember that you must use GNU make to build MySQL. + If you cannot get mysqld to start, you can try to make a trace + file to find the problem by using the --debug option. See MySQL + Internals: Porting + (http://forge.mysql.com/wiki/MySQL_Internals_Porting) -Note +2.13.2. Securing the Initial MySQL Accounts - If you don't start mysqld_safe as root, you should get only the - default 110 open files per process. mysqld writes a note about - this in the log file. + Part of the MySQL installation process is to set up the mysql + database that contains the grant tables: + + * Windows distributions contain preinitialized grant tables that + are installed automatically. - With SCO 3.2V4.2, you should use FSU Pthreads version 3.14 or - newer. The following configure command should work: -CFLAGS="-D_XOPEN_XPG4" CXX=gcc CXXFLAGS="-D_XOPEN_XPG4" \ -./configure \ - --prefix=/usr/local/mysql \ - --with-named-thread-libs="-lgthreads -lsocket -lgen -lgthreads" \ - --with-named-curses-libs="-lcurses" + * On Unix, the grant tables are populated by the + mysql_install_db program. Some installation methods run this + program for you. Others require that you execute it manually. + For details, see Section 2.13.1, "Unix Post-Installation + Procedures." - You may have problems with some include files. In this case, you - can find new SCO-specific include files at - ftp://ftp.zenez.com/pub/zenez/prgms/SCO-3.2v4.2-includes.tar.gz. + The grant tables define the initial MySQL user accounts and their + access privileges. These accounts are set up as follows: - You should unpack this file in the include directory of your MySQL - source tree. + * Accounts with the user name root are created. These are + superuser accounts that can do anything. The initial root + account passwords are empty, so anyone can connect to the + MySQL server as root --- without a password --- and be granted + all privileges. - SCO development notes: + + On Windows, one root account is created; this account + allows connecting from the local host only. The Windows + installer will optionally create an account allowing for + connections from any host only if the user selects the + Enable root access from remote machines option during + installation. - * MySQL should automatically detect FSU Pthreads and link mysqld - with -lgthreads -lsocket -lgthreads. + + On Unix, both root accounts are for connections from the + local host. Connections must be made from the local host + by specifying a host name of localhost for one of the + accounts, or the actual host name or IP number for the + other. - * The SCO development libraries are re-entrant in FSU Pthreads. - SCO claims that its library functions are re-entrant, so they - must be re-entrant with FSU Pthreads. FSU Pthreads on - OpenServer tries to use the SCO scheme to make re-entrant - libraries. + * Two anonymous-user accounts are created, each with an empty + user name. The anonymous accounts have no password, so anyone + can use them to connect to the MySQL server. - * FSU Pthreads (at least the version at ftp://ftp.zenez.com) - comes linked with GNU malloc. If you encounter problems with - memory usage, make sure that gmalloc.o is included in - libgthreads.a and libgthreads.so. + + On Windows, one anonymous account is for connections from + the local host. It has no global privileges. (Before + MySQL 5.1.16, it has all global privileges, just like the + root accounts.) The other is for connections from any + host and has all privileges for the test database and for + other databases with names that start with test. - * In FSU Pthreads, the following system calls are - pthreads-aware: read(), write(), getmsg(), connect(), - accept(), select(), and wait(). + + On Unix, both anonymous accounts are for connections from + the local host. Connections must be made from the local + host by specifying a host name of localhost for one of + the accounts, or the actual host name or IP number for + the other. These accounts have all privileges for the + test database and for other databases with names that + start with test_. - * The CSSA-2001-SCO.35.2 (the patch is listed in custom as - erg711905-dscr_remap security patch (version 2.0.0)) breaks - FSU threads and makes mysqld unstable. You have to remove this - one if you want to run mysqld on an OpenServer 5.0.6 machine. + As noted, none of the initial accounts have passwords. This means + that your MySQL installation is unprotected until you do something + about it: - * If you use SCO OpenServer 5, you may need to recompile FSU - pthreads with -DDRAFT7 in CFLAGS. Otherwise, InnoDB may hang - at a mysqld startup. + * If you want to prevent clients from connecting as anonymous + users without a password, you should either assign a password + to each anonymous account or else remove the accounts. - * SCO provides operating system patches at - ftp://ftp.sco.com/pub/openserver5 for OpenServer 5.0.x. + * You should assign a password to each MySQL root account. - * SCO provides security fixes and libsocket.so.2 at - ftp://ftp.sco.com/pub/security/OpenServer and - ftp://ftp.sco.com/pub/security/sse for OpenServer 5.0.x. + The following instructions describe how to set up passwords for + the initial MySQL accounts, first for the anonymous accounts and + then for the root accounts. Replace "newpwd" in the examples with + the actual password that you want to use. The instructions also + cover how to remove the anonymous accounts, should you prefer not + to allow anonymous access at all. - * Pre-OSR506 security fixes. Also, the telnetd fix at - ftp://stage.caldera.com/pub/security/openserver/ or - ftp://stage.caldera.com/pub/security/openserver/CSSA-2001-SCO. - 10/ as both libsocket.so.2 and libresolv.so.1 with - instructions for installing on pre-OSR506 systems. - It is probably a good idea to install these patches before - trying to compile/use MySQL. + You might want to defer setting the passwords until later, so that + you don't need to specify them while you perform additional setup + or testing. However, be sure to set them before using your + installation for production purposes. - Beginning with Legend/OpenServer 6.0.0, there are native threads - and no 2GB file size limit. + Anonymous Account Password Assignment -2.13.5.9. SCO OpenServer 6.0.x Notes + To assign passwords to the anonymous accounts, connect to the + server as root and then use either SET PASSWORD or UPDATE. In + either case, be sure to encrypt the password using the PASSWORD() + function. - OpenServer 6 includes these key improvements: + To use SET PASSWORD on Windows, do this: +shell> mysql -u root +mysql> SET PASSWORD FOR ''@'localhost' = PASSWORD('newpwd'); +mysql> SET PASSWORD FOR ''@'%' = PASSWORD('newpwd'); - * Larger file support up to 1 TB + To use SET PASSWORD on Unix, do this: +shell> mysql -u root +mysql> SET PASSWORD FOR ''@'localhost' = PASSWORD('newpwd'); +mysql> SET PASSWORD FOR ''@'host_name' = PASSWORD('newpwd'); - * Multiprocessor support increased from 4 to 32 processors + In the second SET PASSWORD statement, replace host_name with the + name of the server host. This is the name that is specified in the + Host column of the non-localhost record for root in the user + table. If you don't know what host name this is, issue the + following statement before using SET PASSWORD: +mysql> SELECT Host, User FROM mysql.user; - * Increased memory support up to 64GB + Look for the record that has root in the User column and something + other than localhost in the Host column. Then use that Host value + in the second SET PASSWORD statement. - * Extending the power of UnixWare into OpenServer 6 + Anonymous Account Removal - * Dramatic performance improvement + If you prefer to remove the anonymous accounts instead, do so as + follows: +shell> mysql -u root +mysql> DROP USER ''; - OpenServer 6.0.0 commands are organized as follows: + The DROP statement applies both to Windows and to Unix. On + Windows, if you want to remove only the anonymous account that has + the same privileges as root, do this instead: +shell> mysql -u root +mysql> DROP USER ''@'localhost'; - * /bin is for commands that behave exactly the same as on - OpenServer 5.0.x. + That account allows anonymous access but has full privileges, so + removing it improves security. - * /u95/bin is for commands that have better standards - conformance, for example Large File System (LFS) support. + root Account Password Assignment - * /udk/bin is for commands that behave the same as on UnixWare - 7.1.4. The default is for the LFS support. + You can assign passwords to the root accounts in several ways. The + following discussion demonstrates three methods: - The following is a guide to setting PATH on OpenServer 6. If the - user wants the traditional OpenServer 5.0.x then PATH should be - /bin first. If the user wants LFS support, the path should be - /u95/bin:/bin. If the user wants UnixWare 7 support first, the - path would be /udk/bin:/u95/bin:/bin:. + * Use the SET PASSWORD statement - Use the latest production release of MySQL. Should you choose to - use an older release of MySQL on OpenServer 6.0.x, you must use a - version of MySQL at least as recent as 3.22.13 to get fixes for - some portability and OS problems. + * Use the mysqladmin command-line client program - MySQL distribution files with names of the following form are tar - archives of media are tar archives of media images suitable for - installation with the SCO Software Manager (/etc/custom) on SCO - OpenServer 6: -mysql-PRODUCT-5.1.39-sco-osr6-i686.VOLS.tar + * Use the UPDATE statement - A distribution where PRODUCT is pro-cert is the Commercially - licensed MySQL Pro Certified server. A distribution where PRODUCT - is pro-gpl-cert is the MySQL Pro Certified server licensed under - the terms of the General Public License (GPL). + To assign passwords using SET PASSWORD, connect to the server as + root and issue SET PASSWORD statements. Be sure to encrypt the + password using the PASSWORD() function. - Select whichever distribution you wish to install and, after - download, extract the tar archive into an empty directory. For - example: -shell> mkdir /tmp/mysql-pro -shell> cd /tmp/mysql-pro -shell> tar xf /tmp/mysql-pro-cert-5.1.39-sco-osr6-i686.VOLS.tar + For Windows, do this: +shell> mysql -u root +mysql> SET PASSWORD FOR 'root'@'localhost' = PASSWORD('newpwd'); +mysql> SET PASSWORD FOR 'root'@'%' = PASSWORD('newpwd'); - Prior to installation, back up your data in accordance with the - procedures outlined in Section 2.12.1, "Upgrading MySQL." + For Unix, do this: +shell> mysql -u root +mysql> SET PASSWORD FOR 'root'@'localhost' = PASSWORD('newpwd'); +mysql> SET PASSWORD FOR 'root'@'host_name' = PASSWORD('newpwd'); - Remove any previously installed pkgadd version of MySQL: -shell> pkginfo mysql 2>&1 > /dev/null && pkgrm mysql + In the second SET PASSWORD statement, replace host_name with the + name of the server host. This is the same host name that you used + when you assigned the anonymous account passwords. - Install MySQL Pro from media images using the SCO Software - Manager: -shell> /etc/custom -p SCO:MySQL -i -z /tmp/mysql-pro + If the user table contains an account with User and Host values of + 'root' and '127.0.0.1', use an additional SET PASSWORD statement + to set that account's password: +mysql> SET PASSWORD FOR 'root'@'127.0.0.1' = PASSWORD('newpwd'); - Alternatively, the SCO Software Manager can be displayed - graphically by clicking on the Software Manager icon on the - desktop, selecting Software -> Install New, selecting the host, - selecting Media Images for the Media Device, and entering - /tmp/mysql-pro as the Image Directory. + To assign passwords to the root accounts using mysqladmin, execute + the following commands: +shell> mysqladmin -u root password "newpwd" +shell> mysqladmin -u root -h host_name password "newpwd" - After installation, run mkdev mysql as the root user to configure - your newly installed MySQL Pro Certified server. + These commands apply both to Windows and to Unix. In the second + command, replace host_name with the name of the server host. The + double quotes around the password are not always necessary, but + you should use them if the password contains spaces or other + characters that are special to your command interpreter. -Note + The mysqladmin method of setting the root account passwords does + not set the password for the 'root'@'127.0.0.1' account. To do so, + use SET PASSWORD as shown earlier. - The installation procedure for VOLS packages does not create the - mysql user and group that the package uses by default. You should - either create the mysql user and group, or else select a different - user and group using an option in mkdev mysql. - - If you wish to configure your MySQL Pro server to interface with - the Apache Web server via PHP, download and install the PHP update - from SCO at - ftp://ftp.sco.com/pub/updates/OpenServer/SCOSA-2006.17/. - - We have been able to compile MySQL with the following configure - command on OpenServer 6.0.x: -CC=cc CFLAGS="-D_FILE_OFFSET_BITS=64 -O3" \ -CXX=CC CXXFLAGS="-D_FILE_OFFSET_BITS=64 -O3" \ -./configure --prefix=/usr/local/mysql \ - --enable-thread-safe-client \ - --with-extra-charsets=complex \ - --build=i686-unknown-sysv5SCO_SV6.0.0 - - If you use gcc, you must use gcc 2.95.3 or newer. -CC=gcc CXX=g++ ... ./configure ... - - SCO provides OpenServer 6 operating system patches at - ftp://ftp.sco.com/pub/openserver6. - - SCO provides information about security fixes at - ftp://ftp.sco.com/pub/security/OpenServer. - - By default, the maximum file size on a OpenServer 6.0.0 system is - 1TB. Some operating system utilities have a limitation of 2GB. The - maximum possible file size on UnixWare 7 is 1TB with VXFS or HTFS. + You can also use UPDATE to modify the user table directly. The + following UPDATE statement assigns a password to all root + accounts: +shell> mysql -u root +mysql> UPDATE mysql.user SET Password = PASSWORD('newpwd') + -> WHERE User = 'root'; +mysql> FLUSH PRIVILEGES; - OpenServer 6 can be configured for large file support (file sizes - greater than 2GB) by tuning the UNIX kernel. + The UPDATE statement applies both to Windows and to Unix. - By default, the entries in /etc/conf/cf.d/mtune are set as - follows: -Value Default Min Max ------ ------- --- --- -SVMMLIM 0x9000000 0x1000000 0x7FFFFFFF -HVMMLIM 0x9000000 0x1000000 0x7FFFFFFF - - To make changes to the kernel, use the idtune name parameter - command. idtune modifies the /etc/conf/cf.d/stune file for you. To - set the kernel values, execute the following commands as root: -# /etc/conf/bin/idtune SDATLIM 0x7FFFFFFF -# /etc/conf/bin/idtune HDATLIM 0x7FFFFFFF -# /etc/conf/bin/idtune SVMMLIM 0x7FFFFFFF -# /etc/conf/bin/idtune HVMMLIM 0x7FFFFFFF -# /etc/conf/bin/idtune SFNOLIM 2048 -# /etc/conf/bin/idtune HFNOLIM 2048 - - Then rebuild and reboot the kernel by issuing this command: -# /etc/conf/bin/idbuild -B && init 6 - - To tune the system, the proper parameter values to use depend on - the number of users accessing the application or database and size - the of the database (that is, the used buffer pool). The following - kernel parameters can be set with idtune: - - * SHMMAX (recommended setting: 128MB) and SHMSEG (recommended - setting: 15). These parameters have an influence on the MySQL - database engine to create user buffer pools. - - * SFNOLIM and HFNOLIM should be at maximum 2048. - - * NPROC should be set to at least 3000/4000 (depends on number - of users). - - * The following formulas are recommended to calculate values for - SEMMSL, SEMMNS, and SEMMNU: -SEMMSL = 13 - 13 is what has been found to be the best for both Progress and - MySQL. -SEMMNS = SEMMSL x number of db servers to be run on the system - Set SEMMNS to the value of SEMMSL multiplied by the number of - database servers (maximum) that you are running on the system - at one time. -SEMMNU = SEMMNS - Set the value of SEMMNU to equal the value of SEMMNS. You - could probably set this to 75% of SEMMNS, but this is a - conservative estimate. - -2.13.5.10. SCO UnixWare 7.1.x and OpenUNIX 8.0.0 Notes - - Use the latest production release of MySQL. Should you choose to - use an older release of MySQL on UnixWare 7.1.x, you must use a - version of MySQL at least as recent as 3.22.13 to get fixes for - some portability and OS problems. - - We have been able to compile MySQL with the following configure - command on UnixWare 7.1.x: -CC="cc" CFLAGS="-I/usr/local/include" \ -CXX="CC" CXXFLAGS="-I/usr/local/include" \ -./configure --prefix=/usr/local/mysql \ - --enable-thread-safe-client \ - --with-innodb --with-openssl --with-extra-charsets=complex - - If you want to use gcc, you must use gcc 2.95.3 or newer. -CC=gcc CXX=g++ ... ./configure ... - - SCO provides operating system patches at - ftp://ftp.sco.com/pub/unixware7 for UnixWare 7.1.1, - ftp://ftp.sco.com/pub/unixware7/713/ for UnixWare 7.1.3, - ftp://ftp.sco.com/pub/unixware7/714/ for UnixWare 7.1.4, and - ftp://ftp.sco.com/pub/openunix8 for OpenUNIX 8.0.0. - - SCO provides information about security fixes at - ftp://ftp.sco.com/pub/security/OpenUNIX for OpenUNIX and - ftp://ftp.sco.com/pub/security/UnixWare for UnixWare. - - The UnixWare 7 file size limit is 1 TB with VXFS. Some OS - utilities have a limitation of 2GB. - - On UnixWare 7.1.4 you do not need to do anything to get large file - support, but to enable large file support on prior versions of - UnixWare 7.1.x, run fsadm. -# fsadm -Fvxfs -o largefiles / -# fsadm / * Note -# ulimit unlimited -# /etc/conf/bin/idtune SFSZLIM 0x7FFFFFFF ** Note -# /etc/conf/bin/idtune HFSZLIM 0x7FFFFFFF ** Note -# /etc/conf/bin/idbuild -B + After the passwords have been set, you must supply the appropriate + password whenever you connect to the server. For example, if you + want to use mysqladmin to shut down the server, you can do so + using this command: +shell> mysqladmin -u root -p shutdown +Enter password: (enter root password here) -* This should report "largefiles". -** 0x7FFFFFFF represents infinity for these values. +Note - Reboot the system using shutdown. + If you forget your root password after setting it up, Section + B.5.4.1, "How to Reset the Root Password," covers the procedure + for resetting it. - By default, the entries in /etc/conf/cf.d/mtune are set as - follows: -Value Default Min Max ------ ------- --- --- -SVMMLIM 0x9000000 0x1000000 0x7FFFFFFF -HVMMLIM 0x9000000 0x1000000 0x7FFFFFFF - - To make changes to the kernel, use the idtune name parameter - command. idtune modifies the /etc/conf/cf.d/stune file for you. To - set the kernel values, execute the following commands as root: -# /etc/conf/bin/idtune SDATLIM 0x7FFFFFFF -# /etc/conf/bin/idtune HDATLIM 0x7FFFFFFF -# /etc/conf/bin/idtune SVMMLIM 0x7FFFFFFF -# /etc/conf/bin/idtune HVMMLIM 0x7FFFFFFF -# /etc/conf/bin/idtune SFNOLIM 2048 -# /etc/conf/bin/idtune HFNOLIM 2048 - - Then rebuild and reboot the kernel by issuing this command: -# /etc/conf/bin/idbuild -B && init 6 - - To tune the system, the proper parameter values to use depend on - the number of users accessing the application or database and size - the of the database (that is, the used buffer pool). The following - kernel parameters can be set with idtune: - - * SHMMAX (recommended setting: 128MB) and SHMSEG (recommended - setting: 15). These parameters have an influence on the MySQL - database engine to create user buffer pools. - - * SFNOLIM and HFNOLIM should be at maximum 2048. - - * NPROC should be set to at least 3000/4000 (depends on number - of users). - - * The following formulas are recommended to calculate values for - SEMMSL, SEMMNS, and SEMMNU: -SEMMSL = 13 - 13 is what has been found to be the best for both Progress and - MySQL. -SEMMNS = SEMMSL x number of db servers to be run on the system - Set SEMMNS to the value of SEMMSL multiplied by the number of - database servers (maximum) that you are running on the system - at one time. -SEMMNU = SEMMNS - Set the value of SEMMNU to equal the value of SEMMNS. You - could probably set this to 75% of SEMMNS, but this is a - conservative estimate. + To set up additional accounts, you can use the GRANT statement. + For instructions, see Section 5.5.2, "Adding User Accounts." 2.14. Environment Variables @@ -9003,7 +7980,7 @@ SEMMNU = SEMMNS PATH Used by the shell to find MySQL programs. TMPDIR The directory where temporary files are created. TZ This should be set to your local time zone. See Section - B.1.4.6, "Time Zone Problems." + B.5.4.6, "Time Zone Problems." UMASK The user-file creation mode when creating files. See note following table. UMASK_DIR The user-directory creation mode when creating @@ -9047,9 +8024,9 @@ SEMMNU = SEMMNS sections describe how to do this. Perl support for MySQL must be installed if you want to run the - MySQL benchmark scripts; see Section 7.1.4, "The MySQL Benchmark + MySQL benchmark scripts; see Section 7.1.3, "The MySQL Benchmark Suite." It is also required for the MySQL Cluster ndb_size.pl - utility; see Section 17.6.21, "ndb_size.pl --- NDBCLUSTER Size + utility; see Section 17.4.21, "ndb_size.pl --- NDBCLUSTER Size Requirement Estimator." 2.15.1. Installing Perl on Unix === modified file 'INSTALL-WIN-SOURCE' --- a/INSTALL-WIN-SOURCE 2009-09-16 12:03:18 +0000 +++ b/INSTALL-WIN-SOURCE 2009-12-01 07:24:05 +0000 @@ -1,5 +1,5 @@ -2.10.6. Installing MySQL from Source on Windows +2.5.10. Installing MySQL from Source on Windows These instructions describe how to build binaries from source for MySQL 5.1 on Windows. Instructions are provided for building @@ -15,7 +15,7 @@ Note to use precompiled binary distributions of MySQL that are built specifically for optimal performance on Windows by Sun Microsystems, Inc. Instructions for installing binary - distributions are available in Section 2.3, "Installing MySQL on + distributions are available in Section 2.5, "Installing MySQL on Windows." To build MySQL on Windows from source, you must satisfy the @@ -68,7 +68,7 @@ Note * 3GB to 5GB of disk space. - The exact system requirements can be found here: + The exact system requirements for Visual Studio can be found here: http://msdn.microsoft.com/vstudio/Previous/2003/sysreqs/default.as px and http://msdn.microsoft.com/vstudio/products/sysreqs/default.aspx @@ -81,7 +81,7 @@ Note * Package a source distribution yourself from the latest Bazaar developer source tree. For instructions on pulling the latest - source files, see Section 2.10.3, "Installing from the + source files, see Section 2.3.3, "Installing from the Development Source Tree." If you find something not working as expected, or you have @@ -89,7 +89,7 @@ Note Windows, please send a message to the win32 mailing list. See Section 1.5.1, "MySQL Mailing Lists." -2.10.6.1. Building MySQL from Source Using CMake and Visual Studio +2.5.10.1. Building MySQL from Source Using CMake and Visual Studio You can build MySQL on Windows by using a combination of cmake and Microsoft Visual Studio .NET 2003 (7.1), Microsoft Visual Studio @@ -99,7 +99,7 @@ Note Note To compile from the source code on Windows you must use the - standard source distribution (for example, mysql-5.0.45.tar.gz). + standard source distribution (for example, mysql-5.1.41.tar.gz). You build from the same distribution as used to build MySQL on Unix, Linux and other platforms. Do not use the Windows Source distributions as they do not contain the necessary configuration @@ -264,5 +264,5 @@ C:\workdir> copy libmysql\libmysql.def C C:\workdir> xcopy sql-bench\*.* C:\mysql\bench /E After installation, set up and start the server in the same way as - for binary Windows distributions. See Section 2.3, "Installing + for binary Windows distributions. See Section 2.5, "Installing MySQL on Windows." === modified file 'man/comp_err.1' --- a/man/comp_err.1 2009-09-16 12:03:18 +0000 +++ b/man/comp_err.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBcomp_err\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBCOMP_ERR\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBCOMP_ERR\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -91,7 +91,8 @@ Display a help message and exit\&. .\} .\" comp_err: charset option .\" charset option: comp_err -\fB\-\-charset=\fR\fB\fIpath\fR\fR\fB, \-C \fR\fB\fIpath\fR\fR +\fB\-\-charset=\fR\fB\fIpath\fR\fR, +\fB\-C \fR\fB\fIpath\fR\fR .sp The character set directory\&. The default is \&.\&./sql/share/charsets\&. @@ -107,7 +108,8 @@ The character set directory\&. The defau .\} .\" comp_err: debug option .\" debug option: comp_err -\fB\-\-debug=\fR\fB\fIdebug_options\fR\fR\fB, \-# \fR\fB\fIdebug_options\fR\fR +\fB\-\-debug=\fR\fB\fIdebug_options\fR\fR, +\fB\-# \fR\fB\fIdebug_options\fR\fR .sp Write a debugging log\&. A typical \fIdebug_options\fR @@ -142,7 +144,8 @@ Print some debugging information when th .\} .\" comp_err: header_file option .\" header_file option: comp_err -\fB\-\-header_file=\fR\fB\fIfile_name\fR\fR\fB, \-H \fR\fB\fIfile_name\fR\fR +\fB\-\-header_file=\fR\fB\fIfile_name\fR\fR, +\fB\-H \fR\fB\fIfile_name\fR\fR .sp The name of the error header file\&. The default is mysqld_error\&.h\&. @@ -158,7 +161,8 @@ mysqld_error\&.h\&. .\} .\" comp_err: in_file option .\" in_file option: comp_err -\fB\-\-in_file=\fR\fB\fIfile_name\fR\fR\fB, \-F \fR\fB\fIfile_name\fR\fR +\fB\-\-in_file=\fR\fB\fIfile_name\fR\fR, +\fB\-F \fR\fB\fIfile_name\fR\fR .sp The name of the input file\&. The default is \&.\&./sql/share/errmsg\&.txt\&. @@ -174,7 +178,8 @@ The name of the input file\&. The defaul .\} .\" comp_err: name_file option .\" name_file option: comp_err -\fB\-\-name_file=\fR\fB\fIfile_name\fR\fR\fB, \-N \fR\fB\fIfile_name\fR\fR +\fB\-\-name_file=\fR\fB\fIfile_name\fR\fR, +\fB\-N \fR\fB\fIfile_name\fR\fR .sp The name of the error name file\&. The default is mysqld_ername\&.h\&. @@ -190,7 +195,8 @@ mysqld_ername\&.h\&. .\} .\" comp_err: out_dir option .\" out_dir option: comp_err -\fB\-\-out_dir=\fR\fB\fIpath\fR\fR\fB, \-D \fR\fB\fIpath\fR\fR +\fB\-\-out_dir=\fR\fB\fIpath\fR\fR, +\fB\-D \fR\fB\fIpath\fR\fR .sp The name of the output base directory\&. The default is \&.\&./sql/share/\&. @@ -206,7 +212,8 @@ The name of the output base directory\&. .\} .\" comp_err: out_file option .\" out_file option: comp_err -\fB\-\-out_file=\fR\fB\fIfile_name\fR\fR\fB, \-O \fR\fB\fIfile_name\fR\fR +\fB\-\-out_file=\fR\fB\fIfile_name\fR\fR, +\fB\-O \fR\fB\fIfile_name\fR\fR .sp The name of the output file\&. The default is errmsg\&.sys\&. @@ -222,7 +229,8 @@ errmsg\&.sys\&. .\} .\" comp_err: statefile option .\" statefile option: comp_err -\fB\-\-statefile=\fR\fB\fIfile_name\fR\fR\fB, \-S \fR\fB\fIfile_name\fR\fR +\fB\-\-statefile=\fR\fB\fIfile_name\fR\fR, +\fB\-S \fR\fB\fIfile_name\fR\fR .sp The name for the SQLSTATE header file\&. The default is sql_state\&.h\&. === modified file 'man/innochecksum.1' --- a/man/innochecksum.1 2009-09-16 12:03:18 +0000 +++ b/man/innochecksum.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBinnochecksum\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBINNOCHECKSUM\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBINNOCHECKSUM\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -29,7 +29,20 @@ innochecksum \- offline InnoDB file chec \fBinnochecksum\fR prints checksums for InnoDB -files\&. +files\&. This tool reads an +InnoDB +tablespace file, calculates the checksum for each page, compares the calculated checksum to the stored checksum, and reports mismatches, which indicate damaged pages\&. It was originally developed to speed up verifying the integrity of tablespace files after power outages but can also be used after file copies\&. Because checksum mismatches will cause +InnoDB +to deliberately shut down a running server, it can be preferable to use this tool rather than waiting for a server in production usage to encounter the damaged pages\&. +.PP +\fBinnochecksum\fR +cannot be used on tablespace files that the server already has open\&. For such files, you should use +CHECK TABLE +to check tables within the tablespace\&. +.PP +If checksum mismatches are found, you would normally restore the tablespace from backup or start the server and attempt to use +\fBmysqldump\fR +to make a backup of the tables within the tablespace\&. .PP Invoke \fBinnochecksum\fR === modified file 'man/make_win_bin_dist.1' --- a/man/make_win_bin_dist.1 2009-09-16 12:03:18 +0000 +++ b/man/make_win_bin_dist.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmake_win_bin_dist\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMAKE_WIN_BIN_DIST" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMAKE_WIN_BIN_DIST" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/msql2mysql.1' --- a/man/msql2mysql.1 2009-09-16 12:03:18 +0000 +++ b/man/msql2mysql.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmsql2mysql\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMSQL2MYSQL\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMSQL2MYSQL\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/my_print_defaults.1' --- a/man/my_print_defaults.1 2009-09-16 12:03:18 +0000 +++ b/man/my_print_defaults.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmy_print_defaults\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMY_PRINT_DEFAULTS" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMY_PRINT_DEFAULTS" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -100,7 +100,8 @@ Read only the given option file\&. .\} .\" my_print_defaults: debug option .\" debug option: my_print_defaults -\fB\-\-debug=\fR\fB\fIdebug_options\fR\fR\fB, \-# \fR\fB\fIdebug_options\fR\fR +\fB\-\-debug=\fR\fB\fIdebug_options\fR\fR, +\fB\-# \fR\fB\fIdebug_options\fR\fR .sp Write a debugging log\&. A typical \fIdebug_options\fR === modified file 'man/myisam_ftdump.1' --- a/man/myisam_ftdump.1 2009-09-16 12:03:18 +0000 +++ b/man/myisam_ftdump.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmyisam_ftdump\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYISAM_FTDUMP\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYISAM_FTDUMP\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/myisamchk.1' --- a/man/myisamchk.1 2009-09-16 12:03:18 +0000 +++ b/man/myisamchk.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmyisamchk\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYISAMCHK\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYISAMCHK\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -275,7 +275,8 @@ Display a help message and exit\&. Optio .\} .\" myisamchk: debug option .\" debug option: myisamchk -\fB\-\-debug=\fR\fB\fIdebug_options\fR\fR\fB, \-# \fR\fB\fIdebug_options\fR\fR +\fB\-\-debug=\fR\fB\fIdebug_options\fR\fR, +\fB\-# \fR\fB\fIdebug_options\fR\fR .sp Write a debugging log\&. A typical \fIdebug_options\fR @@ -520,7 +521,7 @@ system variable\&. For more information, myisam_stats_method in Section\ \&5.1.4, \(lqServer System Variables\(rq, and -Section\ \&7.4.6, \(lqMyISAM Index Statistics Collection\(rq\&. For MySQL 5\&.1, +Section\ \&7.4.7, \(lqMyISAM Index Statistics Collection\(rq\&. For MySQL 5\&.1, stats_method was added in MySQL 5\&.0\&.14\&. For older versions, the statistics collection method is equivalent to nulls_equal\&. @@ -662,6 +663,9 @@ If you are using and have plenty of memory, setting the key_buffer_size variable to a large value helps the repair operation run faster\&. +.sp +For a description of the output format, see +the section called \(lqMYISAMCHK TABLE INFORMATION\(rq\&. .RE .sp .RS 4 @@ -782,7 +786,11 @@ server is using the table and you are ru .\" files: repairing .PP \fBmyisamchk\fR -supports the following options for table repair operations: +supports the following options for table repair operations (operations performed when an option such as +\fB\-\-recover\fR +or +\fB\-\-safe\-recover\fR +is given): .sp .RS 4 .ie n \{\ @@ -844,7 +852,8 @@ Correct the checksum information for the .\} .\" myisamchk: data-file-length option .\" data-file-length option: myisamchk -\fB\-\-data\-file\-length=\fR\fB\fIlen\fR\fR\fB, \-D \fR\fB\fIlen\fR\fR +\fB\-\-data\-file\-length=\fR\fB\fIlen\fR\fR, +\fB\-D \fR\fB\fIlen\fR\fR .sp The maximum length of the data file (when re\-creating data file when it is \(lqfull\(rq)\&. @@ -864,6 +873,9 @@ The maximum length of the data file (whe \fB\-e\fR .sp Do a repair that tries to recover every possible row from the data file\&. Normally, this also finds a lot of garbage rows\&. Do not use this option unless you are desperate\&. +.sp +For a description of the output format, see +the section called \(lqMYISAMCHK TABLE INFORMATION\(rq\&. .RE .sp .RS 4 @@ -1172,7 +1184,10 @@ Find the record that a block at the give \fB\-\-description\fR, \fB\-d\fR .sp -Print some descriptive information about the table\&. +Print some descriptive information about the table\&. Specifying the +\fB\-\-verbose\fR +option once or twice produces additional information\&. See +the section called \(lqMYISAMCHK TABLE INFORMATION\(rq\&. .RE .sp .RS 4 @@ -1243,47 +1258,187 @@ sorts and moves records, it just overwri \fBmyisamchk\fR must unpack key blocks first, then re\-create indexes and pack the key blocks again\&. (In this case, re\-creating indexes is faster than updating offsets for each index\&.) .RE -.SH "MYISAMCHK MEMORY USAGE" -.\" memory usage: myisamchk +.SH "MYISAMCHK TABLE INFORMATION" +.\" table description: myisamchk +.\" tables: information +.\" examples: myisamchk output +.\" myisamchk: example output .PP -Memory allocation is important when you run -\fBmyisamchk\fR\&. +To obtain a description of a +MyISAM +table or statistics about it, use the commands shown here\&. The output from these commands is explained later in this section\&. +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +\fBmyisamchk \-d \fR\fB\fItbl_name\fR\fR +.sp +Runs \fBmyisamchk\fR -uses no more memory than its memory\-related variables are set to\&. If you are going to use +in +\(lqdescribe mode\(rq +to produce a description of your table\&. If you start the MySQL server with external locking disabled, \fBmyisamchk\fR -on very large tables, you should first decide how much memory you want it to use\&. The default is to use only about 3MB to perform repairs\&. By using larger values, you can get +may report an error for a table that is updated while it runs\&. However, because \fBmyisamchk\fR -to operate faster\&. For example, if you have more than 32MB RAM, you could use options such as these (in addition to any other options you might specify): +does not change the table in describe mode, there is no risk of destroying data\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +\fBmyisamchk \-dv \fR\fB\fItbl_name\fR\fR +.sp +Adding +\fB\-v\fR +runs +\fBmyisamchk\fR +in verbose mode so that it produces more information about the table\&. Adding +\fB\-v\fR +a second time produces even more information\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +\fBmyisamchk \-eis \fR\fB\fItbl_name\fR\fR +.sp +Shows only the most important information from a table\&. This operation is slow because it must read the entire table\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +\fBmyisamchk \-eiv \fR\fB\fItbl_name\fR\fR +.sp +This is like +\fB\-eis\fR, but tells you what is being done\&. +.RE +.PP +The +\fItbl_name\fR +argument can be either the name of a +MyISAM +table or the name of its index file, as described in +\fBmyisamchk\fR(1)\&. Multiple +\fItbl_name\fR +arguments can be given\&. +.PP +Suppose that a table named +person +has the following structure\&. (The +MAX_ROWS +table option is included so that in the example output from +\fBmyisamchk\fR +shown later, some values are smaller and fit the output format more easily\&.) .sp .if n \{\ .RS 4 .\} .nf -shell> \fBmyisamchk \-\-sort_buffer_size=16M \-\-key_buffer_size=16M \e\fR - \fB\-\-read_buffer_size=1M \-\-write_buffer_size=1M \&.\&.\&.\fR +CREATE TABLE person +( + id INT NOT NULL AUTO_INCREMENT, + last_name VARCHAR(20) NOT NULL, + first_name VARCHAR(20) NOT NULL, + birth DATE, + death DATE, + PRIMARY KEY (id), + INDEX (last_name, first_name), + INDEX (birth) +) MAX_ROWS = 1000000; .fi .if n \{\ .RE .\} .PP -Using -\fB\-\-sort_buffer_size=16M\fR -should probably be enough for most cases\&. +Suppose also that the table has these data and index file sizes: +.sp +.if n \{\ +.RS 4 +.\} +.nf +\-rw\-rw\-\-\-\- 1 mysql mysql 9347072 Aug 19 11:47 person\&.MYD +\-rw\-rw\-\-\-\- 1 mysql mysql 6066176 Aug 19 11:47 person\&.MYI +.fi +.if n \{\ +.RE +.\} .PP -Be aware that -\fBmyisamchk\fR -uses temporary files in -TMPDIR\&. If -TMPDIR -points to a memory file system, you may easily get out of memory errors\&. If this happens, run -\fBmyisamchk\fR -with the -\fB\-\-tmpdir=\fR\fB\fIpath\fR\fR -option to specify some directory located on a file system that has more space\&. +Example of +\fBmyisamchk \-dvv\fR +output: +.sp +.if n \{\ +.RS 4 +.\} +.nf +MyISAM file: person +Record format: Packed +Character set: latin1_swedish_ci (8) +File\-version: 1 +Creation time: 2009\-08\-19 16:47:41 +Recover time: 2009\-08\-19 16:47:56 +Status: checked,analyzed,optimized keys +Auto increment key: 1 Last value: 306688 +Data records: 306688 Deleted blocks: 0 +Datafile parts: 306688 Deleted data: 0 +Datafile pointer (bytes): 4 Keyfile pointer (bytes): 3 +Datafile length: 9347072 Keyfile length: 6066176 +Max datafile length: 4294967294 Max keyfile length: 17179868159 +Recordlength: 54 +table description: +Key Start Len Index Type Rec/key Root Blocksize +1 2 4 unique long 1 99328 1024 +2 6 20 multip\&. varchar prefix 512 3563520 1024 + 27 20 varchar 512 +3 48 3 multip\&. uint24 NULL 306688 6065152 1024 +Field Start Length Nullpos Nullbit Type +1 1 1 +2 2 4 no zeros +3 6 21 varchar +4 27 21 varchar +5 48 3 1 1 no zeros +6 51 3 1 2 no zeros +.fi +.if n \{\ +.RE +.\} .PP -When repairing, +Explanations for the types of information \fBmyisamchk\fR -also needs a lot of disk space: +produces are given here\&. +\(lqKeyfile\(rq +refers to the index file\&. +\(lqRecord\(rq +and +\(lqrow\(rq +are synonymous, as are +\(lqfield\(rq +and +\(lqcolumn\&.\(rq +.PP +The initial part of the table description contains these values: .sp .RS 4 .ie n \{\ @@ -1293,9 +1448,11 @@ also needs a lot of disk space: .sp -1 .IP \(bu 2.3 .\} -Double the size of the data file (the original file and a copy)\&. This space is not needed if you do a repair with -\fB\-\-quick\fR; in this case, only the index file is re\-created\&. -\fIThis space must be available on the same file system as the original data file\fR, as the copy is created in the same directory as the original\&. +MyISAM file +.sp +Name of the +MyISAM +(index) file\&. .RE .sp .RS 4 @@ -1306,7 +1463,13 @@ Double the size of the data file (the or .sp -1 .IP \(bu 2.3 .\} -Space for the new index file that replaces the old one\&. The old index file is truncated at the start of the repair operation, so you usually ignore this space\&. This space must be available on the same file system as the original data file\&. +Record format +.sp +The format used to store table rows\&. The preceding examples use +Fixed length\&. Other possible values are +Compressed +and +Packed\&. .RE .sp .RS 4 @@ -1317,30 +1480,964 @@ Space for the new index file that replac .sp -1 .IP \(bu 2.3 .\} -When using -\fB\-\-recover\fR -or -\fB\-\-sort\-recover\fR -(but not when using -\fB\-\-safe\-recover\fR), you need space for a sort buffer\&. The following formula yields the amount of space required: +Chararacter set +.sp +The table default character set\&. +.RE .sp -.if n \{\ .RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c .\} -.nf -(\fIlargest_key\fR + \fIrow_pointer_length\fR) \(mu \fInumber_of_rows\fR \(mu 2 -.fi -.if n \{\ +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +File\-version +.sp +Version of +MyISAM +format\&. Currently always 1\&. .RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 .\} +Creation time .sp -You can check the length of the keys and the -row_pointer_length -with -\fBmyisamchk \-dv \fR\fB\fItbl_name\fR\fR\&. This space is allocated in the temporary directory (specified by -TMPDIR -or -\fB\-\-tmpdir=\fR\fB\fIpath\fR\fR)\&. +When the data file was created\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Recover time +.sp +When the index/data file was last reconstructed\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Status +.sp +Table status flags\&. Possible values are +crashed, +open, +changed, +analyzed, +optimized keys, and +sorted index pages\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Auto increment key, +Last value +.sp +The key number associated the table\'s +AUTO_INCREMENT +column, and the most recently generated value for this column\&. These fields do not appear if there is no such column\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Data records +.sp +The number of rows in the table\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Deleted blocks +.sp +How many deleted blocks still have reserved space\&. You can optimize your table to minimize this space\&. See +Section\ \&6.4.4, \(lqTable Optimization\(rq\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Datafile parts +.sp +For dynamic\-row format, this indicates how many data blocks there are\&. For an optimized table without fragmented rows, this is the same as +Data records\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Deleted data +.sp +How many bytes of unreclaimed deleted data there are\&. You can optimize your table to minimize this space\&. See +Section\ \&6.4.4, \(lqTable Optimization\(rq\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Datafile pointer +.sp +The size of the data file pointer, in bytes\&. It is usually 2, 3, 4, or 5 bytes\&. Most tables manage with 2 bytes, but this cannot be controlled from MySQL yet\&. For fixed tables, this is a row address\&. For dynamic tables, this is a byte address\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Keyfile pointer +.sp +The size of the index file pointer, in bytes\&. It is usually 1, 2, or 3 bytes\&. Most tables manage with 2 bytes, but this is calculated automatically by MySQL\&. It is always a block address\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Max datafile length +.sp +How long the table data file can become, in bytes\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Max keyfile length +.sp +How long the table index file can become, in bytes\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Recordlength +.sp +How much space each row takes, in bytes\&. +.RE +.PP +The +table description +part of the output includes a list of all keys in the table\&. For each key, +\fBmyisamchk\fR +displays some low\-level information: +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Key +.sp +This key\'s number\&. This value is shown only for the first column of the key\&. If this value is missing, the line corresponds to the second or later column of a multiple\-column key\&. For the table shown in the example, there are two +table description +lines for the second index\&. This indicates that it is a multiple\-part index with two parts\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Start +.sp +Where in the row this portion of the index starts\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Len +.sp +How long this portion of the index is\&. For packed numbers, this should always be the full length of the column\&. For strings, it may be shorter than the full length of the indexed column, because you can index a prefix of a string column\&. The total length of a multiple\-part key is the sum of the +Len +values for all key parts\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Index +.sp +Whether a key value can exist multiple times in the index\&. Possible values are +unique +or +multip\&. +(multiple)\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Type +.sp +What data type this portion of the index has\&. This is a +MyISAM +data type with the possible values +packed, +stripped, or +empty\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Root +.sp +Address of the root index block\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Blocksize +.sp +The size of each index block\&. By default this is 1024, but the value may be changed at compile time when MySQL is built from source\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Rec/key +.sp +This is a statistical value used by the optimizer\&. It tells how many rows there are per value for this index\&. A unique index always has a value of 1\&. This may be updated after a table is loaded (or greatly changed) with +\fBmyisamchk \-a\fR\&. If this is not updated at all, a default value of 30 is given\&. +.RE +.PP +The last part of the output provides information about each column: +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Field +.sp +The column number\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Start +.sp +The byte position of the column within table rows\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Length +.sp +The length of the column in bytes\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Nullpos, +Nullbit +.sp +For columns that can be +NULL, +MyISAM +stores +NULL +values as a flag in a byte\&. Depending on how many nullable columns there are, there can be one or more bytes used for this purpose\&. The +Nullpos +and +Nullbit +values, if nonempty, indicate which byte and bit contains that flag indicating whether the column is +NULL\&. +.sp +The position and number of bytes used to store +NULL +flags is shown in the line for field 1\&. This is why there are six +Field +lines for the +person +table even though it has only five columns\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Type +.sp +The data type\&. The value may contain any of the following descriptors: +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +constant +.sp +All rows have the same value\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +no endspace +.sp +Do not store endspace\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +no endspace, not_always +.sp +Do not store endspace and do not do endspace compression for all values\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +no endspace, no empty +.sp +Do not store endspace\&. Do not store empty values\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +table\-lookup +.sp +The column was converted to an +ENUM\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +zerofill(\fIN\fR) +.sp +The most significant +\fIN\fR +bytes in the value are always 0 and are not stored\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +no zeros +.sp +Do not store zeros\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +always zero +.sp +Zero values are stored using one bit\&. +.RE +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Huff tree +.sp +The number of the Huffman tree associated with the column\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Bits +.sp +The number of bits used in the Huffman tree\&. +.RE +.PP +The +Huff tree +and +Bits +fields are displayed if the table has been compressed with +\fBmyisampack\fR\&. See +\fBmyisampack\fR(1), for an example of this information\&. +.PP +Example of +\fBmyisamchk \-eiv\fR +output: +.sp +.if n \{\ +.RS 4 +.\} +.nf +Checking MyISAM file: person +Data records: 306688 Deleted blocks: 0 +\- check file\-size +\- check record delete\-chain +No recordlinks +\- check key delete\-chain +block_size 1024: +\- check index reference +\- check data record references index: 1 +Key: 1: Keyblocks used: 98% Packed: 0% Max levels: 3 +\- check data record references index: 2 +Key: 2: Keyblocks used: 99% Packed: 97% Max levels: 3 +\- check data record references index: 3 +Key: 3: Keyblocks used: 98% Packed: \-14% Max levels: 3 +Total: Keyblocks used: 98% Packed: 89% +\- check records and index references +\fI*** LOTS OF ROW NUMBERS DELETED ***\fR +Records: 306688 M\&.recordlength: 25 Packed: 83% +Recordspace used: 97% Empty space: 2% Blocks/Record: 1\&.00 +Record blocks: 306688 Delete blocks: 0 +Record data: 7934464 Deleted data: 0 +Lost space: 256512 Linkdata: 1156096 +User time 43\&.08, System time 1\&.68 +Maximum resident set size 0, Integral resident set size 0 +Non\-physical pagefaults 0, Physical pagefaults 0, Swaps 0 +Blocks in 0 out 7, Messages in 0 out 0, Signals 0 +Voluntary context switches 0, Involuntary context switches 0 +Maximum memory usage: 1046926 bytes (1023k) +.fi +.if n \{\ +.RE +.\} +.PP +\fBmyisamchk \-eiv\fR +output includes the following information: +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Data records +.sp +The number of rows in the table\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Deleted blocks +.sp +How many deleted blocks still have reserved space\&. You can optimize your table to minimize this space\&. See +Section\ \&6.4.4, \(lqTable Optimization\(rq\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Key +.sp +The key number\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Keyblocks used +.sp +What percentage of the keyblocks are used\&. When a table has just been reorganized with +\fBmyisamchk\fR, the values are very high (very near theoretical maximum)\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Packed +.sp +MySQL tries to pack key values that have a common suffix\&. This can only be used for indexes on +CHAR +and +VARCHAR +columns\&. For long indexed strings that have similar leftmost parts, this can significantly reduce the space used\&. In the preceding example, the second key is 40 bytes long and a 97% reduction in space is achieved\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Max levels +.sp +How deep the B\-tree for this key is\&. Large tables with long key values get high values\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Records +.sp +How many rows are in the table\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +M\&.recordlength +.sp +The average row length\&. This is the exact row length for tables with fixed\-length rows, because all rows have the same length\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Packed +.sp +MySQL strips spaces from the end of strings\&. The +Packed +value indicates the percentage of savings achieved by doing this\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Recordspace used +.sp +What percentage of the data file is used\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Empty space +.sp +What percentage of the data file is unused\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Blocks/Record +.sp +Average number of blocks per row (that is, how many links a fragmented row is composed of)\&. This is always 1\&.0 for fixed\-format tables\&. This value should stay as close to 1\&.0 as possible\&. If it gets too large, you can reorganize the table\&. See +Section\ \&6.4.4, \(lqTable Optimization\(rq\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Recordblocks +.sp +How many blocks (links) are used\&. For fixed\-format tables, this is the same as the number of rows\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Deleteblocks +.sp +How many blocks (links) are deleted\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Recorddata +.sp +How many bytes in the data file are used\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Deleted data +.sp +How many bytes in the data file are deleted (unused)\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Lost space +.sp +If a row is updated to a shorter length, some space is lost\&. This is the sum of all such losses, in bytes\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Linkdata +.sp +When the dynamic table format is used, row fragments are linked with pointers (4 to 7 bytes each)\&. +Linkdata +is the sum of the amount of storage used by all such pointers\&. +.RE +.SH "MYISAMCHK MEMORY USAGE" +.\" memory usage: myisamchk +.PP +Memory allocation is important when you run +\fBmyisamchk\fR\&. +\fBmyisamchk\fR +uses no more memory than its memory\-related variables are set to\&. If you are going to use +\fBmyisamchk\fR +on very large tables, you should first decide how much memory you want it to use\&. The default is to use only about 3MB to perform repairs\&. By using larger values, you can get +\fBmyisamchk\fR +to operate faster\&. For example, if you have more than 32MB RAM, you could use options such as these (in addition to any other options you might specify): +.sp +.if n \{\ +.RS 4 +.\} +.nf +shell> \fBmyisamchk \-\-sort_buffer_size=16M \e\fR + \fB\-\-key_buffer_size=16M \e\fR + \fB\-\-read_buffer_size=1M \e\fR + \fB\-\-write_buffer_size=1M \&.\&.\&.\fR +.fi +.if n \{\ +.RE +.\} +.PP +Using +\fB\-\-sort_buffer_size=16M\fR +should probably be enough for most cases\&. +.PP +Be aware that +\fBmyisamchk\fR +uses temporary files in +TMPDIR\&. If +TMPDIR +points to a memory file system, out of memory errors can easily occur\&. If this happens, run +\fBmyisamchk\fR +with the +\fB\-\-tmpdir=\fR\fB\fIpath\fR\fR +option to specify a directory located on a file system that has more space\&. +.PP +When performing repair operations, +\fBmyisamchk\fR +also needs a lot of disk space: +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Twice the size of the data file (the original file and a copy)\&. This space is not needed if you do a repair with +\fB\-\-quick\fR; in this case, only the index file is re\-created\&. +\fIThis space must be available on the same file system as the original data file\fR, as the copy is created in the same directory as the original\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Space for the new index file that replaces the old one\&. The old index file is truncated at the start of the repair operation, so you usually ignore this space\&. This space must be available on the same file system as the original data file\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +When using +\fB\-\-recover\fR +or +\fB\-\-sort\-recover\fR +(but not when using +\fB\-\-safe\-recover\fR), you need space on disk for sorting\&. This space is allocated in the temporary directory (specified by +TMPDIR +or +\fB\-\-tmpdir=\fR\fB\fIpath\fR\fR)\&. The following formula yields the amount of space required: +.sp +.if n \{\ +.RS 4 +.\} +.nf +(\fIlargest_key\fR + \fIrow_pointer_length\fR) \(mu \fInumber_of_rows\fR \(mu 2 +.fi +.if n \{\ +.RE +.\} +.sp +You can check the length of the keys and the +\fIrow_pointer_length\fR +with +\fBmyisamchk \-dv \fR\fB\fItbl_name\fR\fR +(see +the section called \(lqMYISAMCHK TABLE INFORMATION\(rq)\&. The +\fIrow_pointer_length\fR +and +\fInumber_of_rows\fR +values are the +Datafile pointer +and +Data records +values in the table description\&. To determine the +\fIlargest_key\fR +value, check the +Key +lines in the table description\&. The +Len +column indicates the number of bytes for each key part\&. For a multiple\-column index, the key size is the sum of the +Len +values for all key parts\&. .RE .PP If you have a problem with disk space during repair, you can try === modified file 'man/myisamlog.1' --- a/man/myisamlog.1 2009-09-16 12:03:18 +0000 +++ b/man/myisamlog.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmyisamlog\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYISAMLOG\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYISAMLOG\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/myisampack.1' --- a/man/myisampack.1 2009-09-16 12:03:18 +0000 +++ b/man/myisampack.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmyisampack\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYISAMPACK\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYISAMPACK\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -652,9 +652,11 @@ After join The number of distinct Huffman trees left after joining trees to save some header space\&. .RE .PP -After a table has been compressed, +After a table has been compressed, the +Field +lines displayed by \fBmyisamchk \-dvv\fR -prints additional information about each column: +include additional information about each column: .sp .RS 4 .ie n \{\ === modified file 'man/mysql-stress-test.pl.1' --- a/man/mysql-stress-test.pl.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql-stress-test.pl.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql-stress-test.pl\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/08/2009 +.\" Date: 10/29/2009 .\" Manual: MySQL Database System .\" Source: MySQL .\" Language: English .\" -.TH "\FBMYSQL\-STRESS\-TE" "1" "08/08/2009" "MySQL" "MySQL Database System" +.TH "\FBMYSQL\-STRESS\-TE" "1" "10/29/2009" "MySQL" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysql-test-run.pl.1' --- a/man/mysql-test-run.pl.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql-test-run.pl.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql-test-run.pl\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/08/2009 +.\" Date: 10/29/2009 .\" Manual: MySQL Database System .\" Source: MySQL .\" Language: English .\" -.TH "\FBMYSQL\-TEST\-RUN\" "1" "08/08/2009" "MySQL" "MySQL Database System" +.TH "\FBMYSQL\-TEST\-RUN\" "1" "10/29/2009" "MySQL" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -700,6 +700,9 @@ Specify a file that contains a list of t code rather than [ fail ] if they fail\&. This option was added in MySQL 5\&.1\&.33/6\&.0\&.11\&. +.sp +For an example of a file that might be specified via this option, see +mysql\-test/collections/default\&.experimental\&. .RE .sp .RS 4 @@ -1132,6 +1135,26 @@ not to generate a timing file\&. .sp -1 .IP \(bu 2.3 .\} +.\" mysql-test-run.pl: parallel option +.\" parallel option: mysql-test-run.pl +\fB\-\-parallel={\fR\fB\fIN\fR\fR\fB|auto}\fR +.sp +Run tests using +\fIN\fR +parallel threads\&. By default, 1 thread is used\&. Use +\fB\-\-parallel=auto\fR +for auto\-setting of +\fIN\fR\&. This option was added in MySQL 5\&.1\&.36\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} .\" mysql-test-run.pl: ps-protocol option .\" ps-protocol option: mysql-test-run.pl \fB\-\-ps\-protocol\fR === modified file 'man/mysql.1' --- a/man/mysql.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -334,8 +334,18 @@ to display result set metadata\&. .sp Use \fIcharset_name\fR -as the default character set\&. See -Section\ \&9.2, \(lqThe Character Set Used for Data and Sorting\(rq\&. +as the default character set for the client and connection\&. +.sp +A common issue that can occur when the operating system uses +utf8 +or another multi\-byte character set is that output from the +\fBmysql\fR +client is formatted incorrectly, due to the fact that the MySQL client uses the +latin1 +character set by default\&. You can usually fix such issues by using this option to force the client to use the system character set instead\&. +.sp +See +Section\ \&9.2, \(lqThe Character Set Used for Data and Sorting\(rq, for more information\&. .RE .sp .RS 4 @@ -648,7 +658,7 @@ PAGER environment variable\&. Valid pagers are \fBless\fR, \fBmore\fR, -\fBcat [> filename]\fR, and so forth\&. This option works only on Unix\&. It does not work in batch mode\&. To disable paging, use +\fBcat [> filename]\fR, and so forth\&. This option works only on Unix and only in interactive mode\&. To disable paging, use \fB\-\-skip\-pager\fR\&. the section called \(lqMYSQL COMMANDS\(rq, discusses output paging further\&. .RE @@ -1026,7 +1036,7 @@ Display output in table format\&. This i .\" tee option: mysql \fB\-\-tee=\fR\fB\fIfile_name\fR\fR .sp -Append a copy of output to the given file\&. This option does not work in batch mode\&. +Append a copy of output to the given file\&. This option works only in interactive mode\&. the section called \(lqMYSQL COMMANDS\(rq, discusses tee files further\&. .RE .sp @@ -1523,7 +1533,7 @@ is set to something other than the defau \(lq;\(rq, instances of that character are sent to the server without interpretation\&. However, the server itself still interprets \(lq;\(rq as a statement delimiter and processes statements accordingly\&. This behavior on the server side comes into play for multiple\-statement execution (see -Section\ \&21.10.12, \(lqC API Support for Multiple Statement Execution\(rq), and for parsing the body of stored procedures and functions, triggers, and events (see +Section\ \&21.9.12, \(lqC API Support for Multiple Statement Execution\(rq), and for parsing the body of stored procedures and functions, triggers, and events (see Section\ \&19.1, \(lqDefining Stored Programs\(rq)\&. .RE .sp @@ -1680,7 +1690,7 @@ option when you invoke \fBmysql\fR checks the value of the PAGER -environment variable and sets the pager to that\&. +environment variable and sets the pager to that\&. Pager functionality works only in interactive mode\&. .sp Output paging can be enabled interactively with the \fBpager\fR @@ -1853,7 +1863,7 @@ By using the option when you invoke \fBmysql\fR, you can log statements and their output\&. All the data displayed on the screen is appended into a given file\&. This can be very useful for debugging purposes also\&. \fBmysql\fR -flushes results to the file after each statement, just before it prints its next prompt\&. +flushes results to the file after each statement, just before it prints its next prompt\&. Tee functionality works only in interactive mode\&. .sp You can enable this feature interactively with the \fBtee\fR @@ -2334,7 +2344,7 @@ prompt=(\e\eu@\e\eh) [\e\ed]>\e\e_ .sp In this example, note that the backslashes are doubled\&. If you set the prompt using the prompt -option in an option file, it is advisable to double the backslashes when using the special prompt options\&. There is some overlap in the set of allowable prompt options and the set of special escape sequences that are recognized in option files\&. (These sequences are listed in +option in an option file, it is advisable to double the backslashes when using the special prompt options\&. There is some overlap in the set of allowable prompt options and the set of special escape sequences that are recognized in option files\&. (The rules for escape sequences in option files are listed in Section\ \&4.2.3.3, \(lqUsing Option Files\(rq\&.) The overlap may cause you problems if you use single backslashes\&. For example, \es is interpreted as a space rather than as the current seconds value\&. The following example shows how to define a prompt within an option file to include the current time in @@ -2586,6 +2596,12 @@ SELECT \'<info_to_display>\' AS \' \'; The statement shown outputs <info_to_display>\&. .PP +You can also invoke +\fBmysql\fR +with the +\fB\-\-verbose\fR +option, which causes each statement to be displayed before the result that it produces\&. +.PP As of MySQL 5\&.1\&.23, \fBmysql\fR ignores Unicode byte order mark (BOM) characters at the beginning of input files\&. Previously, it read them and sent them to the server, resulting in a syntax error\&. Presence of a BOM does not cause @@ -2785,7 +2801,7 @@ client with the option\&. .PP For more information about auto\-reconnect and its effect on state information when a reconnection occurs, see -Section\ \&21.10.11, \(lqControlling Automatic Reconnection Behavior\(rq\&. +Section\ \&21.9.11, \(lqControlling Automatic Reconnection Behavior\(rq\&. .SH "COPYRIGHT" .br .PP === modified file 'man/mysql.server.1' --- a/man/mysql.server.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql.server.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql.server\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL\&.SERVER\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL\&.SERVER\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -40,7 +40,7 @@ script will be installed in the /etc/init\&.d directory with the name mysql\&. You need not install it manually\&. See -Section\ \&2.4, \(lqInstalling MySQL from RPM Packages on Linux\(rq, for more information on the Linux RPM packages\&. +Section\ \&2.6.1, \(lqInstalling MySQL from RPM Packages on Linux\(rq, for more information on the Linux RPM packages\&. .PP Some vendors provide RPM packages that install a startup script under a different name such as \fBmysqld\fR\&. @@ -48,7 +48,7 @@ Some vendors provide RPM packages that i If you install MySQL from a source distribution or using a binary distribution format that does not install \fBmysql\&.server\fR automatically, you can install it manually\&. Instructions are provided in -Section\ \&2.11.2.2, \(lqStarting and Stopping MySQL Automatically\(rq\&. +Section\ \&2.13.1.2, \(lqStarting and Stopping MySQL Automatically\(rq\&. .PP \fBmysql\&.server\fR reads options from the === modified file 'man/mysql_client_test.1' --- a/man/mysql_client_test.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_client_test.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_client_test\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/08/2009 +.\" Date: 10/29/2009 .\" Manual: MySQL Database System .\" Source: MySQL .\" Language: English .\" -.TH "\FBMYSQL_CLIENT_TEST" "1" "08/08/2009" "MySQL" "MySQL Database System" +.TH "\FBMYSQL_CLIENT_TEST" "1" "10/29/2009" "MySQL" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysql_config.1' --- a/man/mysql_config.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_config.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_config\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL_CONFIG\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL_CONFIG\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysql_convert_table_format.1' --- a/man/mysql_convert_table_format.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_convert_table_format.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_convert_table_format\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL_CONVERT_TAB" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL_CONVERT_TAB" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysql_find_rows.1' --- a/man/mysql_find_rows.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_find_rows.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_find_rows\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL_FIND_ROWS\F" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL_FIND_ROWS\F" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysql_fix_extensions.1' --- a/man/mysql_fix_extensions.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_fix_extensions.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_fix_extensions\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL_FIX_EXTENSI" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL_FIX_EXTENSI" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysql_fix_privilege_tables.1' --- a/man/mysql_fix_privilege_tables.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_fix_privilege_tables.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_fix_privilege_tables\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL_FIX_PRIVILE" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL_FIX_PRIVILE" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysql_install_db.1' --- a/man/mysql_install_db.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_install_db.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_install_db\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL_INSTALL_DB\" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL_INSTALL_DB\" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -81,7 +81,7 @@ with the and \fB\-\-skip\-grant\-tables\fR options (see -Section\ \&2.10.2, \(lqTypical configure Options\(rq)\&. If MySQL was configured with the +Section\ \&2.3.2, \(lqTypical configure Options\(rq)\&. If MySQL was configured with the \fB\-\-disable\-grant\-options\fR option, \fB\-\-bootstrap\fR === modified file 'man/mysql_secure_installation.1' --- a/man/mysql_secure_installation.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_secure_installation.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_secure_installation\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL_SECURE_INST" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL_SECURE_INST" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysql_setpermission.1' --- a/man/mysql_setpermission.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_setpermission.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_setpermission\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL_SETPERMISSI" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL_SETPERMISSI" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysql_tzinfo_to_sql.1' --- a/man/mysql_tzinfo_to_sql.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_tzinfo_to_sql.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_tzinfo_to_sql\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL_TZINFO_TO_S" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL_TZINFO_TO_S" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysql_upgrade.1' --- a/man/mysql_upgrade.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_upgrade.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_upgrade\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL_UPGRADE\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL_UPGRADE\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -41,7 +41,7 @@ script, which should no longer be used\& If a table is found to have a possible incompatibility, \fBmysql_upgrade\fR performs a table check\&. If any problems are found, a table repair is attempted\&. If the table cannot be repaired, see -Section\ \&2.12.4, \(lqRebuilding or Repairing Tables or Indexes\(rq +Section\ \&2.4.4, \(lqRebuilding or Repairing Tables or Indexes\(rq for manual table repair strategies\&. .if n \{\ .sp @@ -59,11 +59,11 @@ for manual table repair strategies\&. You should always back up your current MySQL installation \fIbefore\fR performing an upgrade\&. See -Section\ \&6.1, \(lqDatabase Backups\(rq\&. +Section\ \&6.1, \(lqDatabase Backup Methods\(rq\&. .PP Some upgrade incompatibilities may require special handling before you upgrade your MySQL installation and run \fBmysql_upgrade\fR\&. See -Section\ \&2.12.1, \(lqUpgrading MySQL\(rq, for instructions on determining whether any such incompatibilities apply to your installation and how to handle them\&. +Section\ \&2.4.1, \(lqUpgrading MySQL\(rq, for instructions on determining whether any such incompatibilities apply to your installation and how to handle them\&. .sp .5v .RE .PP @@ -189,7 +189,7 @@ If you install MySQL from RPM packages o \fBmysql_upgrade\fR is included in the server RPM but requires the client RPM because the latter includes \fBmysqlcheck\fR\&. (See -Section\ \&2.4, \(lqInstalling MySQL from RPM Packages on Linux\(rq\&.) +Section\ \&2.6.1, \(lqInstalling MySQL from RPM Packages on Linux\(rq\&.) .PP In MySQL 5\&.1\&.7, \fBmysql_upgrade \fR @@ -352,6 +352,26 @@ root\&. .sp Verbose mode\&. Print more information about what the program does\&. .RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +.\" mysql_upgrade: write-binlog option +.\" write-binlog option: mysql_upgrade +\fB\-\-write\-binlog\fR +.sp +Cause binary logging to be enabled while +\fBmysql_upgrade\fR +runs\&. This is the default behavior; to disable binary logging during the upgrade, use the inverse of this option (that is, start the program with +\fB\-\-skip\-write\-binlog\fR)\&. +.sp +This option was introduced in MySQL 5\&.1\&.40\&. +.RE .SH "COPYRIGHT" .br .PP === modified file 'man/mysql_waitpid.1' --- a/man/mysql_waitpid.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_waitpid.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_waitpid\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL_WAITPID\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL_WAITPID\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysql_zap.1' --- a/man/mysql_zap.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_zap.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_zap\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL_ZAP\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL_ZAP\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysqlaccess.1' --- a/man/mysqlaccess.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqlaccess.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqlaccess\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLACCESS\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLACCESS\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysqladmin.1' --- a/man/mysqladmin.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqladmin.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqladmin\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLADMIN\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLADMIN\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysqlbinlog.1' --- a/man/mysqlbinlog.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqlbinlog.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqlbinlog\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLBINLOG\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLBINLOG\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -855,7 +855,7 @@ You can pipe the output of into the \fBmysql\fR client to execute the statements contained in the binary log\&. This is used to recover from a crash when you have an old backup (see -Section\ \&6.1, \(lqDatabase Backups\(rq)\&. For example: +Section\ \&6.1, \(lqDatabase Backup Methods\(rq)\&. For example: .sp .if n \{\ .RS 4 === modified file 'man/mysqlbug.1' --- a/man/mysqlbug.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqlbug.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqlbug\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLBUG\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLBUG\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysqlcheck.1' --- a/man/mysqlcheck.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqlcheck.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqlcheck\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLCHECK\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLCHECK\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -34,7 +34,14 @@ The \fBmysqlcheck\fR client performs table maintenance: It checks, repairs, optimizes, or analyzes tables\&. .PP -Each table is locked and therefore unavailable to other sessions while it is being processed\&. Table maintenance operations can be time\-consuming, particularly for large tables\&. If you use the +Each table is locked and therefore unavailable to other sessions while it is being processed, although for check operations, the table is locked with a +READ +lock only (see +Section\ \&12.4.5, \(lqLOCK TABLES and UNLOCK TABLES Syntax\(rq, for more information about +READ +and +WRITE +locks)\&. Table maintenance operations can be time\-consuming, particularly for large tables\&. If you use the \fB\-\-databases\fR or \fB\-\-all\-databases\fR @@ -94,7 +101,7 @@ note : The storage engine for the ta If \fBmysqlcheck\fR is unable to repair a table, see -Section\ \&2.12.4, \(lqRebuilding or Repairing Tables or Indexes\(rq +Section\ \&2.4.4, \(lqRebuilding or Repairing Tables or Indexes\(rq for manual table repair strategies\&. This will be the case, for example, for InnoDB tables, which can be checked with === modified file 'man/mysqld.8' --- a/man/mysqld.8 2009-09-16 12:03:18 +0000 +++ b/man/mysqld.8 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqld\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLD\FR" "8" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLD\FR" "8" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysqld_multi.1' --- a/man/mysqld_multi.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqld_multi.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqld_multi\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLD_MULTI\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLD_MULTI\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -51,7 +51,7 @@ to specify which servers you want to sta [mysqld] group used for starting \fBmysqld\fR\&. (See, for example, -Section\ \&2.11.2.2, \(lqStarting and Stopping MySQL Automatically\(rq\&.) However, when using multiple servers, it is necessary that each one use its own value for options such as the Unix socket file and TCP/IP port number\&. For more information on which options must be unique per server in a multiple\-server environment, see +Section\ \&2.13.1.2, \(lqStarting and Stopping MySQL Automatically\(rq\&.) However, when using multiple servers, it is necessary that each one use its own value for options such as the Unix socket file and TCP/IP port number\&. For more information on which options must be unique per server in a multiple\-server environment, see Section\ \&5.6, \(lqRunning Multiple MySQL Servers on the Same Machine\(rq\&. .PP To invoke @@ -183,12 +183,6 @@ Otherwise, option files in the standard option, if one is given\&. (If the option is given multiple times, the last value is used\&.) .RE .PP -Option files read are searched for -[mysqld_multi] -and -[mysqld\fIN\fR] -option groups\&. -.PP Before MySQL 5\&.1\&.18, the preceding options are not recognized\&. Files in the standard locations are read, and any file named by the \fB\-\-config\-file=\fR\fB\fIfile_name\fR\fR option, if one is given\&. A file named by @@ -199,6 +193,39 @@ option groups, not the [mysqld_multi] group\&. .PP +Option files read are searched for +[mysqld_multi] +and +[mysqld\fIN\fR] +option groups\&. The +[mysqld_multi] +group can be used for options to +\fBmysqld_multi\fR +itself\&. +[mysqld\fIN\fR] +groups can be used for options passed to specific +\fBmysqld\fR +instances\&. +.PP +As of MySQL 5\&.1\&.35, the +[mysqld] +or +[mysqld_safe] +groups can be used for common options read by all instances of +\fBmysqld\fR +or +\fBmysqld_safe\fR\&. You can specify a +\fB\-\-defaults\-file=\fR\fB\fIfile_name\fR\fR +option to use a different configuration file for that instance, in which case the +[mysqld] +or +[mysqld_safe] +groups from that file will be used for that instance\&. Before MySQL 5\&.1\&.35, some versions of +\fBmysqld_multi\fR +pass the +\fB\-\-no\-defaults\fR +options to instances, so these techniques are inapplicable\&. +.PP \fBmysqld_multi\fR supports the following options: .sp === modified file 'man/mysqld_safe.1' --- a/man/mysqld_safe.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqld_safe.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqld_safe\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLD_SAFE\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLD_SAFE\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysqldump.1' --- a/man/mysqldump.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqldump.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqldump\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLDUMP\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLDUMP\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -575,7 +575,7 @@ uses utf8, and earlier versions use latin1\&. .sp -This option has no effect for output data files produced by using the +Prior to MySQL 5\&.1\&.38, this option has no effect for output data files produced by using the \fB\-\-tab\fR option\&. See the description for that option\&. .RE @@ -1232,7 +1232,8 @@ statement for the table\&. \fB\-\-no\-set\-names\fR .sp This option is deprecated\&. Use -\fB\-\-skip\-set\-charset\fR\&. instead\&. +\fB\-\-skip\-set\-charset\fR +instead\&. .RE .sp .RS 4 @@ -1657,11 +1658,11 @@ and \fB\-\-lines\-terminated\-by\fR options\&. .sp -Column values are dumped using the -binary -character set and the +As of MySQL 5\&.1\&.38, column values are written converted to the character set specified by the \fB\-\-default\-character\-set\fR -option is ignored\&. In effect, there is no character set conversion\&. If a table contains columns in several character sets, the output data file will as well and you may not be able to reload the file correctly\&. +option\&. Prior to 5\&.1\&.38 or if no such option is present, values are dumped using the +binary +character set\&. In effect, there is no character set conversion\&. If a table contains columns in several character sets, the output data file will as well and you may not be able to reload the file correctly\&. .if n \{\ .sp .\} @@ -2110,7 +2111,7 @@ InnoDB storage engine\&. .PP For more information on making backups, see -Section\ \&6.1, \(lqDatabase Backups\(rq, and +Section\ \&6.1, \(lqDatabase Backup Methods\(rq, and Section\ \&6.2, \(lqExample Backup and Recovery Strategy\(rq\&. .\" mysqldump: views .\" mysqldump: problems === modified file 'man/mysqldumpslow.1' --- a/man/mysqldumpslow.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqldumpslow.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqldumpslow\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLDUMPSLOW\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLDUMPSLOW\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysqlhotcopy.1' --- a/man/mysqlhotcopy.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqlhotcopy.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqlhotcopy\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLHOTCOPY\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLHOTCOPY\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysqlimport.1' --- a/man/mysqlimport.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqlimport.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqlimport\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLIMPORT\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLIMPORT\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysqlmanager.8' --- a/man/mysqlmanager.8 2009-09-16 12:03:18 +0000 +++ b/man/mysqlmanager.8 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqlmanager\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLMANAGER\FR" "8" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLMANAGER\FR" "8" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -274,7 +274,8 @@ Drop all users from the password file\&. .\} .\" mysqlmanager: debug option .\" debug option: mysqlmanager -\fB\-\-debug=\fR\fB\fIdebug_options\fR\fR\fB, \-# \fR\fB\fIdebug_options\fR\fR +\fB\-\-debug=\fR\fB\fIdebug_options\fR\fR, +\fB\-# \fR\fB\fIdebug_options\fR\fR .sp Write a debugging log\&. A typical \fIdebug_options\fR === modified file 'man/mysqlshow.1' --- a/man/mysqlshow.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqlshow.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqlshow\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLSHOW\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLSHOW\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysqlslap.1' --- a/man/mysqlslap.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqlslap.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqlslap\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLSLAP\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLSLAP\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysqltest.1' --- a/man/mysqltest.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqltest.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqltest\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/08/2009 +.\" Date: 10/29/2009 .\" Manual: MySQL Database System .\" Source: MySQL .\" Language: English .\" -.TH "\FBMYSQLTEST\FR" "1" "08/08/2009" "MySQL" "MySQL Database System" +.TH "\FBMYSQLTEST\FR" "1" "10/29/2009" "MySQL" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/ndbd.8' --- a/man/ndbd.8 2009-09-16 12:03:18 +0000 +++ b/man/ndbd.8 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBndbd\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBNDBD\FR" "8" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBNDBD\FR" "8" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -41,8 +41,9 @@ processes cooperate in handling data\&. .\" command options (MySQL Cluster): ndbd .\" MySQL Cluster: ndbd process .PP -The following list describes command options specific to the MySQL Cluster data node program -\fBndbd\fR\&. +The following table includes command options specific to the MySQL Cluster data node program +\fBndbd\fR\&. Additional descriptions follow the table\&. For options common to all MySQL Cluster programs, see +Section\ \&17.4.2, \(lqOptions Common to MySQL Cluster Programs\(rq\&. .if n \{\ .sp .\} @@ -68,7 +69,7 @@ wherever the latter occurs in this secti For options common to all NDBCLUSTER programs, see -Section\ \&17.6.2, \(lqOptions Common to MySQL Cluster Programs\(rq\&. +Section\ \&17.4.2, \(lqOptions Common to MySQL Cluster Programs\(rq\&. .sp .RS 4 .ie n \{\ @@ -81,9 +82,11 @@ Section\ \&17.6.2, \(lqOptions Common to \fB\-\-bind\-address\fR .TS allbox tab(:); -l l -l l -l l. +l l s +l l s +l l s +^ l l +^ l l. T{ \fBVersion Introduced\fR T}:T{ @@ -95,28 +98,16 @@ T}:T{ \-\-bind\-address=name T} T{ -\fBPermitted Values \fR +\ \& T}:T{ -[\fInested\ table\fR]* +\fBPermitted Values \fR T} -.TE -.sp 1 -.sp -.it 1 an-trap -.nr an-no-space-flag 1 -.nr an-break-flag 1 -.br -.B *[nested\ table] -.TS -allbox tab(:); -l l -l l. -T{ +:T{ \fBType\fR T}:T{ string T} -T{ +:T{ \fBDefault\fR T}:T{ T} @@ -141,36 +132,26 @@ This option was added in MySQL 5\&.1\&.1 \fB\-d\fR .TS allbox tab(:); -l l -l l. +l l s +l l s +^ l l +^ l l. T{ \fBCommand Line Format\fR T}:T{ \-\-daemon T} T{ -\fBPermitted Values \fR +\ \& T}:T{ -[\fInested\ table\fR]* +\fBPermitted Values \fR T} -.TE -.sp 1 -.sp -.it 1 an-trap -.nr an-no-space-flag 1 -.nr an-break-flag 1 -.br -.B *[nested\ table] -.TS -allbox tab(:); -l l -l l. -T{ +:T{ \fBType\fR T}:T{ boolean T} -T{ +:T{ \fBDefault\fR T}:T{ TRUE @@ -182,6 +163,12 @@ Instructs to execute as a daemon process\&. This is the default behavior\&. \fB\-\-nodaemon\fR can be used to prevent the process from running as a daemon\&. +.sp +This option has no effect when running +\fBndbd\fR +or +\fBndbmtd\fR +on Windows platforms\&. .RE .sp .RS 4 @@ -197,36 +184,26 @@ can be used to prevent the process from \fB\-\-initial\fR .TS allbox tab(:); -l l -l l. +l l s +l l s +^ l l +^ l l. T{ \fBCommand Line Format\fR T}:T{ \-\-initial T} T{ -\fBPermitted Values \fR +\ \& T}:T{ -[\fInested\ table\fR]* +\fBPermitted Values \fR T} -.TE -.sp 1 -.sp -.it 1 an-trap -.nr an-no-space-flag 1 -.nr an-break-flag 1 -.br -.B *[nested\ table] -.TS -allbox tab(:); -l l -l l. -T{ +:T{ \fBType\fR T}:T{ boolean T} -T{ +:T{ \fBDefault\fR T}:T{ FALSE @@ -321,9 +298,11 @@ Backup files that have already been crea .IP \(bu 2.3 .\} MySQL Cluster Disk Data files (see -Section\ \&17.10, \(lqMySQL Cluster Disk Data Tables\(rq)\&. +Section\ \&17.5.9, \(lqMySQL Cluster Disk Data Tables\(rq)\&. .RE .RS 4 +.sp +This option also has no effect on recovery of data by a data node that is just starting (or restarting) from data nodes that are already running\&. This recovery of data occurs automatically, and requires no user intervention in a MySQL Cluster that is running normally\&. .sp .5v .RE It is permissible to use this option when starting the cluster for the very first time (that is, before any data node files have been created); however, it is @@ -344,9 +323,11 @@ necessary to do so\&. \fB\-\-initial\-start\fR .TS allbox tab(:); -l l -l l -l l. +l l s +l l s +l l s +^ l l +^ l l. T{ \fBVersion Introduced\fR T}:T{ @@ -358,28 +339,16 @@ T}:T{ \-\-initial\-start T} T{ -\fBPermitted Values \fR +\ \& T}:T{ -[\fInested\ table\fR]* +\fBPermitted Values \fR T} -.TE -.sp 1 -.sp -.it 1 an-trap -.nr an-no-space-flag 1 -.nr an-break-flag 1 -.br -.B *[nested\ table] -.TS -allbox tab(:); -l l -l l. -T{ +:T{ \fBType\fR T}:T{ boolean T} -T{ +:T{ \fBDefault\fR T}:T{ FALSE @@ -435,9 +404,11 @@ Prior to MySQL 5\&.1\&.19, it was not po \fB\-\-nowait\-nodes=\fR\fB\fInode_id_1\fR\fR\fB[, \fR\fB\fInode_id_2\fR\fR\fB[, \&.\&.\&.]]\fR .TS allbox tab(:); -l l -l l -l l. +l l s +l l s +l l s +^ l l +^ l l. T{ \fBVersion Introduced\fR T}:T{ @@ -449,28 +420,16 @@ T}:T{ \-\-nowait\-nodes=list T} T{ -\fBPermitted Values \fR +\ \& T}:T{ -[\fInested\ table\fR]* +\fBPermitted Values \fR T} -.TE -.sp 1 -.sp -.it 1 an-trap -.nr an-no-space-flag 1 -.nr an-break-flag 1 -.br -.B *[nested\ table] -.TS -allbox tab(:); -l l -l l. -T{ +:T{ \fBType\fR T}:T{ string T} -T{ +:T{ \fBDefault\fR T}:T{ T} @@ -508,36 +467,26 @@ This option was added in MySQL 5\&.1\&.9 \fB\-\-nodaemon\fR .TS allbox tab(:); -l l -l l. +l l s +l l s +^ l l +^ l l. T{ \fBCommand Line Format\fR T}:T{ \-\-nodaemon T} T{ -\fBPermitted Values \fR +\ \& T}:T{ -[\fInested\ table\fR]* +\fBPermitted Values \fR T} -.TE -.sp 1 -.sp -.it 1 an-trap -.nr an-no-space-flag 1 -.nr an-break-flag 1 -.br -.B *[nested\ table] -.TS -allbox tab(:); -l l -l l. -T{ +:T{ \fBType\fR T}:T{ boolean T} -T{ +:T{ \fBDefault\fR T}:T{ FALSE @@ -549,6 +498,12 @@ Instructs not to start as a daemon process\&. This is useful when \fBndbd\fR is being debugged and you want output to be redirected to the screen\&. +.sp +As of MySQL Cluster NDB 7\&.0\&.8, the default behavior for +\fBndbd\fR +and +\fBndbmtd\fR +on Windows is to run in the foreground, making this option unnecessary on Windows platforms\&. (\m[blue]\fBBug#45588\fR\m[]\&\s-2\u[2]\d\s+2) .RE .sp .RS 4 @@ -567,36 +522,26 @@ is being debugged and you want output to \fB\-n\fR .TS allbox tab(:); -l l -l l. +l l s +l l s +^ l l +^ l l. T{ \fBCommand Line Format\fR T}:T{ \-\-nostart T} T{ -\fBPermitted Values \fR +\ \& T}:T{ -[\fInested\ table\fR]* +\fBPermitted Values \fR T} -.TE -.sp 1 -.sp -.it 1 an-trap -.nr an-no-space-flag 1 -.nr an-break-flag 1 -.br -.B *[nested\ table] -.TS -allbox tab(:); -l l -l l. -T{ +:T{ \fBType\fR T}:T{ boolean T} -T{ +:T{ \fBDefault\fR T}:T{ FALSE @@ -610,7 +555,7 @@ not to start automatically\&. When this connects to the management server, obtains configuration data from it, and initializes communication objects\&. However, it does not actually start the execution engine until specifically requested to do so by the management server\&. This can be accomplished by issuing the proper START command in the management client (see -Section\ \&17.7.2, \(lqCommands in the MySQL Cluster Management Client\(rq)\&. +Section\ \&17.5.2, \(lqCommands in the MySQL Cluster Management Client\(rq)\&. .RE .\" MySQL Cluster: log files .\" log files (MySQL Cluster) @@ -668,7 +613,7 @@ TraceFile: ndb_2_trace\&.log\&.2 Listings of possible \fBndbd\fR exit codes and messages generated when a data node process shuts down prematurely can be found in -\m[blue]\fBndbd Error Messages\fR\m[]\&\s-2\u[2]\d\s+2\&. +\m[blue]\fBndbd Error Messages\fR\m[]\&\s-2\u[3]\d\s+2\&. .if n \{\ .sp .\} @@ -784,7 +729,7 @@ shell> \fBndbd \-\-connect\-string="node .\} .PP See -Section\ \&17.3.4.3, \(lqThe MySQL Cluster Connectstring\(rq, for additional information about this issue\&. +Section\ \&17.3.2.3, \(lqThe MySQL Cluster Connectstring\(rq, for additional information about this issue\&. \fBndbd\fR(8), describes other options for \fBndbd\fR\&. .PP @@ -810,7 +755,7 @@ process can consume up to 2 CPUs if perm For a machine with many CPUs it is possible to use several \fBndbd\fR processes which belong to different node groups; however, such a configuration is still considered experimental and is not supported for MySQL 5\&.1 in a production setting\&. See -Section\ \&17.12, \(lqKnown Limitations of MySQL Cluster\(rq\&. +Section\ \&17.1.5, \(lqKnown Limitations of MySQL Cluster\(rq\&. .SH "COPYRIGHT" .br .PP @@ -829,6 +774,11 @@ Bug#24631 \%http://bugs.mysql.com/24631 .RE .IP " 2." 4 +Bug#45588 +.RS 4 +\%http://bugs.mysql.com/45588 +.RE +.IP " 3." 4 ndbd Error Messages .RS 4 \%http://dev.mysql.com/doc/ndbapi/en/ndbd-error-messages.html === modified file 'man/ndbd_redo_log_reader.1' --- a/man/ndbd_redo_log_reader.1 2009-09-16 12:03:18 +0000 +++ b/man/ndbd_redo_log_reader.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBndbd_redo_log_reader\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBNDBD_REDO_LOG_REA" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBNDBD_REDO_LOG_REA" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -37,6 +37,10 @@ The C++ source files for can be found in the directory /storage/ndb/src/kernel/blocks/dblqh/redoLogReader\&. .PP +The following table includes options that are specific to the MySQL Cluster program +\fBndbd_redo_log_reader\fR\&. Additional descriptions follow the table\&. For options common to all MySQL Cluster programs, see +Section\ \&17.4.2, \(lqOptions Common to MySQL Cluster Programs\(rq\&. +.PP \fBUsage\fR: .sp .if n \{\ @@ -56,105 +60,83 @@ ndb_\fI#\fR_fs/D\fI#\fR/LCP/\fI#\fR/T\fI represents a number (not necessarily the same number)\&. For more information, see \m[blue]\fBCluster Data Node FileSystemDir Files\fR\m[]\&\s-2\u[1]\d\s+2\&. .PP -\fBAdditional Options\fR: +The name of the file to be read may be followed by one or more of the options listed here: +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} .TS allbox tab(:); -l l -l l. +l l s +l l s +^ l l +^ l l. T{ \fBCommand Line Format\fR T}:T{ \-noprint T} T{ -\fBPermitted Values \fR +\ \& T}:T{ -[\fInested\ table\fR]* +\fBPermitted Values \fR T} -.TE -.sp 1 -.sp -.it 1 an-trap -.nr an-no-space-flag 1 -.nr an-break-flag 1 -.br -.B *[nested\ table] -.TS -allbox tab(:); -l l -l l. -T{ +:T{ \fBType\fR T}:T{ boolean T} -T{ +:T{ \fBDefault\fR T}:T{ FALSE T} .TE .sp 1 +\fB\-noprint\fR: Do not print the contents of the log file\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} .TS allbox tab(:); -l l -l l. +l l s +l l s +^ l l +^ l l. T{ \fBCommand Line Format\fR T}:T{ \-nocheck T} T{ -\fBPermitted Values \fR +\ \& T}:T{ -[\fInested\ table\fR]* +\fBPermitted Values \fR T} -.TE -.sp 1 -.sp -.it 1 an-trap -.nr an-no-space-flag 1 -.nr an-break-flag 1 -.br -.B *[nested\ table] -.TS -allbox tab(:); -l l -l l. -T{ +:T{ \fBType\fR T}:T{ boolean T} -T{ +:T{ \fBDefault\fR T}:T{ FALSE T} .TE .sp 1 -.PP -The name of the file to be read may be followed by one or more of the options listed here: -.sp -.RS 4 -.ie n \{\ -\h'-04'\(bu\h'+03'\c -.\} -.el \{\ -.sp -1 -.IP \(bu 2.3 -.\} -\fB\-noprint\fR: Do not print the contents of the log file\&. -.RE -.sp -.RS 4 -.ie n \{\ -\h'-04'\(bu\h'+03'\c -.\} -.el \{\ -.sp -1 -.IP \(bu 2.3 -.\} \fB\-nocheck\fR: Do not check the log file for errors\&. .RE .sp @@ -184,7 +166,7 @@ You should have received a copy of the G .IP " 1." 4 Cluster Data Node FileSystemDir Files .RS 4 -\%http://dev.mysql.com/doc/ndbapi/en/ndb-internals-ndbd-filesystem.html#ndb-internals-ndbd-filesystemdir-files +\%http://dev.mysql.com/doc/ndbapi/en/ndb-internals-ndbd-filesystemdir-files.html .RE .SH "SEE ALSO" For more information, please refer to the MySQL Reference Manual, === modified file 'man/ndbmtd.8' --- a/man/ndbmtd.8 2009-09-16 12:03:18 +0000 +++ b/man/ndbmtd.8 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBndbmtd\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBNDBMTD\FR" "8" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBNDBMTD\FR" "8" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -49,8 +49,8 @@ Command\-line options and configuration \fBndbd\fR also apply to \fBndbmtd\fR\&. For more information about these options and parameters, see -Section\ \&17.6.3.2, \(lqProgram Options for ndbd and ndbmtd\(rq, and -Section\ \&17.3.4.6, \(lqDefining MySQL Cluster Data Nodes\(rq, respectively\&. +\fBndbd\fR(8), and +Section\ \&17.3.2.6, \(lqDefining MySQL Cluster Data Nodes\(rq, respectively\&. .PP \fBndbmtd\fR is also file system\-compatible with @@ -69,25 +69,13 @@ simply by stopping the node and then sta \fBndbd\fR in place of the multi\-threaded binary\&. It is not necessary when switching between the two to start the data node binary using \fB\-\-initial\fR\&. -.if n \{\ -.sp -.\} -.RS 4 -.it 1 an-trap -.nr an-no-space-flag 1 -.nr an-break-flag 1 -.br -.ps +1 -\fBImportant\fR -.ps -1 -.br .PP -We do not currently recommend using +Prior to MySQL Cluster NDB 7\&.0\&.6, there were known issues when using \fBndbmtd\fR -with MySQL Cluster Disk Data tables in production, due to known issues which we are working to fix in a future MySQL Cluster release\&. (\m[blue]\fBBug#41915\fR\m[]\&\s-2\u[1]\d\s+2, +with MySQL Cluster Disk Data tables\&. If you wish to use multi\-threaded data nodes with disk\-based +NDB +tables, you should insure that you are running MySQL Cluster NDB 7\&.0\&.6 or later\&. (\m[blue]\fBBug#41915\fR\m[]\&\s-2\u[1]\d\s+2, \m[blue]\fBBug#44915\fR\m[]\&\s-2\u[2]\d\s+2) -.sp .5v -.RE .PP Using \fBndbmtd\fR @@ -129,6 +117,8 @@ failures\&. .RE .PP These differences are discussed in more detail in the next few paragraphs\&. +.\" execution threads (MySQL Cluster) +.\" MySQL Cluster: execution threads .\" ndbmtd: MaxNoOfExecutionThreads .\" MaxNoOfExecutionThreads: ndbmtd .\" ndbmtd: trace files @@ -148,7 +138,7 @@ file, it is exclusive to and does not apply to \fBndbd\fR\&. .PP -This parameter takes an integer value from 2 to 8 inclusive\&. Generally, you should set this to the number of CPU cores on the data node host, as shown in the following table: +This parameter takes an integer value from 2 to 8 inclusive\&. Generally, you should set this parameter equal to the number of CPU cores on the data node host, as shown in the following table: .TS allbox tab(:); lB lB. === modified file 'man/perror.1' --- a/man/perror.1 2009-09-16 12:03:18 +0000 +++ b/man/perror.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBperror\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBPERROR\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBPERROR\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/replace.1' --- a/man/replace.1 2009-09-16 12:03:18 +0000 +++ b/man/replace.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBreplace\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBREPLACE\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBREPLACE\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/resolve_stack_dump.1' --- a/man/resolve_stack_dump.1 2009-09-16 12:03:18 +0000 +++ b/man/resolve_stack_dump.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBresolve_stack_dump\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBRESOLVE_STACK_DUM" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBRESOLVE_STACK_DUM" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/resolveip.1' --- a/man/resolveip.1 2009-09-16 12:03:18 +0000 +++ b/man/resolveip.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBresolveip\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBRESOLVEIP\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBRESOLVEIP\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'scripts/fill_help_tables.sql' --- a/scripts/fill_help_tables.sql 2009-09-16 12:03:18 +0000 +++ b/scripts/fill_help_tables.sql 2009-12-01 07:24:05 +0000 @@ -87,7 +87,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (18,25,'SHOW CREATE PROCEDURE','Syntax:\nSHOW CREATE PROCEDURE proc_name\n\nThis statement is a MySQL extension. It returns the exact string that\ncan be used to re-create the named stored procedure. A similar\nstatement, SHOW CREATE FUNCTION, displays information about stored\nfunctions (see [HELP SHOW CREATE FUNCTION]).\n\nBoth statements require that you be the owner of the routine or have\nSELECT access to the mysql.proc table. If you do not have privileges\nfor the routine itself, the value displayed for the Create Procedure or\nCreate Function field will be NULL.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/show-create-procedure.html\n\n','mysql> SHOW CREATE PROCEDURE test.simpleproc\\G\n*************************** 1. row ***************************\n Procedure: simpleproc\n sql_mode:\n Create Procedure: CREATE PROCEDURE `simpleproc`(OUT param1 INT)\n BEGIN\n SELECT COUNT(*) INTO param1 FROM t;\n END\ncharacter_set_client: latin1\ncollation_connection: latin1_swedish_ci\n Database Collation: latin1_swedish_ci\n\nmysql> SHOW CREATE FUNCTION test.hello\\G\n*************************** 1. row ***************************\n Function: hello\n sql_mode:\n Create Function: CREATE FUNCTION `hello`(s CHAR(20))\n RETURNS CHAR(50)\n RETURN CONCAT(\'Hello, \',s,\'!\')\ncharacter_set_client: latin1\ncollation_connection: latin1_swedish_ci\n Database Collation: latin1_swedish_ci\n','http://dev.mysql.com/doc/refman/5.1/en/show-create-procedure.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (19,20,'INTEGER','INTEGER[(M)] [UNSIGNED] [ZEROFILL]\n\nThis type is a synonym for INT.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/numeric-type-overview.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/numeric-type-overview.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (20,35,'LOWER','Syntax:\nLOWER(str)\n\nReturns the string str with all characters changed to lowercase\naccording to the current character set mapping. The default is latin1\n(cp1252 West European).\n\nmysql> SELECT LOWER(\'QUADRATICALLY\');\n -> \'quadratically\'\n\nLOWER() (and UPPER()) are ineffective when applied to binary strings\n(BINARY, VARBINARY, BLOB). To perform lettercase conversion, convert\nthe string to a nonbinary string:\n\nmysql> SET @str = BINARY \'New York\';\nmysql> SELECT LOWER(@str), LOWER(CONVERT(@str USING latin1));\n+-------------+-----------------------------------+\n| LOWER(@str) | LOWER(CONVERT(@str USING latin1)) |\n+-------------+-----------------------------------+\n| New York | new york |\n+-------------+-----------------------------------+\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/string-functions.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/string-functions.html'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (21,25,'SHOW COLUMNS','Syntax:\nSHOW [FULL] COLUMNS {FROM | IN} tbl_name [{FROM | IN} db_name]\n [LIKE \'pattern\' | WHERE expr]\n\nSHOW COLUMNS displays information about the columns in a given table.\nIt also works for views. The LIKE clause, if present, indicates which\ncolumn names to match. The WHERE clause can be given to select rows\nusing more general conditions, as discussed in\nhttp://dev.mysql.com/doc/refman/5.1/en/extended-show.html.\n\nmysql> SHOW COLUMNS FROM City;\n+------------+----------+------+-----+---------+----------------+\n| Field | Type | Null | Key | Default | Extra |\n+------------+----------+------+-----+---------+----------------+\n| Id | int(11) | NO | PRI | NULL | auto_increment |\n| Name | char(35) | NO | | | |\n| Country | char(3) | NO | UNI | | |\n| District | char(20) | YES | MUL | | |\n| Population | int(11) | NO | | 0 | |\n+------------+----------+------+-----+---------+----------------+\n5 rows in set (0.00 sec)\n\nIf the data types differ from what you expect them to be based on a\nCREATE TABLE statement, note that MySQL sometimes changes data types\nwhen you create or alter a table. The conditions under which this\noccurs are described in\nhttp://dev.mysql.com/doc/refman/5.1/en/silent-column-changes.html.\n\nT… FULL keyword causes the output to include the column collation and\ncomments, as well as the privileges you have for each column.\n\nYou can use db_name.tbl_name as an alternative to the tbl_name FROM\ndb_name syntax. In other words, these two statements are equivalent:\n\nmysql> SHOW COLUMNS FROM mytable FROM mydb;\nmysql> SHOW COLUMNS FROM mydb.mytable;\n\nSHOW COLUMNS displays the following values for each table column:\n\nField indicates the column name.\n\nType indicates the column data type.\n\nCollation indicates the collation for nonbinary string columns, or NULL\nfor other columns. This value is displayed only if you use the FULL\nkeyword.\n\nThe Null field contains YES if NULL values can be stored in the column,\nNO if not.\n\nThe Key field indicates whether the column is indexed:\n\no If Key is empty, the column either is not indexed or is indexed only\n as a secondary column in a multiple-column, nonunique index.\n\no If Key is PRI, the column is a PRIMARY KEY or is one of the columns\n in a multiple-column PRIMARY KEY.\n\no If Key is UNI, the column is the first column of a unique-valued\n index that cannot contain NULL values.\n\no If Key is MUL, multiple occurrences of a given value are allowed\n within the column. The column is the first column of a nonunique\n index or a unique-valued index that can contain NULL values.\n\nIf more than one of the Key values applies to a given column of a\ntable, Key displays the one with the highest priority, in the order\nPRI, UNI, MUL.\n\nA UNIQUE index may be displayed as PRI if it cannot contain NULL values\nand there is no PRIMARY KEY in the table. A UNIQUE index may display as\nMUL if several columns form a composite UNIQUE index; although the\ncombination of the columns is unique, each column can still hold\nmultiple occurrences of a given value.\n\nThe Default field indicates the default value that is assigned to the\ncolumn.\n\nThe Extra field contains any additional information that is available\nabout a given column. The value is auto_increment if the column was\ncreated with the AUTO_INCREMENT keyword and empty otherwise.\n\nPrivileges indicates the privileges you have for the column. This value\nis displayed only if you use the FULL keyword.\n\nComment indicates any comment the column has. This value is displayed\nonly if you use the FULL keyword.\n\nSHOW FIELDS is a synonym for SHOW COLUMNS. You can also list a table\'s\ncolumns with the mysqlshow db_name tbl_name command.\n\nThe DESCRIBE statement provides information similar to SHOW COLUMNS.\nSee [HELP DESCRIBE].\n\nThe SHOW CREATE TABLE, SHOW TABLE STATUS, and SHOW INDEX statements\nalso provide information about tables. See [HELP SHOW].\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/show-columns.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/show-columns.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (21,25,'SHOW COLUMNS','Syntax:\nSHOW [FULL] COLUMNS {FROM | IN} tbl_name [{FROM | IN} db_name]\n [LIKE \'pattern\' | WHERE expr]\n\nSHOW COLUMNS displays information about the columns in a given table.\nIt also works for views. The LIKE clause, if present, indicates which\ncolumn names to match. The WHERE clause can be given to select rows\nusing more general conditions, as discussed in\nhttp://dev.mysql.com/doc/refman/5.1/en/extended-show.html.\n\nmysql> SHOW COLUMNS FROM City;\n+------------+----------+------+-----+---------+----------------+\n| Field | Type | Null | Key | Default | Extra |\n+------------+----------+------+-----+---------+----------------+\n| Id | int(11) | NO | PRI | NULL | auto_increment |\n| Name | char(35) | NO | | | |\n| Country | char(3) | NO | UNI | | |\n| District | char(20) | YES | MUL | | |\n| Population | int(11) | NO | | 0 | |\n+------------+----------+------+-----+---------+----------------+\n5 rows in set (0.00 sec)\n\nIf the data types differ from what you expect them to be based on a\nCREATE TABLE statement, note that MySQL sometimes changes data types\nwhen you create or alter a table. The conditions under which this\noccurs are described in\nhttp://dev.mysql.com/doc/refman/5.1/en/silent-column-changes.html.\n\nT… FULL keyword causes the output to include the column collation and\ncomments, as well as the privileges you have for each column.\n\nYou can use db_name.tbl_name as an alternative to the tbl_name FROM\ndb_name syntax. In other words, these two statements are equivalent:\n\nmysql> SHOW COLUMNS FROM mytable FROM mydb;\nmysql> SHOW COLUMNS FROM mydb.mytable;\n\nSHOW COLUMNS displays the following values for each table column:\n\nField indicates the column name.\n\nType indicates the column data type.\n\nCollation indicates the collation for nonbinary string columns, or NULL\nfor other columns. This value is displayed only if you use the FULL\nkeyword.\n\nThe Null field contains YES if NULL values can be stored in the column,\nNO if not.\n\nThe Key field indicates whether the column is indexed:\n\no If Key is empty, the column either is not indexed or is indexed only\n as a secondary column in a multiple-column, nonunique index.\n\no If Key is PRI, the column is a PRIMARY KEY or is one of the columns\n in a multiple-column PRIMARY KEY.\n\no If Key is UNI, the column is the first column of a unique-valued\n index that cannot contain NULL values.\n\no If Key is MUL, multiple occurrences of a given value are allowed\n within the column. The column is the first column of a nonunique\n index or a unique-valued index that can contain NULL values.\n\nIf more than one of the Key values applies to a given column of a\ntable, Key displays the one with the highest priority, in the order\nPRI, UNI, MUL.\n\nA UNIQUE index may be displayed as PRI if it cannot contain NULL values\nand there is no PRIMARY KEY in the table. A UNIQUE index may display as\nMUL if several columns form a composite UNIQUE index; although the\ncombination of the columns is unique, each column can still hold\nmultiple occurrences of a given value.\n\nThe Default field indicates the default value that is assigned to the\ncolumn.\n\nThe Extra field contains any additional information that is available\nabout a given column. The value is nonempty in these cases:\nauto_increment for columns that have the AUTO_INCREMENT attribute; as\nof MySQL 5.1.23, on update CURRENT_TIMESTAMP for TIMESTAMP columns that\nhave the ON UPDATE CURRENT_TIMESTAMP attribute.\n\nPrivileges indicates the privileges you have for the column. This value\nis displayed only if you use the FULL keyword.\n\nComment indicates any comment the column has. This value is displayed\nonly if you use the FULL keyword.\n\nSHOW FIELDS is a synonym for SHOW COLUMNS. You can also list a table\'s\ncolumns with the mysqlshow db_name tbl_name command.\n\nThe DESCRIBE statement provides information similar to SHOW COLUMNS.\nSee [HELP DESCRIBE].\n\nThe SHOW CREATE TABLE, SHOW TABLE STATUS, and SHOW INDEX statements\nalso provide information about tables. See [HELP SHOW].\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/show-columns.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/show-columns.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (22,37,'CREATE TRIGGER','Syntax:\nCREATE\n [DEFINER = { user | CURRENT_USER }]\n TRIGGER trigger_name trigger_time trigger_event\n ON tbl_name FOR EACH ROW trigger_stmt\n\nThis statement creates a new trigger. A trigger is a named database\nobject that is associated with a table, and that activates when a\nparticular event occurs for the table. The trigger becomes associated\nwith the table named tbl_name, which must refer to a permanent table.\nYou cannot associate a trigger with a TEMPORARY table or a view.\n\nCREATE TRIGGER requires the TRIGGER privilege for the table associated\nwith the trigger. (Before MySQL 5.1.6, this statement requires the\nSUPER privilege.)\n\nThe DEFINER clause determines the security context to be used when\nchecking access privileges at trigger activation time.\n\ntrigger_time is the trigger action time. It can be BEFORE or AFTER to\nindicate that the trigger activates before or after each row to be\nmodified.\n\ntrigger_event indicates the kind of statement that activates the\ntrigger. The trigger_event can be one of the following:\n\no INSERT: The trigger is activated whenever a new row is inserted into\n the table; for example, through INSERT, LOAD DATA, and REPLACE\n statements.\n\no UPDATE: The trigger is activated whenever a row is modified; for\n example, through UPDATE statements.\n\no DELETE: The trigger is activated whenever a row is deleted from the\n table; for example, through DELETE and REPLACE statements. However,\n DROP TABLE and TRUNCATE statements on the table do not activate this\n trigger, because they do not use DELETE. Dropping a partition does\n not activate DELETE triggers, either. See [HELP TRUNCATE TABLE].\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/create-trigger.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/create-trigger.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (23,30,'MONTH','Syntax:\nMONTH(date)\n\nReturns the month for date, in the range 1 to 12 for January to\nDecember, or 0 for dates such as \'0000-00-00\' or \'2008-00-00\' that have\na zero month part.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','mysql> SELECT MONTH(\'2008-02-03\');\n -> 2\n','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (24,20,'TINYINT','TINYINT[(M)] [UNSIGNED] [ZEROFILL]\n\nA very small integer. The signed range is -128 to 127. The unsigned\nrange is 0 to 255.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/numeric-type-overview.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/numeric-type-overview.html'); @@ -153,7 +153,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (84,25,'EXECUTE STATEMENT','Syntax:\nEXECUTE stmt_name\n [USING @var_name [, @var_name] ...]\n\nAfter preparing a statement with PREPARE, you execute it with an\nEXECUTE statement that refers to the prepared statement name. If the\nprepared statement contains any parameter markers, you must supply a\nUSING clause that lists user variables containing the values to be\nbound to the parameters. Parameter values can be supplied only by user\nvariables, and the USING clause must name exactly as many variables as\nthe number of parameter markers in the statement.\n\nYou can execute a given prepared statement multiple times, passing\ndifferent variables to it or setting the variables to different values\nbefore each execution.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/execute.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/execute.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (85,37,'DROP INDEX','Syntax:\nDROP [ONLINE|OFFLINE] INDEX index_name ON tbl_name\n\nDROP INDEX drops the index named index_name from the table tbl_name.\nThis statement is mapped to an ALTER TABLE statement to drop the index.\nSee [HELP ALTER TABLE].\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/drop-index.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/drop-index.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (86,35,'MATCH AGAINST','Syntax:\nMATCH (col1,col2,...) AGAINST (expr [search_modifier])\n\nMySQL has support for full-text indexing and searching:\n\no A full-text index in MySQL is an index of type FULLTEXT.\n\no Full-text indexes can be used only with MyISAM tables, and can be\n created only for CHAR, VARCHAR, or TEXT columns.\n\no A FULLTEXT index definition can be given in the CREATE TABLE\n statement when a table is created, or added later using ALTER TABLE\n or CREATE INDEX.\n\no For large data sets, it is much faster to load your data into a table\n that has no FULLTEXT index and then create the index after that, than\n to load data into a table that has an existing FULLTEXT index.\n\nFull-text searching is performed using MATCH() ... AGAINST syntax.\nMATCH() takes a comma-separated list that names the columns to be\nsearched. AGAINST takes a string to search for, and an optional\nmodifier that indicates what type of search to perform. The search\nstring must be a literal string, not a variable or a column name. There\nare three types of full-text searches:\n\no A boolean search interprets the search string using the rules of a\n special query language. The string contains the words to search for.\n It can also contain operators that specify requirements such that a\n word must be present or absent in matching rows, or that it should be\n weighted higher or lower than usual. Common words such as "some" or\n "then" are stopwords and do not match if present in the search\n string. The IN BOOLEAN MODE modifier specifies a boolean search. For\n more information, see\n http://dev.mysql.com/doc/refman/5.1/en/fulltext-boolean.html.\n\no A natural language search interprets the search string as a phrase in\n natural human language (a phrase in free text). There are no special\n operators. The stopword list applies. In addition, words that are\n present in 50% or more of the rows are considered common and do not\n match. Full-text searches are natural language searches if the IN\n NATURAL LANGUAGE MODE modifier is given or if no modifier is given.\n\no A query expansion search is a modification of a natural language\n search. The search string is used to perform a natural language\n search. Then words from the most relevant rows returned by the search\n are added to the search string and the search is done again. The\n query returns the rows from the second search. The IN NATURAL\n LANGUAGE MODE WITH QUERY EXPANSION or WITH QUERY EXPANSION modifier\n specifies a query expansion search. For more information, see\n http://dev.mysql.com/doc/refman/5.1/en/fulltext-query-expansion.html.\n\nThe IN NATURAL LANGUAGE MODE and IN NATURAL LANGUAGE MODE WITH QUERY\nEXPANSION modifiers were added in MySQL 5.1.7.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/fulltext-search.html\n\n','mysql> SELECT id, body, MATCH (title,body) AGAINST\n -> (\'Security implications of running MySQL as root\'\n -> IN NATURAL LANGUAGE MODE) AS score\n -> FROM articles WHERE MATCH (title,body) AGAINST\n -> (\'Security implications of running MySQL as root\'\n -> IN NATURAL LANGUAGE MODE);\n+----+-------------------------------------+-----------------+\n| id | body | score |\n+----+-------------------------------------+-----------------+\n| 4 | 1. Never run mysqld as root. 2. ... | 1.5219271183014 |\n| 6 | When configured properly, MySQL ... | 1.3114095926285 |\n+----+-------------------------------------+-----------------+\n2 rows in set (0.00 sec)\n','http://dev.mysql.com/doc/refman/5.1/en/fulltext-search.html'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (87,37,'CREATE EVENT','Syntax:\nCREATE\n [DEFINER = { user | CURRENT_USER }]\n EVENT\n [IF NOT EXISTS]\n event_name\n ON SCHEDULE schedule\n [ON COMPLETION [NOT] PRESERVE]\n [ENABLE | DISABLE | DISABLE ON SLAVE]\n [COMMENT \'comment\']\n DO sql_statement;\n\nschedule:\n AT timestamp [+ INTERVAL interval] ...\n | EVERY interval\n [STARTS timestamp [+ INTERVAL interval] ...]\n [ENDS timestamp [+ INTERVAL interval] ...]\n\ninterval:\n quantity {YEAR | QUARTER | MONTH | DAY | HOUR | MINUTE |\n WEEK | SECOND | YEAR_MONTH | DAY_HOUR | DAY_MINUTE |\n DAY_SECOND | HOUR_MINUTE | HOUR_SECOND | MINUTE_SECOND}\n\nThis statement creates and schedules a new event. It requires the EVENT\nprivilege for the schema in which the event is to be created.\n\nThe minimum requirements for a valid CREATE EVENT statement are as\nfollows:\n\no The keywords CREATE EVENT plus an event name, which uniquely\n identifies the event in the current schema. (Prior to MySQL 5.1.12,\n the event name needed to be unique only among events created by the\n same user on a given database.)\n\no An ON SCHEDULE clause, which determines when and how often the event\n executes.\n\no A DO clause, which contains the SQL statement to be executed by an\n event.\n\nThis is an example of a minimal CREATE EVENT statement:\n\nCREATE EVENT myevent\n ON SCHEDULE AT CURRENT_TIMESTAMP + INTERVAL 1 HOUR\n DO\n UPDATE myschema.mytable SET mycol = mycol + 1;\n\nThe previous statement creates an event named myevent. This event\nexecutes once --- one hour following its creation --- by running an SQL\nstatement that increments the value of the myschema.mytable table\'s\nmycol column by 1.\n\nThe event_name must be a valid MySQL identifier with a maximum length\nof 64 characters. It may be delimited using back ticks, and may be\nqualified with the name of a database schema. An event is associated\nwith both a MySQL user (the definer) and a schema, and its name must be\nunique among names of events within that schema. In general, the rules\ngoverning event names are the same as those for names of stored\nroutines. See http://dev.mysql.com/doc/refman/5.1/en/identifiers.html.\n\nIf no schema is indicated as part of event_name, the default (current)\nschema is assumed.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/create-event.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/create-event.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (87,37,'CREATE EVENT','Syntax:\nCREATE\n [DEFINER = { user | CURRENT_USER }]\n EVENT\n [IF NOT EXISTS]\n event_name\n ON SCHEDULE schedule\n [ON COMPLETION [NOT] PRESERVE]\n [ENABLE | DISABLE | DISABLE ON SLAVE]\n [COMMENT \'comment\']\n DO sql_statement;\n\nschedule:\n AT timestamp [+ INTERVAL interval] ...\n | EVERY interval\n [STARTS timestamp [+ INTERVAL interval] ...]\n [ENDS timestamp [+ INTERVAL interval] ...]\n\ninterval:\n quantity {YEAR | QUARTER | MONTH | DAY | HOUR | MINUTE |\n WEEK | SECOND | YEAR_MONTH | DAY_HOUR | DAY_MINUTE |\n DAY_SECOND | HOUR_MINUTE | HOUR_SECOND | MINUTE_SECOND}\n\nThis statement creates and schedules a new event. It requires the EVENT\nprivilege for the schema in which the event is to be created.\n\nThe minimum requirements for a valid CREATE EVENT statement are as\nfollows:\n\no The keywords CREATE EVENT plus an event name, which uniquely\n identifies the event within a database schema. (Prior to MySQL\n 5.1.12, the event name needed to be unique only among events created\n by the same user within a schema.)\n\no An ON SCHEDULE clause, which determines when and how often the event\n executes.\n\no A DO clause, which contains the SQL statement to be executed by an\n event.\n\nThis is an example of a minimal CREATE EVENT statement:\n\nCREATE EVENT myevent\n ON SCHEDULE AT CURRENT_TIMESTAMP + INTERVAL 1 HOUR\n DO\n UPDATE myschema.mytable SET mycol = mycol + 1;\n\nThe previous statement creates an event named myevent. This event\nexecutes once --- one hour following its creation --- by running an SQL\nstatement that increments the value of the myschema.mytable table\'s\nmycol column by 1.\n\nThe event_name must be a valid MySQL identifier with a maximum length\nof 64 characters. Event names are not case sensitive, so you cannot\nhave two events named myevent and MyEvent in the same schema. In\ngeneral, the rules governing event names are the same as those for\nnames of stored routines. See\nhttp://dev.mysql.com/doc/refman/5.1/en/identifiers.html.\n\nAn event is associated with a schema. If no schema is indicated as part\nof event_name, the default (current) schema is assumed. To create an\nevent in a specific schema, qualify the event name with a schema using\nschema_name.event_name syntax.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/create-event.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/create-event.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (88,4,'ABS','Syntax:\nABS(X)\n\nReturns the absolute value of X.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/mathematical-functions.html\n\n','mysql> SELECT ABS(2);\n -> 2\nmysql> SELECT ABS(-32);\n -> 32\n','http://dev.mysql.com/doc/refman/5.1/en/mathematical-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (89,31,'POLYFROMWKB','PolyFromWKB(wkb[,srid]), PolygonFromWKB(wkb[,srid])\n\nConstructs a POLYGON value using its WKB representation and SRID.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/creating-spatial-values.html#gis-wkb…','','http://dev.mysql.com/doc/refman/5.1/en/creating-spatial-values.html#gis-wkb…'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (90,35,'NOT LIKE','Syntax:\nexpr NOT LIKE pat [ESCAPE \'escape_char\']\n\nThis is the same as NOT (expr LIKE pat [ESCAPE \'escape_char\']).\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/string-comparison-functions.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/string-comparison-functions.html'); @@ -163,7 +163,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (94,16,'MAX','Syntax:\nMAX([DISTINCT] expr)\n\nReturns the maximum value of expr. MAX() may take a string argument; in\nsuch cases, it returns the maximum string value. See\nhttp://dev.mysql.com/doc/refman/5.1/en/mysql-indexes.html. The DISTINCT\nkeyword can be used to find the maximum of the distinct values of expr,\nhowever, this produces the same result as omitting DISTINCT.\n\nMAX() returns NULL if there were no matching rows.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/group-by-functions.html\n\n','mysql> SELECT student_name, MIN(test_score), MAX(test_score)\n -> FROM student\n -> GROUP BY student_name;\n','http://dev.mysql.com/doc/refman/5.1/en/group-by-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (95,21,'CREATE FUNCTION UDF','Syntax:\nCREATE [AGGREGATE] FUNCTION function_name RETURNS {STRING|INTEGER|REAL|DECIMAL}\n SONAME shared_library_name\n\nA user-defined function (UDF) is a way to extend MySQL with a new\nfunction that works like a native (built-in) MySQL function such as\nABS() or CONCAT().\n\nfunction_name is the name that should be used in SQL statements to\ninvoke the function. The RETURNS clause indicates the type of the\nfunction\'s return value. DECIMAL is a legal value after RETURNS, but\ncurrently DECIMAL functions return string values and should be written\nlike STRING functions.\n\nshared_library_name is the basename of the shared object file that\ncontains the code that implements the function. The file must be\nlocated in the plugin directory. This directory is given by the value\nof the plugin_dir system variable.\n\n*Note*: This is a change in MySQL 5.1. For earlier versions of MySQL,\nthe shared object can be located in any directory that is searched by\nyour system\'s dynamic linker.\n\nTo create a function, you must have the INSERT privilege for the mysql\ndatabase. This is necessary because CREATE FUNCTION adds a row to the\nmysql.func system table that records the function\'s name, type, and\nshared library name. If you do not have this table, you should run the\nmysql_upgrade command to create it. See\nhttp://dev.mysql.com/doc/refman/5.1/en/mysql-upgrade.html.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/create-function-udf.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/create-function-udf.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (96,4,'*','Syntax:\n*\n\nMultiplication:\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/arithmetic-functions.html\n\n','mysql> SELECT 3*5;\n -> 15\nmysql> SELECT 18014398509481984*18014398509481984.0;\n -> 324518553658426726783156020576256.0\nmysql> SELECT 18014398509481984*18014398509481984;\n -> 0\n','http://dev.mysql.com/doc/refman/5.1/en/arithmetic-functions.html'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (97,20,'TIMESTAMP','TIMESTAMP\n\nA timestamp. The range is \'1970-01-01 00:00:01\' UTC to \'2038-01-09\n03:14:07\' UTC. TIMESTAMP values are stored as the number of seconds\nsince the epoch (\'1970-01-01 00:00:00\' UTC). A TIMESTAMP cannot\nrepresent the value \'1970-01-01 00:00:00\' because that is equivalent to\n0 seconds from the epoch and the value 0 is reserved for representing\n\'0000-00-00 00:00:00\', the "zero" TIMESTAMP value.\n\nA TIMESTAMP column is useful for recording the date and time of an\nINSERT or UPDATE operation. By default, the first TIMESTAMP column in a\ntable is automatically set to the date and time of the most recent\noperation if you do not assign it a value yourself. You can also set\nany TIMESTAMP column to the current date and time by assigning it a\nNULL value. Variations on automatic initialization and update\nproperties are described in\nhttp://dev.mysql.com/doc/refman/5.1/en/timestamp.html.\n\nA TIMESTAMP value is returned as a string in the format \'YYYY-MM-DD\nHH:MM:SS\' with a display width fixed at 19 characters. To obtain the\nvalue as a number, you should add +0 to the timestamp column.\n\n*Note*: The TIMESTAMP format that was used prior to MySQL 4.1 is not\nsupported in MySQL 5.1; see MySQL 3.23, 4.0, 4.1 Reference Manual for\ninformation regarding the old format.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-type-overview.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-type-overview.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (97,20,'TIMESTAMP','TIMESTAMP\n\nA timestamp. The range is \'1970-01-01 00:00:01\' UTC to \'2038-01-19\n03:14:07\' UTC. TIMESTAMP values are stored as the number of seconds\nsince the epoch (\'1970-01-01 00:00:00\' UTC). A TIMESTAMP cannot\nrepresent the value \'1970-01-01 00:00:00\' because that is equivalent to\n0 seconds from the epoch and the value 0 is reserved for representing\n\'0000-00-00 00:00:00\', the "zero" TIMESTAMP value.\n\nA TIMESTAMP column is useful for recording the date and time of an\nINSERT or UPDATE operation. By default, the first TIMESTAMP column in a\ntable is automatically set to the date and time of the most recent\noperation if you do not assign it a value yourself. You can also set\nany TIMESTAMP column to the current date and time by assigning it a\nNULL value. Variations on automatic initialization and update\nproperties are described in\nhttp://dev.mysql.com/doc/refman/5.1/en/timestamp.html.\n\nA TIMESTAMP value is returned as a string in the format \'YYYY-MM-DD\nHH:MM:SS\' with a display width fixed at 19 characters. To obtain the\nvalue as a number, you should add +0 to the timestamp column.\n\n*Note*: The TIMESTAMP format that was used prior to MySQL 4.1 is not\nsupported in MySQL 5.1; see MySQL 3.23, 4.0, 4.1 Reference Manual for\ninformation regarding the old format.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-type-overview.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-type-overview.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (98,11,'DES_DECRYPT','Syntax:\nDES_DECRYPT(crypt_str[,key_str])\n\nDecrypts a string encrypted with DES_ENCRYPT(). If an error occurs,\nthis function returns NULL.\n\nThis function works only if MySQL has been configured with SSL support.\nSee http://dev.mysql.com/doc/refman/5.1/en/secure-connections.html.\n\nIf no key_str argument is given, DES_DECRYPT() examines the first byte\nof the encrypted string to determine the DES key number that was used\nto encrypt the original string, and then reads the key from the DES key\nfile to decrypt the message. For this to work, the user must have the\nSUPER privilege. The key file can be specified with the --des-key-file\nserver option.\n\nIf you pass this function a key_str argument, that string is used as\nthe key for decrypting the message.\n\nIf the crypt_str argument does not appear to be an encrypted string,\nMySQL returns the given crypt_str.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/encryption-functions.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/encryption-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (99,25,'CACHE INDEX','Syntax:\nCACHE INDEX\n tbl_index_list [, tbl_index_list] ...\n IN key_cache_name\n\ntbl_index_list:\n tbl_name [[INDEX|KEY] (index_name[, index_name] ...)]\n\nThe CACHE INDEX statement assigns table indexes to a specific key\ncache. It is used only for MyISAM tables.\n\nThe following statement assigns indexes from the tables t1, t2, and t3\nto the key cache named hot_cache:\n\nmysql> CACHE INDEX t1, t2, t3 IN hot_cache;\n+---------+--------------------+----------+----------+\n| Table | Op | Msg_type | Msg_text |\n+---------+--------------------+----------+----------+\n| test.t1 | assign_to_keycache | status | OK |\n| test.t2 | assign_to_keycache | status | OK |\n| test.t3 | assign_to_keycache | status | OK |\n+---------+--------------------+----------+----------+\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/cache-index.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/cache-index.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (100,12,'ENDPOINT','EndPoint(ls)\n\nReturns the Point that is the endpoint of the LineString value ls.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/geometry-property-functions.html#lin…','mysql> SET @ls = \'LineString(1 1,2 2,3 3)\';\nmysql> SELECT AsText(EndPoint(GeomFromText(@ls)));\n+-------------------------------------+\n| AsText(EndPoint(GeomFromText(@ls))) |\n+-------------------------------------+\n| POINT(3 3) |\n+-------------------------------------+\n','http://dev.mysql.com/doc/refman/5.1/en/geometry-property-functions.html#lin…'); @@ -202,7 +202,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (133,25,'SHOW STATUS','Syntax:\nSHOW [GLOBAL | SESSION] STATUS\n [LIKE \'pattern\' | WHERE expr]\n\nSHOW STATUS provides server status information. This information also\ncan be obtained using the mysqladmin extended-status command. The LIKE\nclause, if present, indicates which variable names to match. The WHERE\nclause can be given to select rows using more general conditions, as\ndiscussed in http://dev.mysql.com/doc/refman/5.1/en/extended-show.html.\nThis statement does not require any privilege. It requires only the\nability to connect to the server.\nWith a LIKE clause, the statement displays only rows for those\nvariables with names that match the pattern:\n\nmysql> SHOW STATUS LIKE \'Key%\';\n+--------------------+----------+\n| Variable_name | Value |\n+--------------------+----------+\n| Key_blocks_used | 14955 |\n| Key_read_requests | 96854827 |\n| Key_reads | 162040 |\n| Key_write_requests | 7589728 |\n| Key_writes | 3813196 |\n+--------------------+----------+\n\nWith the GLOBAL modifier, SHOW STATUS displays the status values for\nall connections to MySQL. With SESSION, it displays the status values\nfor the current connection. If no modifier is present, the default is\nSESSION. LOCAL is a synonym for SESSION.\n\nSome status variables have only a global value. For these, you get the\nsame value for both GLOBAL and SESSION. The scope for each status\nvariable is listed at\nhttp://dev.mysql.com/doc/refman/5.1/en/server-status-variables.html.\n\…: http://dev.mysql.com/doc/refman/5.1/en/show-status.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/show-status.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (134,35,'EXTRACTVALUE','Syntax:\nExtractValue(xml_frag, xpath_expr)\n\nExtractValue() takes two string arguments, a fragment of XML markup\nxml_frag and an XPath expression xpath_expr (also known as a locator);\nit returns the text (CDATA) of the first text node which is a child of\nthe element(s) matched by the XPath expression. It is the equivalent of\nperforming a match using the xpath_expr after appending /text(). In\nother words, ExtractValue(\'<a>Sakila</a>\', \'/a/b\') and\nExtractValue(\'<a>Sakila</a>\', \'/a/b/text()\') produce the same\nresult.\n\nIf multiple matches are found, then the content of the first child text\nnode of each matching element is returned (in the order matched) as a\nsingle, space-delimited string.\n\nIf no matching text node is found for the expression (including the\nimplicit /text()) --- for whatever reason, as long as xpath_expr is\nvalid, and xml_frag consists of elements which are properly nested and\nclosed --- an empty string is returned. No distinction is made between\na match on an empty element and no match at all. This is by design.\n\nIf you need to determine whether no matching element was found in\nxml_frag or such an element was found but contained no child text\nnodes, you should test the result of an expression that uses the XPath\ncount() function. For example, both of these statements return an empty\nstring, as shown here:\n\nmysql> SELECT ExtractValue(\'<a></a>\', \'/a/b\');\n+-------------------------------------+\n| ExtractValue(\'<a></a>\', \'/a/b\') |\n+-------------------------------------+\n| |\n+-------------------------------------+\n1 row in set (0.00 sec)\n\nmysql> SELECT ExtractValue(\'<a><c/></a>\', \'/a/b\');\n+-------------------------------------+\n| ExtractValue(\'<a><c/></a>\', \'/a/b\') |\n+-------------------------------------+\n| |\n+-------------------------------------+\n1 row in set (0.00 sec)\n\nHowever, you can determine whether there was actually a matching\nelement using the following:\n\nmysql> SELECT ExtractValue(\'<a></a>\', \'count(/a/b)\');\n+-------------------------------------+\n| ExtractValue(\'<a></a>\', \'count(/a/b)\') |\n+-------------------------------------+\n| 1 |\n+-------------------------------------+\n1 row in set (0.00 sec)\n\nmysql> SELECT ExtractValue(\'<a><c/></a>\', \'count(/a/b)\');\n+-------------------------------------+\n| ExtractValue(\'<a><c/></a>\', \'count(/a/b)\') |\n+-------------------------------------+\n| 0 |\n+-------------------------------------+\n1 row in set (0.01 sec)\n\n*Important*: ExtractValue() returns only CDATA, and does not return any\ntags that might be contained within a matching tag, nor any of their\ncontent (see the result returned as val1 in the following example).\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/xml-functions.html\n\n','mysql> SELECT\n -> ExtractValue(\'<a>cccddd</a>\', \'/a\') AS val1,\n -> ExtractValue(\'<a>cccddd</a>\', \'/a/b\') AS val2,\n -> ExtractValue(\'<a>cccddd</a>\', \'//b\') AS val3,\n -> ExtractValue(\'<a>cccddd</a>\', \'/b\') AS val4,\n -> ExtractValue(\'<a>cccdddeee</a>\', \'//b\') AS val5;\n\n+------+------+------+------+---------+\n| val1 | val2 | val3 | val4 | val5 |\n+------+------+------+------+---------+\n| ccc | ddd | ddd | | ddd eee |\n+------+------+------+------+---------+\n','http://dev.mysql.com/doc/refman/5.1/en/xml-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (135,11,'OLD_PASSWORD','Syntax:\nOLD_PASSWORD(str)\n\nOLD_PASSWORD() was added to MySQL when the implementation of PASSWORD()\nwas changed to improve security. OLD_PASSWORD() returns the value of\nthe old (pre-4.1) implementation of PASSWORD() as a binary string, and\nis intended to permit you to reset passwords for any pre-4.1 clients\nthat need to connect to your version 5.1 MySQL server without locking\nthem out. See\nhttp://dev.mysql.com/doc/refman/5.1/en/password-hashing.html.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/encryption-functions.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/encryption-functions.html'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (136,22,'SET VARIABLE','Syntax:\nSET var_name = expr [, var_name = expr] ...\n\nThe SET statement in stored programs is an extended version of the\ngeneral SET statement (see [HELP SET]). Referenced variables may be\nones declared inside a stored program, global system variables, or\nuser-defined variables.\n\nThe SET statement in stored programs is implemented as part of the\npre-existing SET syntax. This allows an extended syntax of SET a=x,\nb=y, ... where different variable types (locally declared variables,\nglobal and session server variables, user-defined variables) can be\nmixed. This also allows combinations of local variables and some\noptions that make sense only for system variables; in that case, the\noptions are recognized but ignored.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/set-statement.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/set-statement.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (136,22,'SET VARIABLE','Syntax:\nSET var_name = expr [, var_name = expr] ...\n\nThe SET statement in stored programs is an extended version of the\ngeneral SET statement (see [HELP SET]). Each var_name may refer to a\nlocal variable declared inside a stored program, a system variable, or\na user-defined variable.\n\nThe SET statement in stored programs is implemented as part of the\npre-existing SET syntax. This allows an extended syntax of SET a=x,\nb=y, ... where different variable types (locally declared variables,\nglobal and session system variables, user-defined variables) can be\nmixed. This also allows combinations of local variables and some\noptions that make sense only for system variables; in that case, the\noptions are recognized but ignored.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/set-statement.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/set-statement.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (137,35,'FORMAT','Syntax:\nFORMAT(X,D)\n\nFormats the number X to a format like \'#,###,###.##\', rounded to D\ndecimal places, and returns the result as a string. If D is 0, the\nresult has no decimal point or fractional part.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/string-functions.html\n\n','mysql> SELECT FORMAT(12332.123456, 4);\n -> \'12,332.1235\'\nmysql> SELECT FORMAT(12332.1,4);\n -> \'12,332.1000\'\nmysql> SELECT FORMAT(12332.2,0);\n -> \'12,332\'\n','http://dev.mysql.com/doc/refman/5.1/en/string-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (138,13,'||','Syntax:\nOR, ||\n\nLogical OR. When both operands are non-NULL, the result is 1 if any\noperand is nonzero, and 0 otherwise. With a NULL operand, the result is\n1 if the other operand is nonzero, and NULL otherwise. If both operands\nare NULL, the result is NULL.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/logical-operators.html\n\n','mysql> SELECT 1 || 1;\n -> 1\nmysql> SELECT 1 || 0;\n -> 1\nmysql> SELECT 0 || 0;\n -> 0\nmysql> SELECT 0 || NULL;\n -> NULL\nmysql> SELECT 1 || NULL;\n -> 1\n','http://dev.mysql.com/doc/refman/5.1/en/logical-operators.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (139,35,'BIT_LENGTH','Syntax:\nBIT_LENGTH(str)\n\nReturns the length of the string str in bits.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/string-functions.html\n\n','mysql> SELECT BIT_LENGTH(\'text\');\n -> 32\n','http://dev.mysql.com/doc/refman/5.1/en/string-functions.html'); @@ -233,7 +233,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (164,30,'SEC_TO_TIME','Syntax:\nSEC_TO_TIME(seconds)\n\nReturns the seconds argument, converted to hours, minutes, and seconds,\nas a TIME value. The range of the result is constrained to that of the\nTIME data type. A warning occurs if the argument corresponds to a value\noutside that range.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','mysql> SELECT SEC_TO_TIME(2378);\n -> \'00:39:38\'\nmysql> SELECT SEC_TO_TIME(2378) + 0;\n -> 3938\n','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (165,20,'FLOAT','FLOAT[(M,D)] [UNSIGNED] [ZEROFILL]\n\nA small (single-precision) floating-point number. Allowable values are\n-3.402823466E+38 to -1.175494351E-38, 0, and 1.175494351E-38 to\n3.402823466E+38. These are the theoretical limits, based on the IEEE\nstandard. The actual range might be slightly smaller depending on your\nhardware or operating system.\n\nM is the total number of digits and D is the number of digits following\nthe decimal point. If M and D are omitted, values are stored to the\nlimits allowed by the hardware. A single-precision floating-point\nnumber is accurate to approximately 7 decimal places.\n\nUNSIGNED, if specified, disallows negative values.\n\nUsing FLOAT might give you some unexpected problems because all\ncalculations in MySQL are done with double precision. See\nhttp://dev.mysql.com/doc/refman/5.1/en/no-matching-rows.html.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/numeric-type-overview.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/numeric-type-overview.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (166,35,'LOCATE','Syntax:\nLOCATE(substr,str), LOCATE(substr,str,pos)\n\nThe first syntax returns the position of the first occurrence of\nsubstring substr in string str. The second syntax returns the position\nof the first occurrence of substring substr in string str, starting at\nposition pos. Returns 0 if substr is not in str.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/string-functions.html\n\n','mysql> SELECT LOCATE(\'bar\', \'foobarbar\');\n -> 4\nmysql> SELECT LOCATE(\'xbar\', \'foobar\');\n -> 0\nmysql> SELECT LOCATE(\'bar\', \'foobarbar\', 5);\n -> 7\n','http://dev.mysql.com/doc/refman/5.1/en/string-functions.html'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (167,25,'SHOW EVENTS','Syntax:\nSHOW EVENTS [{FROM | IN} schema_name]\n [LIKE \'pattern\' | WHERE expr]\n\nIn its simplest form, SHOW EVENTS lists all of the events in the\ncurrent schema:\n\nmysql> SELECT CURRENT_USER(), SCHEMA();\n+----------------+----------+\n| CURRENT_USER() | SCHEMA() |\n+----------------+----------+\n| jon@ghidora | myschema |\n+----------------+----------+\n1 row in set (0.00 sec)\n\nmysql> SHOW EVENTS\\G\n*************************** 1. row ***************************\n Db: myschema\n Name: e_daily\n Definer: jon@ghidora\n Time zone: SYSTEM\n Type: RECURRING\n Execute at: NULL\n Interval value: 10\n Interval field: SECOND\n Starts: 2006-02-09 10:41:23\n Ends: 0000-00-00 00:00:00\n Status: ENABLED\n Originator: 0\ncharacter_set_client: latin1\ncollation_connection: latin1_swedish_ci\n Database Collation: latin1_swedish_ci\n\nThe LIKE clause, if present, indicates which event names to match. The\nWHERE clause can be given to select rows using more general conditions,\nas discussed in\nhttp://dev.mysql.com/doc/refman/5.1/en/extended-show.html.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/show-events.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/show-events.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (167,25,'SHOW EVENTS','Syntax:\nSHOW EVENTS [{FROM | IN} schema_name]\n [LIKE \'pattern\' | WHERE expr]\n\nIn its simplest form, SHOW EVENTS lists all of the events in the\ncurrent schema:\n\nmysql> SELECT CURRENT_USER(), SCHEMA();\n+----------------+----------+\n| CURRENT_USER() | SCHEMA() |\n+----------------+----------+\n| jon@ghidora | myschema |\n+----------------+----------+\n1 row in set (0.00 sec)\n\nmysql> SHOW EVENTS\\G\n*************************** 1. row ***************************\n Db: myschema\n Name: e_daily\n Definer: jon@ghidora\n Time zone: SYSTEM\n Type: RECURRING\n Execute at: NULL\n Interval value: 10\n Interval field: SECOND\n Starts: 2006-02-09 10:41:23\n Ends: NULL\n Status: ENABLED\n Originator: 0\ncharacter_set_client: latin1\ncollation_connection: latin1_swedish_ci\n Database Collation: latin1_swedish_ci\n\nTo see events for a specific schema, use the FROM clause. For example,\nto see events for the test schema, use the following statement:\n\nSHOW EVENTS FROM test;\n\nThe LIKE clause, if present, indicates which event names to match. The\nWHERE clause can be given to select rows using more general conditions,\nas discussed in\nhttp://dev.mysql.com/doc/refman/5.1/en/extended-show.html.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/show-events.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/show-events.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (168,15,'CHARSET','Syntax:\nCHARSET(str)\n\nReturns the character set of the string argument.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/information-functions.html\n\n','mysql> SELECT CHARSET(\'abc\');\n -> \'latin1\'\nmysql> SELECT CHARSET(CONVERT(\'abc\' USING utf8));\n -> \'utf8\'\nmysql> SELECT CHARSET(USER());\n -> \'utf8\'\n','http://dev.mysql.com/doc/refman/5.1/en/information-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (169,30,'SUBDATE','Syntax:\nSUBDATE(date,INTERVAL expr unit), SUBDATE(expr,days)\n\nWhen invoked with the INTERVAL form of the second argument, SUBDATE()\nis a synonym for DATE_SUB(). For information on the INTERVAL unit\nargument, see the discussion for DATE_ADD().\n\nmysql> SELECT DATE_SUB(\'2008-01-02\', INTERVAL 31 DAY);\n -> \'2007-12-02\'\nmysql> SELECT SUBDATE(\'2008-01-02\', INTERVAL 31 DAY);\n -> \'2007-12-02\'\n\nThe second form allows the use of an integer value for days. In such\ncases, it is interpreted as the number of days to be subtracted from\nthe date or datetime expression expr.\n\nmysql> SELECT SUBDATE(\'2008-01-02 12:00:00\', 31);\n -> \'2007-12-02 12:00:00\'\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (170,30,'DAYOFYEAR','Syntax:\nDAYOFYEAR(date)\n\nReturns the day of the year for date, in the range 1 to 366.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','mysql> SELECT DAYOFYEAR(\'2007-02-03\');\n -> 34\n','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); @@ -250,7 +250,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (181,24,'NUMGEOMETRIES','NumGeometries(gc)\n\nReturns the number of geometries in the GeometryCollection value gc.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/geometry-property-functions.html#geo…','mysql> SET @gc = \'GeometryCollection(Point(1 1),LineString(2 2, 3 3))\';\nmysql> SELECT NumGeometries(GeomFromText(@gc));\n+----------------------------------+\n| NumGeometries(GeomFromText(@gc)) |\n+----------------------------------+\n| 2 |\n+----------------------------------+\n','http://dev.mysql.com/doc/refman/5.1/en/geometry-property-functions.html#geo…'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (182,30,'MONTHNAME','Syntax:\nMONTHNAME(date)\n\nReturns the full name of the month for date. As of MySQL 5.1.12, the\nlanguage used for the name is controlled by the value of the\nlc_time_names system variable\n(http://dev.mysql.com/doc/refman/5.1/en/locale-support.html).\n\n…: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','mysql> SELECT MONTHNAME(\'2008-02-03\');\n -> \'February\'\n','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (183,36,'PROCEDURE ANALYSE','Syntax:\nANALYSE([max_elements[,max_memory]])\n\nANALYSE() is defined in the sql/sql_analyse.cc source file, which\nserves as an example of how to create a procedure for use with the\nPROCEDURE clause of SELECT statements. ANALYSE() is built in and is\navailable by default; other procedures can be created using the format\ndemonstrated in the source file.\n\nANALYSE() examines the result from a query and returns an analysis of\nthe results that suggests optimal data types for each column that may\nhelp reduce table sizes. To obtain this analysis, append PROCEDURE\nANALYSE to the end of a SELECT statement:\n\nSELECT ... FROM ... WHERE ... PROCEDURE ANALYSE([max_elements,[max_memory]])\n\nFor example:\n\nSELECT col1, col2 FROM table1 PROCEDURE ANALYSE(10, 2000);\n\nThe results show some statistics for the values returned by the query,\nand propose an optimal data type for the columns. This can be helpful\nfor checking your existing tables, or after importing new data. You may\nneed to try different settings for the arguments so that PROCEDURE\nANALYSE() does not suggest the ENUM data type when it is not\nappropriate.\n\nThe arguments are optional and are used as follows:\n\no max_elements (default 256) is the maximum number of distinct values\n that ANALYSE() notices per column. This is used by ANALYSE() to check\n whether the optimal data type should be of type ENUM; if there are\n more than max_elements distinct values, then ENUM is not a suggested\n type.\n\no max_memory (default 8192) is the maximum amount of memory that\n ANALYSE() should allocate per column while trying to find all\n distinct values.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/procedure-analyse.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/procedure-analyse.html'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (184,25,'CHANGE MASTER TO','Syntax:\nCHANGE MASTER TO master_def [, master_def] ...\n\nmaster_def:\n MASTER_BIND = \'interface_name\'\n | MASTER_HOST = \'host_name\'\n | MASTER_USER = \'user_name\'\n | MASTER_PASSWORD = \'password\'\n | MASTER_PORT = port_num\n | MASTER_CONNECT_RETRY = interval\n | MASTER_HEARTBEAT_PERIOD = interval\n | MASTER_LOG_FILE = \'master_log_name\'\n | MASTER_LOG_POS = master_log_pos\n | RELAY_LOG_FILE = \'relay_log_name\'\n | RELAY_LOG_POS = relay_log_pos\n | MASTER_SSL = {0|1}\n | MASTER_SSL_CA = \'ca_file_name\'\n | MASTER_SSL_CAPATH = \'ca_directory_name\'\n | MASTER_SSL_CERT = \'cert_file_name\'\n | MASTER_SSL_KEY = \'key_file_name\'\n | MASTER_SSL_CIPHER = \'cipher_list\'\n | MASTER_SSL_VERIFY_SERVER_CERT = {0|1}\n\nCHANGE MASTER TO changes the parameters that the slave server uses for\nconnecting to and communicating with the master server. It also updates\nthe contents of the master.info and relay-log.info files.\n\nMASTER_USER, MASTER_PASSWORD, MASTER_SSL, MASTER_SSL_CA,\nMASTER_SSL_CAPATH, MASTER_SSL_CERT, MASTER_SSL_KEY, MASTER_SSL_CIPHER,\nand MASTER_SSL_VERIFY_SERVER_CERT provide information to the slave\nabout how to connect to its master. MASTER_SSL_VERIFY_SERVER_CERT was\nadded in MySQL 5.1.18. It is used as described for the\n--ssl-verify-server-cert option in\nhttp://dev.mysql.com/doc/refman/5.1/en/ssl-options.html.\n\nMASTER_CONN… specifies how many seconds to wait between connect\nretries. The default is 60. The number of reconnection attempts is\nlimited by the --master-retry-count server option; for more\ninformation, see\nhttp://dev.mysql.com/doc/refman/5.1/en/replication-options.html.\n\nThe SSL options (MASTER_SSL, MASTER_SSL_CA, MASTER_SSL_CAPATH,\nMASTER_SSL_CERT, MASTER_SSL_KEY, MASTER_SSL_CIPHER), and\nMASTER_SSL_VERIFY_SERVER_CERT can be changed even on slaves that are\ncompiled without SSL support. They are saved to the master.info file,\nbut are ignored unless you use a server that has SSL support enabled.\n\nIf you don\'t specify a given parameter, it keeps its old value, except\nas indicated in the following discussion. For example, if the password\nto connect to your MySQL master has changed, you just need to issue\nthese statements to tell the slave about the new password:\n\nSTOP SLAVE; -- if replication was running\nCHANGE MASTER TO MASTER_PASSWORD=\'new3cret\';\nSTART SLAVE; -- if you want to restart replication\n\nThere is no need to specify the parameters that do not change (host,\nport, user, and so forth).\n\nMASTER_HOST and MASTER_PORT are the host name (or IP address) of the\nmaster host and its TCP/IP port.\n\nThe next two options are available only in MySQL Cluster NDB 6.3 and\n6.4; they are not supported in mainline MySQL 5.1:\n\no MASTER_BIND is for use on replication slaves having multiple network\n interfaces, and determines which of the slave\'s network interfaces is\n chosen for connecting to the master. It is also possible to determine\n which network interface is to be used in such cases by starting the\n slave mysqld process with the --master-bind option.\n\n The ability to bind a replication slave to specific network interface\n was added in MySQL Cluster NDB 6.3.4.\n\no MASTER_HEARTBEAT_PERIOD is used to set the interval in seconds\n between replication heartbeats. Whenever the master\'s binlog is\n updated with an event, the waiting period for the next heartbeat is\n reset. interval is a decimal value having the range 0 to 4294967\n seconds and a resolution to hundredths of a second; the smallest\n nonzero value is 0.001. Heartbeats are sent by the master only if\n there are no unsent events in the binlog file for a period longer\n than interval.\n\n Setting interval to 0 disables heartbeats altogether. The default\n value for interval is equal to the value of slave_net_timeout divided\n by 2.\n\n *Note*: Setting @@global.slave_net_timeout to a value less than that\n of the current heartbeat interval results in a warning being issued.\n\n Issuing RESET SLAVE resets the heartbeat interval to the default.\n\n MASTER_HEARTBEAT_PERIOD was added in MySQL Cluster NDB 6.3.4.\n\n*Note*: Replication cannot use Unix socket files. You must be able to\nconnect to the master MySQL server using TCP/IP.\n\nIf you specify MASTER_HOST or MASTER_PORT, the slave assumes that the\nmaster server is different from before (even if you specify a host or\nport value that is the same as the current value.) In this case, the\nold values for the master binary log name and position are considered\nno longer applicable, so if you do not specify MASTER_LOG_FILE and\nMASTER_LOG_POS in the statement, MASTER_LOG_FILE=\'\' and\nMASTER_LOG_POS=4 are silently appended to it.\n\nSetting MASTER_HOST=\'\' --- that is, setting its value explicitly to an\nempty string --- is not the same as not setting it at all. Setting this\noption to an empty string causes START SLAVE subsequently to fail. This\nissue is addressed in MySQL 6.0. (Bug#28796\n(http://bugs.mysql.com/28796))\n\nMASTER_LOG_FILE and MASTER_LOG_POS are the coordinates at which the\nslave I/O thread should begin reading from the master the next time the\nthread starts. If you specify either of them, you cannot specify\nRELAY_LOG_FILE or RELAY_LOG_POS. If neither of MASTER_LOG_FILE or\nMASTER_LOG_POS are specified, the slave uses the last coordinates of\nthe slave SQL thread before CHANGE MASTER TO was issued. This ensures\nthat there is no discontinuity in replication, even if the slave SQL\nthread was late compared to the slave I/O thread, when you merely want\nto change, say, the password to use.\n\nCHANGE MASTER TO deletes all relay log files and starts a new one,\nunless you specify RELAY_LOG_FILE or RELAY_LOG_POS. In that case, relay\nlogs are kept; the relay_log_purge global variable is set silently to\n0.\n\nCHANGE MASTER TO is useful for setting up a slave when you have the\nsnapshot of the master and have recorded the log and the offset\ncorresponding to it. After loading the snapshot into the slave, you can\nrun CHANGE MASTER TO MASTER_LOG_FILE=\'log_name_on_master\',\nMASTER_LOG_POS=log_offset_on_master on the slave.\n\nThe following example changes the master and master\'s binary log\ncoordinates. This is used when you want to set up the slave to\nreplicate the master:\n\nCHANGE MASTER TO\n MASTER_HOST=\'master2.mycompany.com\',\n MASTER_USER=\'replication\',\n MASTER_PASSWORD=\'bigs3cret\',\n MASTER_PORT=3306,\n MASTER_LOG_FILE=\'master2-bin.001\',\n MASTER_LOG_POS=4,\n MASTER_CONNECT_RETRY=10;\n\nThe next example shows an operation that is less frequently employed.\nIt is used when the slave has relay logs that you want it to execute\nagain for some reason. To do this, the master need not be reachable.\nYou need only use CHANGE MASTER TO and start the SQL thread (START\nSLAVE SQL_THREAD):\n\nCHANGE MASTER TO\n RELAY_LOG_FILE=\'slave-relay-bin.006\',\n RELAY_LOG_POS=4025;\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/change-master-to.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/change-master-to.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (184,25,'CHANGE MASTER TO','Syntax:\nCHANGE MASTER TO master_def [, master_def] ...\n\nmaster_def:\n MASTER_BIND = \'interface_name\'\n | MASTER_HOST = \'host_name\'\n | MASTER_USER = \'user_name\'\n | MASTER_PASSWORD = \'password\'\n | MASTER_PORT = port_num\n | MASTER_CONNECT_RETRY = interval\n | MASTER_HEARTBEAT_PERIOD = interval\n | MASTER_LOG_FILE = \'master_log_name\'\n | MASTER_LOG_POS = master_log_pos\n | RELAY_LOG_FILE = \'relay_log_name\'\n | RELAY_LOG_POS = relay_log_pos\n | MASTER_SSL = {0|1}\n | MASTER_SSL_CA = \'ca_file_name\'\n | MASTER_SSL_CAPATH = \'ca_directory_name\'\n | MASTER_SSL_CERT = \'cert_file_name\'\n | MASTER_SSL_KEY = \'key_file_name\'\n | MASTER_SSL_CIPHER = \'cipher_list\'\n | MASTER_SSL_VERIFY_SERVER_CERT = {0|1}\n\nCHANGE MASTER TO changes the parameters that the slave server uses for\nconnecting to and communicating with the master server. It also updates\nthe contents of the master.info and relay-log.info files.\n\nMASTER_USER, MASTER_PASSWORD, MASTER_SSL, MASTER_SSL_CA,\nMASTER_SSL_CAPATH, MASTER_SSL_CERT, MASTER_SSL_KEY, MASTER_SSL_CIPHER,\nand MASTER_SSL_VERIFY_SERVER_CERT provide information to the slave\nabout how to connect to its master. MASTER_SSL_VERIFY_SERVER_CERT was\nadded in MySQL 5.1.18. It is used as described for the\n--ssl-verify-server-cert option in\nhttp://dev.mysql.com/doc/refman/5.1/en/ssl-options.html.\n\nMASTER_CONN… specifies how many seconds to wait between connect\nretries. The default is 60. The number of reconnection attempts is\nlimited by the --master-retry-count server option; for more\ninformation, see\nhttp://dev.mysql.com/doc/refman/5.1/en/replication-options.html.\n\nThe SSL options (MASTER_SSL, MASTER_SSL_CA, MASTER_SSL_CAPATH,\nMASTER_SSL_CERT, MASTER_SSL_KEY, MASTER_SSL_CIPHER), and\nMASTER_SSL_VERIFY_SERVER_CERT can be changed even on slaves that are\ncompiled without SSL support. They are saved to the master.info file,\nbut are ignored unless you use a server that has SSL support enabled.\n\nIf you do not specify a given parameter, it keeps its old value, except\nas indicated in the following discussion. For example, if the password\nto connect to your MySQL master has changed, you just need to issue\nthese statements to tell the slave about the new password:\n\nSTOP SLAVE; -- if replication was running\nCHANGE MASTER TO MASTER_PASSWORD=\'new3cret\';\nSTART SLAVE; -- if you want to restart replication\n\nThere is no need to specify the parameters that do not change (host,\nport, user, and so forth).\n\nMASTER_HOST and MASTER_PORT are the host name (or IP address) of the\nmaster host and its TCP/IP port.\n\nThe next two options (MASTER_BIND and MASTER_HEARTBEAT_PERIOD) are\navailable in MySQL Cluster NDB 6.3 and later, but are not supported in\nmainline MySQL 5.1:\n\no MASTER_BIND is for use on replication slaves having multiple network\n interfaces, and determines which of the slave\'s network interfaces is\n chosen for connecting to the master. It is also possible to determine\n which network interface is to be used in such cases by starting the\n slave mysqld process with the --master-bind option.\n\n The ability to bind a replication slave to specific network interface\n was added in MySQL Cluster NDB 6.3.4.\n\no MASTER_HEARTBEAT_PERIOD is used to set the interval in seconds\n between replication heartbeats. Whenever the master\'s binlog is\n updated with an event, the waiting period for the next heartbeat is\n reset. interval is a decimal value having the range 0 to 4294967\n seconds and a resolution to hundredths of a second; the smallest\n nonzero value is 0.001. Heartbeats are sent by the master only if\n there are no unsent events in the binlog file for a period longer\n than interval.\n\n Setting interval to 0 disables heartbeats altogether. The default\n value for interval is equal to the value of slave_net_timeout divided\n by 2.\n\n Setting @@global.slave_net_timeout to a value less than that of the\n current heartbeat interval results in a warning being issued. The\n effect of issuing RESET SLAVE on the heartbeat interval is to reset\n it to the default value.\n\n MASTER_HEARTBEAT_PERIOD was added in MySQL Cluster NDB 6.3.4.\n\n*Note*: Replication cannot use Unix socket files. You must be able to\nconnect to the master MySQL server using TCP/IP.\n\nIf you specify MASTER_HOST or MASTER_PORT, the slave assumes that the\nmaster server is different from before (even if you specify a host or\nport value that is the same as the current value.) In this case, the\nold values for the master binary log name and position are considered\nno longer applicable, so if you do not specify MASTER_LOG_FILE and\nMASTER_LOG_POS in the statement, MASTER_LOG_FILE=\'\' and\nMASTER_LOG_POS=4 are silently appended to it.\n\nSetting MASTER_HOST=\'\' --- that is, setting its value explicitly to an\nempty string --- is not the same as not setting it at all. Setting this\noption to an empty string causes START SLAVE subsequently to fail. This\nissue is addressed in MySQL 5.5. (Bug#28796\n(http://bugs.mysql.com/28796))\n\nMASTER_LOG_FILE and MASTER_LOG_POS are the coordinates at which the\nslave I/O thread should begin reading from the master the next time the\nthread starts. If you specify either of them, you cannot specify\nRELAY_LOG_FILE or RELAY_LOG_POS. If neither of MASTER_LOG_FILE or\nMASTER_LOG_POS are specified, the slave uses the last coordinates of\nthe slave SQL thread before CHANGE MASTER TO was issued. This ensures\nthat there is no discontinuity in replication, even if the slave SQL\nthread was late compared to the slave I/O thread, when you merely want\nto change, say, the password to use.\n\nCHANGE MASTER TO deletes all relay log files and starts a new one,\nunless you specify RELAY_LOG_FILE or RELAY_LOG_POS. In that case, relay\nlogs are kept; the relay_log_purge global variable is set silently to\n0.\n\nCHANGE MASTER TO is useful for setting up a slave when you have the\nsnapshot of the master and have recorded the log and the offset\ncorresponding to it. After loading the snapshot into the slave, you can\nrun CHANGE MASTER TO MASTER_LOG_FILE=\'log_name_on_master\',\nMASTER_LOG_POS=log_offset_on_master on the slave.\n\nThe following example changes the master and master\'s binary log\ncoordinates. This is used when you want to set up the slave to\nreplicate the master:\n\nCHANGE MASTER TO\n MASTER_HOST=\'master2.mycompany.com\',\n MASTER_USER=\'replication\',\n MASTER_PASSWORD=\'bigs3cret\',\n MASTER_PORT=3306,\n MASTER_LOG_FILE=\'master2-bin.001\',\n MASTER_LOG_POS=4,\n MASTER_CONNECT_RETRY=10;\n\nThe next example shows an operation that is less frequently employed.\nIt is used when the slave has relay logs that you want it to execute\nagain for some reason. To do this, the master need not be reachable.\nYou need only use CHANGE MASTER TO and start the SQL thread (START\nSLAVE SQL_THREAD):\n\nCHANGE MASTER TO\n RELAY_LOG_FILE=\'slave-relay-bin.006\',\n RELAY_LOG_POS=4025;\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/change-master-to.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/change-master-to.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (185,37,'DROP DATABASE','Syntax:\nDROP {DATABASE | SCHEMA} [IF EXISTS] db_name\n\nDROP DATABASE drops all tables in the database and deletes the\ndatabase. Be very careful with this statement! To use DROP DATABASE,\nyou need the DROP privilege on the database. DROP SCHEMA is a synonym\nfor DROP DATABASE.\n\n*Important*: When a database is dropped, user privileges on the\ndatabase are not automatically dropped. See [HELP GRANT].\n\nIF EXISTS is used to prevent an error from occurring if the database\ndoes not exist.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/drop-database.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/drop-database.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (186,6,'MBREQUAL','MBREqual(g1,g2)\n\nReturns 1 or 0 to indicate whether the Minimum Bounding Rectangles of\nthe two geometries g1 and g2 are the same.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/relations-on-geometry-mbr.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/relations-on-geometry-mbr.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (187,30,'TIMESTAMP FUNCTION','Syntax:\nTIMESTAMP(expr), TIMESTAMP(expr1,expr2)\n\nWith a single argument, this function returns the date or datetime\nexpression expr as a datetime value. With two arguments, it adds the\ntime expression expr2 to the date or datetime expression expr1 and\nreturns the result as a datetime value.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','mysql> SELECT TIMESTAMP(\'2003-12-31\');\n -> \'2003-12-31 00:00:00\'\nmysql> SELECT TIMESTAMP(\'2003-12-31 12:00:00\',\'12:00:00\');\n -> \'2004-01-01 00:00:00\'\n','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); @@ -322,7 +322,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (253,20,'VARCHAR','[NATIONAL] VARCHAR(M) [CHARACTER SET charset_name] [COLLATE\ncollation_name]\n\nA variable-length string. M represents the maximum column length in\ncharacters. The range of M is 0 to 65,535. The effective maximum length\nof a VARCHAR is subject to the maximum row size (65,535 bytes, which is\nshared among all columns) and the character set used. For example, utf8\ncharacters can require up to three bytes per character, so a VARCHAR\ncolumn that uses the utf8 character set can be declared to be a maximum\nof 21,844 characters.\n\nMySQL stores VARCHAR values as a one-byte or two-byte length prefix\nplus data. The length prefix indicates the number of bytes in the\nvalue. A VARCHAR column uses one length byte if values require no more\nthan 255 bytes, two length bytes if values may require more than 255\nbytes.\n\n*Note*: MySQL 5.1 follows the standard SQL specification, and does not\nremove trailing spaces from VARCHAR values.\n\nVARCHAR is shorthand for CHARACTER VARYING. NATIONAL VARCHAR is the\nstandard SQL way to define that a VARCHAR column should use some\npredefined character set. MySQL 4.1 and up uses utf8 as this predefined\ncharacter set.\nhttp://dev.mysql.com/doc/refman/5.1/en/charset-national.html. NVARCHAR\nis shorthand for NATIONAL VARCHAR.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/string-type-overview.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/string-type-overview.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (254,35,'UNHEX','Syntax:\n\nUNHEX(str)\n\nPerforms the inverse operation of HEX(str). That is, it interprets each\npair of hexadecimal digits in the argument as a number and converts it\nto the character represented by the number. The resulting characters\nare returned as a binary string.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/string-functions.html\n\n','mysql> SELECT UNHEX(\'4D7953514C\');\n -> \'MySQL\'\nmysql> SELECT 0x4D7953514C;\n -> \'MySQL\'\nmysql> SELECT UNHEX(HEX(\'string\'));\n -> \'string\'\nmysql> SELECT HEX(UNHEX(\'1267\'));\n -> \'1267\'\n','http://dev.mysql.com/doc/refman/5.1/en/string-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (255,4,'- UNARY','Syntax:\n-\n\nUnary minus. This operator changes the sign of the argument.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/arithmetic-functions.html\n\n','mysql> SELECT - 2;\n -> -2\n','http://dev.mysql.com/doc/refman/5.1/en/arithmetic-functions.html'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (256,22,'SELECT INTO','Syntax:\nSELECT col_name [, col_name] ...\n INTO var_name [, var_name] ...\n table_expr\n\nSELECT ... INTO syntax enables selected columns to be stored directly\ninto variables. The query should return a single row. If the query\nreturns no rows, a warning with error code 1329 occurs (No data), and\nthe variable values remain unchanged. If the query returns multiple\nrows, error 1172 occurs (Result consisted of more than one row). If it\nis possible that the statement may retrieve multiple rows, you can use\nLIMIT 1 to limit the result set to a single row.\n\nIn the context of such statements that occur as part of events executed\nby the Event Scheduler, diagnostics messages (not only errors, but also\nwarnings) are written to the error log, and, on Windows, to the\napplication event log. For additional information, see\nhttp://dev.mysql.com/doc/refman/5.1/en/events-status-info.html.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/select-into-statement.html\n\n','SELECT id,data INTO x,y FROM test.t1 LIMIT 1;\n','http://dev.mysql.com/doc/refman/5.1/en/select-into-statement.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (256,22,'SELECT INTO','Syntax:\nSELECT col_name [, col_name] ...\n INTO var_name [, var_name] ...\n table_expr\n\nSELECT ... INTO syntax enables selected columns to be stored directly\ninto variables. The query should return a single row. If the query\nreturns no rows, a warning with error code 1329 occurs (No data), and\nthe variable values remain unchanged. If the query returns multiple\nrows, error 1172 occurs (Result consisted of more than one row). If it\nis possible that the statement may retrieve multiple rows, you can use\nLIMIT 1 to limit the result set to a single row.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/select-into-statement.html\n\n','SELECT id,data INTO x,y FROM test.t1 LIMIT 1;\n','http://dev.mysql.com/doc/refman/5.1/en/select-into-statement.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (257,16,'STD','Syntax:\nSTD(expr)\n\nReturns the population standard deviation of expr. This is an extension\nto standard SQL. The standard SQL function STDDEV_POP() can be used\ninstead.\n\nThis function returns NULL if there were no matching rows.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/group-by-functions.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/group-by-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (258,4,'COS','Syntax:\nCOS(X)\n\nReturns the cosine of X, where X is given in radians.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/mathematical-functions.html\n\n','mysql> SELECT COS(PI());\n -> -1\n','http://dev.mysql.com/doc/refman/5.1/en/mathematical-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (259,30,'DATE FUNCTION','Syntax:\nDATE(expr)\n\nExtracts the date part of the date or datetime expression expr.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','mysql> SELECT DATE(\'2003-12-31 01:02:03\');\n -> \'2003-12-31\'\n','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); @@ -384,7 +384,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (315,15,'FOUND_ROWS','Syntax:\nFOUND_ROWS()\n\nA SELECT statement may include a LIMIT clause to restrict the number of\nrows the server returns to the client. In some cases, it is desirable\nto know how many rows the statement would have returned without the\nLIMIT, but without running the statement again. To obtain this row\ncount, include a SQL_CALC_FOUND_ROWS option in the SELECT statement,\nand then invoke FOUND_ROWS() afterward:\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/information-functions.html\n\n','mysql> SELECT SQL_CALC_FOUND_ROWS * FROM tbl_name\n -> WHERE id > 100 LIMIT 10;\nmysql> SELECT FOUND_ROWS();\n','http://dev.mysql.com/doc/refman/5.1/en/information-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (316,15,'SYSTEM_USER','Syntax:\nSYSTEM_USER()\n\nSYSTEM_USER() is a synonym for USER().\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/information-functions.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/information-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (317,29,'CROSSES','Crosses(g1,g2)\n\nReturns 1 if g1 spatially crosses g2. Returns NULL if g1 is a Polygon\nor a MultiPolygon, or if g2 is a Point or a MultiPoint. Otherwise,\nreturns 0.\n\nThe term spatially crosses denotes a spatial relation between two given\ngeometries that has the following properties:\n\no The two geometries intersect\n\no Their intersection results in a geometry that has a dimension that is\n one less than the maximum dimension of the two given geometries\n\no Their intersection is not equal to either of the two given geometries\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/functions-that-test-spatial-relation…','','http://dev.mysql.com/doc/refman/5.1/en/functions-that-test-spatial-relation…'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (318,26,'TRUNCATE TABLE','Syntax:\nTRUNCATE [TABLE] tbl_name\n\nTRUNCATE TABLE empties a table completely. Logically, this is\nequivalent to a DELETE statement that deletes all rows, but there are\npractical differences under some circumstances.\n\nFor an InnoDB table, InnoDB processes TRUNCATE TABLE by deleting rows\none by one if there are any FOREIGN KEY constraints that reference the\ntable. If there are no FOREIGN KEY constraints, InnoDB performs fast\ntruncation by dropping the original table and creating an empty one\nwith the same definition, which is much faster than deleting rows one\nby one. The AUTO_INCREMENT counter is reset by TRUNCATE TABLE,\nregardless of whether there is a FOREIGN KEY constraint.\n\nIn the case that FOREIGN KEY constraints reference the table, InnoDB\ndeletes rows one by one and processes the constraints on each one. If\nthe FOREIGN KEY constraint specifies DELETE CASCADE, rows from the\nchild (referenced) table are deleted, and the truncated table becomes\nempty. If the FOREIGN KEY constraint does not specify CASCADE, the\nTRUNCATE statement deletes rows one by one and stops if it encounters a\nparent row that is referenced by the child, returning this error:\n\nERROR 1451 (23000): Cannot delete or update a parent row: a foreign\nkey constraint fails (`test`.`child`, CONSTRAINT `child_ibfk_1`\nFOREIGN KEY (`parent_id`) REFERENCES `parent` (`id`))\n\nThis is the same as a DELETE statement with no WHERE clause.\n\nBeginning with MySQL 5.1.32, TRUNCATE is treated for purposes of binary\nlogging and replication as DROP TABLE followed by CREATE TABLE --- that\nis, as DDL rather than DML. This is due to the fact that, when using\nInnoDB and other transactional storage engines where the transaction\nisolation level does not allow for statement-based logging (READ\nCOMMITTED or READ UNCOMMITTED), the statement was not logged and\nreplicated when using STATEMENT or MIXED logging mode. (Bug#36763\n(http://bugs.mysql.com/36763)) However, it is still applied on\nreplication slaves using InnoDB in the manner described previously.\n\nThe count of rows affected by TRUNCATE TABLE is accurate only when it\nis mapped to a DELETE statement.\n\nFor other storage engines, TRUNCATE TABLE differs from DELETE in the\nfollowing ways in MySQL 5.1:\n\no Truncate operations drop and re-create the table, which is much\n faster than deleting rows one by one, particularly for large tables.\n\no Truncate operations cause an implicit commit.\n\no Truncation operations cannot be performed if the session holds an\n active table lock.\n\no Truncation operations do not return a meaningful value for the number\n of deleted rows. The usual result is "0 rows affected," which should\n be interpreted as "no information."\n\no As long as the table format file tbl_name.frm is valid, the table can\n be re-created as an empty table with TRUNCATE TABLE, even if the data\n or index files have become corrupted.\n\no The table handler does not remember the last used AUTO_INCREMENT\n value, but starts counting from the beginning. This is true even for\n MyISAM and InnoDB, which normally do not reuse sequence values.\n\no When used with partitioned tables, TRUNCATE TABLE preserves the\n partitioning; that is, the data and index files are dropped and\n re-created, while the partition definitions (.par) file is\n unaffected.\n\no Since truncation of a table does not make any use of DELETE, the\n TRUNCATE statement does not invoke ON DELETE triggers.\n\nTRUNCATE TABLE requires the DROP privilege as of MySQL 5.1.16. (Before\n5.1.16, it requires the DELETE privilege.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/truncate.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/truncate.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (318,26,'TRUNCATE TABLE','Syntax:\nTRUNCATE [TABLE] tbl_name\n\nTRUNCATE TABLE empties a table completely. It requires the DROP\nprivilege as of MySQL 5.1.16. (Before 5.1.16, it requires the DELETE\nprivilege.\n\nLogically, TRUNCATE TABLE is equivalent to a DELETE statement that\ndeletes all rows, but there are practical differences under some\ncircumstances.\n\nFor an InnoDB table, InnoDB processes TRUNCATE TABLE by deleting rows\none by one if there are any FOREIGN KEY constraints that reference the\ntable. If there are no FOREIGN KEY constraints, InnoDB performs fast\ntruncation by dropping the original table and creating an empty one\nwith the same definition, which is much faster than deleting rows one\nby one. The AUTO_INCREMENT counter is reset by TRUNCATE TABLE,\nregardless of whether there is a FOREIGN KEY constraint.\n\nIn the case that FOREIGN KEY constraints reference the table, InnoDB\ndeletes rows one by one and processes the constraints on each one. If\nthe FOREIGN KEY constraint specifies DELETE CASCADE, rows from the\nchild (referenced) table are deleted, and the truncated table becomes\nempty. If the FOREIGN KEY constraint does not specify CASCADE, the\nTRUNCATE statement deletes rows one by one and stops if it encounters a\nparent row that is referenced by the child, returning this error:\n\nERROR 1451 (23000): Cannot delete or update a parent row: a foreign\nkey constraint fails (`test`.`child`, CONSTRAINT `child_ibfk_1`\nFOREIGN KEY (`parent_id`) REFERENCES `parent` (`id`))\n\nThis is the same as a DELETE statement with no WHERE clause.\n\nThe count of rows affected by TRUNCATE TABLE is accurate only when it\nis mapped to a DELETE statement.\n\nFor other storage engines, TRUNCATE TABLE differs from DELETE in the\nfollowing ways in MySQL 5.1:\n\no Truncate operations drop and re-create the table, which is much\n faster than deleting rows one by one, particularly for large tables.\n\no Truncate operations cause an implicit commit.\n\no Truncation operations cannot be performed if the session holds an\n active table lock.\n\no Truncation operations do not return a meaningful value for the number\n of deleted rows. The usual result is "0 rows affected," which should\n be interpreted as "no information."\n\no As long as the table format file tbl_name.frm is valid, the table can\n be re-created as an empty table with TRUNCATE TABLE, even if the data\n or index files have become corrupted.\n\no The table handler does not remember the last used AUTO_INCREMENT\n value, but starts counting from the beginning. This is true even for\n MyISAM and InnoDB, which normally do not reuse sequence values.\n\no When used with partitioned tables, TRUNCATE TABLE preserves the\n partitioning; that is, the data and index files are dropped and\n re-created, while the partition definitions (.par) file is\n unaffected.\n\no Since truncation of a table does not make any use of DELETE, the\n TRUNCATE statement does not invoke ON DELETE triggers.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/truncate.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/truncate.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (319,16,'BIT_XOR','Syntax:\nBIT_XOR(expr)\n\nReturns the bitwise XOR of all bits in expr. The calculation is\nperformed with 64-bit (BIGINT) precision.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/group-by-functions.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/group-by-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (320,30,'CURRENT_DATE','Syntax:\nCURRENT_DATE, CURRENT_DATE()\n\nCURRENT_DATE and CURRENT_DATE() are synonyms for CURDATE().\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (321,25,'START SLAVE','Syntax:\nSTART SLAVE [thread_type [, thread_type] ... ]\nSTART SLAVE [SQL_THREAD] UNTIL\n MASTER_LOG_FILE = \'log_name\', MASTER_LOG_POS = log_pos\nSTART SLAVE [SQL_THREAD] UNTIL\n RELAY_LOG_FILE = \'log_name\', RELAY_LOG_POS = log_pos\n\nthread_type: IO_THREAD | SQL_THREAD\n\nSTART SLAVE with no thread_type options starts both of the slave\nthreads. The I/O thread reads queries from the master server and stores\nthem in the relay log. The SQL thread reads the relay log and executes\nthe queries. START SLAVE requires the SUPER privilege.\n\nIf START SLAVE succeeds in starting the slave threads, it returns\nwithout any error. However, even in that case, it might be that the\nslave threads start and then later stop (for example, because they do\nnot manage to connect to the master or read its binary logs, or some\nother problem). START SLAVE does not warn you about this. You must\ncheck the slave\'s error log for error messages generated by the slave\nthreads, or check that they are running satisfactorily with SHOW SLAVE\nSTATUS.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/start-slave.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/start-slave.html'); @@ -410,7 +410,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (341,22,'LOOP','Syntax:\n[begin_label:] LOOP\n statement_list\nEND LOOP [end_label]\n\nLOOP implements a simple loop construct, enabling repeated execution of\nthe statement list, which consists of one or more statements, each\nterminated by a semicolon (;) statement delimiter. The statements\nwithin the loop are repeated until the loop is exited; usually this is\naccomplished with a LEAVE statement.\n\nA LOOP statement can be labeled. end_label cannot be given unless\nbegin_label also is present. If both are present, they must be the\nsame.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/loop-statement.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/loop-statement.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (342,4,'TRUNCATE','Syntax:\nTRUNCATE(X,D)\n\nReturns the number X, truncated to D decimal places. If D is 0, the\nresult has no decimal point or fractional part. D can be negative to\ncause D digits left of the decimal point of the value X to become zero.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/mathematical-functions.html\n\n','mysql> SELECT TRUNCATE(1.223,1);\n -> 1.2\nmysql> SELECT TRUNCATE(1.999,1);\n -> 1.9\nmysql> SELECT TRUNCATE(1.999,0);\n -> 1\nmysql> SELECT TRUNCATE(-1.999,1);\n -> -1.9\nmysql> SELECT TRUNCATE(122,-2);\n -> 100\nmysql> SELECT TRUNCATE(10.28*100,0);\n -> 1028\n','http://dev.mysql.com/doc/refman/5.1/en/mathematical-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (343,30,'TIMESTAMPADD','Syntax:\nTIMESTAMPADD(unit,interval,datetime_expr)\n\nAdds the integer expression interval to the date or datetime expression\ndatetime_expr. The unit for interval is given by the unit argument,\nwhich should be one of the following values: FRAC_SECOND\n(microseconds), SECOND, MINUTE, HOUR, DAY, WEEK, MONTH, QUARTER, or\nYEAR.\n\nBeginning with MySQL 5.1.24, it is possible to use MICROSECOND in place\nof FRAC_SECOND with this function, and FRAC_SECOND is deprecated.\n\nThe unit value may be specified using one of keywords as shown, or with\na prefix of SQL_TSI_. For example, DAY and SQL_TSI_DAY both are legal.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','mysql> SELECT TIMESTAMPADD(MINUTE,1,\'2003-01-02\');\n -> \'2003-01-02 00:01:00\'\nmysql> SELECT TIMESTAMPADD(WEEK,1,\'2003-01-02\');\n -> \'2003-01-09\'\n','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (344,25,'SHOW','SHOW has many forms that provide information about databases, tables,\ncolumns, or status information about the server. This section describes\nthose following:\n\nSHOW AUTHORS\nSHOW CHARACTER SET [like_or_where]\nSHOW COLLATION [like_or_where]\nSHOW [FULL] COLUMNS FROM tbl_name [FROM db_name] [like_or_where]\nSHOW CONTRIBUTORS\nSHOW CREATE DATABASE db_name\nSHOW CREATE EVENT event_name\nSHOW CREATE FUNCTION func_name\nSHOW CREATE PROCEDURE proc_name\nSHOW CREATE TABLE tbl_name\nSHOW CREATE TRIGGER trigger_name\nSHOW CREATE VIEW view_name\nSHOW DATABASES [like_or_where]\nSHOW ENGINE engine_name {STATUS | MUTEX}\nSHOW [STORAGE] ENGINES\nSHOW ERRORS [LIMIT [offset,] row_count]\nSHOW [FULL] EVENTS\nSHOW FUNCTION CODE func_name\nSHOW FUNCTION STATUS [like_or_where]\nSHOW GRANTS FOR user\nSHOW INDEX FROM tbl_name [FROM db_name]\nSHOW INNODB STATUS\nSHOW OPEN TABLES [FROM db_name] [like_or_where]\nSHOW PLUGINS\nSHOW PROCEDURE CODE proc_name\nSHOW PROCEDURE STATUS [like_or_where]\nSHOW PRIVILEGES\nSHOW [FULL] PROCESSLIST\nSHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n]\nSHOW PROFILES\nSHOW SCHEDULER STATUS\nSHOW [GLOBAL | SESSION] STATUS [like_or_where]\nSHOW TABLE STATUS [FROM db_name] [like_or_where]\nSHOW TABLES [FROM db_name] [like_or_where]\nSHOW TRIGGERS [FROM db_name] [like_or_where]\nSHOW [GLOBAL | SESSION] VARIABLES [like_or_where]\nSHOW WARNINGS [LIMIT [offset,] row_count]\n\nlike_or_where:\n LIKE \'pattern\'\n | WHERE expr\n\nIf the syntax for a given SHOW statement includes a LIKE \'pattern\'\npart, \'pattern\' is a string that can contain the SQL "%" and "_"\nwildcard characters. The pattern is useful for restricting statement\noutput to matching values.\n\nSeveral SHOW statements also accept a WHERE clause that provides more\nflexibility in specifying which rows to display. See\nhttp://dev.mysql.com/doc/refman/5.1/en/extended-show.html.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/show.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/show.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (344,25,'SHOW','SHOW has many forms that provide information about databases, tables,\ncolumns, or status information about the server. This section describes\nthose following:\n\nSHOW AUTHORS\nSHOW CHARACTER SET [like_or_where]\nSHOW COLLATION [like_or_where]\nSHOW [FULL] COLUMNS FROM tbl_name [FROM db_name] [like_or_where]\nSHOW CONTRIBUTORS\nSHOW CREATE DATABASE db_name\nSHOW CREATE EVENT event_name\nSHOW CREATE FUNCTION func_name\nSHOW CREATE PROCEDURE proc_name\nSHOW CREATE TABLE tbl_name\nSHOW CREATE TRIGGER trigger_name\nSHOW CREATE VIEW view_name\nSHOW DATABASES [like_or_where]\nSHOW ENGINE engine_name {STATUS | MUTEX}\nSHOW [STORAGE] ENGINES\nSHOW ERRORS [LIMIT [offset,] row_count]\nSHOW EVENTS\nSHOW FUNCTION CODE func_name\nSHOW FUNCTION STATUS [like_or_where]\nSHOW GRANTS FOR user\nSHOW INDEX FROM tbl_name [FROM db_name]\nSHOW INNODB STATUS\nSHOW OPEN TABLES [FROM db_name] [like_or_where]\nSHOW PLUGINS\nSHOW PROCEDURE CODE proc_name\nSHOW PROCEDURE STATUS [like_or_where]\nSHOW PRIVILEGES\nSHOW [FULL] PROCESSLIST\nSHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n]\nSHOW PROFILES\nSHOW SCHEDULER STATUS\nSHOW [GLOBAL | SESSION] STATUS [like_or_where]\nSHOW TABLE STATUS [FROM db_name] [like_or_where]\nSHOW TABLES [FROM db_name] [like_or_where]\nSHOW TRIGGERS [FROM db_name] [like_or_where]\nSHOW [GLOBAL | SESSION] VARIABLES [like_or_where]\nSHOW WARNINGS [LIMIT [offset,] row_count]\n\nlike_or_where:\n LIKE \'pattern\'\n | WHERE expr\n\nIf the syntax for a given SHOW statement includes a LIKE \'pattern\'\npart, \'pattern\' is a string that can contain the SQL "%" and "_"\nwildcard characters. The pattern is useful for restricting statement\noutput to matching values.\n\nSeveral SHOW statements also accept a WHERE clause that provides more\nflexibility in specifying which rows to display. See\nhttp://dev.mysql.com/doc/refman/5.1/en/extended-show.html.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/show.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/show.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (345,17,'GREATEST','Syntax:\nGREATEST(value1,value2,...)\n\nWith two or more arguments, returns the largest (maximum-valued)\nargument. The arguments are compared using the same rules as for\nLEAST().\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/comparison-operators.html\n\n','mysql> SELECT GREATEST(2,0);\n -> 2\nmysql> SELECT GREATEST(34.0,3.0,5.0,767.0);\n -> 767.0\nmysql> SELECT GREATEST(\'B\',\'A\',\'C\');\n -> \'C\'\n','http://dev.mysql.com/doc/refman/5.1/en/comparison-operators.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (346,25,'SHOW VARIABLES','Syntax:\nSHOW [GLOBAL | SESSION] VARIABLES\n [LIKE \'pattern\' | WHERE expr]\n\nSHOW VARIABLES shows the values of MySQL system variables. This\ninformation also can be obtained using the mysqladmin variables\ncommand. The LIKE clause, if present, indicates which variable names to\nmatch. The WHERE clause can be given to select rows using more general\nconditions, as discussed in\nhttp://dev.mysql.com/doc/refman/5.1/en/extended-show.html. This\nstatement does not require any privilege. It requires only the ability\nto connect to the server.\n\nWith the GLOBAL modifier, SHOW VARIABLES displays the values that are\nused for new connections to MySQL. With SESSION, it displays the values\nthat are in effect for the current connection. If no modifier is\npresent, the default is SESSION. LOCAL is a synonym for SESSION.\nWith a LIKE clause, the statement displays only rows for those\nvariables with names that match the pattern. To obtain the row for a\nspecific variable, use a LIKE clause as shown:\n\nSHOW VARIABLES LIKE \'max_join_size\';\nSHOW SESSION VARIABLES LIKE \'max_join_size\';\n\nTo get a list of variables whose name match a pattern, use the "%"\nwildcard character in a LIKE clause:\n\nSHOW VARIABLES LIKE \'%size%\';\nSHOW GLOBAL VARIABLES LIKE \'%size%\';\n\nWildcard characters can be used in any position within the pattern to\nbe matched. Strictly speaking, because "_" is a wildcard that matches\nany single character, you should escape it as "\\_" to match it\nliterally. In practice, this is rarely necessary.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/show-variables.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/show-variables.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (347,25,'BINLOG','Syntax:\nBINLOG \'str\'\n\nBINLOG is an internal-use statement. It is generated by the mysqlbinlog\nprogram as the printable representation of certain events in binary log\nfiles. (See http://dev.mysql.com/doc/refman/5.1/en/mysqlbinlog.html.)\nThe \'str\' value is a base 64-encoded string the that server decodes to\ndetermine the data change indicated by the corresponding event. This\nstatement requires the SUPER privilege. It was added in MySQL 5.1.5.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/binlog.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/binlog.html'); @@ -419,7 +419,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (350,4,'ATAN2','Syntax:\nATAN(Y,X), ATAN2(Y,X)\n\nReturns the arc tangent of the two variables X and Y. It is similar to\ncalculating the arc tangent of Y / X, except that the signs of both\narguments are used to determine the quadrant of the result.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/mathematical-functions.html\n\n','mysql> SELECT ATAN(-2,2);\n -> -0.78539816339745\nmysql> SELECT ATAN2(PI(),0);\n -> 1.5707963267949\n','http://dev.mysql.com/doc/refman/5.1/en/mathematical-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (351,6,'MBRCONTAINS','MBRContains(g1,g2)\n\nReturns 1 or 0 to indicate whether the Minimum Bounding Rectangle of g1\ncontains the Minimum Bounding Rectangle of g2. This tests the opposite\nrelationship as MBRWithin().\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/relations-on-geometry-mbr.html\n\n','mysql> SET @g1 = GeomFromText(\'Polygon((0 0,0 3,3 3,3 0,0 0))\');\nmysql> SET @g2 = GeomFromText(\'Point(1 1)\');\nmysql> SELECT MBRContains(@g1,@g2), MBRContains(@g2,@g1);\n----------------------+----------------------+\n| MBRContains(@g1,@g2) | MBRContains(@g2,@g1) |\n+----------------------+----------------------+\n| 1 | 0 |\n+----------------------+----------------------+\n','http://dev.mysql.com/doc/refman/5.1/en/relations-on-geometry-mbr.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (352,30,'HOUR','Syntax:\nHOUR(time)\n\nReturns the hour for time. The range of the return value is 0 to 23 for\ntime-of-day values. However, the range of TIME values actually is much\nlarger, so HOUR can return values greater than 23.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','mysql> SELECT HOUR(\'10:05:03\');\n -> 10\nmysql> SELECT HOUR(\'272:59:59\');\n -> 272\n','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (353,26,'SELECT','Syntax:\nSELECT\n [ALL | DISTINCT | DISTINCTROW ]\n [HIGH_PRIORITY]\n [STRAIGHT_JOIN]\n [SQL_SMALL_RESULT] [SQL_BIG_RESULT] [SQL_BUFFER_RESULT]\n [SQL_CACHE | SQL_NO_CACHE] [SQL_CALC_FOUND_ROWS]\n select_expr [, select_expr ...]\n [FROM table_references\n [WHERE where_condition]\n [GROUP BY {col_name | expr | position}\n [ASC | DESC], ... [WITH ROLLUP]]\n [HAVING where_condition]\n [ORDER BY {col_name | expr | position}\n [ASC | DESC], ...]\n [LIMIT {[offset,] row_count | row_count OFFSET offset}]\n [PROCEDURE procedure_name(argument_list)]\n [INTO OUTFILE \'file_name\' export_options\n | INTO DUMPFILE \'file_name\'\n | INTO var_name [, var_name]]\n [FOR UPDATE | LOCK IN SHARE MODE]]\n\nSELECT is used to retrieve rows selected from one or more tables, and\ncan include UNION statements and subqueries. See [HELP UNION], and\nhttp://dev.mysql.com/doc/refman/5.1/en/subqueries.html.\n\nThe most commonly used clauses of SELECT statements are these:\n\no Each select_expr indicates a column that you want to retrieve. There\n must be at least one select_expr.\n\no table_references indicates the table or tables from which to retrieve\n rows. Its syntax is described in [HELP JOIN].\n\no The WHERE clause, if given, indicates the condition or conditions\n that rows must satisfy to be selected. where_condition is an\n expression that evaluates to true for each row to be selected. The\n statement selects all rows if there is no WHERE clause.\n\n In the WHERE clause, you can use any of the functions and operators\n that MySQL supports, except for aggregate (summary) functions. See\n http://dev.mysql.com/doc/refman/5.1/en/functions.html.\n\nSELECT can also be used to retrieve rows computed without reference to\nany table.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/select.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/select.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (353,26,'SELECT','Syntax:\nSELECT\n [ALL | DISTINCT | DISTINCTROW ]\n [HIGH_PRIORITY]\n [STRAIGHT_JOIN]\n [SQL_SMALL_RESULT] [SQL_BIG_RESULT] [SQL_BUFFER_RESULT]\n [SQL_CACHE | SQL_NO_CACHE] [SQL_CALC_FOUND_ROWS]\n select_expr [, select_expr ...]\n [FROM table_references\n [WHERE where_condition]\n [GROUP BY {col_name | expr | position}\n [ASC | DESC], ... [WITH ROLLUP]]\n [HAVING where_condition]\n [ORDER BY {col_name | expr | position}\n [ASC | DESC], ...]\n [LIMIT {[offset,] row_count | row_count OFFSET offset}]\n [PROCEDURE procedure_name(argument_list)]\n [INTO OUTFILE \'file_name\'\n [CHARACTER SET charset_name]\n export_options\n | INTO DUMPFILE \'file_name\'\n | INTO var_name [, var_name]]\n [FOR UPDATE | LOCK IN SHARE MODE]]\n\nSELECT is used to retrieve rows selected from one or more tables, and\ncan include UNION statements and subqueries. See [HELP UNION], and\nhttp://dev.mysql.com/doc/refman/5.1/en/subqueries.html.\n\nThe most commonly used clauses of SELECT statements are these:\n\no Each select_expr indicates a column that you want to retrieve. There\n must be at least one select_expr.\n\no table_references indicates the table or tables from which to retrieve\n rows. Its syntax is described in [HELP JOIN].\n\no The WHERE clause, if given, indicates the condition or conditions\n that rows must satisfy to be selected. where_condition is an\n expression that evaluates to true for each row to be selected. The\n statement selects all rows if there is no WHERE clause.\n\n In the WHERE clause, you can use any of the functions and operators\n that MySQL supports, except for aggregate (summary) functions. See\n http://dev.mysql.com/doc/refman/5.1/en/functions.html.\n\nSELECT can also be used to retrieve rows computed without reference to\nany table.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/select.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/select.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (354,4,'COT','Syntax:\nCOT(X)\n\nReturns the cotangent of X.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/mathematical-functions.html\n\n','mysql> SELECT COT(12);\n -> -1.5726734063977\nmysql> SELECT COT(0);\n -> NULL\n','http://dev.mysql.com/doc/refman/5.1/en/mathematical-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (355,25,'SHOW CREATE EVENT','Syntax:\nSHOW CREATE EVENT event_name\n\nThis statement displays the CREATE EVENT statement needed to re-create\na given event. For example (using the same event e_daily defined and\nthen altered in [HELP SHOW EVENTS]):\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/show-create-event.html\n\n','mysql> SHOW CREATE EVENT test.e_daily\\G\n*************************** 1. row ***************************\n Event: e_daily\n sql_mode:\n time_zone: SYSTEM\n Create Event: CREATE EVENT `e_daily`\n ON SCHEDULE EVERY 1 DAY\n STARTS CURRENT_TIMESTAMP + INTERVAL 6 HOUR\n ON COMPLETION NOT PRESERVE\n ENABLE\n COMMENT \'Saves total number of sessions then\n clears the table each day\'\n DO BEGIN\n INSERT INTO site_activity.totals (time, total)\n SELECT CURRENT_TIMESTAMP, COUNT(*)\n FROM site_activity.sessions;\n DELETE FROM site_activity.sessions;\n END\ncharacter_set_client: latin1\ncollation_connection: latin1_swedish_ci\n Database Collation: latin1_swedish_ci\n','http://dev.mysql.com/doc/refman/5.1/en/show-create-event.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (356,19,'BACKUP TABLE','Syntax:\nBACKUP TABLE tbl_name [, tbl_name] ... TO \'/path/to/backup/directory\'\n\n*Note*: This statement is deprecated. We are working on a better\nreplacement for it that will provide online backup capabilities. In the\nmeantime, the mysqlhotcopy script can be used instead.\n\nBACKUP TABLE copies to the backup directory the minimum number of table\nfiles needed to restore the table, after flushing any buffered changes\nto disk. The statement works only for MyISAM tables. It copies the .frm\ndefinition and .MYD data files. The .MYI index file can be rebuilt from\nthose two files. The directory should be specified as a full path name.\nTo restore the table, use RESTORE TABLE.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/backup-table.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/backup-table.html'); @@ -486,7 +486,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (417,35,'BIN','Syntax:\nBIN(N)\n\nReturns a string representation of the binary value of N, where N is a\nlonglong (BIGINT) number. This is equivalent to CONV(N,10,2). Returns\nNULL if N is NULL.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/string-functions.html\n\n','mysql> SELECT BIN(12);\n -> \'1100\'\n','http://dev.mysql.com/doc/refman/5.1/en/string-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (418,5,'INSTALL PLUGIN','Syntax:\nINSTALL PLUGIN plugin_name SONAME \'plugin_library\'\n\nThis statement installs a plugin.\n\nplugin_name is the name of the plugin as defined in the plugin\ndeclaration structure contained in the library file. Plugin names are\nnot case sensitive. For maximal compatibility, plugin names should be\nlimited to ASCII letters, digits, and underscore, because they are used\nin C source files, shell command lines, M4 and Bourne shell scripts,\nand SQL environments.\n\nplugin_library is the name of the shared library that contains the\nplugin code. The name includes the file name extension (for example,\nlibmyplugin.so or libmyplugin.dylib).\n\nThe shared library must be located in the plugin directory (that is,\nthe directory named by the plugin_dir system variable). The library\nmust be in the plugin directory itself, not in a subdirectory. By\ndefault, plugin_dir is plugin directory under the directory named by\nthe pkglibdir configuration variable, but it can be changed by setting\nthe value of plugin_dir at server startup. For example, set its value\nin a my.cnf file:\n\n[mysqld]\nplugin_dir=/path/to/plugin/directory\n\nIf the value of plugin_dir is a relative path name, it is taken to be\nrelative to the MySQL base directory (the value of the basedir system\nvariable).\n\nINSTALL PLUGIN adds a line to the mysql.plugin table that describes the\nplugin. This table contains the plugin name and library file name.\n\nAs of MySQL 5.1.33, INSTALL PLUGIN causes the server to read option\n(my.cnf) files just as during server startup. This enables the plugin\nto pick up any relevant options from those files. It is possible to add\nplugin options to an option file even before loading a plugin (if the\nloose prefix is used). It is also possible to uninstall a plugin, edit\nmy.cnf, and install the plugin again. Restarting the plugin this way\nenables it to the new option values without a server restart.\n\nBefore MySQL 5.1.33, a plugin is started with each option set to its\ndefault value.\n\nINSTALL PLUGIN also loads and initializes the plugin code to make the\nplugin available for use. A plugin is initialized by executing its\ninitialization function, which handles any setup that the plugin must\nperform before it can be used.\n\nTo use INSTALL PLUGIN, you must have the INSERT privilege for the\nmysql.plugin table.\n\nAt server startup, the server loads and initializes any plugin that is\nlisted in the mysql.plugin table. This means that a plugin is installed\nwith INSTALL PLUGIN only once, not every time the server starts. Plugin\nloading at startup does not occur if the server is started with the\n--skip-grant-tables option.\n\nWhen the server shuts down, it executes the deinitialization function\nfor each plugin that is loaded so that the plugin has a change to\nperform any final cleanup.\n\nFor options that control individual plugin loading at server startup,\nsee http://dev.mysql.com/doc/refman/5.1/en/server-plugin-options.html.\nIf you need to load plugins for a single server startup when the\n--skip-grant-tables option is given (which tells the server not to read\nsystem tables), use the --plugin-load option. See\nhttp://dev.mysql.com/doc/refman/5.1/en/server-options.html.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/install-plugin.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/install-plugin.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (419,22,'DECLARE CURSOR','Syntax:\nDECLARE cursor_name CURSOR FOR select_statement\n\nThis statement declares a cursor. Multiple cursors may be declared in a\nstored program, but each cursor in a given block must have a unique\nname.\n\nThe SELECT statement cannot have an INTO clause.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/declare-cursor.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/declare-cursor.html'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (420,26,'LOAD DATA','Syntax:\nLOAD DATA [LOW_PRIORITY | CONCURRENT] [LOCAL] INFILE \'file_name\'\n [REPLACE | IGNORE]\n INTO TABLE tbl_name\n [CHARACTER SET charset_name]\n [{FIELDS | COLUMNS}\n [TERMINATED BY \'string\']\n [[OPTIONALLY] ENCLOSED BY \'char\']\n [ESCAPED BY \'char\']\n ]\n [LINES\n [STARTING BY \'string\']\n [TERMINATED BY \'string\']\n ]\n [IGNORE number LINES]\n [(col_name_or_user_var,...)]\n [SET col_name = expr,...]\n\nThe LOAD DATA INFILE statement reads rows from a text file into a table\nat a very high speed. The file name must be given as a literal string.\n\nLOAD DATA INFILE is the complement of SELECT ... INTO OUTFILE. (See\nhttp://dev.mysql.com/doc/refman/5.1/en/select.html.) To write data from\na table to a file, use SELECT ... INTO OUTFILE. To read the file back\ninto a table, use LOAD DATA INFILE. The syntax of the FIELDS and LINES\nclauses is the same for both statements. Both clauses are optional, but\nFIELDS must precede LINES if both are specified.\n\nFor more information about the efficiency of INSERT versus LOAD DATA\nINFILE and speeding up LOAD DATA INFILE, see\nhttp://dev.mysql.com/doc/refman/5.1/en/insert-speed.html.\n\nThe character set indicated by the character_set_database system\nvariable is used to interpret the information in the file. SET NAMES\nand the setting of character_set_client do not affect interpretation of\ninput. If the contents of the input file use a character set that\ndiffers from the default, it is usually preferable to specify the\ncharacter set of the file by using the CHARACTER SET clause, which is\navailable as of MySQL 5.1.17. A character set of binary specifies "no\nconversion."\n\nLOAD DATA INFILE interprets all fields in the file as having the same\ncharacter set, regardless of the data types of the columns into which\nfield values are loaded. For proper interpretation of file contents,\nyou must ensure that it was written with the correct character set. For\nexample, if you write a data file with mysqldump -T or by issuing a\nSELECT ... INTO OUTFILE statement in mysql, be sure to use a\n--default-character-set option with mysqldump or mysql so that output\nis written in the character set to be used when the file is loaded with\nLOAD DATA INFILE.\n\nNote that it is currently not possible to load data files that use the\nucs2, utf16, or utf32 character set.\n\nAs of MySQL 5.1.6, the character_set_filesystem system variable\ncontrols the interpretation of the file name.\n\nYou can also load data files by using the mysqlimport utility; it\noperates by sending a LOAD DATA INFILE statement to the server. The\n--local option causes mysqlimport to read data files from the client\nhost. You can specify the --compress option to get better performance\nover slow networks if the client and server support the compressed\nprotocol. See http://dev.mysql.com/doc/refman/5.1/en/mysqlimport.html.\n\nIf you use LOW_PRIORITY, execution of the LOAD DATA statement is\ndelayed until no other clients are reading from the table. This affects\nonly storage engines that use only table-level locking (MyISAM, MEMORY,\nMERGE).\n\nIf you specify CONCURRENT with a MyISAM table that satisfies the\ncondition for concurrent inserts (that is, it contains no free blocks\nin the middle), other threads can retrieve data from the table while\nLOAD DATA is executing. Using this option affects the performance of\nLOAD DATA a bit, even if no other thread is using the table at the same\ntime.\n\nCONCURRENT is not replicated when using statement-based replication;\nhowever, it is replicated when using row-based replication. See\nhttp://dev.mysql.com/doc/refman/5.1/en/replication-features-load-data.…, for more information.\n\n*Note*: Prior to MySQL 5.1.23, LOAD DATA performed very poorly when\nimporting into partitioned tables. The statement now uses buffering to\nimprove performance; however, the buffer uses 130 KB memory per\npartition to achieve this. (Bug#26527 (http://bugs.mysql.com/26527))\n\nThe LOCAL keyword, if specified, is interpreted with respect to the\nclient end of the connection:\n\no If LOCAL is specified, the file is read by the client program on the\n client host and sent to the server. The file can be given as a full\n path name to specify its exact location. If given as a relative path\n name, the name is interpreted relative to the directory in which the\n client program was started.\n\no If LOCAL is not specified, the file must be located on the server\n host and is read directly by the server. The server uses the\n following rules to locate the file:\n\n o If the file name is an absolute path name, the server uses it as\n given.\n\n o If the file name is a relative path name with one or more leading\n components, the server searches for the file relative to the\n server\'s data directory.\n\n o If a file name with no leading components is given, the server\n looks for the file in the database directory of the default\n database.\n\nNote that, in the non-LOCAL case, these rules mean that a file named as\n./myfile.txt is read from the server\'s data directory, whereas the file\nnamed as myfile.txt is read from the database directory of the default\ndatabase. For example, if db1 is the default database, the following\nLOAD DATA statement reads the file data.txt from the database directory\nfor db1, even though the statement explicitly loads the file into a\ntable in the db2 database:\n\nLOAD DATA INFILE \'data.txt\' INTO TABLE db2.my_table;\n\nWindows path names are specified using forward slashes rather than\nbackslashes. If you do use backslashes, you must double them.\n\nFor security reasons, when reading text files located on the server,\nthe files must either reside in the database directory or be readable\nby all. Also, to use LOAD DATA INFILE on server files, you must have\nthe FILE privilege. See\nhttp://dev.mysql.com/doc/refman/5.1/en/privileges-provided.html. For\nnon-LOCAL load operations, if the secure_file_priv system variable is\nset to a nonempty directory name, the file to be loaded must be located\nin that directory.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/load-data.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/load-data.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (420,26,'LOAD DATA','Syntax:\nLOAD DATA [LOW_PRIORITY | CONCURRENT] [LOCAL] INFILE \'file_name\'\n [REPLACE | IGNORE]\n INTO TABLE tbl_name\n [CHARACTER SET charset_name]\n [{FIELDS | COLUMNS}\n [TERMINATED BY \'string\']\n [[OPTIONALLY] ENCLOSED BY \'char\']\n [ESCAPED BY \'char\']\n ]\n [LINES\n [STARTING BY \'string\']\n [TERMINATED BY \'string\']\n ]\n [IGNORE number LINES]\n [(col_name_or_user_var,...)]\n [SET col_name = expr,...]\n\nThe LOAD DATA INFILE statement reads rows from a text file into a table\nat a very high speed. The file name must be given as a literal string.\n\nLOAD DATA INFILE is the complement of SELECT ... INTO OUTFILE. (See\nhttp://dev.mysql.com/doc/refman/5.1/en/select.html.) To write data from\na table to a file, use SELECT ... INTO OUTFILE. To read the file back\ninto a table, use LOAD DATA INFILE. The syntax of the FIELDS and LINES\nclauses is the same for both statements. Both clauses are optional, but\nFIELDS must precede LINES if both are specified.\n\nFor more information about the efficiency of INSERT versus LOAD DATA\nINFILE and speeding up LOAD DATA INFILE, see\nhttp://dev.mysql.com/doc/refman/5.1/en/insert-speed.html.\n\nThe character set indicated by the character_set_database system\nvariable is used to interpret the information in the file. SET NAMES\nand the setting of character_set_client do not affect interpretation of\ninput. If the contents of the input file use a character set that\ndiffers from the default, it is usually preferable to specify the\ncharacter set of the file by using the CHARACTER SET clause, which is\navailable as of MySQL 5.1.17. A character set of binary specifies "no\nconversion."\n\nLOAD DATA INFILE interprets all fields in the file as having the same\ncharacter set, regardless of the data types of the columns into which\nfield values are loaded. For proper interpretation of file contents,\nyou must ensure that it was written with the correct character set. For\nexample, if you write a data file with mysqldump -T or by issuing a\nSELECT ... INTO OUTFILE statement in mysql, be sure to use a\n--default-character-set option with mysqldump or mysql so that output\nis written in the character set to be used when the file is loaded with\nLOAD DATA INFILE.\n\nNote that it is currently not possible to load data files that use the\nucs2, utf16, or utf32 character set.\n\nAs of MySQL 5.1.6, the character_set_filesystem system variable\ncontrols the interpretation of the file name.\n\nYou can also load data files by using the mysqlimport utility; it\noperates by sending a LOAD DATA INFILE statement to the server. The\n--local option causes mysqlimport to read data files from the client\nhost. You can specify the --compress option to get better performance\nover slow networks if the client and server support the compressed\nprotocol. See http://dev.mysql.com/doc/refman/5.1/en/mysqlimport.html.\n\nIf you use LOW_PRIORITY, execution of the LOAD DATA statement is\ndelayed until no other clients are reading from the table. This affects\nonly storage engines that use only table-level locking (MyISAM, MEMORY,\nMERGE).\n\nIf you specify CONCURRENT with a MyISAM table that satisfies the\ncondition for concurrent inserts (that is, it contains no free blocks\nin the middle), other threads can retrieve data from the table while\nLOAD DATA is executing. Using this option affects the performance of\nLOAD DATA a bit, even if no other thread is using the table at the same\ntime.\n\nCONCURRENT is not replicated when using statement-based replication;\nhowever, it is replicated when using row-based replication. See\nhttp://dev.mysql.com/doc/refman/5.1/en/replication-features-load-data.…, for more information.\n\n*Note*: Prior to MySQL 5.1.23, LOAD DATA performed very poorly when\nimporting into partitioned tables. The statement now uses buffering to\nimprove performance; however, the buffer uses 130 KB memory per\npartition to achieve this. (Bug#26527 (http://bugs.mysql.com/26527))\n\nThe LOCAL keyword, if specified, is interpreted with respect to the\nclient end of the connection:\n\no If LOCAL is specified, the file is read by the client program on the\n client host and sent to the server. The file can be given as a full\n path name to specify its exact location. If given as a relative path\n name, the name is interpreted relative to the directory in which the\n client program was started.\n\no If LOCAL is not specified, the file must be located on the server\n host and is read directly by the server. The server uses the\n following rules to locate the file:\n\n o If the file name is an absolute path name, the server uses it as\n given.\n\n o If the file name is a relative path name with one or more leading\n components, the server searches for the file relative to the\n server\'s data directory.\n\n o If a file name with no leading components is given, the server\n looks for the file in the database directory of the default\n database.\n\nNote that, in the non-LOCAL case, these rules mean that a file named as\n./myfile.txt is read from the server\'s data directory, whereas the file\nnamed as myfile.txt is read from the database directory of the default\ndatabase. For example, if db1 is the default database, the following\nLOAD DATA statement reads the file data.txt from the database directory\nfor db1, even though the statement explicitly loads the file into a\ntable in the db2 database:\n\nLOAD DATA INFILE \'data.txt\' INTO TABLE db2.my_table;\n\nWindows path names are specified using forward slashes rather than\nbackslashes. If you do use backslashes, you must double them.\n\n*Note*: A regression in MySQL 5.1.40 caused the database referenced in\na fully qualified table name to be ignored by LOAD DATA when using\nreplication with either STATEMENT or MIXED as the binary logging\nformat; this could lead to problems if the table was not in the current\ndatabase. As a workaround, you can specify the correct database with\nthe USE statement prior to executing LOAD DATA. If necessary, you can\nreset the current database with a second USE statement following the\nLOAD DATA statement. This issue was fixed in MySQL 5.1.41. (Bug#48297\n(http://bugs.mysql.com/48297))\n\nFor security reasons, when reading text files located on the server,\nthe files must either reside in the database directory or be readable\nby all. Also, to use LOAD DATA INFILE on server files, you must have\nthe FILE privilege. See\nhttp://dev.mysql.com/doc/refman/5.1/en/privileges-provided.html. For\nnon-LOCAL load operations, if the secure_file_priv system variable is\nset to a nonempty directory name, the file to be loaded must be located\nin that directory.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/load-data.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/load-data.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (421,23,'MULTILINESTRING','MultiLineString(ls1,ls2,...)\n\nConstructs a MultiLineString value using LineString or WKB LineString\narguments.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/creating-spatial-values.html#gis-mys…','','http://dev.mysql.com/doc/refman/5.1/en/creating-spatial-values.html#gis-mys…'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (422,30,'LOCALTIME','Syntax:\nLOCALTIME, LOCALTIME()\n\nLOCALTIME and LOCALTIME() are synonyms for NOW().\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (423,3,'MPOINTFROMTEXT','MPointFromText(wkt[,srid]), MultiPointFromText(wkt[,srid])\n\nConstructs a MULTIPOINT value using its WKT representation and SRID.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/creating-spatial-values.html#gis-wkt…','','http://dev.mysql.com/doc/refman/5.1/en/creating-spatial-values.html#gis-wkt…'); @@ -497,7 +497,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (428,20,'CHAR','[NATIONAL] CHAR[(M)] [CHARACTER SET charset_name] [COLLATE\ncollation_name]\n\nA fixed-length string that is always right-padded with spaces to the\nspecified length when stored. M represents the column length in\ncharacters. The range of M is 0 to 255. If M is omitted, the length is\n1.\n\n*Note*: Trailing spaces are removed when CHAR values are retrieved\nunless the PAD_CHAR_TO_FULL_LENGTH SQL mode is enabled.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/string-type-overview.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/string-type-overview.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (429,30,'UTC_DATE','Syntax:\nUTC_DATE, UTC_DATE()\n\nReturns the current UTC date as a value in \'YYYY-MM-DD\' or YYYYMMDD\nformat, depending on whether the function is used in a string or\nnumeric context.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','mysql> SELECT UTC_DATE(), UTC_DATE() + 0;\n -> \'2003-08-14\', 20030814\n','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (430,34,'DIMENSION','Dimension(g)\n\nReturns the inherent dimension of the geometry value g. The result can\nbe -1, 0, 1, or 2. The meaning of these values is given in\nhttp://dev.mysql.com/doc/refman/5.1/en/gis-class-geometry.html.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/geometry-property-functions.html#gen…','mysql> SELECT Dimension(GeomFromText(\'LineString(1 1,2 2)\'));\n+------------------------------------------------+\n| Dimension(GeomFromText(\'LineString(1 1,2 2)\')) |\n+------------------------------------------------+\n| 1 |\n+------------------------------------------------+\n','http://dev.mysql.com/doc/refman/5.1/en/geometry-property-functions.html#gen…'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (431,16,'COUNT DISTINCT','Syntax:\nCOUNT(DISTINCT expr,[expr...])\n\nReturns a count of the number of different non-NULL values.\n\nCOUNT(DISTINCT) returns 0 if there were no matching rows.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/group-by-functions.html\n\n','mysql> SELECT COUNT(DISTINCT results) FROM student;\n','http://dev.mysql.com/doc/refman/5.1/en/group-by-functions.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (431,16,'COUNT DISTINCT','Syntax:\nCOUNT(DISTINCT expr,[expr...])\n\nReturns a count of the number of rows with different non-NULL expr\nvalues.\n\nCOUNT(DISTINCT) returns 0 if there were no matching rows.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/group-by-functions.html\n\n','mysql> SELECT COUNT(DISTINCT results) FROM student;\n','http://dev.mysql.com/doc/refman/5.1/en/group-by-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (432,20,'BIT','BIT[(M)]\n\nA bit-field type. M indicates the number of bits per value, from 1 to\n64. The default is 1 if M is omitted.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/numeric-type-overview.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/numeric-type-overview.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (433,29,'EQUALS','Equals(g1,g2)\n\nReturns 1 or 0 to indicate whether g1 is spatially equal to g2.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/functions-that-test-spatial-relation…','','http://dev.mysql.com/doc/refman/5.1/en/functions-that-test-spatial-relation…'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (434,25,'SHOW CREATE VIEW','Syntax:\nSHOW CREATE VIEW view_name\n\nThis statement shows a CREATE VIEW statement that creates the given\nview.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/show-create-view.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/show-create-view.html'); @@ -532,7 +532,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (463,37,'MERGE','The MERGE storage engine, also known as the MRG_MyISAM engine, is a\ncollection of identical MyISAM tables that can be used as one.\n"Identical" means that all tables have identical column and index\ninformation. You cannot merge MyISAM tables in which the columns are\nlisted in a different order, do not have exactly the same columns, or\nhave the indexes in different order. However, any or all of the MyISAM\ntables can be compressed with myisampack. See\nhttp://dev.mysql.com/doc/refman/5.1/en/myisampack.html. Differences in\ntable options such as AVG_ROW_LENGTH, MAX_ROWS, or PACK_KEYS do not\nmatter.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/merge-storage-engine.html\n\n','mysql> CREATE TABLE t1 (\n -> a INT NOT NULL AUTO_INCREMENT PRIMARY KEY,\n -> message CHAR(20)) ENGINE=MyISAM;\nmysql> CREATE TABLE t2 (\n -> a INT NOT NULL AUTO_INCREMENT PRIMARY KEY,\n -> message CHAR(20)) ENGINE=MyISAM;\nmysql> INSERT INTO t1 (message) VALUES (\'Testing\'),(\'table\'),(\'t1\');\nmysql> INSERT INTO t2 (message) VALUES (\'Testing\'),(\'table\'),(\'t2\');\nmysql> CREATE TABLE total (\n -> a INT NOT NULL AUTO_INCREMENT,\n -> message CHAR(20), INDEX(a))\n -> ENGINE=MERGE UNION=(t1,t2) INSERT_METHOD=LAST;\n','http://dev.mysql.com/doc/refman/5.1/en/merge-storage-engine.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (464,37,'CREATE TABLE','Syntax:\nCREATE [TEMPORARY] TABLE [IF NOT EXISTS] tbl_name\n (create_definition,...)\n [table_options]\n [partition_options]\n\nOr:\n\nCREATE [TEMPORARY] TABLE [IF NOT EXISTS] tbl_name\n [(create_definition,...)]\n [table_options]\n [partition_options]\n select_statement\n\nOr:\n\nCREATE [TEMPORARY] TABLE [IF NOT EXISTS] tbl_name\n { LIKE old_tbl_name | (LIKE old_tbl_name) }\n\ncreate_definition:\n col_name column_definition\n | [CONSTRAINT [symbol]] PRIMARY KEY [index_type] (index_col_name,...)\n [index_option] ...\n | {INDEX|KEY} [index_name] [index_type] (index_col_name,...)\n [index_option] ...\n | [CONSTRAINT [symbol]] UNIQUE [INDEX|KEY]\n [index_name] [index_type] (index_col_name,...)\n [index_option] ...\n | {FULLTEXT|SPATIAL} [INDEX|KEY] [index_name] (index_col_name,...)\n [index_option] ...\n | [CONSTRAINT [symbol]] FOREIGN KEY\n [index_name] (index_col_name,...) reference_definition\n | CHECK (expr)\n\ncolumn_definition:\n data_type [NOT NULL | NULL] [DEFAULT default_value]\n [AUTO_INCREMENT] [UNIQUE [KEY] | [PRIMARY] KEY]\n [COMMENT \'string\']\n [COLUMN_FORMAT {FIXED|DYNAMIC|DEFAULT}]\n [STORAGE {DISK|MEMORY|DEFAULT}]\n [reference_definition]\n\ndata_type:\n BIT[(length)]\n | TINYINT[(length)] [UNSIGNED] [ZEROFILL]\n | SMALLINT[(length)] [UNSIGNED] [ZEROFILL]\n | MEDIUMINT[(length)] [UNSIGNED] [ZEROFILL]\n | INT[(length)] [UNSIGNED] [ZEROFILL]\n | INTEGER[(length)] [UNSIGNED] [ZEROFILL]\n | BIGINT[(length)] [UNSIGNED] [ZEROFILL]\n | REAL[(length,decimals)] [UNSIGNED] [ZEROFILL]\n | DOUBLE[(length,decimals)] [UNSIGNED] [ZEROFILL]\n | FLOAT[(length,decimals)] [UNSIGNED] [ZEROFILL]\n | DECIMAL[(length[,decimals])] [UNSIGNED] [ZEROFILL]\n | NUMERIC[(length[,decimals])] [UNSIGNED] [ZEROFILL]\n | DATE\n | TIME\n | TIMESTAMP\n | DATETIME\n | YEAR\n | CHAR[(length)]\n [CHARACTER SET charset_name] [COLLATE collation_name]\n | VARCHAR(length)\n [CHARACTER SET charset_name] [COLLATE collation_name]\n | BINARY[(length)]\n | VARBINARY(length)\n | TINYBLOB\n | BLOB\n | MEDIUMBLOB\n | LONGBLOB\n | TINYTEXT [BINARY]\n [CHARACTER SET charset_name] [COLLATE collation_name]\n | TEXT [BINARY]\n [CHARACTER SET charset_name] [COLLATE collation_name]\n | MEDIUMTEXT [BINARY]\n [CHARACTER SET charset_name] [COLLATE collation_name]\n | LONGTEXT [BINARY]\n [CHARACTER SET charset_name] [COLLATE collation_name]\n | ENUM(value1,value2,value3,...)\n [CHARACTER SET charset_name] [COLLATE collation_name]\n | SET(value1,value2,value3,...)\n [CHARACTER SET charset_name] [COLLATE collation_name]\n | spatial_type\n\nindex_col_name:\n col_name [(length)] [ASC | DESC]\n\nindex_type:\n USING {BTREE | HASH | RTREE}\n\nindex_option:\n KEY_BLOCK_SIZE [=] value\n | index_type\n | WITH PARSER parser_name\n\nreference_definition:\n REFERENCES tbl_name (index_col_name,...)\n [MATCH FULL | MATCH PARTIAL | MATCH SIMPLE]\n [ON DELETE reference_option]\n [ON UPDATE reference_option]\n\nreference_option:\n RESTRICT | CASCADE | SET NULL | NO ACTION\n\ntable_options:\n table_option [[,] table_option] ...\n\ntable_option:\n ENGINE [=] engine_name\n | AUTO_INCREMENT [=] value\n | AVG_ROW_LENGTH [=] value\n | [DEFAULT] CHARACTER SET [=] charset_name\n | CHECKSUM [=] {0 | 1}\n | [DEFAULT] COLLATE [=] collation_name\n | COMMENT [=] \'string\'\n | CONNECTION [=] \'connect_string\'\n | DATA DIRECTORY [=] \'absolute path to directory\'\n | DELAY_KEY_WRITE [=] {0 | 1}\n | INDEX DIRECTORY [=] \'absolute path to directory\'\n | INSERT_METHOD [=] { NO | FIRST | LAST }\n | KEY_BLOCK_SIZE [=] value\n | MAX_ROWS [=] value\n | MIN_ROWS [=] value\n | PACK_KEYS [=] {0 | 1 | DEFAULT}\n | PASSWORD [=] \'string\'\n | ROW_FORMAT [=] {DEFAULT|DYNAMIC|FIXED|COMPRESSED|REDUNDANT|COMPACT}\n | TABLESPACE tablespace_name [STORAGE {DISK|MEMORY|DEFAULT}]\n | UNION [=] (tbl_name[,tbl_name]...)\n\npartition_options:\n PARTITION BY\n { [LINEAR] HASH(expr)\n | [LINEAR] KEY(column_list)\n | RANGE(expr)\n | LIST(expr) }\n [PARTITIONS num]\n [SUBPARTITION BY\n { [LINEAR] HASH(expr)\n | [LINEAR] KEY(column_list) }\n [SUBPARTITIONS num]\n ]\n [(partition_definition [, partition_definition] ...)]\n\npartition_definition:\n PARTITION partition_name\n [VALUES {LESS THAN {(expr) | MAXVALUE} | IN (value_list)}]\n [[STORAGE] ENGINE [=] engine_name]\n [COMMENT [=] \'comment_text\' ]\n [DATA DIRECTORY [=] \'data_dir\']\n [INDEX DIRECTORY [=] \'index_dir\']\n [MAX_ROWS [=] max_number_of_rows]\n [MIN_ROWS [=] min_number_of_rows]\n [TABLESPACE [=] tablespace_name]\n [NODEGROUP [=] node_group_id]\n [(subpartition_definition [, subpartition_definition] ...)]\n\nsubpartition_definition:\n SUBPARTITION logical_name\n [[STORAGE] ENGINE [=] engine_name]\n [COMMENT [=] \'comment_text\' ]\n [DATA DIRECTORY [=] \'data_dir\']\n [INDEX DIRECTORY [=] \'index_dir\']\n [MAX_ROWS [=] max_number_of_rows]\n [MIN_ROWS [=] min_number_of_rows]\n [TABLESPACE [=] tablespace_name]\n [NODEGROUP [=] node_group_id]\n\nselect_statement:\n [IGNORE | REPLACE] [AS] SELECT ... (Some legal select statement)\n\nCREATE TABLE creates a table with the given name. You must have the\nCREATE privilege for the table.\n\nRules for allowable table names are given in\nhttp://dev.mysql.com/doc/refman/5.1/en/identifiers.html. By default,\nthe table is created in the default database. An error occurs if the\ntable exists, if there is no default database, or if the database does\nnot exist.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/create-table.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/create-table.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (465,17,'>','Syntax:\n>\n\nGreater than:\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/comparison-operators.html\n\n','mysql> SELECT 2 > 2;\n -> 0\n','http://dev.mysql.com/doc/refman/5.1/en/comparison-operators.html'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (466,19,'ANALYZE TABLE','Syntax:\nANALYZE [NO_WRITE_TO_BINLOG | LOCAL] TABLE\n tbl_name [, tbl_name] ...\n\nANALYZE TABLE analyzes and stores the key distribution for a table.\nDuring the analysis, the table is locked with a read lock for MyISAM.\nFor InnoDB the table is locked with a write lock. This statement works\nwith MyISAM, and InnoDB tables. For MyISAM tables, this statement is\nequivalent to using myisamchk --analyze.\n\nFor more information on how the analysis works within InnoDB, see\nhttp://dev.mysql.com/doc/refman/5.1/en/innodb-restrictions.html.\n\nMy… uses the stored key distribution to decide the order in which\ntables should be joined when you perform a join on something other than\na constant. In addition, key distributions can be used when deciding\nwhich indexes to use for a specific table within a query.\n\nThis statement requires SELECT and INSERT privileges for the table.\n\nBeginning with MySQL 5.1.27, ANALYZE TABLE is also supported for\npartitioned tables. Also beginning with MySQL 5.1.27, you can use ALTER\nTABLE ... ANALYZE PARTITION to analyze one or more partitions; for more\ninformation, see [HELP ALTER TABLE], and\nhttp://dev.mysql.com/doc/refman/5.1/en/partitioning-maintenance.html.\…: http://dev.mysql.com/doc/refman/5.1/en/analyze-table.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/analyze-table.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (466,19,'ANALYZE TABLE','Syntax:\nANALYZE [NO_WRITE_TO_BINLOG | LOCAL] TABLE\n tbl_name [, tbl_name] ...\n\nANALYZE TABLE analyzes and stores the key distribution for a table.\nDuring the analysis, the table is locked with a read lock for MyISAM.\nFor InnoDB the table is locked with a write lock. This statement works\nwith MyISAM and InnoDB tables. For MyISAM tables, this statement is\nequivalent to using myisamchk --analyze.\n\nFor more information on how the analysis works within InnoDB, see\nhttp://dev.mysql.com/doc/refman/5.1/en/innodb-restrictions.html.\n\nMy… uses the stored key distribution to decide the order in which\ntables should be joined when you perform a join on something other than\na constant. In addition, key distributions can be used when deciding\nwhich indexes to use for a specific table within a query.\n\nThis statement requires SELECT and INSERT privileges for the table.\n\nBeginning with MySQL 5.1.27, ANALYZE TABLE is also supported for\npartitioned tables. Also beginning with MySQL 5.1.27, you can use ALTER\nTABLE ... ANALYZE PARTITION to analyze one or more partitions; for more\ninformation, see [HELP ALTER TABLE], and\nhttp://dev.mysql.com/doc/refman/5.1/en/partitioning-maintenance.html.\…: http://dev.mysql.com/doc/refman/5.1/en/analyze-table.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/analyze-table.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (467,30,'MICROSECOND','Syntax:\nMICROSECOND(expr)\n\nReturns the microseconds from the time or datetime expression expr as a\nnumber in the range from 0 to 999999.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','mysql> SELECT MICROSECOND(\'12:00:00.123456\');\n -> 123456\nmysql> SELECT MICROSECOND(\'2009-12-31 23:59:59.000010\');\n -> 10\n','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (468,37,'CONSTRAINT','InnoDB supports foreign key constraints. The syntax for a foreign key\nconstraint definition in InnoDB looks like this:\n\n[CONSTRAINT [symbol]] FOREIGN KEY\n [index_name] (index_col_name, ...)\n REFERENCES tbl_name (index_col_name,...)\n [ON DELETE reference_option]\n [ON UPDATE reference_option]\n\nreference_option:\n RESTRICT | CASCADE | SET NULL | NO ACTION\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/innodb-foreign-key-constraints.html\…','CREATE TABLE product (category INT NOT NULL, id INT NOT NULL,\n price DECIMAL,\n PRIMARY KEY(category, id)) ENGINE=INNODB;\nCREATE TABLE customer (id INT NOT NULL,\n PRIMARY KEY (id)) ENGINE=INNODB;\nCREATE TABLE product_order (no INT NOT NULL AUTO_INCREMENT,\n product_category INT NOT NULL,\n product_id INT NOT NULL,\n customer_id INT NOT NULL,\n PRIMARY KEY(no),\n INDEX (product_category, product_id),\n FOREIGN KEY (product_category, product_id)\n REFERENCES product(category, id)\n ON UPDATE CASCADE ON DELETE RESTRICT,\n INDEX (customer_id),\n FOREIGN KEY (customer_id)\n REFERENCES customer(id)) ENGINE=INNODB;\n','http://dev.mysql.com/doc/refman/5.1/en/innodb-foreign-key-constraints.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (469,37,'CREATE SERVER','Syntax:\nCREATE SERVER server_name\n FOREIGN DATA WRAPPER wrapper_name\n OPTIONS (option [, option] ...)\n\noption:\n { HOST character-literal\n | DATABASE character-literal\n | USER character-literal\n | PASSWORD character-literal\n | SOCKET character-literal\n | OWNER character-literal\n | PORT numeric-literal }\n\nThis statement creates the definition of a server for use with the\nFEDERATED storage engine. The CREATE SERVER statement creates a new row\nwithin the servers table within the mysql database. This statement\nrequires the SUPER privilege.\n\nThe server_name should be a unique reference to the server. Server\ndefinitions are global within the scope of the server, it is not\npossible to qualify the server definition to a specific database.\nserver_name has a maximum length of 64 characters (names longer than 64\ncharacters are silently truncated), and is case insensitive. You may\nspecify the name as a quoted string.\n\nThe wrapper_name should be mysql, and may be quoted with single quotes.\nOther values for wrapper_name are not currently supported.\n\nFor each option you must specify either a character literal or numeric\nliteral. Character literals are UTF-8, support a maximum length of 64\ncharacters and default to a blank (empty) string. String literals are\nsilently truncated to 64 characters. Numeric literals must be a number\nbetween 0 and 9999, default value is 0.\n\n*Note*: Note that the OWNER option is currently not applied, and has no\neffect on the ownership or operation of the server connection that is\ncreated.\n\nThe CREATE SERVER statement creates an entry in the mysql.server table\nthat can later be used with the CREATE TABLE statement when creating a\nFEDERATED table. The options that you specify will be used to populate\nthe columns in the mysql.server table. The table columns are\nServer_name, Host, Db, Username, Password, Port and Socket.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/create-server.html\n\n','CREATE SERVER s\nFOREIGN DATA WRAPPER mysql\nOPTIONS (USER \'Remote\', HOST \'192.168.1.106\', DATABASE \'test\');\n','http://dev.mysql.com/doc/refman/5.1/en/create-server.html'); @@ -567,7 +567,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (498,4,'RADIANS','Syntax:\nRADIANS(X)\n\nReturns the argument X, converted from degrees to radians. (Note that\nπ radians equals 180 degrees.)\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/mathematical-functions.html\n\n','mysql> SELECT RADIANS(90);\n -> 1.5707963267949\n','http://dev.mysql.com/doc/refman/5.1/en/mathematical-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (499,15,'COLLATION','Syntax:\nCOLLATION(str)\n\nReturns the collation of the string argument.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/information-functions.html\n\n','mysql> SELECT COLLATION(\'abc\');\n -> \'latin1_swedish_ci\'\nmysql> SELECT COLLATION(_utf8\'abc\');\n -> \'utf8_general_ci\'\n','http://dev.mysql.com/doc/refman/5.1/en/information-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (500,17,'COALESCE','Syntax:\nCOALESCE(value,...)\n\nReturns the first non-NULL value in the list, or NULL if there are no\nnon-NULL values.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/comparison-operators.html\n\n','mysql> SELECT COALESCE(NULL,1);\n -> 1\nmysql> SELECT COALESCE(NULL,NULL,NULL);\n -> NULL\n','http://dev.mysql.com/doc/refman/5.1/en/comparison-operators.html'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (501,15,'VERSION','Syntax:\nVERSION()\n\nReturns a string that indicates the MySQL server version. The string\nuses the utf8 character set.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/information-functions.html\n\n','mysql> SELECT VERSION();\n -> \'5.1.39-standard\'\n','http://dev.mysql.com/doc/refman/5.1/en/information-functions.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (501,15,'VERSION','Syntax:\nVERSION()\n\nReturns a string that indicates the MySQL server version. The string\nuses the utf8 character set.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/information-functions.html\n\n','mysql> SELECT VERSION();\n -> \'5.1.41-standard\'\n','http://dev.mysql.com/doc/refman/5.1/en/information-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (502,35,'MAKE_SET','Syntax:\nMAKE_SET(bits,str1,str2,...)\n\nReturns a set value (a string containing substrings separated by ","\ncharacters) consisting of the strings that have the corresponding bit\nin bits set. str1 corresponds to bit 0, str2 to bit 1, and so on. NULL\nvalues in str1, str2, ... are not appended to the result.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/string-functions.html\n\n','mysql> SELECT MAKE_SET(1,\'a\',\'b\',\'c\');\n -> \'a\'\nmysql> SELECT MAKE_SET(1 | 4,\'hello\',\'nice\',\'world\');\n -> \'hello,world\'\nmysql> SELECT MAKE_SET(1 | 4,\'hello\',\'nice\',NULL,\'world\');\n -> \'hello\'\nmysql> SELECT MAKE_SET(0,\'a\',\'b\',\'c\');\n -> \'\'\n','http://dev.mysql.com/doc/refman/5.1/en/string-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (503,35,'FIND_IN_SET','Syntax:\nFIND_IN_SET(str,strlist)\n\nReturns a value in the range of 1 to N if the string str is in the\nstring list strlist consisting of N substrings. A string list is a\nstring composed of substrings separated by "," characters. If the first\nargument is a constant string and the second is a column of type SET,\nthe FIND_IN_SET() function is optimized to use bit arithmetic. Returns\n0 if str is not in strlist or if strlist is the empty string. Returns\nNULL if either argument is NULL. This function does not work properly\nif the first argument contains a comma (",") character.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/string-functions.html\n\n','mysql> SELECT FIND_IN_SET(\'b\',\'a,b,c,d\');\n -> 2\n','http://dev.mysql.com/doc/refman/5.1/en/string-functions.html'); @@ -1293,6 +1293,7 @@ insert into help_relation (help_topic_id insert into help_relation (help_topic_id,help_keyword_id) values (143,115); insert into help_relation (help_topic_id,help_keyword_id) values (468,115); insert into help_relation (help_topic_id,help_keyword_id) values (180,115); +insert into help_relation (help_topic_id,help_keyword_id) values (353,115); insert into help_relation (help_topic_id,help_keyword_id) values (420,115); insert into help_relation (help_topic_id,help_keyword_id) values (473,115); insert into help_relation (help_topic_id,help_keyword_id) values (197,116); @@ -1550,11 +1551,12 @@ insert into help_relation (help_topic_id insert into help_relation (help_topic_id,help_keyword_id) values (383,238); insert into help_relation (help_topic_id,help_keyword_id) values (330,239); insert into help_relation (help_topic_id,help_keyword_id) values (344,239); -insert into help_relation (help_topic_id,help_keyword_id) values (253,239); insert into help_relation (help_topic_id,help_keyword_id) values (152,239); insert into help_relation (help_topic_id,help_keyword_id) values (428,239); insert into help_relation (help_topic_id,help_keyword_id) values (464,239); +insert into help_relation (help_topic_id,help_keyword_id) values (253,239); insert into help_relation (help_topic_id,help_keyword_id) values (209,239); +insert into help_relation (help_topic_id,help_keyword_id) values (353,239); insert into help_relation (help_topic_id,help_keyword_id) values (420,239); insert into help_relation (help_topic_id,help_keyword_id) values (184,240); insert into help_relation (help_topic_id,help_keyword_id) values (273,241); === modified file 'support-files/MacOSX/ReadMe.txt' --- a/support-files/MacOSX/ReadMe.txt 2009-05-25 09:59:47 +0000 +++ b/support-files/MacOSX/ReadMe.txt 2009-12-01 07:24:05 +0000 @@ -1,5 +1,47 @@ -2.5. Installing MySQL on Mac OS X +2.7. Installing MySQL on Mac OS X + + MySQL for Mac OS X is available in a number of different forms: + + * Native Package Installer format, which uses the native Mac OS + X installer to walk you through the installation of MySQL. For + more information, see Section 2.7.1, "Installing MySQL Using + the Installation Package." You can use the package installer + with Mac OS X 10.3 and later, and available for both PowerPC + and Intel architectures, and both 32-bit and 64-bit + architectures. There is no Universal Binary available using + the package installation method. The user you use to perform + the installation must have administrator privileges. + + * Tar package format, which uses a file packaged using the Unix + tar and gzip commands. To use this method, you will need to + open a Terminal window. You do not need administrator + privileges using this method, as you can install the MySQL + server anywhere using this method. For more information on + using this method, you can use the generic instructions for + using a tarball, Section 2.2, "Installing MySQL from Generic + Binaries on Unix/Linux."You can use the package installer with + Mac OS X 10.3 and later, and available for both PowerPC and + Intel architectures, and both 32-bit and 64-bit architectures. + A Universal Binary, incorporating both Power PC and Intel + architectures and 32-bit and 64-bit binaries is available. + In addition to the core installation, the Package Installer + also includes Section 2.7.2, "Installing the MySQL Startup + Item" and Section 2.7.3, "Installing and Using the MySQL + Preference Pane," both of which simplify the management of + your installation. + + * Mac OS X server includes a version of MySQL as standard. If + you want to use a more recent version than that supplied with + the Mac OS X server release, you can make use of the package + or tar formats. For more information on using the MySQL + bundled with Mac OS X, see Section 2.7.4, "Using MySQL on Mac + OS X Server." + + For additional information on using MySQL on Mac OS X, see Section + 2.7.5, "MySQL Installation on Mac OS X Notes." + +2.7.1. Installing MySQL Using the Installation Package You can install MySQL on Mac OS X 10.3.x ("Panther") or newer using a Mac OS X binary package in PKG format instead of the @@ -11,8 +53,6 @@ first need to mount by double-clicking its icon in the Finder. It should then mount the image and display its contents. - To obtain MySQL, see Section 2.1.3, "How to Get MySQL." - Note Before proceeding with the installation, be sure to shut down all @@ -20,144 +60,319 @@ Note Application (on Mac OS X Server) or via mysqladmin shutdown on the command line. - To actually install the MySQL PKG file, double-click on the - package icon. This launches the Mac OS X Package Installer, which - guides you through the installation of MySQL. + When installing from the package version, you should also install + the MySQL Preference Pane, which will allow you to control the + startup and execution of your MySQL server from System + Preferences. For more information, see Section 2.7.3, "Installing + and Using the MySQL Preference Pane." + + When installing using the package installer, the files are + installed into a directory within /usr/local matching the name of + the installation version and platform. For example, the installer + file mysql-5.1.39-osx10.5-x86_64.pkg installs MySQL into + /usr/local/mysql-5.1.39-osx10.5-x86_64 . The installation layout + of the directory is as shown in the following table: + Directory Contents of Directory + bin Client programs and the mysqld server + data Log files, databases + docs Manual in Info format + include Include (header) files + lib Libraries + man Unix manual pages + mysql-test MySQL test suite + scripts Contains the mysql_install_db script + share/mysql Error message files + sql-bench Benchmarks + support-files Scripts and sample configuration files + /tmp/mysql.sock The location of the MySQL Unix socket + + During the package installer process, a symbolic link from + /usr/local/mysql to the version/platform specific directory + created during installation will be created automatically. + + 1. Download and open the MySQL package installer, which is + provided on a disk image (.dmg). Double-click to open the disk + image, which includes the main MySQL installation package, the + MySQLStartupItem.pkg installation package, and the + MySQL.prefPane. + + 2. Double-click on the MySQL installer package. It will be named + according to the version of MySQL you have downloaded. For + example, if you have downloaded MySQL 5.1.39, double-click + mysql-5.1.39-osx10.5-x86.pkg. + + 3. You will be presented with the openin installer dialog. Click + Continue to begihn installation. + MySQL Package Installer: Step 1 + + 4. A copy of the installation instructions and other important + information relevant to this installation are display. Click + Continue . + + 5. If you have downloaded the community version of MySQL, you + will be shown a copy of the relevent GNU General Public + License. Click Continue . + + 6. Select the drive you want to use to install the MySQL Startup + Item. The drive must have a valid, bootable, Mac OS X + operating system installed. Click Continue. + MySQL Package Installer: Step 4 + + 7. You will be asked to confirm the details of the installation, + including the space required for the installation. To change + the drive on which the startup item is installed you can click + either Go Back or Change Install Location.... To install the + startup item, click Install. - Due to a bug in the Mac OS X package installer, you may see this - error message in the destination disk selection dialog: -You cannot install this software on this disk. (null) - - If this error occurs, simply click the Go Back button once to - return to the previous screen. Then click Continue to advance to - the destination disk selection again, and you should be able to - choose the destination disk correctly. We have reported this bug - to Apple and it is investigating this problem. - - The Mac OS X PKG of MySQL installs itself into - /usr/local/mysql-VERSION and also installs a symbolic link, - /usr/local/mysql, that points to the new location. If a directory - named /usr/local/mysql exists, it is renamed to - /usr/local/mysql.bak first. Additionally, the installer creates - the grant tables in the mysql database by executing - mysql_install_db. - - The installation layout is similar to that of a tar file binary - distribution; all MySQL binaries are located in the directory - /usr/local/mysql/bin. The MySQL socket file is created as - /tmp/mysql.sock by default. See Section 2.1.5, "Installation - Layouts." - - MySQL installation requires a Mac OS X user account named mysql. A - user account with this name should exist by default on Mac OS X - 10.2 and up. + 8. Once the installation has been completed successfully, you + will be given an Install Succeeded message. - If you are running Mac OS X Server, a version of MySQL should - already be installed. The following table shows the versions of - MySQL that ship with Mac OS X Server versions. - Mac OS X Server Version MySQL Version - 10.2-10.2.2 3.23.51 - 10.2.3-10.2.6 3.23.53 - 10.3 4.0.14 - 10.3.2 4.0.16 - 10.4.0 4.1.10a + Once you have completed the basic installation, you must complete + the post-installation steps as specifed in Section 2.13, + "Post-Installation Setup and Testing." - This manual section covers the installation of the official MySQL - Mac OS X PKG only. Make sure to read Apple's help information - about installing MySQL: Run the "Help View" application, select - "Mac OS X Server" help, do a search for "MySQL," and read the item - entitled "Installing MySQL." - - If you previously used Marc Liyanage's MySQL packages for Mac OS X - from http://www.entropy.ch, you can simply follow the update - instructions for packages using the binary installation layout as - given on his pages. - - If you are upgrading from Marc's 3.23.x versions or from the Mac - OS X Server version of MySQL to the official MySQL PKG, you also - need to convert the existing MySQL privilege tables to the current - format, because some new security privileges have been added. See - Section 4.4.8, "mysql_upgrade --- Check Tables for MySQL Upgrade." - - If you want MySQL to start automatically during system startup, - you also need to install the MySQL Startup Item. It is part of the - Mac OS X installation disk images as a separate installation - package. Simply double-click the MySQLStartupItem.pkg icon and - follow the instructions to install it. The Startup Item need be - installed only once. There is no need to install it each time you - upgrade the MySQL package later. + For convenience, you may also want to install the Section 2.7.2, + "Installing the MySQL Startup Item" and Section 2.7.3, "Installing + and Using the MySQL Preference Pane." + +2.7.2. Installing the MySQL Startup Item + + The MySQL Installation Package includes a startup item that can be + used to automatically startup and shutdown MySQL during boot. + + To install the MySQL Startup Item: + + 1. Download and open the MySQL package installer, which is + provided on a disk image (.dmg). Double-click to open the disk + image, which includes the main MySQL installation package, the + MySQLStartupItem.pkg installation package, and the + MySQL.prefPane. + + 2. Double-click on the MySQLStartItem.pkg file to start the + installation process. + + 3. You will be presented with the Install MySQL Startup Item + dialog. + MySQL Startup Item Installer: Step 1 + Click Continue to continue the installation process. + + 4. A copy of the installation instructions and other important + information relevant to this installation are display. Click + Continue . + + 5. Select the drive you want to use to install the MySQL Startup + Item. The drive must have a valid, bootable, Mac OS X + operating system installed. Click Continue. + MySQL Startup Item Installer: Step 3 + + 6. You will be asked to confirm the details of the installation. + To change the drive on which the startup item is installed you + can click either Go Back or Change Install Location.... To + install the startup item, click Install. + + 7. Once the installation has been completed successfully, you + will be given an Install Succeeded message. + MySQL Startup Item Installer: Step 5 The Startup Item for MySQL is installed into - /Library/StartupItems/MySQLCOM. (Before MySQL 4.1.2, the location - was /Library/StartupItems/MySQL, but that collided with the MySQL - Startup Item installed by Mac OS X Server.) Startup Item - installation adds a variable MYSQLCOM=-YES- to the system - configuration file /etc/hostconfig. If you want to disable the - automatic startup of MySQL, simply change this variable to - MYSQLCOM=-NO-. - - On Mac OS X Server, the default MySQL installation uses the - variable MYSQL in the /etc/hostconfig file. The MySQL Startup Item - installer disables this variable by setting it to MYSQL=-NO-. This - avoids boot time conflicts with the MYSQLCOM variable used by the - MySQL Startup Item. However, it does not shut down a running MySQL - server. You should do that yourself. + /Library/StartupItems/MySQLCOM. The Startup Item installation adds + a variable MYSQLCOM=-YES- to the system configuration file + /etc/hostconfig. If you want to disable the automatic startup of + MySQL, simply change this variable to MYSQLCOM=-NO-. After the installation, you can start up MySQL by running the following commands in a terminal window. You must have administrator privileges to perform this task. - If you have installed the Startup Item, use this command: + If you have installed the Startup Item, use this command to start + the server: shell> sudo /Library/StartupItems/MySQLCOM/MySQLCOM start -(Enter your password, if necessary) -(Press Control-D or enter "exit" to exit the shell) - If you don't use the Startup Item, enter the following command - sequence: -shell> cd /usr/local/mysql -shell> sudo ./bin/mysqld_safe -(Enter your password, if necessary) -(Press Control-Z) -shell> bg -(Press Control-D or enter "exit" to exit the shell) + You may be prompted for your password to complete the startup. - You should be able to connect to the MySQL server, for example, by - running /usr/local/mysql/bin/mysql. + If you have installed the Startup Item, use this command to stop + the server: +shell> sudo /Library/StartupItems/MySQLCOM/MySQLCOM stop + + You may be prompted for your password to complete the shutdown. + +2.7.3. Installing and Using the MySQL Preference Pane + + The MySQL Package installer disk image also includes a custom + MySQL Preference Pane that enables you to start, stop and control + automated startup during boot of your MySQL installation. + + To install the MySQL Preference Pane: + + 1. Download and open the MySQL package installer package, which + is provided on a disk image (.dmg). Double-click to open the + disk image, which includes the main MySQL installation + package, the MySQLStartupItem.pkg installation package, and + the MySQL.prefPane. + + 2. Double click on MySQL.prefPane. The MySQL System Preferences + will open. + + 3. If this is the first time you have installed the preference + pane, you will be asked to confirm installation and whether + you want to install the preference pane for all users, or only + the current user. To install the preference pane for all users + you will need administrator privileges. If necessary, you will + be prompted for the username and password for a user with + administrator privileges. + + 4. If you already have the MySQL Preference Pane installed, you + will be asked to confirm whether you want to overwrite the + existing MySQL Preference Pane. Note - The accounts that are listed in the MySQL grant tables initially - have no passwords. After starting the server, you should set up - passwords for them using the instructions in Section 2.11, - "Post-Installation Setup and Testing." + The MySQL Preference Pane only starts and stops MySQL installation + installed from the MySQL package installation that have been + installed in the default location. + + Once the MySQL Preference Pane has been installed, you can control + your MySQL server instance using the preference pane. To use the + preference pane, open the System Preferences... from the Apple + menu. Select the MySQL preference pane by clicking on the MySQL + logo within the Other section of the preference panes list. + MySQL Preference Pane + + The MySQL Preference Pane shows the current status of the MySQL + server, showing stopped (in red) if the server is not running and + running (in green) if the server has already been started. The + preference pane will also show the current setting for whether the + MySQL server has been set to start up automatically. + + * To start MySQL using the preference pane: + Click Start MySQL Server. You may be prompted for the username + and password of a user with administrator privileges to start + the MySQL server. + + * To stop MySQL using the preference pane: + Click Stop MySQL Server. You may be prompted for the username + and password of a user with administrator privileges to + shutdown the MySQL server. + + * To automatically start the MySQL server when the system boots: + Check the checkbox next to Automatically Start MySQL Server on + Startup. + + * To disable the automatic starting of the MySQL server when the + system boots: + Uncheck the checkbox next to Automatically Start MySQL Server + on Startup. + + You can close the System Preferences... once you have completed + your settings. - You might want to add aliases to your shell's resource file to - make it easier to access commonly used programs such as mysql and - mysqladmin from the command line. The syntax for bash is: +2.7.4. Using MySQL on Mac OS X Server + + If you are running Mac OS X Server, a version of MySQL should + already be installed. The following table shows the versions of + MySQL that ship with Mac OS X Server versions. + Mac OS X Server Version MySQL Version + 10.2-10.2.2 3.23.51 + 10.2.3-10.2.6 3.23.53 + 10.3 4.0.14 + 10.3.2 4.0.16 + 10.4.0 4.1.10a + 10.5.0 5.0.45 + 10.6.0 5.0.82 + + The installation layout of MySQL on Mac OS X Server is as shown in + the table below: + Directory Contents of Directory + /usr/bin Client programs + /var/mysql Log files, databases + /usr/libexec The mysqld server + /usr/share/man Unix manual pages + /usr/share/mysql/mysql-test MySQL test suite + /usr/share/mysql Contains the mysql_install_db script + /var/mysql/mysql.sock The location of the MySQL Unix socket + +Note + + The MySQL server bundled with Mac OS X Server does not include the + MySQL client libraries and header files required if you want to + access and use MySQL from a third-party driver, such as Perl DBI + or PHP. For more information on obtaining and installing MySQL + libraries, see Mac OS X Server version 10.5: MySQL libraries + available for download (http://support.apple.com/kb/TA25017) + Alternatively, you can ignore the bundled MySQL server and install + MySQL from the package or tarball installation. + + For more information on managing the bundled MySQL instance in Mac + OS X Server 10.5, see Mac OS X Server: Web Technologies + Administration For Version 10.5 Leopard + (http://images.apple.com/server/macosx/docs/Web_Technologies_Admin + _v10.5.pdf). For more information on managing the bundled MySQL + instance in Mac OS X Server 10.6, see Mac OS X Server: Web + Technologies Administration Version 10.6 Snow Leopard + (http://manuals.info.apple.com/en_US/WebTech_v10.6.pdf) + +2.7.5. MySQL Installation on Mac OS X Notes + + You should keep the following issues and notes in mind: + + * The default location for the MySQL Unix socket is different on + Mac OS X and Mac OS X Server depending on the installation + type you chose. The default locations by installation are as + follows: + + Package Installer from MySQL /tmp/mysql.sock + Tarball from MySQL /tmp/mysql.sock + MySQL Bundled with Mac OS X Server /var/mysql/mysql.sock + To prevent issues, you should either change the configuration + of the socket used within your application (for example, + changing php.ini), or you should configure the socket location + using a MySQL configuration file and the socket option. For + more information, see Section 5.1.2, "Server Command Options." + + * You may need (or want) to create a specific mysql user to own + the MySQL directory and data. On Mac OS X 10.4 and lower you + can do this by using the Netinfo Manager application, located + within the Utilities folder within the Applications folder. On + Mac OS X 10.5 and later you can do this through the Directory + Utility. From Mac OS X 10.5 and later (including Mac OS X + Server 10.5) the mysql should already exist. For use in single + user mode, an entry for _mysql (note the underscore prefix) + should already exist within the system /etc/passwd file. + + * Due to a bug in the Mac OS X package installer, you may see + this error message in the destination disk selection dialog: +You cannot install this software on this disk. (null) + If this error occurs, simply click the Go Back button once to + return to the previous screen. Then click Continue to advance + to the destination disk selection again, and you should be + able to choose the destination disk correctly. We have + reported this bug to Apple and it is investigating this + problem. + + * Because the MySQL package installer installs the MySQL + contents into a version and platform specific directory, you + can use this to upgrade and migrate your database between + versions. You will need to either copy the data directory from + the old version to the new version, or alternatively specify + an alternative datadir value to set location of the data + directory. + + * You might want to add aliases to your shell's resource file to + make it easier to access commonly used programs such as mysql + and mysqladmin from the command line. The syntax for bash is: alias mysql=/usr/local/mysql/bin/mysql alias mysqladmin=/usr/local/mysql/bin/mysqladmin - - For tcsh, use: + For tcsh, use: alias mysql /usr/local/mysql/bin/mysql alias mysqladmin /usr/local/mysql/bin/mysqladmin - - Even better, add /usr/local/mysql/bin to your PATH environment - variable. You can do this by modifying the appropriate startup - file for your shell. For more information, see Section 4.2.1, - "Invoking MySQL Programs." - - If you are upgrading an existing installation, note that - installing a new MySQL PKG does not remove the directory of an - older installation. Unfortunately, the Mac OS X Installer does not - yet offer the functionality required to properly upgrade - previously installed packages. - - To use your existing databases with the new installation, you'll - need to copy the contents of the old data directory to the new - data directory. Make sure that neither the old server nor the new - one is running when you do this. After you have copied over the - MySQL database files from the previous installation and have - successfully started the new server, you should consider removing - the old installation files to save disk space. Additionally, you - should also remove older versions of the Package Receipt - directories located in /Library/Receipts/mysql-VERSION.pkg. + Even better, add /usr/local/mysql/bin to your PATH environment + variable. You can do this by modifying the appropriate startup + file for your shell. For more information, see Section 4.2.1, + "Invoking MySQL Programs." + + * After you have copied over the MySQL database files from the + previous installation and have successfully started the new + server, you should consider removing the old installation + files to save disk space. Additionally, you should also remove + older versions of the Package Receipt directories located in + /Library/Receipts/mysql-VERSION.pkg.

1 0

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (knielsen:2708)
by knielsen＠knielsen-hq.org 01 Dec '09

01 Dec '09

#At lp:maria 2708 knielsen(a)knielsen-hq.org 2009-12-01 Imported MySQL documentation files from MySQL 5.1.41 source tarball modified: Docs/INSTALL-BINARY INSTALL-SOURCE INSTALL-WIN-SOURCE man/comp_err.1 man/innochecksum.1 man/make_win_bin_dist.1 man/msql2mysql.1 man/my_print_defaults.1 man/myisam_ftdump.1 man/myisamchk.1 man/myisamlog.1 man/myisampack.1 man/mysql-stress-test.pl.1 man/mysql-test-run.pl.1 man/mysql.1 man/mysql.server.1 man/mysql_client_test.1 man/mysql_config.1 man/mysql_convert_table_format.1 man/mysql_find_rows.1 man/mysql_fix_extensions.1 man/mysql_fix_privilege_tables.1 man/mysql_install_db.1 man/mysql_secure_installation.1 man/mysql_setpermission.1 man/mysql_tzinfo_to_sql.1 man/mysql_upgrade.1 man/mysql_waitpid.1 man/mysql_zap.1 man/mysqlaccess.1 man/mysqladmin.1 man/mysqlbinlog.1 man/mysqlbug.1 man/mysqlcheck.1 man/mysqld.8 man/mysqld_multi.1 man/mysqld_safe.1 man/mysqldump.1 man/mysqldumpslow.1 man/mysqlhotcopy.1 man/mysqlimport.1 man/mysqlmanager.8 man/mysqlshow.1 man/mysqlslap.1 man/mysqltest.1 man/ndbd.8 man/ndbd_redo_log_reader.1 man/ndbmtd.8 man/perror.1 man/replace.1 man/resolve_stack_dump.1 man/resolveip.1 scripts/fill_help_tables.sql support-files/MacOSX/ReadMe.txt === modified file 'Docs/INSTALL-BINARY' --- a/Docs/INSTALL-BINARY 2009-09-16 12:03:18 +0000 +++ b/Docs/INSTALL-BINARY 2009-12-01 07:24:05 +0000 @@ -1,23 +1,35 @@ -2.9. Installing MySQL from tar.gz Packages on Other Unix-Like Systems +2.2. Installing MySQL from Generic Binaries on Unix/Linux This section covers the installation of MySQL binary distributions that are provided for various platforms in the form of compressed - tar files (files with a .tar.gz extension). See Section 2.1.2.4, - "MySQL Binaries Compiled by Sun Microsystems, Inc.," for a + tar files (files with a .tar.gz extension). See Section 2.2, + "Installing MySQL from Generic Binaries on Unix/Linux," for a detailed list. To obtain MySQL, see Section 2.1.3, "How to Get MySQL." + Sun Microsystems, Inc. provides a set of binary distributions of + MySQL. In addition to binaries provided in platform-specific + package formats, we offer binary distributions for a number of + platforms in the form of compressed tar files (.tar.gz files). For + Windows distributions, see Section 2.5, "Installing MySQL on + Windows." + + If you want to compile a debug version of MySQL from a source + distribution, you should add --with-debug or --with-debug=full to + the configure command used to configure the distribution and + remove any -fomit-frame-pointer options. + MySQL tar file binary distributions have names of the form mysql-VERSION-OS.tar.gz, where VERSION is a number (for example, - 5.1.39), and OS indicates the type of operating system for which + 5.1.41), and OS indicates the type of operating system for which the distribution is intended (for example, pc-linux-i686). In addition to these generic packages, we also offer binaries in - platform-specific package formats for selected platforms. See - Section 2.2, "Standard MySQL Installation Using a Binary - Distribution," for more information on how to install these. + platform-specific package formats for selected platforms. See the + platform specific sections for more information, for more + information on how to install these. You need the following tools to install a MySQL tar file binary distribution: @@ -27,11 +39,13 @@ * A reasonable tar to unpack the distribution. GNU tar is known to work. Some operating systems come with a preinstalled version of tar that is known to have problems. For example, - the tar provided with early versions of Mac OS X, SunOS 4.x - and Solaris 8 and earlier are known to have problems with long - file names. On Mac OS X, you can use the preinstalled gnutar - program. On other systems with a deficient tar, you should - install GNU tar first. + the tar provided with early versions of Mac OS X, SunOS 4.x, + Solaris 8, Solaris 9, Solaris 10 and OpenSolaris, and HP-UX + are known to have problems with long file names. On Mac OS X, + you can use the preinstalled gnutar program. On Solaris 10 and + OpenSolaris you can use the preinstalled gtar. On other + systems with a deficient tar, you should install GNU tar + first. If you run into problems and need to file a bug report, please use the instructions in Section 1.6, "How to Report Bugs or Problems." @@ -54,7 +68,7 @@ shell> bin/mysqld_safe --user=mysql & Note This procedure does not set up any passwords for MySQL accounts. - After following the procedure, proceed to Section 2.11, + After following the procedure, proceed to Section 2.13, "Post-Installation Setup and Testing." A more detailed version of the preceding description for @@ -148,7 +162,7 @@ shell> chown -R mysql data machine, you can copy support-files/mysql.server to the location where your system has its startup files. More information can be found in the support-files/mysql.server - script itself and in Section 2.11.2.2, "Starting and Stopping + script itself and in Section 2.13.1.2, "Starting and Stopping MySQL Automatically." 10. You can set up new accounts using the bin/mysql_setpermission script if you install the DBI and DBD::mysql Perl modules. See @@ -187,5 +201,5 @@ Note The accounts that are listed in the MySQL grant tables initially have no passwords. After starting the server, you should set up - passwords for them using the instructions in Section 2.11, + passwords for them using the instructions in Section 2.13, "Post-Installation Setup and Testing." === modified file 'INSTALL-SOURCE' --- a/INSTALL-SOURCE 2009-09-16 12:03:18 +0000 +++ b/INSTALL-SOURCE 2009-12-01 07:24:05 +0000 @@ -5,9 +5,8 @@ Chapter 2. Installing and Upgrading MySQ of the procedure follows and later sections provide the details. If you plan to upgrade an existing version of MySQL to a newer version rather than install MySQL for the first time, see Section - 2.12.1, "Upgrading MySQL," for information about upgrade - procedures and about issues that you should consider before - upgrading. + 2.4.1, "Upgrading MySQL," for information about upgrade procedures + and about issues that you should consider before upgrading. If you are interested in migrating to MySQL from another database system, you may wish to read Section A.8, "MySQL 5.1 FAQ --- @@ -15,96 +14,61 @@ Chapter 2. Installing and Upgrading MySQ concerning migration issues. 1. Determine whether MySQL runs and is supported on your - platform. Please note that not all platforms are equally - suitable for running MySQL, and that not all platforms on - which MySQL is known to run are officially supported by Sun - Microsystems, Inc.: - - + For MySQL Enterprise Server, the officially supported - platforms are listed at - http://www.mysql.com/support/supportedplatforms.html. - - + MySQL Community Server runs on the platforms listed at - Section 2.1.1, "Operating Systems Supported by MySQL - Community Server." - - 2. Choose which distribution to install. Several versions of - MySQL are available, and most are available in several - distribution formats. You can choose from pre-packaged - distributions containing binary (precompiled) programs or - source code. When in doubt, use a binary distribution. We also - provide public access to our current source tree for those who - want to see our most recent developments and help us test new - code. To determine which version and type of distribution you - should use, see Section 2.1.2, "Choosing Which MySQL - Distribution to Install." - - 3. Download the distribution that you want to install. For - instructions, see Section 2.1.3, "How to Get MySQL." To verify - the integrity of the distribution, use the instructions in - Section 2.1.4, "Verifying Package Integrity Using MD5 + platform. + Please note that not all platforms are equally suitable for + running MySQL, and that not all platforms on which MySQL is + known to run are officially supported by Sun Microsystems, + Inc.: + + 2. Choose which distribution to install. + Several versions of MySQL are available, and most are + available in several distribution formats. You can choose from + pre-packaged distributions containing binary (precompiled) + programs or source code. When in doubt, use a binary + distribution. We also provide public access to our current + source tree for those who want to see our most recent + developments and help us test new code. To determine which + version and type of distribution you should use, see Section + 2.1.2, "Choosing Which MySQL Distribution to Install." + + 3. Download the distribution that you want to install. + For instructions, see Section 2.1.3, "How to Get MySQL." To + verify the integrity of the distribution, use the instructions + in Section 2.1.4, "Verifying Package Integrity Using MD5 Checksums or GnuPG." - 4. Install the distribution. To install MySQL from a binary - distribution, use the instructions in Section 2.2, "Standard - MySQL Installation Using a Binary Distribution." To install - MySQL from a source distribution or from the current - development source tree, use the instructions in Section 2.10, - "MySQL Installation Using a Source Distribution." - If you encounter installation difficulties, see Section 2.13, - "Operating System-Specific Notes," for information on solving - problems for particular platforms. - - 5. Perform any necessary post-installation setup. After - installing MySQL, read Section 2.11, "Post-Installation Setup - and Testing." This section contains important information - about making sure the MySQL server is working properly. It - also describes how to secure the initial MySQL user accounts, - which have no passwords until you assign passwords. The - section applies whether you install MySQL using a binary or - source distribution. + 4. Install the distribution. + To install MySQL from a binary distribution, use the + instructions in Section 2.2, "Installing MySQL from Generic + Binaries on Unix/Linux." + To install MySQL from a source distribution or from the + current development source tree, use the instructions in + Section 2.3, "MySQL Installation Using a Source Distribution." + + 5. Perform any necessary post-installation setup. + After installing MySQL, read Section 2.13, "Post-Installation + Setup and Testing." This section contains important + information about making sure the MySQL server is working + properly. It also describes how to secure the initial MySQL + user accounts, which have no passwords until you assign + passwords. The section applies whether you install MySQL using + a binary or source distribution. 6. If you want to run the MySQL benchmark scripts, Perl support for MySQL must be available. See Section 2.15, "Perl Installation Notes." -2.1. General Installation Issues - - The MySQL installation procedure depends on whether you will - install MySQL Enterprise Server or MySQL Community Server. The set - of applicable platforms depends on which distribution you will - install: - - * For MySQL Enterprise Server, the officially supported - platforms are listed at - http://www.mysql.com/support/supportedplatforms.html. - - * MySQL Community Server runs on the platforms listed at Section - 2.1.1, "Operating Systems Supported by MySQL Community - Server." - - For MySQL Enterprise Server, install the main distribution plus - any service packs or hotfixes that you wish to apply using the - Enterprise Installer. For platforms that do not yet have an - Enterprise Installer, use the Community Server instructions. - - For MySQL Community Server, install the main distribution plus any - hotfixes and updates: - - * Download a binary release, or download a source release and - build MySQL yourself from the source code. - - * Retrieve MySQL from the Bazaar tree and build it from source. - The Bazaar tree contains the latest developer code. +2.1. General Installation Guidance The immediately following sections contain the information necessary to choose, download, and verify your distribution. The instructions in later sections of the chapter describe how to install the distribution that you choose. For binary - distributions, see the instructions at Section 2.2, "Standard - MySQL Installation Using a Binary Distribution." To build MySQL - from source, use the instructions at Section 2.10, "MySQL - Installation Using a Source Distribution." + distributions, see the instructions at Section 2.2, "Installing + MySQL from Generic Binaries on Unix/Linux" or the corresponding + section for your platform if available. To build MySQL from + source, use the instructions in Section 2.3, "MySQL Installation + Using a Source Distribution." 2.1.1. Operating Systems Supported by MySQL Community Server @@ -129,58 +93,31 @@ Important MySQL has been reported to compile successfully on the following combinations of operating system and thread package. - * AIX 4.x, 5.x with native threads. See Section 2.13.5.3, - "IBM-AIX notes." - - * Amiga. + * AIX 4.x, 5.x with native threads. See Section 2.12, + "Installing MySQL on AIX." AIX 5.3 should be upgraded to + technology level 7 (5300-07). - * FreeBSD 5.x and up with native threads. + * FreeBSD 5.x and up with native threads. See Section 2.10, + "Installing MySQL on FreeBSD." - * HP-UX 11.x with the native threads. See Section 2.13.5.2, - "HP-UX Version 11.x Notes." + * HP-UX 11.x with the native threads. See Section 2.11, + "Installing MySQL on HP-UX." * Linux, builds on all fairly recent Linux distributions with - glibc 2.3. See Section 2.13.1, "Linux Notes." - - * Mac OS X. See Section 2.13.2, "Mac OS X Notes." - - * NetBSD 1.3/1.4 Intel and NetBSD 1.3 Alpha. See Section - 2.13.4.2, "NetBSD Notes." + glibc 2.3. See Section 2.6, "Installing MySQL on Linux." - * Novell NetWare 6.0 and 6.5. See Section 2.8, "Installing MySQL - on NetWare." + * Mac OS X. See Section 2.7, "Installing MySQL on Mac OS X." - * OpenBSD 2.5 and with native threads. OpenBSD earlier than 2.5 - with the MIT-pthreads package. See Section 2.13.4.3, "OpenBSD - 2.5 Notes." - - * SCO OpenServer 5.0.X with a recent port of the FSU Pthreads - package. See Section 2.13.5.8, "SCO UNIX and OpenServer 5.0.x - Notes." - - * SCO Openserver 6.0.x. See Section 2.13.5.9, "SCO OpenServer - 6.0.x Notes." - - * SCO UnixWare 7.1.x. See Section 2.13.5.10, "SCO UnixWare 7.1.x - and OpenUNIX 8.0.0 Notes." - - * SGI Irix 6.x with native threads. See Section 2.13.5.7, "SGI - Irix Notes." - - * Solaris 2.5 and above with native threads on SPARC and x86. - See Section 2.13.3, "Solaris Notes." - - * Tru64 Unix. See Section 2.13.5.5, "Alpha-DEC-UNIX Notes - (Tru64)." + * Solaris 2.8 on SPARC and x86, including support for native + threads. See Section 2.8.1, "Solaris Notes." * Windows 2000, Windows XP, Windows Vista, Windows Server 2003, - and Windows Server 2008. See Section 2.3, "Installing MySQL on + and Windows Server 2008. See Section 2.5, "Installing MySQL on Windows." MySQL has also been known to run on other systems in the past. See - Section 2.13, "Operating System-Specific Notes." Some porting - effort might be required for current versions of MySQL on these - systems. + Section 2.1, "General Installation Guidance." Some porting effort + might be required for current versions of MySQL on these systems. Not all platforms are equally well-suited for running MySQL. How well a certain platform is suited for a high-load mission-critical @@ -208,8 +145,8 @@ Important * General file system stability and performance. * Table size. If your tables are large, performance is affected - by the ability of the file system to deal with large files at - all and to deal with them efficiently. + by the ability of the file system to deal with large files and + dealing with them efficiently. * Our level of expertise here at Sun Microsystems, Inc. with the platform. If we know a platform well, we enable @@ -240,7 +177,7 @@ Important development process, multiple release series co-exist, each at a different stage of maturity: - * MySQL 5.4 and 6.0 are the current development release series. + * MySQL 5.5 is the current development release series. * MySQL 5.1 is the current General Availability (Production) release series. New releases are issued for bugfixes only; no @@ -255,9 +192,9 @@ Important has ended. Extended support for MySQL 4.1 remains available. According to the MySQL Lifecycle Policy - (http://www.mysql.com/company/legal/lifecycle/#policy) only - Security and Severity Level 1 issues are still being fixed for - MySQL 4.1. + (http://www.mysql.com/about/legal/lifecycle/) only Security + and Severity Level 1 issues are still being fixed for MySQL + 4.1. We do not believe in a complete code freeze because this prevents us from making bugfixes and other fixes that must be done. By @@ -367,13 +304,13 @@ Important * The MySQL benchmark suite This suite runs a range of common queries. It is also a test to determine whether the latest batch of optimizations - actually made the code faster. See Section 7.1.4, "The MySQL + actually made the code faster. See Section 7.1.3, "The MySQL Benchmark Suite." * The crash-me test This test tries to determine what features the database supports and what its capabilities and limitations are. See - Section 7.1.4, "The MySQL Benchmark Suite." + Section 7.1.3, "The MySQL Benchmark Suite." We also test the newest MySQL version in our internal production environment, on at least one machine. We have more than 100GB of @@ -492,21 +429,6 @@ Important as soon as possible. (We would like other companies to do this, too!) -2.1.2.4. MySQL Binaries Compiled by Sun Microsystems, Inc. - - Sun Microsystems, Inc. provides a set of binary distributions of - MySQL. In addition to binaries provided in platform-specific - package formats, we offer binary distributions for a number of - platforms in the form of compressed tar files (.tar.gz files). See - Section 2.2, "Standard MySQL Installation Using a Binary - Distribution." For Windows distributions, see Section 2.3, - "Installing MySQL on Windows." - - If you want to compile a debug version of MySQL from a source - distribution, you should add --with-debug or --with-debug=full to - the configure command used to configure the distribution and - remove any -fomit-frame-pointer options. - 2.1.3. How to Get MySQL Check our downloads page at http://dev.mysql.com/downloads/ for @@ -553,8 +475,8 @@ Important shell> md5sum package_name Example: -shell> md5sum mysql-standard-5.1.39-linux-i686.tar.gz -aaab65abbec64d5e907dcd41b8699945 mysql-standard-5.1.39-linux-i686.ta +shell> md5sum mysql-standard-5.1.41-linux-i686.tar.gz +aaab65abbec64d5e907dcd41b8699945 mysql-standard-5.1.41-linux-i686.ta r.gz You should verify that the resulting checksum (the string of @@ -728,8 +650,8 @@ pg-signature.html signature, which also is available from the download page. The signature file has the same name as the distribution file with an .asc extension, as shown by the examples in the following table. - Distribution file mysql-standard-5.1.39-linux-i686.tar.gz - Signature file mysql-standard-5.1.39-linux-i686.tar.gz.asc + Distribution file mysql-standard-5.1.41-linux-i686.tar.gz + Signature file mysql-standard-5.1.41-linux-i686.tar.gz.asc Make sure that both files are stored in the same directory and then run the following command to verify the signature for the @@ -737,7 +659,7 @@ pg-signature.html shell> gpg --verify package_name.asc Example: -shell> gpg --verify mysql-standard-5.1.39-linux-i686.tar.gz.asc +shell> gpg --verify mysql-standard-5.1.41-linux-i686.tar.gz.asc gpg: Signature made Tue 12 Jul 2005 23:35:41 EST using DSA key ID 507 2E1F5 gpg: Good signature from "MySQL Package signing key (www.mysql.com) < @@ -757,8 +679,8 @@ build(a)mysql.com>" shell> rpm --checksig package_name.rpm Example: -shell> rpm --checksig MySQL-server-5.1.39-0.glibc23.i386.rpm -MySQL-server-5.1.39-0.glibc23.i386.rpm: md5 gpg OK +shell> rpm --checksig MySQL-server-5.1.41-0.glibc23.i386.rpm +MySQL-server-5.1.41-0.glibc23.i386.rpm: md5 gpg OK Note @@ -786,22 +708,6 @@ shell> rpm --import mysql_pubkey.asc Sun Microsystems, Inc. A distribution provided by another vendor might use a layout different from those shown here. - For MySQL 5.1 on Windows, the default installation directory is - C:\Program Files\MySQL\MySQL Server 5.1. (Some Windows users - prefer to install in C:\mysql, the directory that formerly was - used as the default. However, the layout of the subdirectories - remains the same.) The installation directory has the following - subdirectories. - Directory Contents of Directory - bin Client programs and the mysqld server - data Log files, databases - Docs Manual in CHM format - examples Example programs and scripts - include Include (header) files - lib Libraries - scripts Utility scripts - share Error message files - Installations created from our Linux RPM distributions result in files under the following system directories. Directory Contents of Directory @@ -863,8100 +769,7171 @@ shell> rpm --import mysql_pubkey.asc distribution by executing the scripts/make_binary_distribution script from the top directory of the source distribution. -2.2. Standard MySQL Installation Using a Binary Distribution +2.2. Installing MySQL from Generic Binaries on Unix/Linux - The next several sections cover the installation of MySQL on - platforms where we offer packages using the native packaging - format of the respective platform. (This is also known as - performing a "binary install.") However, binary distributions of - MySQL are available for many other platforms as well. See Section - 2.9, "Installing MySQL from tar.gz Packages on Other Unix-Like - Systems," for generic installation instructions for these packages - that apply to all platforms. - - See Section 2.1, "General Installation Issues," for more - information on what other binary distributions are available and - how to obtain them. - -2.3. Installing MySQL on Windows - - A native Windows distribution of MySQL has been available since - version 3.21 and represents a sizable percentage of the daily - downloads of MySQL. This section describes the process for - installing MySQL on Windows. + This section covers the installation of MySQL binary distributions + that are provided for various platforms in the form of compressed + tar files (files with a .tar.gz extension). See Section 2.2, + "Installing MySQL from Generic Binaries on Unix/Linux," for a + detailed list. -Note + To obtain MySQL, see Section 2.1.3, "How to Get MySQL." - If you are upgrading MySQL from an existing installation older - than MySQL 4.1.5, you must first perform the procedure described - in Section 2.3.14, "Upgrading MySQL on Windows." + Sun Microsystems, Inc. provides a set of binary distributions of + MySQL. In addition to binaries provided in platform-specific + package formats, we offer binary distributions for a number of + platforms in the form of compressed tar files (.tar.gz files). For + Windows distributions, see Section 2.5, "Installing MySQL on + Windows." - To run MySQL on Windows, you need the following: + If you want to compile a debug version of MySQL from a source + distribution, you should add --with-debug or --with-debug=full to + the configure command used to configure the distribution and + remove any -fomit-frame-pointer options. - * A Windows operating system such as Windows 2000, Windows XP, - Windows Vista, Windows Server 2003, or Windows Server 2008. - Both 32-bit and 64-bit versions are supported. - A Windows operating system permits you to run the MySQL server - as a service. See Section 2.3.11, "Starting MySQL as a Windows - Service." - Generally, you should install MySQL on Windows using an - account that has administrator rights. Otherwise, you may - encounter problems with certain operations such as editing the - PATH environment variable or accessing the Service Control - Manager. Once installed, MySQL does not need to be executed - using a user with Administrator privileges. + MySQL tar file binary distributions have names of the form + mysql-VERSION-OS.tar.gz, where VERSION is a number (for example, + 5.1.41), and OS indicates the type of operating system for which + the distribution is intended (for example, pc-linux-i686). - * TCP/IP protocol support. + In addition to these generic packages, we also offer binaries in + platform-specific package formats for selected platforms. See the + platform specific sections for more information, for more + information on how to install these. - * Enough space on the hard drive to unpack, install, and create - the databases in accordance with your requirements (generally - a minimum of 200 megabytes is recommended.) + You need the following tools to install a MySQL tar file binary + distribution: - For a list of limitations within the Windows version of MySQL, see - Section D.7.3, "Windows Platform Limitations." + * GNU gunzip to uncompress the distribution. - There may also be other requirements, depending on how you plan to - use MySQL: + * A reasonable tar to unpack the distribution. GNU tar is known + to work. Some operating systems come with a preinstalled + version of tar that is known to have problems. For example, + the tar provided with early versions of Mac OS X, SunOS 4.x, + Solaris 8, Solaris 9, Solaris 10 and OpenSolaris, and HP-UX + are known to have problems with long file names. On Mac OS X, + you can use the preinstalled gnutar program. On Solaris 10 and + OpenSolaris you can use the preinstalled gtar. On other + systems with a deficient tar, you should install GNU tar + first. - * If you plan to connect to the MySQL server via ODBC, you need - a Connector/ODBC driver. See Section 21.1, "MySQL - Connector/ODBC." + If you run into problems and need to file a bug report, please use + the instructions in Section 1.6, "How to Report Bugs or Problems." - * If you plan to use MySQL server with ADO.NET applications, you - need the Connector/NET driver. See Section 21.2, "MySQL - Connector/NET." + The basic commands that you must execute to install and use a + MySQL binary distribution are: +shell> groupadd mysql +shell> useradd -g mysql mysql +shell> cd /usr/local +shell> gunzip < /path/to/mysql-VERSION-OS.tar.gz | tar xvf - +shell> ln -s full-path-to-mysql-VERSION-OS mysql +shell> cd mysql +shell> chown -R mysql . +shell> chgrp -R mysql . +shell> scripts/mysql_install_db --user=mysql +shell> chown -R root . +shell> chown -R mysql data +shell> bin/mysqld_safe --user=mysql & - * If you need tables with a size larger than 4GB, install MySQL - on an NTFS or newer file system. Don't forget to use MAX_ROWS - and AVG_ROW_LENGTH when you create tables. See Section - 12.1.17, "CREATE TABLE Syntax." +Note - MySQL for Windows is available in several distribution formats: + This procedure does not set up any passwords for MySQL accounts. + After following the procedure, proceed to Section 2.13, + "Post-Installation Setup and Testing." - * Binary distributions are available that contain a setup - program that installs everything you need so that you can - start the server immediately. Another binary distribution - format contains an archive that you simply unpack in the - installation location and then configure yourself. For - details, see Section 2.3.1, "Choosing An Installation - Package." + A more detailed version of the preceding description for + installing a binary distribution follows: - * The source distribution contains all the code and support - files for building the executables using the Visual Studio - compiler system. + 1. Add a login user and group for mysqld to run as: +shell> groupadd mysql +shell> useradd -g mysql mysql + These commands add the mysql group and the mysql user. The + syntax for useradd and groupadd may differ slightly on + different versions of Unix, or they may have different names + such as adduser and addgroup. + You might want to call the user and group something else + instead of mysql. If so, substitute the appropriate name in + the following steps. - Generally speaking, you should use a binary distribution that - includes an installer. It is simpler to use than the others, and - you need no additional tools to get MySQL up and running. The - installer for the Windows version of MySQL, combined with a GUI - Configuration Wizard, automatically installs MySQL, creates an - option file, starts the server, and secures the default user - accounts. + 2. Pick the directory under which you want to unpack the + distribution and change location into it. In the following + example, we unpack the distribution under /usr/local. (The + instructions, therefore, assume that you have permission to + create files and directories in /usr/local. If that directory + is protected, you must perform the installation as root.) +shell> cd /usr/local -Caution + 3. Obtain a distribution file using the instructions in Section + 2.1.3, "How to Get MySQL." For a given release, binary + distributions for all platforms are built from the same MySQL + source distribution. - Using virus scanning software such as Norton/Symantec Anti-Virus - on directories containing MySQL data and temporary tables can - cause issues, both in terms of the performance of MySQL and the - virus-scanning software mis-identifying the contents of the files - as containing spam. This is because of the fingerprinting - mechanism used by the virus scanning software, and the way in - which MySQL rapidly updates different files, which may be - identified as a potential security risk. - - After installing MySQL Server, it is recommended that you disable - virus scanning on the main directory (datadir) being used to store - your MySQL table data. There is usually a system built into the - virus scanning software to allow certain directories to be - specifically ignored during virus scanning. - - In addition, by default, MySQL creates temporary files in the - standard Windows temporary directory. To prevent the temporary - files also being scanned, you should configure a separate - temporary directory for MySQL temporary files and add this to the - virus scanning exclusion list. To do this, add a configuration - option for the tmpdir parameter to your my.ini configuration file. - For more information, see Section 2.3.7, "Creating an Option - File." - - The following section describes how to install MySQL on Windows - using a binary distribution. To use an installation package that - does not include an installer, follow the procedure described in - Section 2.3.5, "Installing MySQL from a Noinstall Zip Archive." To - install using a source distribution, see Section 2.10.6, - "Installing MySQL from Source on Windows." + 4. Unpack the distribution, which creates the installation + directory. Then create a symbolic link to that directory: +shell> gunzip < /path/to/mysql-VERSION-OS.tar.gz | tar xvf - +shell> ln -s full-path-to-mysql-VERSION-OS mysql + The tar command creates a directory named mysql-VERSION-OS. + The ln command makes a symbolic link to that directory. This + lets you refer more easily to the installation directory as + /usr/local/mysql. + With GNU tar, no separate invocation of gunzip is necessary. + You can replace the first line with the following alternative + command to uncompress and extract the distribution: +shell> tar zxvf /path/to/mysql-VERSION-OS.tar.gz - MySQL distributions for Windows can be downloaded from - http://dev.mysql.com/downloads/. See Section 2.1.3, "How to Get - MySQL." + 5. Change location into the installation directory: +shell> cd mysql + You will find several files and subdirectories in the mysql + directory. The most important for installation purposes are + the bin and scripts subdirectories: -2.3.1. Choosing An Installation Package + + The bin directory contains client programs and the + server. You should add the full path name of this + directory to your PATH environment variable so that your + shell finds the MySQL programs properly. See Section + 2.14, "Environment Variables." - For MySQL 5.1, there are three installation packages to choose - from when installing MySQL on Windows: + + The scripts directory contains the mysql_install_db + script used to initialize the mysql database containing + the grant tables that store the server access + permissions. - * The Essentials Package: This package has a file name similar - to mysql-essential-5.1.39-win32.msi and contains the minimum - set of files needed to install MySQL on Windows, including the - Configuration Wizard. This package does not include optional - components such as the embedded server and benchmark suite. - - * The Complete Package: This package has a file name similar to - mysql-5.1.39-win32.zip and contains all files needed for a - complete Windows installation, including the Configuration - Wizard. This package includes optional components such as the - embedded server and benchmark suite. + 6. Ensure that the distribution contents are accessible to mysql. + If you unpacked the distribution as mysql, no further action + is required. If you unpacked the distribution as root, its + contents will be owned by root. Change its ownership to mysql + by executing the following commands as root in the + installation directory: +shell> chown -R mysql . +shell> chgrp -R mysql . + The first command changes the owner attribute of the files to + the mysql user. The second changes the group attribute to the + mysql group. - * The Noinstall Archive: This package has a file name similar to - mysql-noinstall-5.1.39-win32.zip and contains all the files - found in the Complete install package, with the exception of - the Configuration Wizard. This package does not include an - automated installer, and must be manually installed and - configured. - - The Essentials package is recommended for most users. It is - provided as an .msi file for use with the Windows Installer. The - Complete and Noinstall distributions are packaged as Zip archives. - To use them, you must have a tool that can unpack .zip files. + 7. If you have not installed MySQL before, you must create the + MySQL data directory and initialize the grant tables: +shell> scripts/mysql_install_db --user=mysql + If you run the command as root, include the --user option as + shown. If you run the command while logged in as that user, + you can omit the --user option. + The command should create the data directory and its contents + with mysql as the owner. + After creating or updating the grant tables, you need to + restart the server manually. - Your choice of install package affects the installation process - you must follow. If you choose to install either the Essentials or - Complete install packages, see Section 2.3.2, "Installing MySQL - with the Automated Installer." If you choose to install MySQL from - the Noinstall archive, see Section 2.3.5, "Installing MySQL from a - Noinstall Zip Archive." - -2.3.2. Installing MySQL with the Automated Installer - - New MySQL users can use the MySQL Installation Wizard and MySQL - Configuration Wizard to install MySQL on Windows. These are - designed to install and configure MySQL in such a way that new - users can immediately get started using MySQL. + 8. Most of the MySQL installation can be owned by root if you + like. The exception is that the data directory must be owned + by mysql. To accomplish this, run the following commands as + root in the installation directory: +shell> chown -R root . +shell> chown -R mysql data - The MySQL Installation Wizard and MySQL Configuration Wizard are - available in the Essentials and Complete install packages. They - are recommended for most standard MySQL installations. Exceptions - include users who need to install multiple instances of MySQL on a - single server host and advanced users who want complete control of - server configuration. + 9. If you want MySQL to start automatically when you boot your + machine, you can copy support-files/mysql.server to the + location where your system has its startup files. More + information can be found in the support-files/mysql.server + script itself and in Section 2.13.1.2, "Starting and Stopping + MySQL Automatically." + 10. You can set up new accounts using the bin/mysql_setpermission + script if you install the DBI and DBD::mysql Perl modules. See + Section 4.6.14, "mysql_setpermission --- Interactively Set + Permissions in Grant Tables." For Perl module installation + instructions, see Section 2.15, "Perl Installation Notes." + 11. If you would like to use mysqlaccess and have the MySQL + distribution in some nonstandard location, you must change the + location where mysqlaccess expects to find the mysql client. + Edit the bin/mysqlaccess script at approximately line 18. + Search for a line that looks like this: +$MYSQL = '/usr/local/bin/mysql'; # path to mysql executable + Change the path to reflect the location where mysql actually + is stored on your system. If you do not do this, a Broken pipe + error will occur when you run mysqlaccess. -2.3.3. Using the MySQL Installation Wizard + After everything has been unpacked and installed, you should test + your distribution. To start the MySQL server, use the following + command: +shell> bin/mysqld_safe --user=mysql & - MySQL Installation Wizard is an installer for the MySQL server - that uses the latest installer technologies for Microsoft Windows. - The MySQL Installation Wizard, in combination with the MySQL - Configuration Wizard, allows a user to install and configure a - MySQL server that is ready for use immediately after installation. + If you run the command as root, you must use the --user option as + shown. The value of the option is the name of the login account + that you created in the first step to use for running the server. + If you run the command while logged in as mysql, you can omit the + --user option. - The MySQL Installation Wizard is the standard installer for all - MySQL server distributions, version 4.1.5 and higher. Users of - previous versions of MySQL need to shut down and remove their - existing MySQL installations manually before installing MySQL with - the MySQL Installation Wizard. See Section 2.3.3.6, "Upgrading - MySQL with the Installation Wizard," for more information on - upgrading from a previous version. + If the command fails immediately and prints mysqld ended, you can + find some information in the host_name.err file in the data + directory. - Microsoft has included an improved version of their Microsoft - Windows Installer (MSI) in the recent versions of Windows. MSI has - become the de-facto standard for application installations on - Windows 2000, Windows XP, and Windows Server 2003. The MySQL - Installation Wizard makes use of this technology to provide a - smoother and more flexible installation process. + More information about mysqld_safe is given in Section 4.3.2, + "mysqld_safe --- MySQL Server Startup Script." - The Microsoft Windows Installer Engine was updated with the - release of Windows XP; those using a previous version of Windows - can reference this Microsoft Knowledge Base article - (http://support.microsoft.com/default.aspx?scid=kb;EN-US;292539) - for information on upgrading to the latest version of the Windows - Installer Engine. +Note - In addition, Microsoft has introduced the WiX (Windows Installer - XML) toolkit recently. This is the first highly acknowledged Open - Source project from Microsoft. We have switched to WiX because it - is an Open Source project and it allows us to handle the complete - Windows installation process in a flexible manner using scripts. + The accounts that are listed in the MySQL grant tables initially + have no passwords. After starting the server, you should set up + passwords for them using the instructions in Section 2.13, + "Post-Installation Setup and Testing." - Improving the MySQL Installation Wizard depends on the support and - feedback of users like you. If you find that the MySQL - Installation Wizard is lacking some feature important to you, or - if you discover a bug, please report it in our bugs database using - the instructions given in Section 1.6, "How to Report Bugs or - Problems." +2.3. MySQL Installation Using a Source Distribution -2.3.3.1. Downloading and Starting the MySQL Installation Wizard + Before you proceed with an installation from source, first check + whether our binary is available for your platform and whether it + works for you. We put a great deal of effort into ensuring that + our binaries are built with the best possible options. - The MySQL installation packages can be downloaded from - http://dev.mysql.com/downloads/. If the package you download is - contained within a Zip archive, you need to extract the archive - first. - -Note + To obtain a source distribution for MySQL, Section 2.1.3, "How to + Get MySQL." If you want to build MySQL from source on Windows, see + Section 2.5.10, "Installing MySQL from Source on Windows." - If you are installing on Windows Vista it is best to open a - network port before beginning the installation. To do this, first - ensure that you are logged in as an Administrator, go to the - Control Panel, and double click the Windows Firewall icon. Choose - the Allow a program through Windows Firewall option and click the - Add port button. Enter MySQL into the Name text box and 3306 (or - the port of your choice) into the Port number text box. Also - ensure that the TCP protocol radio button is selected. If you - wish, you can also limit access to the MySQL server by choosing - the Change scope button. Confirm your choices by clicking the OK - button. If you do not open a port prior to installation, you - cannot configure the MySQL server immediately after installation. - Additionally, when running the MySQL Installation Wizard on - Windows Vista, ensure that you are logged in as a user with - administrative rights. + MySQL source distributions are provided as compressed tar archives + and have names of the form mysql-VERSION.tar.gz, where VERSION is + a number like 5.1.41. - The process for starting the wizard depends on the contents of the - installation package you download. If there is a setup.exe file - present, double-click it to start the installation process. If - there is an .msi file present, double-click it to start the - installation process. + You need the following tools to build and install MySQL from + source: -2.3.3.2. Choosing an Install Type + * GNU gunzip to uncompress the distribution. - There are three installation types available: Typical, Complete, - and Custom. + * A reasonable tar to unpack the distribution. GNU tar is known + to work. Some operating systems come with a preinstalled + version of tar that is known to have problems. For example, + the tar provided with early versions of Mac OS X, SunOS 4.x, + Solaris 8, Solaris 9, Solaris 10 and OpenSolaris, and HP-UX + are known to have problems with long file names. On Mac OS X, + you can use the preinstalled gnutar program. On Solaris 10 and + OpenSolaris you can use the preinstalled gtar. On other + systems with a deficient tar, you should install GNU tar + first. - The Typical installation type installs the MySQL server, the mysql - command-line client, and the command-line utilities. The - command-line clients and utilities include mysqldump, myisamchk, - and several other tools to help you manage the MySQL server. + * A working ANSI C++ compiler. gcc 2.95.2 or later, SGI C++, and + SunPro C++ are some of the compilers that are known to work. + libg++ is not needed when using gcc. gcc 2.7.x has a bug that + makes it impossible to compile some perfectly legal C++ files, + such as sql/sql_base.cc. If you have only gcc 2.7.x, you must + upgrade your gcc to be able to compile MySQL. gcc 2.8.1 is + also known to have problems on some platforms, so it should be + avoided if a newer compiler exists for the platform. gcc + 2.95.2 or later is recommended. - The Complete installation type installs all components included in - the installation package. The full installation package includes - components such as the embedded server library, the benchmark - suite, support scripts, and documentation. + * A good make program. GNU make is always recommended and is + sometimes required. (BSD make fails, and vendor-provided make + implementations may fail as well.) If you have problems, use + GNU make 3.75 or newer. - The Custom installation type gives you complete control over which - packages you wish to install and the installation path that is - used. See Section 2.3.3.3, "The Custom Install Dialog," for more - information on performing a custom install. + * libtool 1.5.24 or later is also recommended. - If you choose the Typical or Complete installation types and click - the Next button, you advance to the confirmation screen to verify - your choices and begin the installation. If you choose the Custom - installation type and click the Next button, you advance to the - custom installation dialog, described in Section 2.3.3.3, "The - Custom Install Dialog." + If you are using a version of gcc recent enough to understand the + -fno-exceptions option, it is very important that you use this + option. Otherwise, you may compile a binary that crashes randomly. + Also use -felide-constructors and -fno-rtti along with + -fno-exceptions. When in doubt, do the following: +CFLAGS="-O3" CXX=gcc CXXFLAGS="-O3 -felide-constructors \ + -fno-exceptions -fno-rtti" ./configure \ + --prefix=/usr/local/mysql --enable-assembler \ + --with-mysqld-ldflags=-all-static -2.3.3.3. The Custom Install Dialog + On most systems, this gives you a fast and stable binary. - If you wish to change the installation path or the specific - components that are installed by the MySQL Installation Wizard, - choose the Custom installation type. + If you run into problems and need to file a bug report, please use + the instructions in Section 1.6, "How to Report Bugs or Problems." - A tree view on the left side of the custom install dialog lists - all available components. Components that are not installed have a - red X icon; components that are installed have a gray icon. To - change whether a component is installed, click on that component's - icon and choose a new option from the drop-down list that appears. +2.3.1. Source Installation Overview - You can change the default installation path by clicking the - Change... button to the right of the displayed installation path. + The basic commands that you must execute to install a MySQL source + distribution are: +shell> groupadd mysql +shell> useradd -g mysql mysql +shell> gunzip < mysql-VERSION.tar.gz | tar -xvf - +shell> cd mysql-VERSION +shell> ./configure --prefix=/usr/local/mysql +shell> make +shell> make install +shell> cp support-files/my-medium.cnf /etc/my.cnf +shell> cd /usr/local/mysql +shell> chown -R mysql . +shell> chgrp -R mysql . +shell> bin/mysql_install_db --user=mysql +shell> chown -R root . +shell> chown -R mysql var +shell> bin/mysqld_safe --user=mysql & - After choosing your installation components and installation path, - click the Next button to advance to the confirmation dialog. + If you start from a source RPM, do the following: +shell> rpmbuild --rebuild --clean MySQL-VERSION.src.rpm -2.3.3.4. The Confirmation Dialog + This makes a binary RPM that you can install. For older versions + of RPM, you may have to replace the command rpmbuild with rpm + instead. - Once you choose an installation type and optionally choose your - installation components, you advance to the confirmation dialog. - Your installation type and installation path are displayed for you - to review. +Note - To install MySQL if you are satisfied with your settings, click - the Install button. To change your settings, click the Back - button. To exit the MySQL Installation Wizard without installing - MySQL, click the Cancel button. + This procedure does not set up any passwords for MySQL accounts. + After following the procedure, proceed to Section 2.13, + "Post-Installation Setup and Testing," for post-installation setup + and testing. - After installation is complete, you have the option of registering - with the MySQL web site. Registration gives you access to post in - the MySQL forums at forums.mysql.com (http://forums.mysql.com) - along with the ability to report bugs at bugs.mysql.com - (http://bugs.mysql.com) and to subscribe to our newsletter. The - final screen of the installer provides a summary of the - installation and gives you the option to launch the MySQL - Configuration Wizard, which you can use to create a configuration - file, install the MySQL service, and configure security settings. + A more detailed version of the preceding description for + installing MySQL from a source distribution follows: -2.3.3.5. Changes Made by MySQL Installation Wizard + 1. Add a login user and group for mysqld to run as: +shell> groupadd mysql +shell> useradd -g mysql mysql + These commands add the mysql group and the mysql user. The + syntax for useradd and groupadd may differ slightly on + different versions of Unix, or they may have different names + such as adduser and addgroup. + You might want to call the user and group something else + instead of mysql. If so, substitute the appropriate name in + the following steps. - Once you click the Install button, the MySQL Installation Wizard - begins the installation process and makes certain changes to your - system which are described in the sections that follow. + 2. Perform the following steps as the mysql user, except as + noted. - Changes to the Registry + 3. Pick the directory under which you want to unpack the + distribution and change location into it. - The MySQL Installation Wizard creates one Windows registry key in - a typical install situation, located in - HKEY_LOCAL_MACHINE\SOFTWARE\MySQL AB. + 4. Obtain a distribution file using the instructions in Section + 2.1.3, "How to Get MySQL." - The MySQL Installation Wizard creates a key named after the major - version of the server that is being installed, such as MySQL - Server 5.1. It contains two string values, Location and Version. - The Location string contains the path to the installation - directory. In a default installation it contains C:\Program - Files\MySQL\MySQL Server 5.1\. The Version string contains the - release number. For example, for an installation of MySQL Server - 5.1.39, the key contains a value of 5.1.39. + 5. Unpack the distribution into the current directory: +shell> gunzip < /path/to/mysql-VERSION.tar.gz | tar xvf - + This command creates a directory named mysql-VERSION. + With GNU tar, no separate invocation of gunzip is necessary. + You can use the following alternative command to uncompress + and extract the distribution: +shell> tar zxvf /path/to/mysql-VERSION-OS.tar.gz - These registry keys are used to help external tools identify the - installed location of the MySQL server, preventing a complete scan - of the hard-disk to determine the installation path of the MySQL - server. The registry keys are not required to run the server, and - if you install MySQL using the noinstall Zip archive, the registry - keys are not created. + 6. Change location into the top-level directory of the unpacked + distribution: +shell> cd mysql-VERSION + Note that currently you must configure and build MySQL from + this top-level directory. You cannot build it in a different + directory. - Changes to the Start Menu + 7. Configure the release and compile everything: +shell> ./configure --prefix=/usr/local/mysql +shell> make + When you run configure, you might want to specify other + options. Run ./configure --help for a list of options. Section + 2.3.2, "Typical configure Options," discusses some of the more + useful options. + If configure fails and you are going to send mail to a MySQL + mailing list to ask for assistance, please include any lines + from config.log that you think can help solve the problem. + Also include the last couple of lines of output from + configure. To file a bug report, please use the instructions + in Section 1.6, "How to Report Bugs or Problems." + If the compile fails, see Section 2.3.4, "Dealing with + Problems Compiling MySQL," for help. - The MySQL Installation Wizard creates a new entry in the Windows - Start menu under a common MySQL menu heading named after the major - version of MySQL that you have installed. For example, if you - install MySQL 5.1, the MySQL Installation Wizard creates a MySQL - Server 5.1 section in the Start menu. + 8. Install the distribution: +shell> make install + You might need to run this command as root. + If you want to set up an option file, use one of those present + in the support-files directory as a template. For example: +shell> cp support-files/my-medium.cnf /etc/my.cnf + You might need to run this command as root. + If you want to configure support for InnoDB tables, you should + edit the /etc/my.cnf file, remove the # character before the + option lines that start with innodb_..., and modify the option + values to be what you want. See Section 4.2.3.3, "Using Option + Files," and Section 13.6.2, "InnoDB Configuration." - The following entries are created within the new Start menu - section: + 9. Change location into the installation directory: +shell> cd /usr/local/mysql + 10. If you ran the make install command as root, the installed + files will be owned by root. Ensure that the installation is + accessible to mysql by executing the following commands as + root in the installation directory: +shell> chown -R mysql . +shell> chgrp -R mysql . + The first command changes the owner attribute of the files to + the mysql user. The second changes the group attribute to the + mysql group. + 11. If you have not installed MySQL before, you must create the + MySQL data directory and initialize the grant tables: +shell> bin/mysql_install_db --user=mysql + If you run the command as root, include the --user option as + shown. If you run the command while logged in as mysql, you + can omit the --user option. + The command should create the data directory and its contents + with mysql as the owner. + After using mysql_install_db to create the grant tables for + MySQL, you must restart the server manually. The mysqld_safe + command to do this is shown in a later step. + 12. Most of the MySQL installation can be owned by root if you + like. The exception is that the data directory must be owned + by mysql. To accomplish this, run the following commands as + root in the installation directory: +shell> chown -R root . +shell> chown -R mysql var + 13. If you want MySQL to start automatically when you boot your + machine, you can copy support-files/mysql.server to the + location where your system has its startup files. More + information can be found in the support-files/mysql.server + script itself; see also Section 2.13.1.2, "Starting and + Stopping MySQL Automatically." + 14. You can set up new accounts using the bin/mysql_setpermission + script if you install the DBI and DBD::mysql Perl modules. See + Section 4.6.14, "mysql_setpermission --- Interactively Set + Permissions in Grant Tables." For Perl module installation + instructions, see Section 2.15, "Perl Installation Notes." - * MySQL Command Line Client: This is a shortcut to the mysql - command-line client and is configured to connect as the root - user. The shortcut prompts for a root user password when you - connect. + After everything has been installed, you should test your + distribution. To start the MySQL server, use the following + command: +shell> /usr/local/mysql/bin/mysqld_safe --user=mysql & - * MySQL Server Instance Config Wizard: This is a shortcut to the - MySQL Configuration Wizard. Use this shortcut to configure a - newly installed server, or to reconfigure an existing server. + If you run the command as root, you should use the --user option + as shown. The value of the option is the name of the login account + that you created in the first step to use for running the server. + If you run the command while logged in as that user, you can omit + the --user option. - * MySQL Documentation: This is a link to the MySQL server - documentation that is stored locally in the MySQL server - installation directory. This option is not available when the - MySQL server is installed using the Essentials installation - package. - - Changes to the File System - - The MySQL Installation Wizard by default installs the MySQL 5.1 - server to C:\Program Files\MySQL\MySQL Server 5.1, where Program - Files is the default location for applications in your system, and - 5.1 is the major version of your MySQL server. This is the - recommended location for the MySQL server, replacing the former - default location C:\mysql. + If the command fails immediately and prints mysqld ended, you can + find some information in the host_name.err file in the data + directory. - By default, all MySQL applications are stored in a common - directory at C:\Program Files\MySQL, where Program Files is the - default location for applications in your Windows installation. A - typical MySQL installation on a developer machine might look like - this: -C:\Program Files\MySQL\MySQL Server 5.1 -C:\Program Files\MySQL\MySQL Workbench 5.1 OSS + More information about mysqld_safe is given in Section 4.3.2, + "mysqld_safe --- MySQL Server Startup Script." - This approach makes it easier to manage and maintain all MySQL - applications installed on a particular system. +Note - In MySQL 5.1.23 and earlier, the default location for the data - files used by MySQL is located within the corresponding MySQL - Server installation directory. For MySQL 5.1.24 and later, the - default location of the data directory is the AppData directory - configured for the user that installed the MySQL application. + The accounts that are listed in the MySQL grant tables initially + have no passwords. After starting the server, you should set up + passwords for them using the instructions in Section 2.13, + "Post-Installation Setup and Testing." -2.3.3.6. Upgrading MySQL with the Installation Wizard +2.3.2. Typical configure Options - The MySQL Installation Wizard can perform server upgrades - automatically using the upgrade capabilities of MSI. That means - you do not need to remove a previous installation manually before - installing a new release. The installer automatically shuts down - and removes the previous MySQL service before installing the new - version. + The configure script gives you a great deal of control over how + you configure a MySQL source distribution. Typically you do this + using options on the configure command line. You can also affect + configure using certain environment variables. See Section 2.14, + "Environment Variables." For a full list of options supported by + configure, run this command: +shell> ./configure --help - Automatic upgrades are available only when upgrading between - installations that have the same major and minor version numbers. - For example, you can upgrade automatically from MySQL 4.1.5 to - MySQL 4.1.6, but not from MySQL 5.0 to MySQL 5.1. + A list of the available configure options is provided in the table + below. - See Section 2.3.14, "Upgrading MySQL on Windows." + Table 2.1. Build (configure) Reference + Formats Description Default Introduced Removed + --bindir=DIR User executables EPREFIX/bin + --build=BUILD Configure for building on BUILD guessed + --cache-file=FILE Cache test results in FILE disabled + -C Alias for `--cache-file=config.cache' + --config-cache + --datadir=DIR Read-only architecture-independent data PREFIX/share -2.3.4. MySQL Server Instance Configuration Wizard + --disable-FEATURE Do not include FEATURE + --disable-dependency-tracking Disable dependency tracking + --disable-grant-options Disable GRANT options + --disable-largefile Omit support for large files + --disable-libtool-lock Disable libtool lock + --disable-thread-safe-client Compile the client without threads + 5.1.7 + --enable-FEATURE Enable FEATURE + --enable-assembler Use assembler versions of some string functions + if available + --enable-debug-sync Compile in Debug Sync facility 5.1.41 + --enable-dependency-tracking Do not reject slow dependency + extractors + --enable-fast-install Optimize for fast installation yes + --enable-local-infile Enable LOAD DATA LOCAL INFILE disabled + --enable-shared Build shared libraries yes + --enable-static Build static libraries yes + --enable-thread-safe-client Compile the client with threads + --exec-prefix=EPREFIX Install architecture-dependent files in + EPREFIX + -h Display this help and exit + --help + --help=short Display options specific to this package + --help=recursive Display the short help of all the included + packages + --host=HOST Cross-compile to build programs to run on HOST + --includedir=DIR C header files PREFIX/include + --infodir=DIR Info documentation PREFIX/info + --libdir=DIR Object code libraries EPREFIX/lib + --libexecdir=DIR Program executables EPREFIX/libexec + --localstatedir=DIR Modifiable single-machine data PREFIX/var + --mandir=DIR man documentation PREFIX/man + -n Do not create output files + --no-create + --oldincludedir=DIR C header files for non-gcc /usr/include + --prefix=PREFIX Install architecture-independent files in PREFIX - The MySQL Server Instance Configuration Wizard helps automate the - process of configuring your server. It creates a custom MySQL - configuration file (my.ini or my.cnf) by asking you a series of - questions and then applying your responses to a template to - generate the configuration file that is tuned to your - installation. + --program-prefix=PREFIX Prepend PREFIX to installed program names - The MySQL Server Instance Configuration Wizard is included with - the MySQL 5.1 server. The MySQL Server Instance Configuration - Wizard is only available for Windows. - -2.3.4.1. Starting the MySQL Server Instance Configuration Wizard - - The MySQL Server Instance Configuration Wizard is normally started - as part of the installation process. You should only need to run - the MySQL Server Instance Configuration Wizard again when you need - to change the configuration parameters of your server. + --program-suffix=SUFFIX Append SUFFIX to installed program names - If you chose not to open a port prior to installing MySQL on - Windows Vista, you can choose to use the MySQL Server - Configuration Wizard after installation. However, you must open a - port in the Windows Firewall. To do this see the instructions - given in Section 2.3.3.1, "Downloading and Starting the MySQL - Installation Wizard." Rather than opening a port, you also have - the option of adding MySQL as a program that bypasses the Windows - Firewall. One or the other option is sufficient --- you need not - do both. Additionally, when running the MySQL Server Configuration - Wizard on Windows Vista ensure that you are logged in as a user - with administrative rights. - MySQL Server Instance Configuration Wizard - - You can launch the MySQL Configuration Wizard by clicking the - MySQL Server Instance Config Wizard entry in the MySQL section of - the Windows Start menu. + --program-transform-name=PROGRAM run sed PROGRAM on installed + program names + -q Do not print `checking...' messages + --quiet + --sbindir=DIR System admin executables EPREFIX/sbin + --sharedstatedir=DIR Modifiable architecture-independent data + PREFIX/com + --srcdir=DIR Find the sources in DIR configure directory or .. + --sysconfdir=DIR Read-only single-machine data PREFIX/etc + --target=TARGET Configure for building compilers for TARGET + -V Display version information and exit + --version + --with-PACKAGE Use PACKAGE + --with-archive-storage-engine Enable the Archive Storage Engine no - Alternatively, you can navigate to the bin directory of your MySQL - installation and launch the MySQLInstanceConfig.exe file directly. + --with-atomic-ops Implement atomic operations using pthread + rwlocks or atomic CPU instructions for multi-processor 5.1.12 + --with-berkeley-db Use BerkeleyDB located in DIR no + --with-berkeley-db-includes Find Berkeley DB headers in DIR + --with-berkeley-db-libs Find Berkeley DB libraries in DIR + --with-big-tables Support tables with more than 4 G rows even on + 32 bit platforms + --with-blackhole-storage-engine Enable the Blackhole Storage + Engine no + --with-charset Default character set + --with-client-ldflags Extra linking arguments for clients + --with-collation Default collation + --with-comment Comment about compilation environment + --with-csv-storage-engine Enable the CSV Storage Engine yes + --with-darwin-mwcc Use Metrowerks CodeWarrior wrappers on OS + X/Darwin + --with-debug Add debug code 5.1.7 + --with-debug=full Add debug code (adds memory checker, very slow) - The MySQL Server Instance Configuration Wizard places the my.ini - file in the installation directory for the MySQL server. This - helps associate configuration files with particular server - instances. + --with-embedded-privilege-control Build parts to check user's + privileges (only affects embedded library) + --with-embedded-server Build the embedded server + --with-error-inject Enable error injection in MySQL Server + 5.1.11 + --with-example-storage-engine Enable the Example Storage Engine no - To ensure that the MySQL server knows where to look for the my.ini - file, an argument similar to this is passed to the MySQL server as - part of the service installation: ---defaults-file="C:\Program Files\MySQL\MySQL Server 5.1\my.ini" + --with-extra-charsets Use charsets in addition to default + --with-fast-mutexes Compile with fast mutexes enabled 5.1.5 + --with-federated-storage-engine Enable federated storage engine no + 5.1.3 5.1.9 + --with-gnu-ld Assume the C compiler uses GNU ld no + --with-innodb Enable innobase storage engine no 5.1.3 5.1.9 + --with-lib-ccflags Extra CC options for libraries + --with-libwrap=DIR Compile in libwrap (tcp_wrappers) support + --with-low-memory Try to use less memory to compile to avoid + memory limitations + --with-machine-type Set the machine type, like "powerpc" + --with-max-indexes=N Sets the maximum number of indexes per table + 64 + --with-mysqld-ldflags Extra linking arguments for mysqld + --with-mysqld-libs Extra libraries to link with for mysqld + --with-mysqld-user What user the mysqld daemon shall be run as - Here, C:\Program Files\MySQL\MySQL Server 5.1 is replaced with the - installation path to the MySQL Server. The --defaults-file option - instructs the MySQL server to read the specified file for - configuration options when it starts. + --with-mysqlmanager Build the mysqlmanager binary Build if server + is built + --with-named-curses-libs Use specified curses libraries + --with-named-thread-libs Use specified thread libraries + --with-ndb-ccflags Extra CC options for ndb compile + --with-ndb-docs Include the NDB Cluster ndbapi and mgmapi + documentation + --with-ndb-port Port for NDB Cluster management server + --with-ndb-port-base Port for NDB Cluster management server + --with-ndb-sci=DIR Provide MySQL with a custom location of sci + library + --with-ndb-test Include the NDB Cluster ndbapi test programs + --with-ndbcluster Include the NDB Cluster table handler no + --with-openssl=DIR Include the OpenSSL support + --with-openssl-includes Find OpenSSL headers in DIR + --with-openssl-libs Find OpenSSL libraries in DIR + --with-other-libc=DIR Link against libc and other standard + libraries installed in the specified nonstandard location + --with-pic Try to use only PIC/non-PIC objects Use both + --with-plugin-PLUGIN Forces the named plugin to be linked into + mysqld statically 5.1.11 + --with-plugins Plugins to include in mysqld none 5.1.11 + --with-pstack Use the pstack backtrace library + --with-pthread Force use of pthread library + --with-row-based-replication Include row-based replication 5.1.5 + 5.1.6 + --with-server-suffix Append value to the version string + --with-ssl=DIR Include SSL support 5.1.11 + --with-system-type Set the system type, like "sun-solaris10" + --with-tags Include additional configurations automatic + --with-tcp-port Which port to use for MySQL services 3306 + --with-unix-socket-path Where to put the unix-domain socket + --with-yassl Include the yaSSL support + --with-zlib-dir=no|bundled|DIR Provide MySQL with a custom + location of compression library + --without-PACKAGE Do not use PACKAGE + --without-bench Skip building of the benchmark suite + --without-debug Build a production version without debugging code - Apart from making changes to the my.ini file by running the MySQL - Server Instance Configuration Wizard again, you can modify it by - opening it with a text editor and making any necessary changes. - You can also modify the server configuration with the MySQL - Administrator (http://www.mysql.com/products/administrator/) - utility. For more information about server configuration, see - Section 5.1.2, "Server Command Options." + --without-docs Skip building of the documentation + --without-extra-tools Skip building utilities in the tools + directory + --without-geometry Do not build geometry-related parts + --without-libedit Use system libedit instead of bundled copy + --without-man Skip building of the man pages + --without-ndb-binlog Disable ndb binlog 5.1.6 + --without-ndb-debug Disable special ndb debug features + --without-plugin-PLUGIN Exclude PLUGIN 5.1.11 + --without-query-cache Do not build query cache + --without-readline Use system readline instead of bundled copy - MySQL clients and utilities such as the mysql and mysqldump - command-line clients are not able to locate the my.ini file - located in the server installation directory. To configure the - client and utility applications, create a new my.ini file in the - Windows installation directory (for example, C:\WINDOWS). + --without-row-based-replication Don't include row-based + replication 5.1.7 5.1.14 + --without-server Only build the client + --without-uca Skip building of the national Unicode collations - Under Windows Server 2003, Windows Server 2000, Windows XP, and - Windows Vista MySQL Server Instance Configuration Wizard will - configure MySQL to work as a Windows service. To start and stop - MySQL you use the Services application that is supplied as part of - the Windows Administrator Tools. - -2.3.4.2. Choosing a Maintenance Option - - If the MySQL Server Instance Configuration Wizard detects an - existing configuration file, you have the option of either - reconfiguring your existing server, or removing the server - instance by deleting the configuration file and stopping and - removing the MySQL service. + Some of the configure options available are described here. For + options that may be of use if you have difficulties building + MySQL, see Section 2.3.4, "Dealing with Problems Compiling MySQL." - To reconfigure an existing server, choose the Re-configure - Instance option and click the Next button. Any existing - configuration file is not overwritten, but renamed (within the - same directory) using a timestamp (Windows) or sequential number - (Linux). To remove the existing server instance, choose the Remove - Instance option and click the Next button. + * To compile just the MySQL client libraries and client programs + and not the server, use the --without-server option: +shell> ./configure --without-server + If you have no C++ compiler, some client programs such as + mysql cannot be compiled because they require C++.. In this + case, you can remove the code in configure that tests for the + C++ compiler and then run ./configure with the + --without-server option. The compile step should still try to + build all clients, but you can ignore any warnings about files + such as mysql.cc. (If make stops, try make -k to tell it to + continue with the rest of the build even if errors occur.) - If you choose the Remove Instance option, you advance to a - confirmation window. Click the Execute button. The MySQL Server - Configuration Wizard stops and removes the MySQL service, and then - deletes the configuration file. The server installation and its - data folder are not removed. + * If you want to build the embedded MySQL library (libmysqld.a), + use the --with-embedded-server option. - If you choose the Re-configure Instance option, you advance to the - Configuration Type dialog where you can choose the type of - installation that you wish to configure. - -2.3.4.3. Choosing a Configuration Type - - When you start the MySQL Server Instance Configuration Wizard for - a new MySQL installation, or choose the Re-configure Instance - option for an existing installation, you advance to the - Configuration Type dialog. - MySQL Server Instance Configuration Wizard: Configuration Type - - There are two configuration types available: Detailed - Configuration and Standard Configuration. The Standard - Configuration option is intended for new users who want to get - started with MySQL quickly without having to make many decisions - about server configuration. The Detailed Configuration option is - intended for advanced users who want more fine-grained control - over server configuration. - - If you are new to MySQL and need a server configured as a - single-user developer machine, the Standard Configuration should - suit your needs. Choosing the Standard Configuration option causes - the MySQL Configuration Wizard to set all configuration options - automatically with the exception of Service Options and Security - Options. + * If you don't want your log files and database directories + located under /usr/local/var, use a configure command + something like one of these: +shell> ./configure --prefix=/usr/local/mysql +shell> ./configure --prefix=/usr/local \ + --localstatedir=/usr/local/mysql/data + The first command changes the installation prefix so that + everything is installed under /usr/local/mysql rather than the + default of /usr/local. The second command preserves the + default installation prefix, but overrides the default + location for database directories (normally /usr/local/var) + and changes it to /usr/local/mysql/data. + You can also specify the installation directory and data + directory locations at server startup time by using the + --basedir and --datadir options. These can be given on the + command line or in an MySQL option file, although it is more + common to use an option file. See Section 4.2.3.3, "Using + Option Files." - The Standard Configuration sets options that may be incompatible - with systems where there are existing MySQL installations. If you - have an existing MySQL installation on your system in addition to - the installation you wish to configure, the Detailed Configuration - option is recommended. + * This option specifies the port number on which the server + listens for TCP/IP connections. The default is port 3306. To + listen on a different port, use a configure command like this: +shell> ./configure --with-tcp-port=3307 - To complete the Standard Configuration, please refer to the - sections on Service Options and Security Options in Section - 2.3.4.10, "The Service Options Dialog," and Section 2.3.4.11, "The - Security Options Dialog," respectively. + * If you are using Unix and you want the MySQL socket file + location to be somewhere other than the default location + (normally in the directory /tmp or /var/run), use a configure + command like this: +shell> ./configure \ + --with-unix-socket-path=/usr/local/mysql/tmp/mysql.sock + The socket file name must be an absolute path name. You can + also change the location of mysql.sock at server startup by + using a MySQL option file. See Section B.5.4.5, "How to + Protect or Change the MySQL Unix Socket File." -2.3.4.4. The Server Type Dialog + * If you want to compile statically linked programs (for + example, to make a binary distribution, to get better + performance, or to work around problems with some Red Hat + Linux distributions), run configure like this: +shell> ./configure --with-client-ldflags=-all-static \ + --with-mysqld-ldflags=-all-static - There are three different server types available to choose from. - The server type that you choose affects the decisions that the - MySQL Server Instance Configuration Wizard makes with regard to - memory, disk, and processor usage. - MySQL Server Instance Configuration Wizard: Server Type + * If you are using gcc and don't have libg++ or libstdc++ + installed, you can tell configure to use gcc as your C++ + compiler: +shell> CC=gcc CXX=gcc ./configure + When you use gcc as your C++ compiler, it does not attempt to + link in libg++ or libstdc++. This may be a good thing to do + even if you have those libraries installed. Some versions of + them have caused strange problems for MySQL users in the past. + The following list indicates some compilers and environment + variable settings that are commonly used with each one. - * Developer Machine: Choose this option for a typical desktop - workstation where MySQL is intended only for personal use. It - is assumed that many other desktop applications are running. - The MySQL server is configured to use minimal system - resources. + + gcc 2.7.2: +CC=gcc CXX=gcc CXXFLAGS="-O3 -felide-constructors" - * Server Machine: Choose this option for a server machine where - the MySQL server is running alongside other server - applications such as FTP, email, and Web servers. The MySQL - server is configured to use a moderate portion of the system - resources. + + gcc 2.95.2: +CFLAGS="-O3 -mpentiumpro" CXX=gcc CXXFLAGS="-O3 -mpentiumpro \ +-felide-constructors -fno-exceptions -fno-rtti" - * Dedicated MySQL Server Machine: Choose this option for a - server machine that is intended to run only the MySQL server. - It is assumed that no other applications are running. The - MySQL server is configured to use all available system - resources. + + pgcc 2.90.29 or newer: +CFLAGS="-O3 -mpentiumpro -mstack-align-double" CXX=gcc \ +CXXFLAGS="-O3 -mpentiumpro -mstack-align-double \ +-felide-constructors -fno-exceptions -fno-rtti" + In most cases, you can get a reasonably optimized MySQL binary + by using the options from the preceding list and adding the + following options to the configure line: +--prefix=/usr/local/mysql --enable-assembler \ +--with-mysqld-ldflags=-all-static + The full configure line would, in other words, be something + like the following for all recent gcc versions: +CFLAGS="-O3 -mpentiumpro" CXX=gcc CXXFLAGS="-O3 -mpentiumpro \ +-felide-constructors -fno-exceptions -fno-rtti" ./configure \ +--prefix=/usr/local/mysql --enable-assembler \ +--with-mysqld-ldflags=-all-static + The binaries we provide on the MySQL Web site at + http://dev.mysql.com/downloads/ are all compiled with full + optimization and should be perfect for most users. See Section + 2.2, "Installing MySQL from Generic Binaries on Unix/Linux." + There are some configuration settings you can tweak to build + an even faster binary, but these are only for advanced users. + See Section 7.5.1, "How Compiling and Linking Affects the + Speed of MySQL." + If the build fails and produces errors about your compiler or + linker not being able to create the shared library + libmysqlclient.so.N (where N is a version number), you can + work around this problem by giving the --disable-shared option + to configure. In this case, configure does not build a shared + libmysqlclient.so.N library. -Note + * By default, MySQL uses the latin1 (cp1252 West European) + character set. To change the default set, use the + --with-charset option: +shell> ./configure --with-charset=CHARSET + CHARSET may be one of binary, armscii8, ascii, big5, cp1250, + cp1251, cp1256, cp1257, cp850, cp852, cp866, cp932, dec8, + eucjpms, euckr, gb2312, gbk, geostd8, greek, hebrew, hp8, + keybcs2, koi8r, koi8u, latin1, latin2, latin5, latin7, macce, + macroman, sjis, swe7, tis620, ucs2, ujis, utf8. See Section + 9.2, "The Character Set Used for Data and Sorting." + (Additional character sets might be available. Check the + output from ./configure --help for the current list.) + The default collation may also be specified. MySQL uses the + latin1_swedish_ci collation by default. To change this, use + the --with-collation option: +shell> ./configure --with-collation=COLLATION + To change both the character set and the collation, use both + the --with-charset and --with-collation options. The collation + must be a legal collation for the character set. (Use the SHOW + COLLATION statement to determine which collations are + available for each character set.) + With the configure option --with-extra-charsets=LIST, you can + define which additional character sets should be compiled into + the server. LIST is one of the following: - By selecting one of the preconfigured configurations, the values - and settings of various options in your my.cnf or my.ini will be - altered accordingly. The default values and options as described - in the reference manual may therefore be different to the options - and values that were created during the execution of the - configuration wizard. + + A list of character set names separated by spaces -2.3.4.5. The Database Usage Dialog + + complex to include all character sets that can't be + dynamically loaded - The Database Usage dialog allows you to indicate the storage - engines that you expect to use when creating MySQL tables. The - option you choose determines whether the InnoDB storage engine is - available and what percentage of the server resources are - available to InnoDB. - MySQL Server Instance Configuration Wizard: Usage Dialog + + all to include all character sets into the binaries + Clients that want to convert characters between the server and + the client should use the SET NAMES statement. See Section + 5.1.5, "Session System Variables," and Section 9.1.4, + "Connection Character Sets and Collations." - * Multifunctional Database: This option enables both the InnoDB - and MyISAM storage engines and divides resources evenly - between the two. This option is recommended for users who use - both storage engines on a regular basis. + * To configure MySQL with debugging code, use the --with-debug + option: +shell> ./configure --with-debug + This causes a safe memory allocator to be included that can + find some errors and that provides output about what is + happening. See MySQL Internals: Porting + (http://forge.mysql.com/wiki/MySQL_Internals_Porting) + As of MySQL 5.1.12, using --with-debug to configure MySQL with + debugging support enables you to use the + --debug="d,parser_debug" option when you start the server. + This causes the Bison parser that is used to process SQL + statements to dump a parser trace to the server's standard + error output. Typically, this output is written to the error + log. - * Transactional Database Only: This option enables both the - InnoDB and MyISAM storage engines, but dedicates most server - resources to the InnoDB storage engine. This option is - recommended for users who use InnoDB almost exclusively and - make only minimal use of MyISAM. + * To cause the Debug Sync facility to be compiled into the + server, use the --enable-debug-sync option. This facility is + used for testing and debugging. When compiled in, Debug Sync + is disabled by default. To enable it, start mysqld with the + --debug-sync-timeout=N option, where N is a timeout value + greater than 0. (The default value is 0, which disables Debug + Sync.) N becomes the default timeout for individual + synchronization points. + Debug Sync is also compiled in if you configure with the + --with-debug option (which implies --enable-debug-sync), + unless you also use the --disable-debug-sync option. + For a description of the Debug Sync facility and how to use + synchronization points, see MySQL Internals: Test + Synchronization + (http://forge.mysql.com/wiki/MySQL_Internals_Test_Synchronizat + ion). + The --enable-debug-sync and --disable-debug-sync options were + added in MySQL 5.1.41. - * Non-Transactional Database Only: This option disables the - InnoDB storage engine completely and dedicates all server - resources to the MyISAM storage engine. This option is - recommended for users who do not use InnoDB. + * If your client programs are using threads, you must compile a + thread-safe version of the MySQL client library with the + --enable-thread-safe-client configure option. This creates a + libmysqlclient_r library with which you should link your + threaded applications. See Section 21.9.16.2, "How to Make a + Threaded Client." - The Configuration Wizard uses a template to generate the server - configuration file. The Database Usage dialog sets one of the - following option strings: -Multifunctional Database: MIXED -Transactional Database Only: INNODB -Non-Transactional Database Only: MYISAM + * Some features require that the server be built with + compression library support, such as the COMPRESS() and + UNCOMPRESS() functions, and compression of the client/server + protocol. The --with-zlib-dir=no|bundled|DIR option provides + control over compression library support. The value no + explicitly disables compression support. bundled causes the + zlib library bundled in the MySQL sources to be used. A DIR + path name specifies the directory in which to find the + compression library sources. - When these options are processed through the default template - (my-template.ini) the result is: -Multifunctional Database: -default-storage-engine=InnoDB -_myisam_pct=50 + * It is possible to build MySQL with large table support using + the --with-big-tables option. + This option causes the variables that store table row counts + to be declared as unsigned long long rather than unsigned + long. This enables tables to hold up to approximately + 1.844E+19 ((2^32)^2) rows rather than 2^32 (~4.295E+09) rows. + Previously it was necessary to pass -DBIG_TABLES to the + compiler manually in order to enable this feature. -Transactional Database Only: -default-storage-engine=InnoDB -_myisam_pct=5 + * Run configure with the --disable-grant-options option to cause + the --bootstrap, --skip-grant-tables, and --init-file options + for mysqld to be disabled. For Windows, the configure.js + script recognizes the DISABLE_GRANT_OPTIONS flag, which has + the same effect. The capability is available as of MySQL + 5.1.15. -Non-Transactional Database Only: -default-storage-engine=MyISAM -_myisam_pct=100 -skip-innodb + * This option allows MySQL Community Server features to be + enabled. Additional options may be required for individual + features, such as --enable-profiling to enable statement + profiling. This option was added in MySQL 5.1.24. It is + enabled by default as of MySQL 5.1.28; to disable it, use + --disable-community-features. - The _myisam_pct value is used to calculate the percentage of - resources dedicated to MyISAM. The remaining resources are - allocated to InnoDB. + * When given with --enable-community-features, the + --enable-profiling option enables the statement profiling + capability exposed by the SHOW PROFILE and SHOW PROFILES + statements. (See Section 12.5.5.33, "SHOW PROFILES Syntax.") + This option was added in MySQL 5.1.24. It is enabled by + default as of MySQL 5.1.28; to disable it, use + --disable-profiling. -2.3.4.6. The InnoDB Tablespace Dialog + * See Section 2.1, "General Installation Guidance," for options + that pertain to particular operating systems. - Some users may want to locate the InnoDB tablespace files in a - different location than the MySQL server data directory. Placing - the tablespace files in a separate location can be desirable if - your system has a higher capacity or higher performance storage - device available, such as a RAID storage system. - MySQL Server Instance Configuration Wizard: InnoDB Data Tablespace + * See Section 5.5.7.2, "Using SSL Connections," for options that + pertain to configuring MySQL to support secure (encrypted) + connections. - To change the default location for the InnoDB tablespace files, - choose a new drive from the drop-down list of drive letters and - choose a new path from the drop-down list of paths. To create a - custom path, click the ... button. - - If you are modifying the configuration of an existing server, you - must click the Modify button before you change the path. In this - situation you must move the existing tablespace files to the new - location manually before starting the server. - -2.3.4.7. The Concurrent Connections Dialog - - To prevent the server from running out of resources, it is - important to limit the number of concurrent connections to the - MySQL server that can be established. The Concurrent Connections - dialog allows you to choose the expected usage of your server, and - sets the limit for concurrent connections accordingly. It is also - possible to set the concurrent connection limit manually. - MySQL Server Instance Configuration Wizard: Connections - - * Decision Support (DSS)/OLAP: Choose this option if your server - does not require a large number of concurrent connections. The - maximum number of connections is set at 100, with an average - of 20 concurrent connections assumed. - - * Online Transaction Processing (OLTP): Choose this option if - your server requires a large number of concurrent connections. - The maximum number of connections is set at 500. + * Several configure options apply to plugin selection and + building: +--with-plugins=PLUGIN[,PLUGIN]... +--with-plugins=GROUP +--with-plugin-PLUGIN +--without-plugin-PLUGIN + PLUGIN is an individual plugin name such as csv or archive. + As shorthand, GROUP is a configuration group name such as none + (select no plugins) or all (select all plugins). + You can build a plugin as static (compiled into the server) or + dynamic (built as a dynamic library that must be installed + using the INSTALL PLUGIN statement before it can be used). + Some plugins might not support static or dynamic build. + configure --help shows the following information pertaining to + plugins: - * Manual Setting: Choose this option to set the maximum number - of concurrent connections to the server manually. Choose the - number of concurrent connections from the drop-down box - provided, or enter the maximum number of connections into the - drop-down box if the number you desire is not listed. + + The plugin-related options -2.3.4.8. The Networking and Strict Mode Options Dialog + + The names of all available plugins - Use the Networking Options dialog to enable or disable TCP/IP - networking and to configure the port number that is used to - connect to the MySQL server. - MySQL Server Instance Configuration Wizard: Network Configuration + + For each plugin, a description of its purpose, which + build types it supports (static or dynamic), and which + plugin groups it is a part of. + --with-plugins can take a list of one or more plugin names + separated by commas, or a plugin group name. The named plugins + are configured to be built as static plugins. + --with-plugin-PLUGIN configures the given plugin to be built + as a static plugin. + --without-plugin-PLUGIN disables the given plugin from being + built. + If a plugin is named both with a --with and --without option, + the result is undefined. + For any plugin that is not explicitly selected or disabled, it + is selected to be built dynamically if it supports dynamic + build, and not built if it does not support dynamic build. + (Thus, in the case that no plugin options are given, all + plugins that support dynamic build are selected to be built as + dynamic plugins. Plugins that do not support dynamic build are + not built.) - TCP/IP networking is enabled by default. To disable TCP/IP - networking, uncheck the box next to the Enable TCP/IP Networking - option. +2.3.3. Installing from the Development Source Tree - Port 3306 is used by default. To change the port used to access - MySQL, choose a new port number from the drop-down box or type a - new port number directly into the drop-down box. If the port - number you choose is in use, you are prompted to confirm your - choice of port number. +Caution - Set the Server SQL Mode to either enable or disable strict mode. - Enabling strict mode (default) makes MySQL behave more like other - database management systems. If you run applications that rely on - MySQL's old "forgiving" behavior, make sure to either adapt those - applications or to disable strict mode. For more information about - strict mode, see Section 5.1.8, "Server SQL Modes." + You should read this section only if you are interested in helping + us test our new code. If you just want to get MySQL up and running + on your system, you should use a standard release distribution + (either a binary or source distribution). -2.3.4.9. The Character Set Dialog + To obtain the most recent development source tree, you must have + Bazaar installed. You can obtain Bazaar from the Bazaar VCS + Website (http://bazaar-vcs.org) Bazaar is supported by any + platform that supports Python, and is therefore compatible with + any Linux, Unix, Windows or Mac OS X host. Instructions for + downloading and installing Bazaar on the different platforms are + available on the Bazaar website. - The MySQL server supports multiple character sets and it is - possible to set a default server character set that is applied to - all tables, columns, and databases unless overridden. Use the - Character Set dialog to change the default character set of the - MySQL server. - MySQL Server Instance Configuration Wizard: Character Set + All MySQL projects are hosted on Launchpad + (http://launchpad.net/) MySQL projects, including MySQL server, + MySQL Workbench, and others are available from the Sun/MySQL + Engineering (http://launchpad.net/~mysql) page. For the + repositories related only to MySQL server, see the MySQL Server + (http://launchpad.net/mysql-server) page. - * Standard Character Set: Choose this option if you want to use - latin1 as the default server character set. latin1 is used for - English and many Western European languages. + To build under Unix/Linux, you must have the following tools + installed: - * Best Support For Multilingualism: Choose this option if you - want to use utf8 as the default server character set. This is - a Unicode character set that can store characters from many - different languages. + * GNU make, available from http://www.gnu.org/software/make/. + Although some platforms come with their own make + implementations, it is highly recommended that you use GNU + make. It may already be available on your system as gmake. - * Manual Selected Default Character Set / Collation: Choose this - option if you want to pick the server's default character set - manually. Choose the desired character set from the provided - drop-down list. + * autoconf 2.58 (or newer), available from + http://www.gnu.org/software/autoconf/. -2.3.4.10. The Service Options Dialog + * automake 1.8.1, available from + http://www.gnu.org/software/automake/. - On Windows platforms, the MySQL server can be installed as a - Windows service. When installed this way, the MySQL server can be - started automatically during system startup, and even restarted - automatically by Windows in the event of a service failure. + * libtool 1.5, available from + http://www.gnu.org/software/libtool/. - The MySQL Server Instance Configuration Wizard installs the MySQL - server as a service by default, using the service name MySQL. If - you do not wish to install the service, uncheck the box next to - the Install As Windows Service option. You can change the service - name by picking a new service name from the drop-down box provided - or by entering a new service name into the drop-down box. + * m4, available from http://www.gnu.org/software/m4/. -Note + * bison, available from http://www.gnu.org/software/bison/. You + should use the latest version of bison where possible. Version + 1.75 and version 2.1 are known to work. There have been + reported problems with bison 1.875. If you experience + problems, upgrade to a later, rather than earlier, version. + Versions of bison older than 1.75 may report this error: +sql_yacc.yy:#####: fatal error: maximum table size (32767) exceeded + The maximum table size is not actually exceeded; the error is + caused by bugs in older versions of bison. - Service names can include any legal character except forward (/) - or backward (\) slashes, and must be less than 256 characters - long. + To build under Windows you must have Microsoft Visual C++ 2005 + Express Edition, Visual Studio .Net 2003 (7.1), or Visual Studio + 2005 (8.0) compiler system. -Warning + Once the necessary tools are installed, you must create a local + branch of the MySQL source code on your machine: - If you are installing multiple versions of MySQL onto the same - machine, you must choose a different service name for each version - that you install. If you do not choose a different service for - each installed version then the service manager information will - be inconsistent and this will cause problems when you try to - uninstall a previous version. + 1. To obtain a copy of the MySQL source code, you must create a + new Bazaar branch. If you do not already have a Bazaar + repository directory set up, you need to initialize a new + directory: +shell> mkdir mysql-server +shell> bzr init-repo --trees mysql-server - If you have already installed multiple versions using the same - service name, you must manually edit the contents of the - HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services parameters - within the Windows registry to update the association of the - service name with the correct server version. + 2. Once you have an initialized directory, you can branch from + the public MySQL server repositories to create a local source + tree. To create a branch of a specific version: +shell> cd mysql-server +shell> bzr branch lp:mysql-server/5.1 mysql-5.1 - Typically, when installing multiple versions you create a service - name based on the version information. For example, you might - install MySQL 5.x as mysql5, or specific versions such as MySQL - 5.1.30 as mysql50130. + 3. The initial download will take some time to complete, + depending on the speed of your connection. Please be patient. + Once you have downloaded the first tree, additional trees + should take significantly less time to download. - To install the MySQL server as a service but not have it started - automatically at startup, uncheck the box next to the Launch the - MySQL Server Automatically option. + 4. When building from the Bazaar branch, you may want to create a + copy of your active branch so that you can make configuration + and other changes without affecting the original branch + contents. You can achieve this by branching from the original + branch: +shell> bzr branch mysql-5.1 mysql-5.1-build -2.3.4.11. The Security Options Dialog + 5. To obtain changes made after you have set up the branch + initially, update it using the pull option periodically. Use + this command in the top-level directory of the local copy: +shell> bzr pull + You can examine the changeset comments for the tree by using + the log option to bzr: +shell> bzr log + You can also browse changesets, comments, and source code + online. To browse this information for MySQL 5.1, go to the + Launchpad MySQL Server (http://launchpad.net/mysql-server) + page. + If you see diffs (changes) or code that you have a question + about, do not hesitate to send email to the MySQL internals + mailing list. See Section 1.5.1, "MySQL Mailing Lists." Also, + if you think you have a better idea on how to do something, + send an email message to the list with a patch. + + After you have the local branch, you can build MySQL server from + the source code. On Windows, the build process is different from + Unix/Linux: see Section 2.5.10, "Installing MySQL from Source on + Windows." - It is strongly recommended that you set a root password for your - MySQL server, and the MySQL Server Instance Configuration Wizard - requires by default that you do so. If you do not wish to set a - root password, uncheck the box next to the Modify Security - Settings option. - MySQL Server Instance Configuration Wizard: Security + On Unix/Linux, use the autoconf system to create the configure + script so that you can configure the build environment before + building. The following example shows the typical commands + required to build MySQL from a source tree. - To set the root password, enter the desired password into both the - New root password and Confirm boxes. If you are reconfiguring an - existing server, you need to enter the existing root password into - the Current root password box. + 1. Change location to the top-level directory of the source tree; + replace mysql-5.1 with the appropriate directory name. +shell> cd mysql-5.1 - To prevent root logins from across the network, check the box next - to the Root may only connect from localhost option. This increases - the security of your root account. + 2. Prepare the source tree for configuration. + Prior to MySQL 5.1.12, you must separately configure the + InnoDB storage engine. Run the following command from the main + source directory: +shell> (cd storage/innobase; autoreconf --force --install) + You can omit the previous command for MySQL 5.1.12 and later, + or if you do not require InnoDB support. + Prepare the remainder of the source tree: +shell> autoreconf --force --install + As an alternative to the preceding autoreconf command, you can + use BUILD/autorun.sh, which acts as a shortcut for the + following sequence of commands: +shell> aclocal; autoheader +shell> libtoolize --automake --force +shell> automake --force --add-missing; autoconf + If you get some strange errors during this stage, verify that + you have the correct version of libtool installed. - To create an anonymous user account, check the box next to the - Create An Anonymous Account option. Creating an anonymous account - can decrease server security and cause login and permission - difficulties. For this reason, it is not recommended. + 3. Configure the source tree and compile MySQL: +shell> ./configure # Add your favorite options here +shell> make + For a description of some configure options, see Section + 2.3.2, "Typical configure Options." + A collection of our standard configuration scripts is located + in the BUILD/ subdirectory. For example, you may find it more + convenient to use the BUILD/compile-pentium-debug script than + the preceding set of shell commands. To compile on a different + architecture, modify the script by removing flags that are + Pentium-specific, or use another script that may be more + appropriate. These scripts are provided on an "as-is" basis. + They are not officially maintained and their contents may + change from release to release. -2.3.4.12. The Confirmation Dialog + 4. When the build is done, run make install. Be careful with this + on a production machine; the command may overwrite your live + release installation. If you already have MySQL installed and + do not want to overwrite it, run ./configure with values for + the --prefix, --with-tcp-port, and --with-unix-socket-path + options different from those used for your production server. - The final dialog in the MySQL Server Instance Configuration Wizard - is the Confirmation Dialog. To start the configuration process, - click the Execute button. To return to a previous dialog, click - the Back button. To exit the MySQL Server Instance Configuration - Wizard without configuring the server, click the Cancel button. - MySQL Server Instance Configuration Wizard: Confirmation + 5. Play hard with your new installation and try to make the new + features crash. Start by running make test. See Section + 22.1.2, "MySQL Test Suite." - After you click the Execute button, the MySQL Server Instance - Configuration Wizard performs a series of tasks and displays the - progress onscreen as the tasks are performed. + 6. If you have gotten to the make stage, but the distribution + does not compile, please enter the problem into our bugs + database using the instructions given in Section 1.6, "How to + Report Bugs or Problems." If you have installed the latest + versions of the required GNU tools, and they crash trying to + process our configuration files, please report that also. + However, if you get a command not found error or a similar + problem for aclocal, configure, or other required tools, do + not report it. Instead, make sure that all the required tools + are installed and that your PATH variable is set correctly so + that your shell can find them. - The MySQL Server Instance Configuration Wizard first determines - configuration file options based on your choices using a template - prepared by MySQL developers and engineers. This template is named - my-template.ini and is located in your server installation - directory. +2.3.4. Dealing with Problems Compiling MySQL - The MySQL Configuration Wizard then writes these options to the - corresponding configuration file. + All MySQL programs compile cleanly for us with no warnings on + Solaris or Linux using gcc. On other systems, warnings may occur + due to differences in system include files. See Section 2.3.5, + "MIT-pthreads Notes," for warnings that may occur when using + MIT-pthreads. For other problems, check the following list. - If you chose to create a service for the MySQL server, the MySQL - Server Instance Configuration Wizard creates and starts the - service. If you are reconfiguring an existing service, the MySQL - Server Instance Configuration Wizard restarts the service to apply - your configuration changes. - - If you chose to set a root password, the MySQL Configuration - Wizard connects to the server, sets your new root password, and - applies any other security settings you may have selected. - - After the MySQL Server Instance Configuration Wizard has completed - its tasks, it displays a summary. Click the Finish button to exit - the MySQL Server Configuration Wizard. + The solution to many problems involves reconfiguring. If you do + need to reconfigure, take note of the following: -2.3.5. Installing MySQL from a Noinstall Zip Archive + * If configure is run after it has previously been run, it may + use information that was gathered during its previous + invocation. This information is stored in config.cache. When + configure starts up, it looks for that file and reads its + contents if it exists, on the assumption that the information + is still correct. That assumption is invalid when you + reconfigure. - Users who are installing from the Noinstall package can use the - instructions in this section to manually install MySQL. The - process for installing MySQL from a Zip archive is as follows: + * Each time you run configure, you must run make again to + recompile. However, you may want to remove old object files + from previous builds first because they were compiled using + different configuration options. - 1. Extract the archive to the desired install directory + To prevent old configuration information or object files from + being used, run these commands before re-running configure: +shell> rm config.cache +shell> make clean - 2. Create an option file + Alternatively, you can run make distclean. - 3. Choose a MySQL server type + The following list describes some of the problems when compiling + MySQL that have been found to occur most often: - 4. Start the MySQL server + * If you get errors such as the ones shown here when compiling + sql_yacc.cc, you probably have run out of memory or swap + space: +Internal compiler error: program cc1plus got fatal signal 11 +Out of virtual memory +Virtual memory exhausted + The problem is that gcc requires a huge amount of memory to + compile sql_yacc.cc with inline functions. Try running + configure with the --with-low-memory option: +shell> ./configure --with-low-memory + This option causes -fno-inline to be added to the compile line + if you are using gcc and -O0 if you are using something else. + You should try the --with-low-memory option even if you have + so much memory and swap space that you think you can't + possibly have run out. This problem has been observed to occur + even on systems with generous hardware configurations, and the + --with-low-memory option usually fixes it. - 5. Secure the default user accounts + * By default, configure picks c++ as the compiler name and GNU + c++ links with -lg++. If you are using gcc, that behavior can + cause problems during configuration such as this: +configure: error: installation or configuration problem: +C++ compiler cannot create executables. + You might also observe problems during compilation related to + g++, libg++, or libstdc++. + One cause of these problems is that you may not have g++, or + you may have g++ but not libg++, or libstdc++. Take a look at + the config.log file. It should contain the exact reason why + your C++ compiler didn't work. To work around these problems, + you can use gcc as your C++ compiler. Try setting the + environment variable CXX to "gcc -O3". For example: +shell> CXX="gcc -O3" ./configure + This works because gcc compiles C++ source files as well as + g++ does, but does not link in libg++ or libstdc++ by default. + Another way to fix these problems is to install g++, libg++, + and libstdc++. However, do not use libg++ or libstdc++ with + MySQL because this only increases the binary size of mysqld + without providing any benefits. Some versions of these + libraries have also caused strange problems for MySQL users in + the past. - This process is described in the sections that follow. + * If your compile fails with errors such as any of the + following, you must upgrade your version of make to GNU make: +making all in mit-pthreads +make: Fatal error in reader: Makefile, line 18: +Badly formed macro assignment + Or: +make: file `Makefile' line 18: Must be a separator (: + Or: +pthread.h: No such file or directory + Solaris and FreeBSD are known to have troublesome make + programs. + GNU make 3.75 is known to work. -2.3.6. Extracting the Install Archive + * If you want to define flags to be used by your C or C++ + compilers, do so by adding the flags to the CFLAGS and + CXXFLAGS environment variables. You can also specify the + compiler names this way using CC and CXX. For example: +shell> CC=gcc +shell> CFLAGS=-O3 +shell> CXX=gcc +shell> CXXFLAGS=-O3 +shell> export CC CFLAGS CXX CXXFLAGS + See Section 2.2, "Installing MySQL from Generic Binaries on + Unix/Linux," for a list of flag definitions that have been + found to be useful on various systems. - To install MySQL manually, do the following: + * If you get errors such as those shown here when compiling + mysqld, configure did not correctly detect the type of the + last argument to accept(), getsockname(), or getpeername(): +cxx: Error: mysqld.cc, line 645: In this statement, the referenced + type of the pointer value ''length'' is ''unsigned long'', + which is not compatible with ''int''. +new_sock = accept(sock, (struct sockaddr *)&cAddr, &length); + To fix this, edit the config.h file (which is generated by + configure). Look for these lines: +/* Define as the base type of the last arg to accept */ +#define SOCKET_SIZE_TYPE XXX + Change XXX to size_t or int, depending on your operating + system. (You must do this each time you run configure because + configure regenerates config.h.) - 1. If you are upgrading from a previous version please refer to - Section 2.3.14, "Upgrading MySQL on Windows," before beginning - the upgrade process. + * The sql_yacc.cc file is generated from sql_yacc.yy. Normally, + the build process does not need to create sql_yacc.cc because + MySQL comes with a pre-generated copy. However, if you do need + to re-create it, you might encounter this error: +"sql_yacc.yy", line xxx fatal: default action causes potential... + This is a sign that your version of yacc is deficient. You + probably need to install bison (the GNU version of yacc) and + use that instead. - 2. Make sure that you are logged in as a user with administrator - privileges. + * On Debian Linux 3.0, you need to install gawk instead of the + default mawk. - 3. Choose an installation location. Traditionally, the MySQL - server is installed in C:\mysql. The MySQL Installation Wizard - installs MySQL under C:\Program Files\MySQL. If you do not - install MySQL at C:\mysql, you must specify the path to the - install directory during startup or in an option file. See - Section 2.3.7, "Creating an Option File." + * If you need to debug mysqld or a MySQL client, run configure + with the --with-debug option, and then recompile and link your + clients with the new client library. See MySQL Internals: + Porting (http://forge.mysql.com/wiki/MySQL_Internals_Porting) - 4. Extract the install archive to the chosen installation - location using your preferred Zip archive tool. Some tools may - extract the archive to a folder within your chosen - installation location. If this occurs, you can move the - contents of the subfolder into the chosen installation - location. + * If you get a compilation error on Linux (for example, SuSE + Linux 8.1 or Red Hat Linux 7.3) similar to the following one, + you probably do not have g++ installed: +libmysql.c:1329: warning: passing arg 5 of `gethostbyname_r' from +incompatible pointer type +libmysql.c:1329: too few arguments to function `gethostbyname_r' +libmysql.c:1329: warning: assignment makes pointer from integer +without a cast +make[2]: *** [libmysql.lo] Error 1 + By default, the configure script attempts to determine the + correct number of arguments by using g++ (the GNU C++ + compiler). This test yields incorrect results if g++ is not + installed. There are two ways to work around this problem: -2.3.7. Creating an Option File + + Make sure that the GNU C++ g++ is installed. On some + Linux distributions, the required package is called gpp; + on others, it is named gcc-c++. - If you need to specify startup options when you run the server, - you can indicate them on the command line or place them in an - option file. For options that are used every time the server - starts, you may find it most convenient to use an option file to - specify your MySQL configuration. This is particularly true under - the following circumstances: + + Use gcc as your C++ compiler by setting the CXX + environment variable to gcc: +export CXX="gcc" + You must run configure again after making either of those + changes. - * The installation or data directory locations are different - from the default locations (C:\Program Files\MySQL\MySQL - Server 5.1 and C:\Program Files\MySQL\MySQL Server 5.1\data). +2.3.5. MIT-pthreads Notes - * You need to tune the server settings, such as memory, cache, - or InnoDB configuration information. + This section describes some of the issues involved in using + MIT-pthreads. - When the MySQL server starts on Windows, it looks for options in - two files: the my.ini file in the Windows directory, and the - C:\my.cnf file. The Windows directory typically is named something - like C:\WINDOWS. You can determine its exact location from the - value of the WINDIR environment variable using the following - command: -C:\> echo %WINDIR% + On Linux, you should not use MIT-pthreads. Use the installed + LinuxThreads implementation instead. See Section 2.6, "Installing + MySQL on Linux." - MySQL looks for options first in the my.ini file, and then in the - my.cnf file. However, to avoid confusion, it is best if you use - only one file. If your PC uses a boot loader where C: is not the - boot drive, your only option is to use the my.ini file. Whichever - option file you use, it must be a plain text file. + If your system does not provide native thread support, you should + build MySQL using the MIT-pthreads package. This includes older + FreeBSD systems, SunOS 4.x, Solaris 2.4 and earlier, and some + others. See Section 2.1, "General Installation Guidance." - You can also make use of the example option files included with - your MySQL distribution; see Section 4.2.3.3.2, "Preconfigured - Option Files." + MIT-pthreads is not part of the MySQL 5.1 source distribution. If + you require this package, you need to download it separately from + http://dev.mysql.com/Downloads/Contrib/pthreads-1_60_beta6-mysql.t + ar.gz - An option file can be created and modified with any text editor, - such as Notepad. For example, if MySQL is installed in E:\mysql - and the data directory is in E:\mydata\data, you can create an - option file containing a [mysqld] section to specify values for - the basedir and datadir options: -[mysqld] -# set basedir to your installation path -basedir=E:/mysql -# set datadir to the location of your data directory -datadir=E:/mydata/data + After downloading, extract this source archive into the top level + of the MySQL source directory. It creates a new subdirectory named + mit-pthreads. - Note that Windows path names are specified in option files using - (forward) slashes rather than backslashes. If you do use - backslashes, you must double them: -[mysqld] -# set basedir to your installation path -basedir=E:\\mysql -# set datadir to the location of your data directory -datadir=E:\\mydata\\data + * On most systems, you can force MIT-pthreads to be used by + running configure with the --with-mit-threads option: +shell> ./configure --with-mit-threads + Building in a nonsource directory is not supported when using + MIT-pthreads because we want to minimize our changes to this + code. - MySQL Enterprise For expert advice on the start-up options - appropriate to your circumstances, subscribe to the MySQL - Enterprise Monitor. For more information, see - http://www.mysql.com/products/enterprise/advisors.html. + * The checks that determine whether to use MIT-pthreads occur + only during the part of the configuration process that deals + with the server code. If you have configured the distribution + using --without-server to build only the client code, clients + do not know whether MIT-pthreads is being used and use Unix + socket file connections by default. Because Unix socket files + do not work under MIT-pthreads on some platforms, this means + you need to use -h or --host with a value other than localhost + when you run client programs. - In MySQL 5.1.23 and earlier, the MySQL installer places the data - directory directly under the directory where you install MySQL. On - MySQL 5.1.24 and later, the data directory is located within the - AppData directory for the user running MySQL. + * When MySQL is compiled using MIT-pthreads, system locking is + disabled by default for performance reasons. You can tell the + server to use system locking with the --external-locking + option. This is needed only if you want to be able to run two + MySQL servers against the same data files, but that is not + recommended, anyway. - If you would like to use a data directory in a different location, - you should copy the entire contents of the data directory to the - new location. For example, if you want to use E:\mydata as the - data directory instead, you must do two things: + * Sometimes the pthread bind() command fails to bind to a socket + without any error message (at least on Solaris). The result is + that all connections to the server fail. For example: +shell> mysqladmin version +mysqladmin: connect to server at '' failed; +error: 'Can't connect to mysql server on localhost (146)' + The solution to this problem is to kill the mysqld server and + restart it. This has happened to us only when we have forcibly + stopped the server and restarted it immediately. - 1. Move the entire data directory and all of its contents from - the default location (for example C:\Program Files\MySQL\MySQL - Server 5.1\data) to E:\mydata. + * With MIT-pthreads, the sleep() system call isn't interruptible + with SIGINT (break). This is noticeable only when you run + mysqladmin --sleep. You must wait for the sleep() call to + terminate before the interrupt is served and the process + stops. - 2. Use a --datadir option to specify the new data directory - location each time you start the server. + * When linking, you might receive warning messages like these + (at least on Solaris); they can be ignored: +ld: warning: symbol `_iob' has differing sizes: + (file /my/local/pthreads/lib/libpthread.a(findfp.o) value=0x4; +file /usr/lib/libc.so value=0x140); + /my/local/pthreads/lib/libpthread.a(findfp.o) definition taken +ld: warning: symbol `__iob' has differing sizes: + (file /my/local/pthreads/lib/libpthread.a(findfp.o) value=0x4; +file /usr/lib/libc.so value=0x140); + /my/local/pthreads/lib/libpthread.a(findfp.o) definition taken -2.3.8. Selecting a MySQL Server Type + * Some other warnings also can be ignored: +implicit declaration of function `int strtoll(...)' +implicit declaration of function `int strtoul(...)' - The following table shows the available servers for Windows in - MySQL 5.1.20 and earlier. - Binary Description - mysqld-nt Optimized binary with named-pipe support - mysqld Optimized binary without named-pipe support - mysqld-debug Like mysqld-nt, but compiled with full debugging and - automatic memory allocation checking + * We have not been able to make readline work with MIT-pthreads. + (This is not necessary, but may be of interest to some.) - The following table shows the available servers for Windows in - MySQL 5.1.21 and later. - Binary Description - mysqld Optimized binary with named-pipe support - mysqld-debug Like mysqld, but compiled with full debugging and - automatic memory allocation checking +2.4. Upgrading or Downgrading MySQL - All of the preceding binaries are optimized for modern Intel - processors, but should work on any Intel i386-class or higher - processor. +2.4.1. Upgrading MySQL - Each of the servers in a distribution support the same set of - storage engines. The SHOW ENGINES statement displays which engines - a given server supports. - - All Windows MySQL 5.1 servers have support for symbolic linking of - database directories. + As a general rule, to upgrade from one release series to another, + you should go to the next series rather than skipping a series. To + upgrade from a release series previous to MySQL 5.0, upgrade to + each successive release series in turn until you have reached + MySQL 5.0, and then proceed with the upgrade to MySQL 5.1. For + example, if you currently are running MySQL 4.0 and wish to + upgrade to a newer series, upgrade to MySQL 4.1 first before + upgrading to 5.0, and so forth. For information on upgrading to + MySQL 5.0, see the MySQL 5.0 Reference Manual; for earlier + releases, see the MySQL 3.23, 4.0, 4.1 Reference Manual. - MySQL supports TCP/IP on all Windows platforms. MySQL servers on - Windows support named pipes as indicated in the following list. - However, the default is to use TCP/IP regardless of platform. - (Named pipes are slower than TCP/IP in many Windows - configurations.) + If you perform a binary (in-place) upgrade without dumping and + reloading tables, you cannot upgrade directly from MySQL 4.1 to + 5.1. This occurs due to an incompatible change in the MyISAM table + index format in MySQL 5.0. Upgrade from MySQL 4.1 to 5.0 and + repair all MyISAM tables (see Section 2.4.4, "Rebuilding or + Repairing Tables or Indexes"). Then upgrade from MySQL 5.0 to 5.1 + and check and repair your tables. - Use of named pipes is subject to these conditions: + To upgrade from MySQL 5.0 to 5.1, use the items in the following + checklist as a guide: - * Named pipes are enabled only if you start the server with the - --enable-named-pipe option. It is necessary to use this option - explicitly because some users have experienced problems with - shutting down the MySQL server when named pipes were used. + * Before any upgrade, back up your databases, including the + mysql database that contains the grant tables. See Section + 6.1, "Database Backup Methods." - * For MySQL 5.1.20 and earlier, named-pipe connections are - allowed only by the mysqld-nt and mysqld-debug servers. For - MySQL 5.1.21 and later, the mysqld and mysqld-debug servers - both contain support for named-pipe connections. + * Read all the notes in Section 2.4.1.1, "Upgrading from MySQL + 5.0 to 5.1." These notes enable you to identify upgrade issues + that apply to your current MySQL installation. Some + incompatibilities discussed in that section require your + attention before upgrading. Others should be dealt with after + upgrading. -Note + * Read Appendix C, "MySQL Change History" as well, which + provides information about features that are new in MySQL 5.1 + or differ from those found in MySQL 5.0. - Most of the examples in this manual use mysqld as the server name. - If you choose to use a different server, such as mysqld-nt or - mysqld-debug, make the appropriate substitutions in the commands - that are shown in the examples. + * After you upgrade to a new version of MySQL, run mysql_upgrade + (see Section 4.4.8, "mysql_upgrade --- Check Tables for MySQL + Upgrade"). This program checks your tables, and attempts to + repair them if necessary. It also updates your grant tables to + make sure that they have the current structure so that you can + take advantage of any new capabilities. (Some releases of + MySQL introduce changes to the structure of the grant tables + to add new privileges or features.) -2.3.9. Starting the Server for the First Time + * If you are running MySQL Server on Windows, see Section 2.5.7, + "Upgrading MySQL on Windows." - This section gives a general overview of starting the MySQL - server. The following sections provide more specific information - for starting the MySQL server from the command line or as a - Windows service. + * If you are using replication, see Section 16.3.3, "Upgrading a + Replication Setup," for information on upgrading your + replication setup. - The information here applies primarily if you installed MySQL - using the Noinstall version, or if you wish to configure and test - MySQL manually rather than with the GUI tools. + * If you are upgrading an installation originally produced by + installing multiple RPM packages, it is best to upgrade all + the packages, not just some. For example, if you previously + installed the server and client RPMs, do not upgrade just the + server RPM. - The examples in these sections assume that MySQL is installed - under the default location of C:\Program Files\MySQL\MySQL Server - 5.1. Adjust the path names shown in the examples if you have MySQL - installed in a different location. + * As of MySQL 5.1.9, the mysqld-max server is included in binary + distributions. There is no separate MySQL-Max distribution. As + of MySQL 5.1.12, there is no mysqld-max server at all in + binary distributions. They contain a server that includes the + features previously included in mysqld-max. - Clients have two options. They can use TCP/IP, or they can use a - named pipe if the server supports named-pipe connections. + * If you have created a user-defined function (UDF) with a given + name and upgrade MySQL to a version that implements a new + built-in function with the same name, the UDF becomes + inaccessible. To correct this, use DROP FUNCTION to drop the + UDF, and then use CREATE FUNCTION to re-create the UDF with a + different nonconflicting name. The same is true if the new + version of MySQL implements a built-in function with the same + name as an existing stored function. See Section 8.2.4, + "Function Name Parsing and Resolution," for the rules + describing how the server interprets references to different + kinds of functions. - MySQL for Windows also supports shared-memory connections if the - server is started with the --shared-memory option. Clients can - connect through shared memory by using the --protocol=MEMORY - option. + You can always move the MySQL format files and data files between + different versions on systems with the same architecture as long + as you stay within versions for the same release series of MySQL. - For information about which server binary to run, see Section - 2.3.8, "Selecting a MySQL Server Type." + If you are cautious about using new versions, you can always + rename your old mysqld before installing a newer one. For example, + if you are using MySQL 5.0.13 and want to upgrade to 5.1.10, + rename your current server from mysqld to mysqld-5.0.13. If your + new mysqld then does something unexpected, you can simply shut it + down and restart with your old mysqld. - Testing is best done from a command prompt in a console window (or - "DOS window"). In this way you can have the server display status - messages in the window where they are easy to see. If something is - wrong with your configuration, these messages make it easier for - you to identify and fix any problems. + If, after an upgrade, you experience problems with recompiled + client programs, such as Commands out of sync or unexpected core + dumps, you probably have used old header or library files when + compiling your programs. In this case, you should check the date + for your mysql.h file and libmysqlclient.a library to verify that + they are from the new MySQL distribution. If not, recompile your + programs with the new headers and libraries. - To start the server, enter this command: -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" --console + If problems occur, such as that the new mysqld server does not + start or that you cannot connect without a password, verify that + you do not have an old my.cnf file from your previous + installation. You can check this with the --print-defaults option + (for example, mysqld --print-defaults). If this command displays + anything other than the program name, you have an active my.cnf + file that affects server or client operation. - For a server that includes InnoDB support, you should see the - messages similar to those following as it starts (the path names - and sizes may differ): -InnoDB: The first specified datafile c:\ibdata\ibdata1 did not exist: -InnoDB: a new database to be created! -InnoDB: Setting file c:\ibdata\ibdata1 size to 209715200 -InnoDB: Database physically writes the file full: wait... -InnoDB: Log file c:\iblogs\ib_logfile0 did not exist: new to be creat -ed -InnoDB: Setting log file c:\iblogs\ib_logfile0 size to 31457280 -InnoDB: Log file c:\iblogs\ib_logfile1 did not exist: new to be creat -ed -InnoDB: Setting log file c:\iblogs\ib_logfile1 size to 31457280 -InnoDB: Log file c:\iblogs\ib_logfile2 did not exist: new to be creat -ed -InnoDB: Setting log file c:\iblogs\ib_logfile2 size to 31457280 -InnoDB: Doublewrite buffer not found: creating new -InnoDB: Doublewrite buffer created -InnoDB: creating foreign key constraint system tables -InnoDB: foreign key constraint system tables created -011024 10:58:25 InnoDB: Started + If your MySQL installation contains a large amount of data that + might take a long time to convert after an in-place upgrade, you + might find it useful to create a "dummy" database instance for + assessing what conversions might be needed and the work involved + to perform them. Make a copy of your MySQL instance that contains + a full copy of the mysql database, plus all other databases + without data. Run your upgrade procedure on this dummy instance to + see what actions might be needed so that you can better evaluate + the work involved when performing actual data conversion on your + original database instance. - When the server finishes its startup sequence, you should see - something like this, which indicates that the server is ready to - service client connections: -mysqld: ready for connections -Version: '5.1.39' socket: '' port: 3306 + It is a good idea to rebuild and reinstall the Perl DBD::mysql + module whenever you install a new release of MySQL. The same + applies to other MySQL interfaces as well, such as PHP mysql + extensions and the Python MySQLdb module. - The server continues to write to the console any further - diagnostic output it produces. You can open a new console window - in which to run client programs. +2.4.1.1. Upgrading from MySQL 5.0 to 5.1 - If you omit the --console option, the server writes diagnostic - output to the error log in the data directory (C:\Program - Files\MySQL\MySQL Server 5.1\data by default). The error log is - the file with the .err extension. + After upgrading a 5.0 installation to 5.0.10 or above, it is + necessary to upgrade your grant tables. Otherwise, creating stored + procedures and functions might not work. To perform this upgrade, + run mysql_upgrade. Note - The accounts that are listed in the MySQL grant tables initially - have no passwords. After starting the server, you should set up - passwords for them using the instructions in Section 2.11, - "Post-Installation Setup and Testing." + It is good practice to back up your data before installing any new + version of software. Although MySQL works very hard to ensure a + high level of quality, you should protect your data by making a + backup. -2.3.10. Starting MySQL from the Windows Command Line + To upgrade to 5.1 from any previous version, MySQL recommends that + you dump your tables with mysqldump before upgrading and reload + the dump file after upgrading. - The MySQL server can be started manually from the command line. - This can be done on any version of Windows. + If you perform a binary (in-place) upgrade without dumping and + reloading tables, you cannot upgrade directly from MySQL 4.1 to + 5.1. This occurs due to an incompatible change in the MyISAM table + index format in MySQL 5.0. Upgrade from MySQL 4.1 to 5.0 and + repair all MyISAM tables (see Section 2.4.4, "Rebuilding or + Repairing Tables or Indexes"). Then upgrade from MySQL 5.0 to 5.1 + and check and repair your tables. - To start the mysqld server from the command line, you should start - a console window (or "DOS window") and enter this command: -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" + In general, you should do the following when upgrading from MySQL + 5.0 to 5.1: - The path to mysqld may vary depending on the install location of - MySQL on your system. + * Read all the items in the following sections to see whether + any of them might affect your applications: - You can stop the MySQL server by executing this command: -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqladmin" -u root - shutdown + + Section 2.4.1, "Upgrading MySQL," has general update + information. -Note + + The items in the change lists found later in this section + enable you to identify upgrade issues that apply to your + current MySQL installation. - If the MySQL root user account has a password, you need to invoke - mysqladmin with the -p option and supply the password when - prompted. + + The MySQL 5.1 change history describes significant new + features you can use in 5.1 or that differ from those + found in MySQL 5.0. Some of these changes may result in + incompatibilities. See Section C.1, "Changes in Release + 5.1.x (Production)." - This command invokes the MySQL administrative utility mysqladmin - to connect to the server and tell it to shut down. The command - connects as the MySQL root user, which is the default - administrative account in the MySQL grant system. Note that users - in the MySQL grant system are wholly independent from any login - users under Windows. + * Note particularly any changes that are marked Known issue or + Incompatible change. These incompatibilities with earlier + versions of MySQL may require your attention before you + upgrade. + Our aim is to avoid these changes, but occasionally they are + necessary to correct problems that would be worse than an + incompatibility between releases. If any upgrade issue + applicable to your installation involves an incompatibility + that requires special handling, follow the instructions given + in the incompatibility description. Often this will involve a + dump and reload, or use of a statement such as CHECK TABLE or + REPAIR TABLE. + For dump and reload instructions, see Section 2.4.4, + "Rebuilding or Repairing Tables or Indexes." Any procedure + that involves REPAIR TABLE with the USE_FRM option must be + done before upgrading. Use of this statement with a version of + MySQL different from the one used to create the table (that + is, using it after upgrading) may damage the table. See + Section 12.5.2.6, "REPAIR TABLE Syntax." - If mysqld doesn't start, check the error log to see whether the - server wrote any messages there to indicate the cause of the - problem. The error log is located in the C:\Program - Files\MySQL\MySQL Server 5.1\data directory. It is the file with a - suffix of .err. You can also try to start the server as mysqld - --console; in this case, you may get some useful information on - the screen that may help solve the problem. + * After you upgrade to a new version of MySQL, run mysql_upgrade + (see Section 4.4.8, "mysql_upgrade --- Check Tables for MySQL + Upgrade"). This program checks your tables, and attempts to + repair them if necessary. It also updates your grant tables to + make sure that they have the current structure so that you can + take advantage of any new capabilities. (Some releases of + MySQL introduce changes to the structure of the grant tables + to add new privileges or features.) - The last option is to start mysqld with the --standalone and - --debug options. In this case, mysqld writes a log file - C:\mysqld.trace that should contain the reason why mysqld doesn't - start. See MySQL Internals: Porting - (http://forge.mysql.com/wiki/MySQL_Internals_Porting) - - Use mysqld --verbose --help to display all the options that mysqld - supports. - -2.3.11. Starting MySQL as a Windows Service - - On Windows, the recommended way to run MySQL is to install it as a - Windows service, whereby MySQL starts and stops automatically when - Windows starts and stops. A MySQL server installed as a service - can also be controlled from the command line using NET commands, - or with the graphical Services utility. Generally, to install - MySQL as a Windows service you should be logged in using an - account that has administrator rights. + * Check Section 2.4.3, "Checking Whether Tables or Indexes Must + Be Rebuilt," to see whether changes to table formats or to + character sets or collations were made between your current + version of MySQL and the version to which you are upgrading. + If so and these changes result in an incompatibility between + MySQL versions, you will need to upgrade the affected tables + using the instructions in Section 2.4.4, "Rebuilding or + Repairing Tables or Indexes." - The Services utility (the Windows Service Control Manager) can be - found in the Windows Control Panel (under Administrative Tools on - Windows 2000, XP, Vista and Server 2003). To avoid conflicts, it - is advisable to close the Services utility while performing server - installation or removal operations from the command line. + * If you are running MySQL Server on Windows, see Section 2.5.7, + "Upgrading MySQL on Windows." - Before installing MySQL as a Windows service, you should first - stop the current server if it is running by using the following - command: -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqladmin" - -u root shutdown + * If you are using replication, see Section 16.3.3, "Upgrading a + Replication Setup," for information on upgrading your + replication setup. -Note + If your MySQL installation contains a large amount of data that + might take a long time to convert after an in-place upgrade, you + might find it useful to create a "dummy" database instance for + assessing what conversions might be needed and the work involved + to perform them. Make a copy of your MySQL instance that contains + a full copy of the mysql database, plus all other databases + without data. Run your upgrade procedure on this dummy instance to + see what actions might be needed so that you can better evaluate + the work involved when performing actual data conversion on your + original database instance. - If the MySQL root user account has a password, you need to invoke - mysqladmin with the -p option and supply the password when - prompted. + MySQL Enterprise MySQL Enterprise subscribers will find more + information about upgrading in the Knowledge Base articles found + at Upgrading + (https://kb.mysql.com/search.php?cat=search&category=41) Access + to the MySQL Knowledge Base collection of articles is one of the + advantages of subscribing to MySQL Enterprise. For more + information, see + http://www.mysql.com/products/enterprise/advisors.html. - This command invokes the MySQL administrative utility mysqladmin - to connect to the server and tell it to shut down. The command - connects as the MySQL root user, which is the default - administrative account in the MySQL grant system. Note that users - in the MySQL grant system are wholly independent from any login - users under Windows. + The following lists describe changes that may affect applications + and that you should watch out for when upgrading to MySQL 5.1. - Install the server as a service using this command: -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" --install + Configuration Changes: - The service-installation command does not start the server. - Instructions for that are given later in this section. + * Before MySQL 5.1.11, to build MySQL from source with SSL + support enabled, you would invoke configure with either the + --with-openssl or --with-yassl option. In MySQL 5.1.11, those + options both have been replaced by the --with-ssl option. By + default, --with-ssl causes the bundled yaSSL library to be + used. To select OpenSSL instead, give the option as + --with-ssl=path, where path is the directory where the OpenSSL + header files and libraries are located. - To make it easier to invoke MySQL programs, you can add the path - name of the MySQL bin directory to your Windows system PATH - environment variable: + Server Changes: - * On the Windows desktop, right-click on the My Computer icon, - and select Properties. + * Known issue: After a binary upgrade to MySQL 5.1 from a MySQL + 5.0 installation that contains ARCHIVE tables, accessing those + tables will cause the server to crash, even if you have run + mysql_upgrade or CHECK TABLE ... FOR UPGRADE. To work around + this problem, use mysqldump to dump all ARCHIVE tables before + upgrading, and reload them into MySQL 5.1 after upgrading. + + * Known issue: The fix for + Bug#23491: http://bugs.mysql.com/23491 introduced a problem + with SHOW CREATE VIEW, which is used by mysqldump. This causes + an incompatibility when upgrading from versions affected by + that bug fix (MySQL 5.0.40 through 5.0.43, MySQL 5.1.18 + through 5.1.19): If you use mysqldump before upgrading from an + affected version and reload the data after upgrading to a + higher version, you must drop and recreate your views. - * Next select the Advanced tab from the System Properties menu - that appears, and click the Environment Variables button. + * Known issue: Dumps performed by using mysqldump to generate a + dump file before the upgrade and reloading the file after + upgrading are subject to the following problem: + Before MySQL 5.0.40, mysqldump displays SPATIAL index + definitions using prefix lengths for the indexed columns. + These prefix lengths are accepted in MySQL 5.0, but not as of + MySQL 5.1. If you use mysqldump from versions of MySQL older + than 5.0.40, any table containing SPATIAL indexes will cause + an error when the dump file is reloaded into MySQL 5.1 or + higher. + For example, a table definition might look like this when + dumped in MySQL 5.0: +CREATE TABLE `t` ( + `g` geometry NOT NULL, + SPATIAL KEY `g` (`g`(32)) +) ENGINE=MyISAM DEFAULT CHARSET=latin1 + The SPATIAL index definition will not be accepted in MySQL + 5.1. To work around this, edit the dump file to remove the + prefix: +CREATE TABLE `t` ( + `g` geometry NOT NULL, + SPATIAL KEY `g` (`g`) +) ENGINE=MyISAM DEFAULT CHARSET=latin1 + Dump files can be large, so it may be preferable to dump table + definitions and data separately to make it easier to edit the + definitions: +shell> mysqldump --no-data other_args > definitions.sql +shell> mysqldump --no-create-info other_args > data.sql + Then edit definitions.sql before reloading definitions.sql and + data.sql, in that order. + If you upgrade to a version of MySQL 5.0 higher than 5.0.40 + before upgrading to MySQL 5.1, this problem does not occur. - * Under System Variables, select Path, and then click the Edit - button. The Edit System Variable dialogue should appear. + * Known issue: Before MySQL 5.1.30, the CHECK TABLE ... FOR + UPGRADE statement did not check for incompatible collation + changes made in MySQL 5.1.24. (This also affects mysqlcheck + and mysql_upgrade, which cause that statement to be executed.) + Prior to the fix made in 5.1.30, a binary upgrade (performed + without dumping tables with mysqldump before the upgrade and + reloading the dump file after the upgrade) would corrupt + tables. After the fix, CHECK TABLE ... FOR UPGRADE properly + detects the problem and warns about tables that need repair. + However, the fix is not backward compatible and can result in + a downgrading problem under these circumstances: - * Place your cursor at the end of the text appearing in the - space marked Variable Value. (Use the End key to ensure that - your cursor is positioned at the very end of the text in this - space.) Then enter the complete path name of your MySQL bin - directory (for example, C:\Program Files\MySQL\MySQL Server - 5.1\bin), Note that there should be a semicolon separating - this path from any values present in this field. Dismiss this - dialogue, and each dialogue in turn, by clicking OK until all - of the dialogues that were opened have been dismissed. You - should now be able to invoke any MySQL executable program by - typing its name at the DOS prompt from any directory on the - system, without having to supply the path. This includes the - servers, the mysql client, and all MySQL command-line - utilities such as mysqladmin and mysqldump. - You should not add the MySQL bin directory to your Windows - PATH if you are running multiple MySQL servers on the same - machine. + 1. Perform a binary upgrade to a version of MySQL that + includes the fix. -Warning + 2. Run CHECK TABLE ... FOR UPGRADE (or mysqlcheck or + mysql_upgrade) to upgrade tables. - You must exercise great care when editing your system PATH by - hand; accidental deletion or modification of any portion of the - existing PATH value can leave you with a malfunctioning or even - unusable system. + 3. Perform a binary downgrade to a version of MySQL that + does not include the fix. + The solution is to dump tables with mysqldump before the + downgrade and reload the dump file after the downgrade. + Alternatively, drop and recreate affected indexes. - The following additional arguments can be used in MySQL 5.1 when - installing the service: + * Known issue: MySQL introduces encoding for table names that + have non-ASCII characters (see Section 8.2.3, "Mapping of + Identifiers to File Names"). After a binary upgrade from MySQL + 5.0 to 5.1 or higher, the server recognizes names that have + non-ASCII characters and adds a #mysql50# prefix to them. + As of MySQL 5.1.31, mysql_upgrade encodes these names by + executing the following command: +mysqlcheck --all-databases --check-upgrade --fix-db-names --fix-table +-names + Prior to MySQL 5.1.31, mysql_upgrade does not execute this + command, so you should execute it manually if you have + database or table names that contain nonalphanumeric + characters. + Prior to MySQL 5.1.23, the mysqlcheck command does not perform + the name encoding for views. To work around this problem, drop + each affected view and recreate it. + mysqlcheck cannot fix names that contain literal instances of + the @ character that is used for encoding special characters. + If you have databases or tables that contain this character, + use mysqldump to dump them before upgrading to MySQL 5.1, and + then reload the dump file after upgrading. - * You can specify a service name immediately following the - --install option. The default service name is MySQL. + * Known issue: When upgrading from MySQL 5.0 to versions of 5.1 + prior to 5.1.23, running mysqlcheck (or mysql_upgrade, which + runs mysqlcheck) to upgrade tables fails for names that must + be written as quoted identifiers. To work around this problem, + rename each affected table to a name that does not require + quoting: +RENAME TABLE `tab``le_a` TO table_a; +RENAME TABLE `table b` TO table_b; + After renaming the tables, run the mysql_upgrade program. Then + rename the tables back to their original names: +RENAME TABLE table_a TO `tab``le_a`; +RENAME TABLE table_b TO `table b`; - * If a service name is given, it can be followed by a single - option. By convention, this should be - --defaults-file=file_name to specify the name of an option - file from which the server should read options when it starts. - The use of a single option other than --defaults-file is - possible but discouraged. --defaults-file is more flexible - because it enables you to specify multiple startup options for - the server by placing them in the named option file. + * Known issue: In connection with view creation, the server + created arc directories inside database directories and + maintained useless copies of .frm files there. Creation and + renaming procedures of those copies as well as creation of arc + directories has been discontinued in MySQL 5.1.29. + This change does cause a problem when downgrading to older + server versions which manifests itself under these + circumstances: - * You can also specify a --local-service option following the - service name. This causes the server to run using the - LocalService Windows account that has limited system - privileges. This account is available only for Windows XP or - newer. If both --defaults-file and --local-service are given - following the service name, they can be in any order. + 1. Create a view v_orig in MySQL 5.1.29 or higher. - For a MySQL server that is installed as a Windows service, the - following rules determine the service name and option files that - the server uses: + 2. Rename the view to v_new and then back to v_orig. - * If the service-installation command specifies no service name - or the default service name (MySQL) following the --install - option, the server uses the a service name of MySQL and reads - options from the [mysqld] group in the standard option files. + 3. Downgrade to an older 5.1.x server and run mysql_upgrade. - * If the service-installation command specifies a service name - other than MySQL following the --install option, the server - uses that service name. It reads options from the [mysqld] - group and the group that has the same name as the service in - the standard option files. This allows you to use the [mysqld] - group for options that should be used by all MySQL services, - and an option group with the service name for use by the - server installed with that service name. + 4. Try to rename v_orig to v_new again. This operation + fails. + As a workaround to avoid this problem, use either of these + approaches: - * If the service-installation command specifies a - --defaults-file option after the service name, the server - reads options only from the [mysqld] group of the named file - and ignores the standard option files. + + Dump your data using mysqldump before downgrading and + reload the dump file after downgrading. - As a more complex example, consider the following command: -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" - --install MySQL --defaults-file=C:\my-opts.cnf + + Instead of renaming a view after the downgrade, drop it + and recreate it. - Here, the default service name (MySQL) is given after the - --install option. If no --defaults-file option had been given, - this command would have the effect of causing the server to read - the [mysqld] group from the standard option files. However, - because the --defaults-file option is present, the server reads - options from the [mysqld] option group, and only from the named - file. + * Incompatible change: Character set or collation changes were + made in MySQL 5.1.21, 5.1.23, and 5.1.24 that may require + table indexes to be rebuilt. For details, see Section 2.4.3, + "Checking Whether Tables or Indexes Must Be Rebuilt." - You can also specify options as Start parameters in the Windows - Services utility before you start the MySQL service. + * Incompatible change: MySQL 5.1 implements support for a plugin + API that allows the loading and unloading of components at + runtime, without restarting the server. Section 22.2, "The + MySQL Plugin Interface." The plugin API requires the + mysql.plugin table. After upgrading from an older version of + MySQL, you should run the mysql_upgrade command to create this + table. See Section 4.4.8, "mysql_upgrade --- Check Tables for + MySQL Upgrade." + Plugins are installed in the directory named by the plugin_dir + system variable. This variable also controls the location from + which the server loads user-defined functions (UDFs), which is + a change from earlier versions of MySQL. That is, all UDF + library files now must be installed in the plugin directory. + When upgrading from an older version of MySQL, you must + migrate your UDF files to the plugin directory. - Once a MySQL server has been installed as a service, Windows - starts the service automatically whenever Windows starts. The - service also can be started immediately from the Services utility, - or by using a NET START MySQL command. The NET command is not case - sensitive. + * Incompatible change: The table_cache system variable has been + renamed to table_open_cache. Any scripts that refer to + table_cache must be updated to use the new name. - When run as a service, mysqld has no access to a console window, - so no messages can be seen there. If mysqld does not start, check - the error log to see whether the server wrote any messages there - to indicate the cause of the problem. The error log is located in - the MySQL data directory (for example, C:\Program - Files\MySQL\MySQL Server 5.1\data). It is the file with a suffix - of .err. + * Incompatible change: In MySQL 5.1.36, options for loading + plugins such as pluggable storage engines were changed from + boolean to tristate format. The implementations overlap, but + if you previously used options of the form --plugin_name=0 or + --plugin_name=1, you should instead use --plugin_name=OFF or + --plugin_name=ON, respectively. For details, see Section + 5.1.3, "Server Options for Loading Plugins." - When a MySQL server has been installed as a service, and the - service is running, Windows stops the service automatically when - Windows shuts down. The server also can be stopped manually by - using the Services utility, the NET STOP MySQL command, or the - mysqladmin shutdown command. + * Incompatible change: From MySQL 5.1.24 to 5.1.31, the UPDATE + statement was changed such that assigning NULL to a NOT NULL + column caused an error even when strict SQL mode was not + enabled. The original behavior before MySQL 5.1.24 was that + such assignments caused an error only in strict SQL mode, and + otherwise set the column to the implicit default value for the + column data type and generated a warning. (For information + about implicit default values, see Section 10.1.4, "Data Type + Default Values.") + The change caused compatibility problems for applications that + relied on the original behavior. It also caused replication + problems between servers that had the original behavior and + those that did not, for applications that assigned NULL to NOT + NULL columns in UPDATE statements without strict SQL mode + enabled. The change was reverted in MySQL 5.1.32 so that + UPDATE again had the original behavior. Problems can still + occur if you replicate between servers that have the modified + UPDATE behavior and those that do not. - You also have the choice of installing the server as a manual - service if you do not wish for the service to be started - automatically during the boot process. To do this, use the - --install-manual option rather than the --install option: -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" --install-m -anual + * Incompatible change: As of MySQL 5.1.29, the default binary + logging mode has been changed from MIXED to STATEMENT for + compatibility with MySQL 5.0. - To remove a server that is installed as a service, first stop it - if it is running by executing NET STOP MySQL. Then use the - --remove option to remove it: -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" --remove + * Incompatible change: In MySQL 5.1.25, a change was made to the + way that the server handles prepared statements. This affects + prepared statements processed at the SQL level (using the + PREPARE statement) and those processed using the binary + client-server protocol (using the mysql_stmt_prepare() C API + function). + Previously, changes to metadata of tables or views referred to + in a prepared statement could cause a server crash when the + statement was next executed, or perhaps an error at execute + time with a crash occurring later. For example, this could + happen after dropping a table and recreating it with a + different definition. + Now metadata changes to tables or views referred to by + prepared statements are detected and cause automatic + repreparation of the statement when it is next executed. + Metadata changes occur for DDL statements such as those that + create, drop, alter, rename, or truncate tables, or that + analyze, optimize, or repair tables. Repreparation also occurs + after referenced tables or views are flushed from the table + definition cache, either implicitly to make room for new + entries in the cache, or explicitly due to FLUSH TABLES. + Repreparation is automatic, but to the extent that it occurs, + performance of prepared statements is diminished. + Table content changes (for example, with INSERT or UPDATE) do + not cause repreparation, nor do SELECT statements. + An incompatibility with previous versions of MySQL is that a + prepared statement may now return a different set of columns + or different column types from one execution to the next. For + example, if the prepared statement is SELECT * FROM t1, + altering t1 to contain a different number of columns causes + the next execution to return a number of columns different + from the previous execution. + Older versions of the client library cannot handle this change + in behavior. For applications that use prepared statements + with the new server, an upgrade to the new client library is + strongly recommended. + Along with this change to statement repreparation, the default + value of the table_definition_cache system variable has been + increased from 128 to 256. The purpose of this increase is to + lessen the chance that prepared statements will need + repreparation due to referred-to tables/views having been + flushed from the cache to make room for new entries. + A new status variable, Com_stmt_reprepare, has been introduced + to track the number of repreparations. - If mysqld is not running as a service, you can start it from the - command line. For instructions, see Section 2.3.10, "Starting - MySQL from the Windows Command Line." + * Incompatible change: As of MySQL 5.1.23, within a stored + routine, it is no longer allowable to declare a cursor for a + SHOW or DESCRIBE statement. This happened to work in some + instances, but is no longer supported. In many cases, a + workaround for this change is to use the cursor with a SELECT + query to read from an INFORMATION_SCHEMA table that produces + the same information as the SHOW statement. - Please see Section 2.3.13, "Troubleshooting a MySQL Installation - Under Windows," if you encounter difficulties during installation. + * Incompatible change: SHOW CREATE VIEW displays view + definitions using an AS alias_name clause for each column. If + a column is created from an expression, the default alias is + the expression text, which can be quite long. As of MySQL + 5.1.23, aliases for column names in CREATE VIEW statements are + checked against the maximum column length of 64 characters + (not the maximum alias length of 256 characters). As a result, + views created from the output of SHOW CREATE VIEW fail if any + column alias exceeds 64 characters. This can cause problems + for replication or loading dump files. For additional + information and workarounds, see Section D.4, "Restrictions on + Views." -2.3.12. Testing The MySQL Installation + * Incompatible change: Several issues were identified for stored + programs (stored procedures and functions, triggers, and + events) and views containing non-ASCII symbols. These issues + involved conversion errors due to incomplete character set + information when translating these objects to and from stored + format. + To address these problems, the representation for these + objects was changed in MySQL 5.1.21. However, the fixes affect + all stored programs and views. (For example, you will see + warnings about "no creation context.") To avoid warnings from + the server about the use of old definitions from any release + prior to 5.1.21, you should dump stored programs and views + with mysqldump after upgrading to 5.1.21 or higher, and then + reload them to recreate them with new definitions. Invoke + mysqldump with a --default-character-set option that names the + non-ASCII character set that was used for the definitions when + the objects were originally defined. - You can test whether the MySQL server is working by executing any - of the following commands: -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqlshow" -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqlshow" -u root -mysql -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqladmin" version - status proc -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysql" test + * Incompatible change: As of MySQL 5.1.20, mysqld_safe supports + error logging to syslog on systems that support the logger + command. The new --syslog and --skip-syslog options can be + used instead of the --log-error option to control logging + behavior, as described in Section 4.3.2, "mysqld_safe --- + MySQL Server Startup Script." + In 5.1.21 and up, the default is --skip-syslog, which is + compatible with the default behavior of writing an error log + file for releases prior to 5.1.20. + In 5.1.20 only, the following conditions apply: 1) The default + is to use syslog, which is not compatible with releases prior + to 5.1.20. 2) Logging to syslog may fail to operate correctly + in some cases. For these reasons, avoid using MySQL 5.1.20. - If mysqld is slow to respond to TCP/IP connections from client - programs, there is probably a problem with your DNS. In this case, - start mysqld with the --skip-name-resolve option and use only - localhost and IP numbers in the Host column of the MySQL grant - tables. + * Incompatible change: As of MySQL 5.1.18, the plugin interface + and its handling of system variables was changed. Command-line + options such as --skip-innodb now cause an error if InnoDB is + not built-in or plugin-loaded. You should use + --loose-skip-innodb if you do not want any error even if + InnoDB is not available. The --loose prefix modifier should be + used for all command-line options where you are uncertain + whether the plugin exists and when you want the operation to + proceed even if the option is necessarily ignored due to the + absence of the plugin. (For a desecription of how --loose + works, see Section 4.2.3.1, "Using Options on the Command + Line.") - You can force a MySQL client to use a named-pipe connection rather - than TCP/IP by specifying the --pipe or --protocol=PIPE option, or - by specifying . (period) as the host name. Use the --socket option - to specify the name of the pipe if you do not want to use the - default pipe name. + * Incompatible change: As of MySQL 5.1.15, InnoDB rolls back + only the last statement on a transaction timeout. A new + option, --innodb_rollback_on_timeout, causes InnoDB to abort + and roll back the entire transaction if a transaction timeout + occurs (the same behavior as in MySQL 4.1). - Note that if you have set a password for the root account, deleted - the anonymous account, or created a new user account, then you - must use the appropriate -u and -p options with the commands shown - above in order to connect with the MySQL Server. See Section - 4.2.2, "Connecting to the MySQL Server." + * Incompatible change: As of MySQL 5.1.15, the following + conditions apply to enabling the read_only system variable: - For more information about mysqlshow, see Section 4.5.6, - "mysqlshow --- Display Database, Table, and Column Information." + + If you attempt to enable read_only while you have any + explicit locks (acquired with LOCK TABLES or have a + pending transaction, an error will occur. -2.3.13. Troubleshooting a MySQL Installation Under Windows + + If other clients hold explicit table locks or have + pending transactions, the attempt to enable read_only + blocks until the locks are released and the transactions + end. While the attempt to enable read_only is pending, + requests by other clients for table locks or to begin + transactions also block until read_only has been set. - When installing and running MySQL for the first time, you may - encounter certain errors that prevent the MySQL server from - starting. The purpose of this section is to help you diagnose and - correct some of these errors. + + read_only can be enabled while you hold a global read + lock (acquired with FLUSH TABLES WITH READ LOCK) because + that does not involve table locks. + Previously, the attempt to enable read_only would return + immediately even if explicit locks or transactions were + pending, so some data changes could occur for statements + executing in the server at the same time. - Your first resource when troubleshooting server issues is the - error log. The MySQL server uses the error log to record - information relevant to the error that prevents the server from - starting. The error log is located in the data directory specified - in your my.ini file. The default data directory location is - C:\Program Files\MySQL\MySQL Server 5.1\data. See Section 5.2.2, - "The Error Log." + * Incompatible change: The number of function names affected by + IGNORE_SPACE was reduced significantly in MySQL 5.1.13, from + about 200 to about 30. (For details about IGNORE_SPACE, see + Section 8.2.4, "Function Name Parsing and Resolution.") This + change improves the consistency of parser operation. However, + it also introduces the possibility of incompatibility for old + SQL code that relies on the following conditions: - Another source of information regarding possible errors is the - console messages displayed when the MySQL service is starting. Use - the NET START MySQL command from the command line after installing - mysqld as a service to see any error messages regarding the - starting of the MySQL server as a service. See Section 2.3.11, - "Starting MySQL as a Windows Service." + + IGNORE_SPACE is disabled. - The following examples show other common error messages you may - encounter when installing MySQL and starting the server for the - first time: + + The presence or absence of whitespace following a + function name is used to distinguish between a built-in + function and stored function that have the same name (for + example, PI() versus PI ()). + For functions that are no longer affected by IGNORE_SPACE as + of MySQL 5.1.13, that strategy no longer works. Either of the + following approaches can be used if you have code that is + subject to the preceding incompatibility: - * If the MySQL server cannot find the mysql privileges database - or other critical files, you may see these messages: -System error 1067 has occurred. -Fatal error: Can't open privilege tables: Table 'mysql.host' doesn't -exist - These messages often occur when the MySQL base or data - directories are installed in different locations than the - default locations (C:\Program Files\MySQL\MySQL Server 5.1 and - C:\Program Files\MySQL\MySQL Server 5.1\data, respectively). - This situation may occur when MySQL is upgraded and installed - to a new location, but the configuration file is not updated - to reflect the new location. In addition, there may be old and - new configuration files that conflict. Be sure to delete or - rename any old configuration files when upgrading MySQL. - If you have installed MySQL to a directory other than - C:\Program Files\MySQL\MySQL Server 5.1, you need to ensure - that the MySQL server is aware of this through the use of a - configuration (my.ini) file. The my.ini file needs to be - located in your Windows directory, typically C:\WINDOWS. You - can determine its exact location from the value of the WINDIR - environment variable by issuing the following command from the - command prompt: -C:\> echo %WINDIR% - An option file can be created and modified with any text - editor, such as Notepad. For example, if MySQL is installed in - E:\mysql and the data directory is D:\MySQLdata, you can - create the option file and set up a [mysqld] section to - specify values for the basedir and datadir options: -[mysqld] -# set basedir to your installation path -basedir=E:/mysql -# set datadir to the location of your data directory -datadir=D:/MySQLdata - Note that Windows path names are specified in option files - using (forward) slashes rather than backslashes. If you do use - backslashes, you must double them: -[mysqld] -# set basedir to your installation path -basedir=C:\\Program Files\\MySQL\\MySQL Server 5.1 -# set datadir to the location of your data directory -datadir=D:\\MySQLdata - If you change the datadir value in your MySQL configuration - file, you must move the contents of the existing MySQL data - directory before restarting the MySQL server. - See Section 2.3.7, "Creating an Option File." - - * If you reinstall or upgrade MySQL without first stopping and - removing the existing MySQL service and install MySQL using - the MySQL Configuration Wizard, you may see this error: -Error: Cannot create Windows service for MySql. Error: 0 - This occurs when the Configuration Wizard tries to install the - service and finds an existing service with the same name. - One solution to this problem is to choose a service name other - than mysql when using the configuration wizard. This allows - the new service to be installed correctly, but leaves the - outdated service in place. Although this is harmless, it is - best to remove old services that are no longer in use. - To permanently remove the old mysql service, execute the - following command as a user with administrative privileges, on - the command-line: -C:\> sc delete mysql -[SC] DeleteService SUCCESS - If the sc utility is not available for your version of - Windows, download the delsrv utility from - http://www.microsoft.com/windows2000/techinfo/reskit/tools/exi - sting/delsrv-o.asp and use the delsrv mysql syntax. - -2.3.14. Upgrading MySQL on Windows - - This section lists some of the steps you should take when - upgrading MySQL on Windows. - - 1. Review Section 2.12.1, "Upgrading MySQL," for additional - information on upgrading MySQL that is not specific to - Windows. - - 2. You should always back up your current MySQL installation - before performing an upgrade. See Section 6.1, "Database - Backups." - - 3. Download the latest Windows distribution of MySQL from - http://dev.mysql.com/downloads/. - - 4. Before upgrading MySQL, you must stop the server. If the - server is installed as a service, stop the service with the - following command from the command prompt: -C:\> NET STOP MySQL - If you are not running the MySQL server as a service, use the - following command to stop it: -C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqladmin" -u root - shutdown - -Note - If the MySQL root user account has a password, you need to - invoke mysqladmin with the -p option and supply the password - when prompted. - - 5. When upgrading to MySQL 5.1 from a version previous to 4.1.5, - or when upgrading from a version of MySQL installed from a Zip - archive to a version of MySQL installed with the MySQL - Installation Wizard, you must manually remove the previous - installation and MySQL service (if the server is installed as - a service). - To remove the MySQL service, use the following command: -C:\> C:\mysql\bin\mysqld --remove - If you do not remove the existing service, the MySQL - Installation Wizard may fail to properly install the new MySQL - service. - - 6. When upgrading from MySQL 5.1.23 to MySQL 5.1.24, the change - in the default location of the data directory from a directory - within the MySQL installation to the AppData folder means that - you must manually copy the data files from your old - installation to the new location. - - 7. If you are using the MySQL Installation Wizard, start the - wizard as described in Section 2.3.3, "Using the MySQL - Installation Wizard." - - 8. If you are installing MySQL from a Zip archive, extract the - archive. You may either overwrite your existing MySQL - installation (usually located at C:\mysql), or install it into - a different directory, such as C:\mysql5. Overwriting the - existing installation is recommended. - - 9. If you were running MySQL as a Windows service and you had to - remove the service earlier in this procedure, reinstall the - service. (See Section 2.3.11, "Starting MySQL as a Windows - Service.") - 10. Restart the server. For example, use NET START MySQL if you - run MySQL as a service, or invoke mysqld directly otherwise. - 11. If you encounter errors, see Section 2.3.13, "Troubleshooting - a MySQL Installation Under Windows." - -2.3.15. MySQL on Windows Compared to MySQL on Unix - - MySQL for Windows has proven itself to be very stable. The Windows - version of MySQL has the same features as the corresponding Unix - version, with the following exceptions: - - * Limited number of ports - Windows systems have about 4,000 ports available for client - connections, and after a connection on a port closes, it takes - two to four minutes before the port can be reused. In - situations where clients connect to and disconnect from the - server at a high rate, it is possible for all available ports - to be used up before closed ports become available again. If - this happens, the MySQL server appears to be unresponsive even - though it is running. Note that ports may be used by other - applications running on the machine as well, in which case the - number of ports available to MySQL is lower. - For more information about this problem, see - http://support.microsoft.com/default.aspx?scid=kb;en-us;196271 - . - - * Concurrent reads - MySQL depends on the pread() and pwrite() system calls to be - able to mix INSERT and SELECT. Currently, we use mutexes to - emulate pread() and pwrite(). We intend to replace the file - level interface with a virtual interface in the future so that - we can use the readfile()/writefile() interface to get more - speed. The current implementation limits the number of open - files that MySQL 5.1 can use to 2,048, which means that you - cannot run as many concurrent threads on Windows as on Unix. - - * Blocking read - MySQL uses a blocking read for each connection. That has the - following implications if named-pipe connections are enabled: - - + A connection is not disconnected automatically after - eight hours, as happens with the Unix version of MySQL. - - + If a connection hangs, it is not possible to break it - without killing MySQL. - - + mysqladmin kill does not work on a sleeping connection. - - + mysqladmin shutdown cannot abort as long as there are - sleeping connections. - We plan to fix this problem in the future. - - * ALTER TABLE - While you are executing an ALTER TABLE statement, the table is - locked from being used by other threads. This has to do with - the fact that on Windows, you can't delete a file that is in - use by another thread. In the future, we may find some way to - work around this problem. - - * DROP TABLE - DROP TABLE on a table that is in use by a MERGE table does not - work on Windows because the MERGE handler does the table - mapping hidden from the upper layer of MySQL. Because Windows - does not allow dropping files that are open, you first must - flush all MERGE tables (with FLUSH TABLES) or drop the MERGE - table before dropping the table. - - * DATA DIRECTORY and INDEX DIRECTORY - The DATA DIRECTORY and INDEX DIRECTORY options for CREATE - TABLE are ignored on Windows, because Windows doesn't support - symbolic links. These options also are ignored on systems that - have a nonfunctional realpath() call. - - * DROP DATABASE - You cannot drop a database that is in use by some thread. - - * Case-insensitive names - File names are not case sensitive on Windows, so MySQL - database and table names are also not case sensitive on - Windows. The only restriction is that database and table names - must be specified using the same case throughout a given - statement. See Section 8.2.2, "Identifier Case Sensitivity." - - * Directory and file names - On Windows, MySQL Server supports only directory and file - names that are compatible with the current ANSI code pages. - For example, the following Japanese directory name will not - work in the Western locale (code page 1252): -datadir="C:/维基百科关于中文维基百科" - The same limitation applies to directory and file names - referred to in SQL statements, such as the data file path name - in LOAD DATA INFILE. - - * The "\" path name separator character - Path name components in Windows are separated by the "\" - character, which is also the escape character in MySQL. If you - are using LOAD DATA INFILE or SELECT ... INTO OUTFILE, use - Unix-style file names with "/" characters: -mysql> LOAD DATA INFILE 'C:/tmp/skr.txt' INTO TABLE skr; -mysql> SELECT * INTO OUTFILE 'C:/tmp/skr.txt' FROM skr; - Alternatively, you must double the "\" character: -mysql> LOAD DATA INFILE 'C:\\tmp\\skr.txt' INTO TABLE skr; -mysql> SELECT * INTO OUTFILE 'C:\\tmp\\skr.txt' FROM skr; - - * Problems with pipes - Pipes do not work reliably from the Windows command-line - prompt. If the pipe includes the character ^Z / CHAR(24), - Windows thinks that it has encountered end-of-file and aborts - the program. - This is mainly a problem when you try to apply a binary log as - follows: -C:\> mysqlbinlog binary_log_file | mysql --user=root - If you have a problem applying the log and suspect that it is - because of a ^Z / CHAR(24) character, you can use the - following workaround: -C:\> mysqlbinlog binary_log_file --result-file=/tmp/bin.sql -C:\> mysql --user=root --execute "source /tmp/bin.sql" - The latter command also can be used to reliably read in any - SQL file that may contain binary data. - - * Access denied for user error - If MySQL cannot resolve your host name properly, you may get - the following error when you attempt to run a MySQL client - program to connect to a server running on the same machine: -Access denied for user 'some_user'@'unknown' -to database 'mysql' - To fix this problem, you should create a file named - \windows\hosts containing the following information: -127.0.0.1 localhost - - Here are some open issues for anyone who might want to help us - improve MySQL on Windows: - - * Add macros to use the faster thread-safe increment/decrement - methods provided by Windows. - -2.4. Installing MySQL from RPM Packages on Linux - - The recommended way to install MySQL on RPM-based Linux - distributions is by using the RPM packages. The RPMs that we - provide to the community should work on all versions of Linux that - support RPM packages and use glibc 2.3. To obtain RPM packages, - see Section 2.1.3, "How to Get MySQL." - - For non-RPM Linux distributions, you can install MySQL using a - .tar.gz package. See Section 2.9, "Installing MySQL from tar.gz - Packages on Other Unix-Like Systems." - - We do provide some platform-specific RPMs; the difference between - a platform-specific RPM and a generic RPM is that a - platform-specific RPM is built on the targeted platform and is - linked dynamically whereas a generic RPM is linked statically with - LinuxThreads. - -Note - - RPM distributions of MySQL often are provided by other vendors. Be - aware that they may differ in features and capabilities from those - built by us, and that the instructions in this manual do not - necessarily apply to installing them. The vendor's instructions - should be consulted instead. - - If you have problems with an RPM file (for example, if you receive - the error Sorry, the host 'xxxx' could not be looked up), see - Section 2.13.1.2, "Linux Binary Distribution Notes." - - In most cases, you need to install only the MySQL-server and - MySQL-client packages to get a functional MySQL installation. The - other packages are not required for a standard installation. - - RPMs for MySQL Cluster. Beginning with MySQL 5.1.24, standard - MySQL server RPMs built by MySQL no longer provide support for the - NDBCLUSTER storage engine. MySQL Cluster users wanting to upgrade - MySQL 5.1.23 or earlier installations from RPMs built by MySQL - should upgrade to MySQL Cluster NDB 6.2 or MySQL Cluster NDB 6.3; - RPMs that should work with most Linux distributions are available - for both of these release series. - -Important - - When upgrading a MySQL Cluster RPM installation, you must upgrade - all installed RPMs, including the Server and Client RPMs. - - For more information about installing MySQL Cluster from RPMs, see - Section 17.2.2, "MySQL Cluster Multi-Computer Installation." - - For upgrades, if your installation was originally produced by - installing multiple RPM packages, it is best to upgrade all the - packages, not just some. For example, if you previously installed - the server and client RPMs, do not upgrade just the server RPM. - - If you get a dependency failure when trying to install MySQL - packages (for example, error: removing these packages would break - dependencies: libmysqlclient.so.10 is needed by ...), you should - also install the MySQL-shared-compat package, which includes both - the shared libraries for backward compatibility - (libmysqlclient.so.12 for MySQL 4.0 and libmysqlclient.so.10 for - MySQL 3.23). - - Some Linux distributions still ship with MySQL 3.23 and they - usually link applications dynamically to save disk space. If these - shared libraries are in a separate package (for example, - MySQL-shared), it is sufficient to simply leave this package - installed and just upgrade the MySQL server and client packages - (which are statically linked and do not depend on the shared - libraries). For distributions that include the shared libraries in - the same package as the MySQL server (for example, Red Hat Linux), - you could either install our 3.23 MySQL-shared RPM, or use the - MySQL-shared-compat package instead. (Do not install both.) - - The RPM packages shown in the following list are available. The - names shown here use a suffix of .glibc23.i386.rpm, but particular - packages can have different suffixes, as described later. - - * MySQL-server-VERSION.glibc23.i386.rpm - The MySQL server. You need this unless you only want to - connect to a MySQL server running on another machine. - - * MySQL-client-VERSION.glibc23.i386.rpm - The standard MySQL client programs. You probably always want - to install this package. - - * MySQL-devel-VERSION.glibc23.i386.rpm - The libraries and include files that are needed if you want to - compile other MySQL clients, such as the Perl modules. - - * MySQL-debuginfo-VERSION.glibc23.i386.rpm - This package contains debugging information. debuginfo RPMs - are never needed to use MySQL software; this is true both for - the server and for client programs. However, they contain - additional information that might be needed by a debugger to - analyze a crash. - - * MySQL-shared-VERSION.glibc23.i386.rpm - This package contains the shared libraries - (libmysqlclient.so*) that certain languages and applications - need to dynamically load and use MySQL. It contains - single-threaded and thread-safe libraries. If you install this - package, do not install the MySQL-shared-compat package. - - * MySQL-shared-compat-VERSION.glibc23.i386.rpm - This package includes the shared libraries for MySQL 3.23, - 4.0, and so on, up to the current release. It contains - single-threaded and thread-safe libraries. Install this - package instead of MySQL-shared if you have applications - installed that are dynamically linked against older versions - of MySQL but you want to upgrade to the current version - without breaking the library dependencies. - - * MySQL-embedded-VERSION.glibc23.i386.rpm - The embedded MySQL server library. - - * MySQL-ndb-management-VERSION.glibc23.i386.rpm, - MySQL-ndb-storage-VERSION.glibc23.i386.rpm, - MySQL-ndb-tools-VERSION.glibc23.i386.rpm, - MySQL-ndb-extra-VERSION.glibc23.i386.rpm - Packages that contain additional files for MySQL Cluster - installations. - -Note - The MySQL-ndb-tools RPM requires a working installation of - perl. Prior to MySQL 5.1.18, the DBI and HTML::Template - packages were also required. See Section 2.15, "Perl - Installation Notes," and Section 17.6.21, "ndb_size.pl --- - NDBCLUSTER Size Requirement Estimator," for more information. - - * MySQL-test-VERSION.glibc23.i386.rpm - This package includes the MySQL test suite. - - * MySQL-VERSION.src.rpm - This contains the source code for all of the previous - packages. It can also be used to rebuild the RPMs on other - architectures (for example, Alpha or SPARC). - - The suffix of RPM package names (following the VERSION value) has - the following syntax: -.PLATFORM.CPU.rpm - - The PLATFORM and CPU values indicate the type of system for which - the package is built. PLATFORM indicates the platform and CPU - indicates the processor type or family. - - All packages are dynamically linked against glibc 2.3. The - PLATFORM value indicates whether the package is platform - independent or intended for a specific platform, as shown in the - following table. - glibc23 Platform independent, should run on any Linux distribution - that supports glibc 2.3 - rhel3, rhel4 Red Hat Enterprise Linux 3 or 4 - sles9, sles10 SuSE Linux Enterprise Server 9 or 10 - - In MySQL 5.1, only glibc23 packages are available currently. - - The CPU value indicates the processor type or family for which the - package is built. - i386 x86 processor, 386 and up - i586 x86 processor, Pentium and up - x86_64 64-bit x86 processor - ia64 Itanium (IA-64) processor - - To see all files in an RPM package (for example, a MySQL-server - RPM), run a command like this: -shell> rpm -qpl MySQL-server-VERSION.glibc23.i386.rpm - - To perform a standard minimal installation, install the server and - client RPMs: -shell> rpm -i MySQL-server-VERSION.glibc23.i386.rpm -shell> rpm -i MySQL-client-VERSION.glibc23.i386.rpm - - To install only the client programs, install just the client RPM: -shell> rpm -i MySQL-client-VERSION.glibc23.i386.rpm - - RPM provides a feature to verify the integrity and authenticity of - packages before installing them. If you would like to learn more - about this feature, see Section 2.1.4, "Verifying Package - Integrity Using MD5 Checksums or GnuPG." - - The server RPM places data under the /var/lib/mysql directory. The - RPM also creates a login account for a user named mysql (if one - does not exist) to use for running the MySQL server, and creates - the appropriate entries in /etc/init.d/ to start the server - automatically at boot time. (This means that if you have performed - a previous installation and have made changes to its startup - script, you may want to make a copy of the script so that you - don't lose it when you install a newer RPM.) See Section 2.11.2.2, - "Starting and Stopping MySQL Automatically," for more information - on how MySQL can be started automatically on system startup. - - If you want to install the MySQL RPM on older Linux distributions - that do not support initialization scripts in /etc/init.d - (directly or via a symlink), you should create a symbolic link - that points to the location where your initialization scripts - actually are installed. For example, if that location is - /etc/rc.d/init.d, use these commands before installing the RPM to - create /etc/init.d as a symbolic link that points there: -shell> cd /etc -shell> ln -s rc.d/init.d . - - However, all current major Linux distributions should support the - new directory layout that uses /etc/init.d, because it is required - for LSB (Linux Standard Base) compliance. - - If the RPM files that you install include MySQL-server, the mysqld - server should be up and running after installation. You should be - able to start using MySQL. - - If something goes wrong, you can find more information in the - binary installation section. See Section 2.9, "Installing MySQL - from tar.gz Packages on Other Unix-Like Systems." - -Note - - The accounts that are listed in the MySQL grant tables initially - have no passwords. After starting the server, you should set up - passwords for them using the instructions in Section 2.11, - "Post-Installation Setup and Testing." - - During RPM installation, a user named mysql and a group named - mysql are created on the system. This is done using the useradd, - groupadd, and usermod commands. Those commands require appropriate - administrative privileges, which is ensured for locally managed - users and groups (as listed in the /etc/passwd and /etc/group - files) by the RPM installation process being run by root. - - For nonlocal user management (LDAP, NIS, and so forth), the - administrative tools may require additional authentication (such - as a password), and will fail if the installing user does not - provide this authentication. Even if they fail, the RPM - installation will not abort but succeed, and this is intentional. - If they failed, some of the intended transfer of ownership may be - missing, and it is recommended that the system administrator then - manually ensures some appropriate user andgroup exists and - manually transfers ownership following the actions in the RPM spec - file. - -2.5. Installing MySQL on Mac OS X - - You can install MySQL on Mac OS X 10.3.x ("Panther") or newer - using a Mac OS X binary package in PKG format instead of the - binary tarball distribution. Please note that older versions of - Mac OS X (for example, 10.1.x or 10.2.x) are not supported by this - package. - - The package is located inside a disk image (.dmg) file that you - first need to mount by double-clicking its icon in the Finder. It - should then mount the image and display its contents. - - To obtain MySQL, see Section 2.1.3, "How to Get MySQL." - -Note - - Before proceeding with the installation, be sure to shut down all - running MySQL server instances by either using the MySQL Manager - Application (on Mac OS X Server) or via mysqladmin shutdown on the - command line. - - To actually install the MySQL PKG file, double-click on the - package icon. This launches the Mac OS X Package Installer, which - guides you through the installation of MySQL. - - Due to a bug in the Mac OS X package installer, you may see this - error message in the destination disk selection dialog: -You cannot install this software on this disk. (null) + + If a stored function has a name that conflicts with a + built-in function, refer to the stored function with a + schema name qualifier, regardless of whether whitespace + is present. For example, write schema_name.PI() or + schema_name.PI (). - If this error occurs, simply click the Go Back button once to - return to the previous screen. Then click Continue to advance to - the destination disk selection again, and you should be able to - choose the destination disk correctly. We have reported this bug - to Apple and it is investigating this problem. - - The Mac OS X PKG of MySQL installs itself into - /usr/local/mysql-VERSION and also installs a symbolic link, - /usr/local/mysql, that points to the new location. If a directory - named /usr/local/mysql exists, it is renamed to - /usr/local/mysql.bak first. Additionally, the installer creates - the grant tables in the mysql database by executing - mysql_install_db. - - The installation layout is similar to that of a tar file binary - distribution; all MySQL binaries are located in the directory - /usr/local/mysql/bin. The MySQL socket file is created as - /tmp/mysql.sock by default. See Section 2.1.5, "Installation - Layouts." - - MySQL installation requires a Mac OS X user account named mysql. A - user account with this name should exist by default on Mac OS X - 10.2 and up. + + Alternatively, rename the stored function to use a + nonconflicting name and change invocations of the + function to use the new name. - If you are running Mac OS X Server, a version of MySQL should - already be installed. The following table shows the versions of - MySQL that ship with Mac OS X Server versions. - Mac OS X Server Version MySQL Version - 10.2-10.2.2 3.23.51 - 10.2.3-10.2.6 3.23.53 - 10.3 4.0.14 - 10.3.2 4.0.16 - 10.4.0 4.1.10a + * Incompatible change: For utf8 columns, the full-text parser + incorrectly considered several nonword punctuation and + whitespace characters as word characters, causing some + searches to return incorrect results. The fix involves a + change to the full-text parser in MySQL 5.1.12, so as of + 5.1.12, any tables that have FULLTEXT indexes on utf8 columns + must be repaired with REPAIR TABLE: +REPAIR TABLE tbl_name QUICK; - This manual section covers the installation of the official MySQL - Mac OS X PKG only. Make sure to read Apple's help information - about installing MySQL: Run the "Help View" application, select - "Mac OS X Server" help, do a search for "MySQL," and read the item - entitled "Installing MySQL." - - If you previously used Marc Liyanage's MySQL packages for Mac OS X - from http://www.entropy.ch, you can simply follow the update - instructions for packages using the binary installation layout as - given on his pages. - - If you are upgrading from Marc's 3.23.x versions or from the Mac - OS X Server version of MySQL to the official MySQL PKG, you also - need to convert the existing MySQL privilege tables to the current - format, because some new security privileges have been added. See - Section 4.4.8, "mysql_upgrade --- Check Tables for MySQL Upgrade." - - If you want MySQL to start automatically during system startup, - you also need to install the MySQL Startup Item. It is part of the - Mac OS X installation disk images as a separate installation - package. Simply double-click the MySQLStartupItem.pkg icon and - follow the instructions to install it. The Startup Item need be - installed only once. There is no need to install it each time you - upgrade the MySQL package later. + * Incompatible change: Storage engines can be pluggable at + runtime, so the distinction between disabled and invalid + storage engines no longer applies. As of MySQL 5.1.12, this + affects the NO_ENGINE_SUBSTITUTION SQL mode, as described in + Section 5.1.8, "Server SQL Modes." - The Startup Item for MySQL is installed into - /Library/StartupItems/MySQLCOM. (Before MySQL 4.1.2, the location - was /Library/StartupItems/MySQL, but that collided with the MySQL - Startup Item installed by Mac OS X Server.) Startup Item - installation adds a variable MYSQLCOM=-YES- to the system - configuration file /etc/hostconfig. If you want to disable the - automatic startup of MySQL, simply change this variable to - MYSQLCOM=-NO-. - - On Mac OS X Server, the default MySQL installation uses the - variable MYSQL in the /etc/hostconfig file. The MySQL Startup Item - installer disables this variable by setting it to MYSQL=-NO-. This - avoids boot time conflicts with the MYSQLCOM variable used by the - MySQL Startup Item. However, it does not shut down a running MySQL - server. You should do that yourself. + * Incompatible change: The structure of FULLTEXT indexes has + been changed in MySQL 5.1.6. After upgrading to MySQL 5.1.6 or + greater, any tables that have FULLTEXT indexes must be + repaired with REPAIR TABLE: +REPAIR TABLE tbl_name QUICK; - After the installation, you can start up MySQL by running the - following commands in a terminal window. You must have - administrator privileges to perform this task. + * Incompatible change: In MySQL 5.1.6, when log tables were + implemented, the default log destination for the general query + and slow query log was TABLE. As of MySQL 5.1.21, this default + has been changed to FILE, which is compatible with MySQL 5.0, + but incompatible with earlier releases of MySQL 5.1. If you + are upgrading from MySQL 5.0 to 5.1.21 or higher, no logging + option changes should be necessary. However, if you are + upgrading from 5.1.6 through 5.1.20 to 5.1.21 or higher and + were using TABLE logging, use the --log-output=TABLE option + explicitly to preserve your server's table-logging behavior. - If you have installed the Startup Item, use this command: -shell> sudo /Library/StartupItems/MySQLCOM/MySQLCOM start -(Enter your password, if necessary) -(Press Control-D or enter "exit" to exit the shell) + * Incompatible change: For ENUM columns that had enumeration + values containing commas, the commas were mapped to 0xff + internally. However, this rendered the commas + indistinguishable from true 0xff characters in the values. + This no longer occurs. However, the fix requires that you dump + and reload any tables that have ENUM columns containing true + 0xff in their values: Dump the tables using mysqldump with the + current server before upgrading from a version of MySQL 5.1 + older than 5.1.15 to version 5.1.15 or newer. - If you don't use the Startup Item, enter the following command - sequence: -shell> cd /usr/local/mysql -shell> sudo ./bin/mysqld_safe -(Enter your password, if necessary) -(Press Control-Z) -shell> bg -(Press Control-D or enter "exit" to exit the shell) + * As of MySQL 5.1.12, the lc_time_names system variable + specifies the locale that controls the language used to + display day and month names and abbreviations. This variable + affects the output from the DATE_FORMAT(), DAYNAME() and + MONTHNAME() functions. See Section 9.8, "MySQL Server Locale + Support." - You should be able to connect to the MySQL server, for example, by - running /usr/local/mysql/bin/mysql. + * As of MySQL 5.1.9, mysqld_safe no longer implicitly invokes + mysqld-max if it exists. Instead, it invokes mysqld unless a + --mysqld or --mysqld-version option is given to specify + another server explicitly. If you previously relied on the + implicit invocation of mysqld-max, you should use an + appropriate option now. As of MySQL 5.1.12, there is no longer + any separate mysqld-max server, so no change should be + necessary. -Note + SQL Changes: - The accounts that are listed in the MySQL grant tables initially - have no passwords. After starting the server, you should set up - passwords for them using the instructions in Section 2.11, - "Post-Installation Setup and Testing." + * Known issue: Prior to MySQL 5.1.17, the parser accepted + invalid code in SQL condition handlers, leading to server + crashes or unexpected execution behavior in stored programs. + Specifically, the parser allowed a condition handler to refer + to labels for blocks that enclose the handler declaration. + This was incorrect because block label scope does not include + the code for handlers declared within the labeled block. + As of 5.1.17, the parser rejects this invalid construct, but + if you perform a binary upgrade (without dumping and reloading + your databases), existing handlers that contain the construct + still are invalid and should be rewritten even if they appear + to function as you expect. + To find affected handlers, use mysqldump to dump all stored + procedures and functions, triggers, and events. Then attempt + to reload them into an upgraded server. Handlers that contain + illegal label references will be rejected. + For more information about condition handlers and writing them + to avoid invalid jumps, see Section 12.8.4.2, "DECLARE for + Handlers." - You might want to add aliases to your shell's resource file to - make it easier to access commonly used programs such as mysql and - mysqladmin from the command line. The syntax for bash is: -alias mysql=/usr/local/mysql/bin/mysql -alias mysqladmin=/usr/local/mysql/bin/mysqladmin + * Incompatible change: The parser accepted statements that + contained /* ... */ that were not properly closed with */, + such as SELECT 1 /* + 2. As of MySQL 5.1.23, statements that + contain unclosed /*-comments now are rejected with a syntax + error. + This fix has the potential to cause incompatibilities. Because + of Bug#26302: http://bugs.mysql.com/26302, which caused the + trailing */ to be truncated from comments in views, stored + routines, triggers, and events, it is possible that objects of + those types may have been stored with definitions that now + will be rejected as syntactically invalid. Such objects should + be dropped and re-created so that their definitions do not + contain truncated comments. - For tcsh, use: -alias mysql /usr/local/mysql/bin/mysql -alias mysqladmin /usr/local/mysql/bin/mysqladmin + * Incompatible change: Multiple-table DELETE statements + containing ambiguous aliases could have unintended side + effects such as deleting rows from the wrong table. Example: +DELETE FROM t1 AS a2 USING t1 AS a1 INNER JOIN t2 AS a2; + As of MySQL 5.1.23, alias declarations can be declared only in + the table_references part. Elsewhere in the statement, alias + references are allowed but not alias declarations. Statements + containing aliases that are no longer allowed must be + rewritten. - Even better, add /usr/local/mysql/bin to your PATH environment - variable. You can do this by modifying the appropriate startup - file for your shell. For more information, see Section 4.2.1, - "Invoking MySQL Programs." - - If you are upgrading an existing installation, note that - installing a new MySQL PKG does not remove the directory of an - older installation. Unfortunately, the Mac OS X Installer does not - yet offer the functionality required to properly upgrade - previously installed packages. - - To use your existing databases with the new installation, you'll - need to copy the contents of the old data directory to the new - data directory. Make sure that neither the old server nor the new - one is running when you do this. After you have copied over the - MySQL database files from the previous installation and have - successfully started the new server, you should consider removing - the old installation files to save disk space. Additionally, you - should also remove older versions of the Package Receipt - directories located in /Library/Receipts/mysql-VERSION.pkg. + * Incompatible change: As of MySQL 5.1.8, TYPE = engine_name is + still accepted as a synonym for the ENGINE = engine_name table + option but generates a warning. You should note that this + option is not available in MySQL 5.1.7, and is removed + altogether as of MySQL 5.4 and produces a syntax error. + TYPE has been deprecated since MySQL 4.0. -2.6. Installing MySQL on Solaris + * Incompatible change: The namespace for triggers changed in + MySQL 5.0.10. Previously, trigger names had to be unique per + table. Now they must be unique within the schema (database). + An implication of this change is that DROP TRIGGER syntax now + uses a schema name instead of a table name (schema name is + optional and, if omitted, the current schema will be used). + When upgrading from a version of MySQL 5 older than 5.0.10 to + MySQL 5.0.10 or newer, you must drop all triggers and + re-create them or DROP TRIGGER will not work after the + upgrade. Here is a suggested procedure for doing this: - To obtain a binary MySQL distribution for Solaris in tarball or - PKG format, http://dev.mysql.com/downloads/mysql/5.1.html. + 1. Upgrade to MySQL 5.0.10 or later to be able to access + trigger information in the INFORMATION_SCHEMA.TRIGGERS + table. (This should work even for pre-5.0.10 triggers.) - If you install MySQL using a binary tarball distribution on - Solaris, you may run into trouble even before you get the MySQL - distribution unpacked, as the Solaris tar cannot handle long file - names. This means that you may see errors when you try to unpack - MySQL. + 2. Dump all trigger definitions using the following SELECT + statement: +SELECT CONCAT('CREATE TRIGGER ', t.TRIGGER_SCHEMA, '.', t.TRIGGER_NAM +E, + ' ', t.ACTION_TIMING, ' ', t.EVENT_MANIPULATION, ' ON ' +, + t.EVENT_OBJECT_SCHEMA, '.', t.EVENT_OBJECT_TABLE, + ' FOR EACH ROW ', t.ACTION_STATEMENT, '//' ) +INTO OUTFILE '/tmp/triggers.sql' +FROM INFORMATION_SCHEMA.TRIGGERS AS t; + The statement uses INTO OUTFILE, so you must have the + FILE privilege. The file will be created on the server + host. Use a different file name if you like. To be 100% + safe, inspect the trigger definitions in the triggers.sql + file, and perhaps make a backup of the file. - If this occurs, you must use GNU tar (gtar) to unpack the - distribution. + 3. Stop the server and drop all triggers by removing all + .TRG files in your database directories. Change location + to your data directory and issue this command: +shell> rm */*.TRG - You can install MySQL on Solaris using a binary package in PKG - format instead of the binary tarball distribution. Before - installing using the binary PKG format, you should create the - mysql user and group, for example: -groupadd mysql -useradd -g mysql mysql + 4. Start the server and re-create all triggers using the + triggers.sql file: +mysql> delimiter // ; +mysql> source /tmp/triggers.sql // - Some basic PKG-handling commands follow: + 5. Check that all triggers were successfully created using + the SHOW TRIGGERS statement. - * To add a package: -pkgadd -d package_name.pkg + * Incompatible change: MySQL 5.1.6 introduces the TRIGGER + privilege. Previously, the SUPER privilege was needed to + create or drop triggers. Now those operations require the + TRIGGER privilege. This is a security improvement because you + no longer need to grant users the SUPER privilege to enable + them to create triggers. However, the requirement that the + account named in a trigger's DEFINER clause must have the + SUPER privilege has changed to a requirement for the TRIGGER + privilege. When upgrading from a previous version of MySQL 5.0 + or 5.1 to MySQL 5.1.6 or newer, be sure to update your grant + tables by running mysql_upgrade. This will assign the TRIGGER + privilege to all accounts that had the SUPER privilege. If you + fail to update the grant tables, triggers may fail when + activated. After updating the grant tables, you can revoke the + SUPER privilege from those accounts that no longer otherwise + require it. - * To remove a package: -pkgrm package_name + * Some keywords may be reserved in MySQL 5.1 that were not + reserved in MySQL 5.0. See Section 8.3, "Reserved Words." - * To get a full list of installed packages: -pkginfo + * The BACKUP TABLE, and RESTORE TABLE statements are deprecated. + mysqldump or mysqlhotcopy can be used as alternatives. - * To get detailed information for a package: -pkginfo -l package_name + * The LOAD DATA FROM MASTER and LOAD TABLE FROM MASTER + statements are deprecated. See Section 12.6.2.2, "LOAD DATA + FROM MASTER Syntax," for recommended alternatives. - * To list the files belonging to a package: -pkgchk -v package_name + * The INSTALL PLUGIN and UNINSTALL PLUGIN statements that are + used for the plugin API are new. So is the WITH PARSER clause + for FULLTEXT index creation that associates a parser plugin + with a full-text index. Section 22.2, "The MySQL Plugin + Interface." - * To get packaging information for an arbitrary file: -pkgchk -l -p file_name + C API Changes: - For additional information about installing MySQL on Solaris, see - Section 2.13.3, "Solaris Notes." + * Incompatible change: As of MySQL 5.1.7, the + mysql_stmt_attr_get() C API function returns a boolean rather + than an unsigned int for STMT_ATTR_UPDATE_MAX_LENGTH. + (Bug#16144: http://bugs.mysql.com/16144) -2.7. Installing MySQL on i5/OS +2.4.2. Downgrading MySQL - The i5/OS POWER MySQL package was created in cooperation with IBM. - MySQL works within the Portable Application Solution Environment - (PASE) on the System i series of hardware and will also provide - database services for the Zend Core for i5/OS. + This section describes what you should do to downgrade to an older + MySQL version in the unlikely case that the previous version + worked better than the new one. - MySQL for i5/OS is provided as a save file (.savf) package that - can be downloaded and installed directly without any additional - installation steps required. + If you are downgrading within the same release series (for + example, from 5.0.13 to 5.0.12) the general rule is that you just + have to install the new binaries on top of the old ones. There is + no need to do anything with the databases. As always, however, it + is always a good idea to make a backup. - MySQL is only supported on i5/OS V5R4 or later releases. The i5/OS - PASE must be installed for MySQL to operate. You must be able to - login as a user in *SECOFR class. + The following items form a checklist of things you should do + whenever you perform a downgrade: - You should the installation notes and tips for i5/OS before - starting installation. See i5/OS Installation Notes. + * Read the upgrading section for the release series from which + you are downgrading to be sure that it does not have any + features you really need. See Section 2.4.1, "Upgrading + MySQL." -Note + * If there is a downgrading section for that version, you should + read that as well. - The installation package will use an existing configuration if you - have previously installed MySQL (which is identified by looking - for the file /etc/my.cnf). The values for the data directory - (DATADIR) and owner of the MySQL files (USRPRF) specified during - the installation will be ignored, and the values determined from - the /etc/my.cnf will be used instead. + * To see which new features were added between the version to + which you are downgrading and your current version, see the + change logs (Appendix C, "MySQL Change History"). - If you want to change these parameters during a new install, you - should temporarily rename /etc/my.cnf, install MySQL using the new - parameters you want to use, and then merge your previous - /etc/my.cnf configuration settings with the new /etc/my.cnf file - that is created during installation. + * Check Section 2.4.3, "Checking Whether Tables or Indexes Must + Be Rebuilt," to see whether changes to table formats or to + character sets or collations were made between your current + version of MySQL and the version to which you are downgrading. + If so and these changes result in an incompatibility between + MySQL versions, you will need to downgrade the affected tables + using the instructions in Section 2.4.4, "Rebuilding or + Repairing Tables or Indexes." - To install MySQL on i5/OS, follow these steps: + In most cases, you can move the MySQL format files and data files + between different versions on the same architecture as long as you + stay within versions for the same release series of MySQL. - 1. Create a user profile MYSQL. The MYSQL user profile will own - all the MySQL files and databases and be the active user used - when the MySQL server is running. The profile should be - disabled so that you cannot log in as the MySQL user. To - create a user profile, use CRTUSRPRF: -CRTUSRPRF USRPRF(MYSQL) STATUS(*DISABLED) TEXT('MySQL user id') + If you downgrade from one release series to another, there may be + incompatibilities in table storage formats. In this case, use + mysqldump to dump your tables before downgrading. After + downgrading, reload the dump file using mysql or mysqlimport to + re-create your tables. For examples, see Section 2.4.5, "Copying + MySQL Databases to Another Machine." - 2. On the System i machine, create a save file that will be used - to receive the downloaded installation save file. The file - should be located within the General Purpose Library (QGPL): -CRTSAVF FILE(QGPL/MYSQLINST) + A typical symptom of a downward-incompatible table format change + when you downgrade is that you cannot open tables. In that case, + use the following procedure: - 3. Download the MySQL installation save file in 32-bit - (mysql-5.0.42-i5os-power-32bit.savf) or 64-bit - (mysql-5.0.42-i5os-power-64bit.savf) from MySQL Downloads - (http://dev.mysql.com/downloads) + 1. Stop the older MySQL server that you are downgrading to. - 4. You need to FTP the downloaded .savf file directly into the - QGPL/MYSQLINST file on the System i server. You can do this - through FTP using the following steps after logging in to the - System i machine: -ftp> bin -ftp> cd qgpl -ftp> put mysql-5.0.42-i5os-power.savf mysqlinst + 2. Restart the newer MySQL server you are downgrading from. - 5. Log into the System i server using a user in the *SECOFR - class, such as the QSECOFR user ID. + 3. Dump any tables that were inaccessible to the older server by + using mysqldump to create a dump file. - 6. You need to restore the installation library stored in the - .savf save file: -RSTLIB MYSQLINST DEV(*SAVF) SAVF(QGPL/MYSQLINST) + 4. Stop the newer MySQL server and restart the older one. - 7. You need to execute the installation command, - MYSQLINST/INSMYSQL. You can specify three parameter settings - during installation: + 5. Reload the dump file into the older server. Your tables should + be accessible. - + DIR('/opt/mysql') sets the installation location for the - MySQL files. The directory will be created if it does not - already exist. - - + DATADIR('/QOpenSys/mysal/data') sets the location of the - directory that will be used to store the database files - and binary logs. The default setting is - /QOpenSys/mysql/data. Note that if the installer detects - an existing installation (due to the existence of - /etc/my.cnf), then this parameter will be ignored. + It might also be the case that the structure of the system tables + in the mysql database has changed and that downgrading introduces + some loss of functionality or requires some adjustments. Here are + some examples: - + USRPRF(MYSQL) sets the user profile that will own the - files that are installed. The profile will be created if - it does not already exist. - MySQL can be installed anywhere, for this example we will - assume MySQL has been installed into /opt/mysql. The MYSQL - user profile that was created earlier in this sequence should - be used for the profile: -MYSQLINST/INSMYSQL DIR('/opt/mysql') DATADIR('/opt/mysqldata') USRPRF -(MYSQL) - If you are updating an installation over an existing MySQL - installation, you should use the same parameter values that - were used when MySQL was originally installed. - The installation copies all the necessary files into a - directory matching the package version (for example - mysql-5.0.42-i5os-power-32bit), sets the ownership on those - files, sets up the MySQL environment and creates the MySQL - configuration file (in /etc/my.cnf) completing all the steps - in a typical binary installation process automatically. If - this is a new installation of MySQL, or if the installer - detects that this is a new version (because the /etc/my.cnf - file does not exist), then the initial core MySQL databases - will also be created during installation. + * Trigger creation requires the TRIGGER privilege as of MySQL + 5.1. In MySQL 5.0, there is no TRIGGER privilege and SUPER is + required instead. If you downgrade from MySQL 5.1 to 5.0, you + will need to give the SUPER privilege to those accounts that + had the TRIGGER privilege in 5.1. - 8. Once the installation has completed, you can delete the - installation file: -DLTLIB LIB(MYSQLINST) + * Triggers were added in MySQL 5.0, so if you downgrade from 5.0 + to 4.1, you cannot use triggers at all. - To start MySQL: +2.4.2.1. Downgrading to MySQL 5.0 - 1. Log into the System i server using a user within the *SECOFR - class, such as the QSECOFR user ID. + When downgrading to MySQL 5.0 from MySQL 5.1, you should keep in + mind the following issues relating to features found in MySQL 5.1, + but not in MySQL 5.0: -Note - You should start mysqld_safe using a user that in the PASE - environment has the id=0 (the equivalent of the standard Unix - root user). If you do not use a user with this ID then the - system will be unable to change the user when executing mysqld - as set using --user option. If this happens, mysqld may be - unable to read the files located within the MySQL data - directory and the execution will fail. + * Partitioning. MySQL 5.0 does not support user-defined + partitioning. If a table was created as a partitioned table in + 5.1 (or if an table created in a previous version of MySQL was + altered to include partitions after an upgrade to 5.1), the + table is accessible after downgrade only if you do one of the + following: - 2. Enter the PASE environment using call qp2term. + + Export the table using mysqldump and then drop it in + MySQL 5.1; import the table again following the downgrade + to MySQL 5.0. - 3. Start the MySQL server by changing to the installation - directory and running mysqld_safe, specifying the user name - used to install the server. The installer conveniently - installs a symbolic link to the installation directory - (mysql-5.0.42-i5os-power-32bit) as /opt/mysql/mysql: -> cd /opt/mysql/mysql -> bin/mysqld_safe --user=mysql & - You should see a message similar to the following: -Starting mysqld daemon with databases » - from /opt/mysql/mysql-enterprise-5.0.42-i5os-power-32bit/data + + Prior to the downgrade, remove the table's partitioning + using ALTER TABLE table_name REMOVE PARTITIONING. - If you are having problems starting MySQL server, see Section - 2.11.2.3, "Starting and Troubleshooting the MySQL Server." + * Event Scheduler. MySQL 5.0 does not support scheduled events. + If your databases contain scheduled event definitions, you + should prevent them from being dumped when you use mysqldump + by using the --skip-events option. (See Section 4.5.4, + "mysqldump --- A Database Backup Program.") - To stop MySQL: + * Stored routines. MySQL 5.1.21 added a number of new columns + to the mysql.proc table in which stored routine definitions + are stored. If you are downgrading from MySQL 5.1.21 or later + to MySQL 5.0, you cannot import the MySQL 5.1 routine + definitions into MySQL 5.0.46 or earlier using the dump of + mysql.proc created by mysqldump (such as when using the + --all-databases option). Instead, you should run mysqldump + --routines prior to performing the downgrade and run the + stored routines DDL statements following the downgrade. + See Bug#11986: http://bugs.mysql.com/11986, + Bug#30029: http://bugs.mysql.com/30029, and + Bug#30660: http://bugs.mysql.com/30660, for more information. - 1. Log into the System i server using the *SECOFR class, such as - the QSECOFR user ID. + * Triggers. Trigger creation requires the TRIGGER privilege as + of MySQL 5.1. In MySQL 5.0, there is no TRIGGER privilege and + SUPER is required instead. If you downgrade from MySQL 5.1 to + 5.0, you will need to give the SUPER privilege to those + accounts that had the TRIGGER privilege in 5.1. - 2. Enter the PASE environment using call qp2term. +2.4.3. Checking Whether Tables or Indexes Must Be Rebuilt - 3. Stop the MySQL server by changing into the installation - directory and running mysqladmin, specifying the user name - used to install the server: -> cd /opt/mysql/mysql -> bin/mysqladmin -u root shutdown - If the session that you started and stopped MySQL are the - same, you may get the log output from mysqld: - STOPPING server from pid file » - /opt/mysql/mysql-enterprise-5.0.42-i5os-power-32bit/data/I5DBX.R -CHLAND.IBM.COM.pid - 070718 10:34:20 mysqld ended - If the sessions used to start and stop MySQL are different, - you will not receive any confirmation of the shutdown. + A binary upgrade or downgrade is one that installs one version of + MySQL "in place" over an existing version, without dumping and + reloading tables: - Note and tips + 1. Stop the server for the existing version if it is running. - * A problem has been identified with the installation process on - DBCS systems. If you are having problems install MySQL on a - DBCS system, you need to change your job's coded character set - identifier (CSSID) to 37 (EBCDIC) before executing the install - command, INSMYSQL. To do this, determine your existing CSSID - (using DSPJOB and selecting option 2), execute CHGJOB - CSSID(37), run INSMYSQL to install MySQL and then execute - CHGJOB again with your original CSSID. + 2. Install a different version of MySQL. This is an upgrade if + the new version is higher than the original version, a + downgrade if the version is lower. - * If you want to use the Perl scripts that are included with - MySQL, you need to download the iSeries Tools for Developers - (5799-PTL). See - http://www-03.ibm.com/servers/enable/site/porting/tools/. + 3. Start the server for the new version. -2.8. Installing MySQL on NetWare + In many cases, the tables from the previous version of MySQL can + be used without problem by the new version. However, sometimes + changes occur that require tables or table indexes to be rebuilt, + as described in this section. If you have tables that are affected + by any of the issues described here, rebuild the tables or indexes + as necessary using the instructions given in Section 2.4.4, + "Rebuilding or Repairing Tables or Indexes." - Porting MySQL to NetWare was an effort spearheaded by Novell. - Novell customers should be pleased to note that NetWare 6.5 ships - with bundled MySQL binaries, complete with an automatic commercial - use license for all servers running that version of NetWare. - - MySQL for NetWare is compiled using a combination of Metrowerks - CodeWarrior for NetWare and special cross-compilation versions of - the GNU autotools. + Table Incompatibilities - The latest binary packages for NetWare can be obtained at - http://dev.mysql.com/downloads/. See Section 2.1.3, "How to Get - MySQL." + After a binary upgrade to MySQL 5.1 from a MySQL 5.0 installation + that contains ARCHIVE tables, accessing those tables causes the + server to crash, even if you have run mysql_upgrade or CHECK TABLE + ... FOR UPGRADE. To work around this problem, use mysqldump to + dump all ARCHIVE tables before upgrading, and reload them into + MySQL 5.1 after upgrading. The same problem occurs for binary + downgrades from MySQL 5.1 to 5.0. + + Index Incompatibilities + + If you perform a binary upgrade without dumping and reloading + tables, you cannot upgrade directly from MySQL 4.1 to 5.1 or + higher. This occurs due to an incompatible change in the MyISAM + table index format in MySQL 5.0. Upgrade from MySQL 4.1 to 5.0 and + repair all MyISAM tables. Then upgrade from MySQL 5.0 to 5.1 and + check and repair your tables. + + Modifications to the handling of character sets or collations + might change the character sort order, which causes the ordering + of entries in any index that uses an affected character set or + collation to be incorrect. Such changes result in several possible + problems: - To host MySQL, the NetWare server must meet these requirements: + * Comparison results that differ from previous results - * The latest Support Pack of NetWare 6.5 - (http://support.novell.com/filefinder/18197/index.html) must - be installed. + * Inability to find some index values due to misordered index + entries - * The system must meet Novell's minimum requirements to run the - respective version of NetWare. + * Misordered ORDER BY results - * MySQL data and the program binaries must be installed on an - NSS volume; traditional volumes are not supported. + * Tables that CHECK TABLE reports as being in need of repair - To install MySQL for NetWare, use the following procedure: + The solution to these problems is to rebuild any indexes that use + an affected character set or collation, either by dropping and + re-creating the indexes, or by dumping and reloading the entire + table. For information about rebuilding indexes, see Section + 2.4.4, "Rebuilding or Repairing Tables or Indexes." - 1. If you are upgrading from a prior installation, stop the MySQL - server. This is done from the server console, using the - following command: -SERVER: mysqladmin -u root shutdown + To check whether a table has indexes that must be rebuilt, consult + the following list. It indicates which versions of MySQL + introduced character set or collation changes that require indexes + to be rebuilt. Each entry indicates the version in which the + change occurred and the character sets or collations that the + change affects. If the change is associated with a particular bug + report, the bug number is given. -Note - If the MySQL root user account has a password, you need to - invoke mysqladmin with the -p option and supply the password - when prompted. + The list applies both for binary upgrades and downgrades. For + example, Bug#27877: http://bugs.mysql.com/27877 was fixed in MySQL + 5.1.24 and 5.4.0, so it applies to upgrades from versions older + than 5.1.24 to 5.1.24 or newer, and to downgrades from 5.1.24 or + newer to versions older than 5.1.24. - 2. Log on to the target server from a client machine with access - to the location where you are installing MySQL. + In many cases, you can use CHECK TABLE ... FOR UPGRADE to identify + tables for which index rebuilding is required. (It will report: + Table upgrade required. Please do "REPAIR TABLE `tbl_name`" or + dump/reload to fix it!) In these cases, you can also use + mysqlcheck --check-upgrade or mysql_upgrade, which execute CHECK + TABLE. However, the use of CHECK TABLE applies only after + upgrades, not downgrades. Also, CHECK TABLE is not applicable to + all storage engines. For details about which storage engines CHECK + TABLE supports, see Section 12.5.2.3, "CHECK TABLE Syntax." - 3. Extract the binary package Zip file onto the server. Be sure - to allow the paths in the Zip file to be used. It is safe to - simply extract the file to SYS:\. - If you are upgrading from a prior installation, you may need - to copy the data directory (for example, SYS:MYSQL\DATA), as - well as my.cnf, if you have customized it. You can then delete - the old copy of MySQL. - - 4. You might want to rename the directory to something more - consistent and easy to use. The examples in this manual use - SYS:MYSQL to refer to the installation directory. - Note that MySQL installation on NetWare does not detect if a - version of MySQL is already installed outside the NetWare - release. Therefore, if you have installed the latest MySQL - version from the Web (for example, MySQL 4.1 or later) in - SYS:\MYSQL, you must rename the folder before upgrading the - NetWare server; otherwise, files in SYS:\MySQL are overwritten - by the MySQL version present in NetWare Support Pack. - - 5. At the server console, add a search path for the directory - containing the MySQL NLMs. For example: -SERVER: SEARCH ADD SYS:MYSQL\BIN - - 6. Initialize the data directory and the grant tables, if - necessary, by executing mysql_install_db at the server - console. - - 7. Start the MySQL server using mysqld_safe at the server - console. - - 8. To finish the installation, you should also add the following - commands to autoexec.ncf. For example, if your MySQL - installation is in SYS:MYSQL and you want MySQL to start - automatically, you could add these lines: -#Starts the MySQL 5.1.x database server -SEARCH ADD SYS:MYSQL\BIN -MYSQLD_SAFE - If you are running MySQL on NetWare 6.0, we strongly suggest - that you use the --skip-external-locking option on the command - line: -#Starts the MySQL 5.1.x database server -SEARCH ADD SYS:MYSQL\BIN -MYSQLD_SAFE --skip-external-locking - It is also necessary to use CHECK TABLE and REPAIR TABLE - instead of myisamchk, because myisamchk makes use of external - locking. External locking is known to have problems on NetWare - 6.0; the problem has been eliminated in NetWare 6.5. Note that - the use of MySQL on Netware 6.0 is not officially supported. - mysqld_safe on NetWare provides a screen presence. When you - unload (shut down) the mysqld_safe NLM, the screen does not go - away by default. Instead, it prompts for user input: -*<NLM has terminated; Press any key to close the screen>* - If you want NetWare to close the screen automatically instead, - use the --autoclose option to mysqld_safe. For example: -#Starts the MySQL 5.1.x database server -SEARCH ADD SYS:MYSQL\BIN -MYSQLD_SAFE --autoclose - The behavior of mysqld_safe on NetWare is described further in - Section 4.3.2, "mysqld_safe --- MySQL Server Startup Script." - - 9. When installing MySQL, either for the first time or upgrading - from a previous version, download and install the latest and - appropriate Perl module and PHP extensions for NetWare: - - + Perl: - http://forge.novell.com/modules/xfcontent/downloads.php/p - erl/Modules/ - - + PHP: - http://forge.novell.com/modules/xfcontent/downloads.php/p - hp/Modules/ - - If there was an existing installation of MySQL on the NetWare - server, be sure to check for existing MySQL startup commands in - autoexec.ncf, and edit or delete them as necessary. + Changes that cause index rebuilding to be necessary: -Note + * MySQL 5.0.48, 5.1.21 (Bug#29461: http://bugs.mysql.com/29461) + Affects indexes for columns that use any of these character + sets: eucjpms, euc_kr, gb2312, latin7, macce, ujis + Affected tables can be detected by CHECK TABLE ... FOR UPGRADE + as of MySQL 5.1.29, 5.4.0 (see + Bug#39585: http://bugs.mysql.com/39585) - The accounts that are listed in the MySQL grant tables initially - have no passwords. After starting the server, you should set up - passwords for them using the instructions in Section 2.11, - "Post-Installation Setup and Testing." + * MySQL 5.0.48, 5.1.23 (Bug#27562: http://bugs.mysql.com/27562) + Affects indexes that use the ascii_general_ci collation for + columns that contain any of these characters: '`' GRAVE + ACCENT, '[' LEFT SQUARE BRACKET, '\' REVERSE SOLIDUS, ']' + RIGHT SQUARE BRACKET, '~' TILDE + Affected tables can be detected by CHECK TABLE ... FOR UPGRADE + as of MySQL 5.1.29, 5.4.0 (see + Bug#39585: http://bugs.mysql.com/39585) -2.9. Installing MySQL from tar.gz Packages on Other Unix-Like Systems + * MySQL 5.1.24, 5.4.0 (Bug#27877: http://bugs.mysql.com/27877) + Affects indexes that use the utf8_general_ci or + ucs2_general_ci collation for columns that contain 'ß' LATIN + SMALL LETTER SHARP S (German). + Affected tables can be detected by CHECK TABLE ... FOR UPGRADE + as of MySQL 5.1.30, 5.4.0 (see + Bug#40053: http://bugs.mysql.com/40053) - This section covers the installation of MySQL binary distributions - that are provided for various platforms in the form of compressed - tar files (files with a .tar.gz extension). See Section 2.1.2.4, - "MySQL Binaries Compiled by Sun Microsystems, Inc.," for a - detailed list. +2.4.4. Rebuilding or Repairing Tables or Indexes - To obtain MySQL, see Section 2.1.3, "How to Get MySQL." + This section describes how to rebuild a table. This can be + necessitated by changes to MySQL such as how data types are + handled or changes to character set handling. For example, an + error in a collation might have been corrected, necessitating a + table rebuild to rebuild the indexes for character columns that + use the collation. It might also be that a table repair or upgrade + should be done as indicated by a table check operation such as + that performed by CHECK TABLE, mysqlcheck, or mysql_upgrade. - MySQL tar file binary distributions have names of the form - mysql-VERSION-OS.tar.gz, where VERSION is a number (for example, - 5.1.39), and OS indicates the type of operating system for which - the distribution is intended (for example, pc-linux-i686). + Methods for rebuilding a table include dumping and reloading it, + or using ALTER TABLE or REPAIR TABLE. - In addition to these generic packages, we also offer binaries in - platform-specific package formats for selected platforms. See - Section 2.2, "Standard MySQL Installation Using a Binary - Distribution," for more information on how to install these. +Note - You need the following tools to install a MySQL tar file binary - distribution: + If you are rebuilding tables because a different version of MySQL + will not handle them after a binary (in-place) upgrade or + downgrade, you must use the dump-and-reload method. Dump the + tables before upgrading or downgrading (using your original + version of MySQL), and reload the tables after upgrading or + downgrading (after installing the new version). - * GNU gunzip to uncompress the distribution. + If you use the dump-and-reload method of rebuilding tables only + for the purpose of rebuilding indexes, you can perform the dump + either before or after upgrading or downgrading. Reloading still + must be done afterward. - * A reasonable tar to unpack the distribution. GNU tar is known - to work. Some operating systems come with a preinstalled - version of tar that is known to have problems. For example, - the tar provided with early versions of Mac OS X, SunOS 4.x - and Solaris 8 and earlier are known to have problems with long - file names. On Mac OS X, you can use the preinstalled gnutar - program. On other systems with a deficient tar, you should - install GNU tar first. + To re-create a table by dumping and reloading it, use mysqldump to + create a dump file and mysql to reload the file: +shell> mysqldump db_name t1 > dump.sql +shell> mysql db_name < dump.sql - If you run into problems and need to file a bug report, please use - the instructions in Section 1.6, "How to Report Bugs or Problems." + To recreate all the tables in a single database, specify the + database name without any following table name: +shell> mysqldump db_name > dump.sql +shell> mysql db_name < dump.sql - The basic commands that you must execute to install and use a - MySQL binary distribution are: -shell> groupadd mysql -shell> useradd -g mysql mysql -shell> cd /usr/local -shell> gunzip < /path/to/mysql-VERSION-OS.tar.gz | tar xvf - -shell> ln -s full-path-to-mysql-VERSION-OS mysql -shell> cd mysql -shell> chown -R mysql . -shell> chgrp -R mysql . -shell> scripts/mysql_install_db --user=mysql -shell> chown -R root . -shell> chown -R mysql data -shell> bin/mysqld_safe --user=mysql & + To recreate all tables in all databases, use the --all-databases + option: +shell> mysqldump --all-databases > dump.sql +shell> mysql < dump.sql -Note + To rebuild a table with ALTER TABLE, use a "null" alteration; that + is, an ALTER TABLE statement that "changes" the table to use the + storage engine that it already has. For example, if t1 is a MyISAM + table, use this statement: +mysql> ALTER TABLE t1 ENGINE = MyISAM; - This procedure does not set up any passwords for MySQL accounts. - After following the procedure, proceed to Section 2.11, - "Post-Installation Setup and Testing." + If you are not sure which storage engine to specify in the ALTER + TABLE statement, use SHOW CREATE TABLE to display the table + definition. - A more detailed version of the preceding description for - installing a binary distribution follows: + If you must rebuild a table because a table checking operation + indicates that the table is corrupt or needs an upgrade, you can + use REPAIR TABLE if that statement supports the table's storage + engine. For example, to repair a MyISAM table, use this statement: +mysql> REPAIR TABLE t1; - 1. Add a login user and group for mysqld to run as: -shell> groupadd mysql -shell> useradd -g mysql mysql - These commands add the mysql group and the mysql user. The - syntax for useradd and groupadd may differ slightly on - different versions of Unix, or they may have different names - such as adduser and addgroup. - You might want to call the user and group something else - instead of mysql. If so, substitute the appropriate name in - the following steps. + For storage engines such as InnoDB that REPAIR TABLE does not + support, use mysqldump to create a dump file and mysql to reload + the file, as described earlier. - 2. Pick the directory under which you want to unpack the - distribution and change location into it. In the following - example, we unpack the distribution under /usr/local. (The - instructions, therefore, assume that you have permission to - create files and directories in /usr/local. If that directory - is protected, you must perform the installation as root.) -shell> cd /usr/local + For specifics about which storage engines REPAIR TABLE supports, + see Section 12.5.2.6, "REPAIR TABLE Syntax." - 3. Obtain a distribution file using the instructions in Section - 2.1.3, "How to Get MySQL." For a given release, binary - distributions for all platforms are built from the same MySQL - source distribution. + mysqlcheck --repair provides command-line access to the REPAIR + TABLE statement. This can be a more convenient means of repairing + tables because you can use the --databases or --all-databases + option to repair all tables in specific databases or all + databases, respectively: +shell> mysqlcheck --repair --databases db_name ... +shell> mysqlcheck --repair --all-databases - 4. Unpack the distribution, which creates the installation - directory. Then create a symbolic link to that directory: -shell> gunzip < /path/to/mysql-VERSION-OS.tar.gz | tar xvf - -shell> ln -s full-path-to-mysql-VERSION-OS mysql - The tar command creates a directory named mysql-VERSION-OS. - The ln command makes a symbolic link to that directory. This - lets you refer more easily to the installation directory as - /usr/local/mysql. - With GNU tar, no separate invocation of gunzip is necessary. - You can replace the first line with the following alternative - command to uncompress and extract the distribution: -shell> tar zxvf /path/to/mysql-VERSION-OS.tar.gz +2.4.5. Copying MySQL Databases to Another Machine - 5. Change location into the installation directory: -shell> cd mysql - You will find several files and subdirectories in the mysql - directory. The most important for installation purposes are - the bin and scripts subdirectories: + You can copy the .frm, .MYI, and .MYD files for MyISAM tables + between different architectures that support the same + floating-point format. (MySQL takes care of any byte-swapping + issues.) See Section 13.5, "The MyISAM Storage Engine." - + The bin directory contains client programs and the - server. You should add the full path name of this - directory to your PATH environment variable so that your - shell finds the MySQL programs properly. See Section - 2.14, "Environment Variables." + In cases where you need to transfer databases between different + architectures, you can use mysqldump to create a file containing + SQL statements. You can then transfer the file to the other + machine and feed it as input to the mysql client. - + The scripts directory contains the mysql_install_db - script used to initialize the mysql database containing - the grant tables that store the server access - permissions. + Use mysqldump --help to see what options are available. - 6. Ensure that the distribution contents are accessible to mysql. - If you unpacked the distribution as mysql, no further action - is required. If you unpacked the distribution as root, its - contents will be owned by root. Change its ownership to mysql - by executing the following commands as root in the - installation directory: -shell> chown -R mysql . -shell> chgrp -R mysql . - The first command changes the owner attribute of the files to - the mysql user. The second changes the group attribute to the - mysql group. + The easiest (although not the fastest) way to move a database + between two machines is to run the following commands on the + machine on which the database is located: +shell> mysqladmin -h 'other_hostname' create db_name +shell> mysqldump db_name | mysql -h 'other_hostname' db_name - 7. If you have not installed MySQL before, you must create the - MySQL data directory and initialize the grant tables: -shell> scripts/mysql_install_db --user=mysql - If you run the command as root, include the --user option as - shown. If you run the command while logged in as that user, - you can omit the --user option. - The command should create the data directory and its contents - with mysql as the owner. - After creating or updating the grant tables, you need to - restart the server manually. + If you want to copy a database from a remote machine over a slow + network, you can use these commands: +shell> mysqladmin create db_name +shell> mysqldump -h 'other_hostname' --compress db_name | mysql db_na +me - 8. Most of the MySQL installation can be owned by root if you - like. The exception is that the data directory must be owned - by mysql. To accomplish this, run the following commands as - root in the installation directory: -shell> chown -R root . -shell> chown -R mysql data + You can also store the dump in a file, transfer the file to the + target machine, and then load the file into the database there. + For example, you can dump a database to a compressed file on the + source machine like this: +shell> mysqldump --quick db_name | gzip > db_name.gz - 9. If you want MySQL to start automatically when you boot your - machine, you can copy support-files/mysql.server to the - location where your system has its startup files. More - information can be found in the support-files/mysql.server - script itself and in Section 2.11.2.2, "Starting and Stopping - MySQL Automatically." - 10. You can set up new accounts using the bin/mysql_setpermission - script if you install the DBI and DBD::mysql Perl modules. See - Section 4.6.14, "mysql_setpermission --- Interactively Set - Permissions in Grant Tables." For Perl module installation - instructions, see Section 2.15, "Perl Installation Notes." - 11. If you would like to use mysqlaccess and have the MySQL - distribution in some nonstandard location, you must change the - location where mysqlaccess expects to find the mysql client. - Edit the bin/mysqlaccess script at approximately line 18. - Search for a line that looks like this: -$MYSQL = '/usr/local/bin/mysql'; # path to mysql executable - Change the path to reflect the location where mysql actually - is stored on your system. If you do not do this, a Broken pipe - error will occur when you run mysqlaccess. + Transfer the file containing the database contents to the target + machine and run these commands there: +shell> mysqladmin create db_name +shell> gunzip < db_name.gz | mysql db_name - After everything has been unpacked and installed, you should test - your distribution. To start the MySQL server, use the following - command: -shell> bin/mysqld_safe --user=mysql & + You can also use mysqldump and mysqlimport to transfer the + database. For large tables, this is much faster than simply using + mysqldump. In the following commands, DUMPDIR represents the full + path name of the directory you use to store the output from + mysqldump. - If you run the command as root, you must use the --user option as - shown. The value of the option is the name of the login account - that you created in the first step to use for running the server. - If you run the command while logged in as mysql, you can omit the - --user option. + First, create the directory for the output files and dump the + database: +shell> mkdir DUMPDIR +shell> mysqldump --tab=DUMPDIR db_name - If the command fails immediately and prints mysqld ended, you can - find some information in the host_name.err file in the data - directory. + Then transfer the files in the DUMPDIR directory to some + corresponding directory on the target machine and load the files + into MySQL there: +shell> mysqladmin create db_name # create database +shell> cat DUMPDIR/*.sql | mysql db_name # create tables in databas +e +shell> mysqlimport db_name DUMPDIR/*.txt # load data into tables - More information about mysqld_safe is given in Section 4.3.2, - "mysqld_safe --- MySQL Server Startup Script." + Do not forget to copy the mysql database because that is where the + grant tables are stored. You might have to run commands as the + MySQL root user on the new machine until you have the mysql + database in place. -Note + After you import the mysql database on the new machine, execute + mysqladmin flush-privileges so that the server reloads the grant + table information. - The accounts that are listed in the MySQL grant tables initially - have no passwords. After starting the server, you should set up - passwords for them using the instructions in Section 2.11, - "Post-Installation Setup and Testing." +2.5. Installing MySQL on Windows -2.10. MySQL Installation Using a Source Distribution + This section describes the process for installing MySQL on + Windows. - Before you proceed with an installation from source, first check - whether our binary is available for your platform and whether it - works for you. We put a great deal of effort into ensuring that - our binaries are built with the best possible options. + To run MySQL on Windows, you need the following: - To obtain a source distribution for MySQL, Section 2.1.3, "How to - Get MySQL." If you want to build MySQL from source on Windows, see - Section 2.10.6, "Installing MySQL from Source on Windows." + * A Windows operating system such as Windows 2000, Windows XP, + Windows Vista, Windows Server 2003, or Windows Server 2008. + Both 32-bit and 64-bit versions are supported. + In addition to running MySQL as a standard application, you + can also run the MySQL server as a Windows service. By using a + service you can monitor and control the operation of the + server through the standard Windows service management tools. + For more information, see Section 2.5.5.6, "Starting MySQL as + a Windows Service." + Generally, you should install MySQL on Windows using an + account that has administrator rights. Otherwise, you may + encounter problems with certain operations such as editing the + PATH environment variable or accessing the Service Control + Manager. Once installed, MySQL does not need to be executed + using a user with Administrator privileges. - MySQL source distributions are provided as compressed tar archives - and have names of the form mysql-VERSION.tar.gz, where VERSION is - a number like 5.1.39. + * TCP/IP protocol support. - You need the following tools to build and install MySQL from - source: + * Enough space on the hard drive to unpack, install, and create + the databases in accordance with your requirements (generally + a minimum of 200 megabytes is recommended.) - * GNU gunzip to uncompress the distribution. + For a list of limitations within the Windows version of MySQL, see + Section D.7.3, "Windows Platform Limitations." - * A reasonable tar to unpack the distribution. GNU tar is known - to work. Some operating systems come with a preinstalled - version of tar that is known to have problems. For example, - the tar provided with early versions of Mac OS X, SunOS 4.x - and Solaris 8 and earlier are known to have problems with long - file names. On Mac OS X, you can use the preinstalled gnutar - program. On other systems with a deficient tar, you should - install GNU tar first. + In addition to the MySQL Server package, you may need or want + additional components to use MySQL with your application or + development environment. These include, but are not limited to: - * A working ANSI C++ compiler. gcc 2.95.2 or later, SGI C++, and - SunPro C++ are some of the compilers that are known to work. - libg++ is not needed when using gcc. gcc 2.7.x has a bug that - makes it impossible to compile some perfectly legal C++ files, - such as sql/sql_base.cc. If you have only gcc 2.7.x, you must - upgrade your gcc to be able to compile MySQL. gcc 2.8.1 is - also known to have problems on some platforms, so it should be - avoided if a newer compiler exists for the platform. gcc - 2.95.2 or later is recommended. + * If you plan to connect to the MySQL server via ODBC, you need + a Connector/ODBC driver. For more information, including + installation and configuration instructions, see Section 21.1, + "MySQL Connector/ODBC." + + * If you plan to use MySQL server with .NET applications, you + need the Connector/NET driver. For more information, including + installation and configuration instructions, see Section 21.2, + "MySQL Connector/NET." - * A good make program. GNU make is always recommended and is - sometimes required. (BSD make fails, and vendor-provided make - implementations may fail as well.) If you have problems, use - GNU make 3.75 or newer. + MySQL distributions for Windows can be downloaded from + http://dev.mysql.com/downloads/. See Section 2.1.3, "How to Get + MySQL." - * libtool 1.5.24 or later is also recommended. + MySQL for Windows is available in several distribution formats, + detailed below. Generally speaking, you should use a binary + distribution that includes an installer. It is simpler to use than + the others, and you need no additional tools to get MySQL up and + running. The installer for the Windows version of MySQL, combined + with a GUI Config Wizard, automatically installs MySQL, creates an + option file, starts the server, and secures the default user + accounts. - If you are using a version of gcc recent enough to understand the - -fno-exceptions option, it is very important that you use this - option. Otherwise, you may compile a binary that crashes randomly. - Also use -felide-constructors and -fno-rtti along with - -fno-exceptions. When in doubt, do the following: -CFLAGS="-O3" CXX=gcc CXXFLAGS="-O3 -felide-constructors \ - -fno-exceptions -fno-rtti" ./configure \ - --prefix=/usr/local/mysql --enable-assembler \ - --with-mysqld-ldflags=-all-static + * Binary installer distribution. The installable distribution + comes packaged as a Microsoft Windows Installer (MSI) package + that you can install manually or automatically on your + systems. Two formats are available, an essentials package that + contains all the files you need to install and configure + MySQL, but no additional components, and a complete package + that includes MySQL, configuration tools, benchmarks and other + components. For more information on the specific differences, + see Section 2.5.2, "Choosing An Installation Package" + For instructions on installing MySQL using one of the MSI + installation packages, see Section 2.5.3, "Installing MySQL + with the MSI Package." + + * Standard binary distribution format packaged as a Zip file + containing all of the necessary files that you unpack into + your chosen location. This package contains all of the files + in the full Windows MSI Installer package, but does not + including an installation program. + For instructions on installing MySQL using the Zip file, see + Section 2.5.5, "Installing MySQL from a noinstall Zip + Archive." - On most systems, this gives you a fast and stable binary. + * The source distribution contains all the code and support + files for building the executables using the Visual Studio + compiler system. + For instructions on building MySQL from source on Windows, see + Section 2.5.10, "Installing MySQL from Source on Windows." - If you run into problems and need to file a bug report, please use - the instructions in Section 1.6, "How to Report Bugs or Problems." + MySQL on Windows considerations: -2.10.1. Source Installation Overview + * Large Table Support + If you need tables with a size larger than 4GB, install MySQL + on an NTFS or newer file system. Don't forget to use MAX_ROWS + and AVG_ROW_LENGTH when you create tables. See Section + 12.1.17, "CREATE TABLE Syntax." - The basic commands that you must execute to install a MySQL source - distribution are: -shell> groupadd mysql -shell> useradd -g mysql mysql -shell> gunzip < mysql-VERSION.tar.gz | tar -xvf - -shell> cd mysql-VERSION -shell> ./configure --prefix=/usr/local/mysql -shell> make -shell> make install -shell> cp support-files/my-medium.cnf /etc/my.cnf -shell> cd /usr/local/mysql -shell> chown -R mysql . -shell> chgrp -R mysql . -shell> bin/mysql_install_db --user=mysql -shell> chown -R root . -shell> chown -R mysql var -shell> bin/mysqld_safe --user=mysql & + * MySQL and Virus Checking Software + Using virus scanning software such as Norton/Symantec + Anti-Virus on directories containing MySQL data and temporary + tables can cause issues, both in terms of the performance of + MySQL and the virus-scanning software mis-identifying the + contents of the files as containing spam. This is because of + the fingerprinting mechanism used by the virus scanning + software, and the way in which MySQL rapidly updates different + files, which may be identified as a potential security risk. + After installing MySQL Server, it is recommended that you + disable virus scanning on the main directory (datadir) being + used to store your MySQL table data. There is usually a system + built into the virus scanning software to allow certain + directories to be specifically ignored during virus scanning. + In addition, by default, MySQL creates temporary files in the + standard Windows temporary directory. To prevent the temporary + files also being scanned, you should configure a separate + temporary directory for MySQL temporary files and add this to + the virus scanning exclusion list. To do this, add a + configuration option for the tmpdir parameter to your my.ini + configuration file. For more information, see Section 2.5.5.2, + "Creating an Option File." - If you start from a source RPM, do the following: -shell> rpmbuild --rebuild --clean MySQL-VERSION.src.rpm +2.5.1. Windows Installation Layout - This makes a binary RPM that you can install. For older versions - of RPM, you may have to replace the command rpmbuild with rpm - instead. + For MySQL 5.1 on Windows, the default installation directory is + C:\Program Files\MySQL\MySQL Server 5.1. Some Windows users prefer + to install in C:\mysql, the directory that formerly was used as + the default. However, the layout of the subdirectories remains the + same. + + For MySQL 5.1.23 and earlier, all of the files are located within + this parent directory, using the following structure: + + Table 2.2. Installation Layout for Windows using MySQL 5.1.23 and + earlier + Directory Contents of Directory + bin Client programs and the mysqld server + data Log files, databases + Docs Manual in CHM format + examples Example programs and scripts + include Include (header) files + lib Libraries + scripts Utility scripts + share Error message files -Note + For MySQL 5.1.24 and later, the default location of data directory + was changed. The remainder of the directory structure remains the + same: - This procedure does not set up any passwords for MySQL accounts. - After following the procedure, proceed to Section 2.11, - "Post-Installation Setup and Testing," for post-installation setup - and testing. + Table 2.3. Installation Layout for Windows using MySQL 5.1.24 and + later + Directory Contents of Directory + bin Client programs and the mysqld server + C:\Documents and Settings\All Users\Application Data\MySQL Log + files, databases + Docs Manual in CHM format + examples Example programs and scripts + include Include (header) files + lib Libraries + scripts Utility scripts + share Error message files - A more detailed version of the preceding description for - installing MySQL from a source distribution follows: +2.5.2. Choosing An Installation Package - 1. Add a login user and group for mysqld to run as: -shell> groupadd mysql -shell> useradd -g mysql mysql - These commands add the mysql group and the mysql user. The - syntax for useradd and groupadd may differ slightly on - different versions of Unix, or they may have different names - such as adduser and addgroup. - You might want to call the user and group something else - instead of mysql. If so, substitute the appropriate name in - the following steps. + For MySQL 5.1, there are three installation packages to choose + from when installing MySQL on Windows: + Packaging + Feature Essentials Complete Zip (No-install) + Installer Yes Yes No + Directory-only + MySQL Server Instance Config Wizard Yes Yes No + Test Suite No Yes Yes + MySQL Server Yes Yes Yes + MySQL Client Programs Yes Yes Yes + C Headers/Libraries Yes Yes Yes + Embedded Server No Optional Yes + Scripts and Examples No Optional Yes + + In the above table: + + * Yes indiciates that the component is installed by default. + + * No indicates that the component is not installed or included. + + * Optional indicates that the component is included with the + package, but not installed unless explicitly requested using + the Custom installation mode. + + The workflow for installing using the MSI installer is shown + below: - 2. Perform the following steps as the mysql user, except as - noted. + Figure 2.1. Installation Workflow for Windows using MSI + Installation Workflow for Windows using MSI - 3. Pick the directory under which you want to unpack the - distribution and change location into it. + The workflow for installing using the MSI installer is shown + below: - 4. Obtain a distribution file using the instructions in Section - 2.1.3, "How to Get MySQL." + Figure 2.2. Installation Workflow for Windows using Zip + Installation Workflow for Windows using Zip - 5. Unpack the distribution into the current directory: -shell> gunzip < /path/to/mysql-VERSION.tar.gz | tar xvf - - This command creates a directory named mysql-VERSION. - With GNU tar, no separate invocation of gunzip is necessary. - You can use the following alternative command to uncompress - and extract the distribution: -shell> tar zxvf /path/to/mysql-VERSION-OS.tar.gz +Note - 6. Change location into the top-level directory of the unpacked - distribution: -shell> cd mysql-VERSION - Note that currently you must configure and build MySQL from - this top-level directory. You cannot build it in a different - directory. + For the Essentials and Complete packages in the MSI installer, you + can select individual components to be installed by using the + Custom mode, including disable the components confiurated for + installation by default. + + Full details on the components are suggested uses are provided + below for reference: + + * Windows Essentials --- this package has a file name similar to + mysql-essential-5.1.41-win32.msi and is supplied as a + Microsoft Installer (MSI) package. The package includes the + minimum set of files needed to install MySQL on Windows, + including the MySQL Server Instance Config Wizard. This + package does not include optional components such as the + embedded server, developer headers and libraries or benchmark + suite. + To install using this package, see Section 2.5.3, "Installing + MySQL with the MSI Package." + + * Windows MSI Installer (Complete) --- this package has a file + name similar to mysql-5.1.41-win32.zip and contains all files + needed for a complete Windows installation, including the + MySQL Server Instance Config Wizard. This package includes + optional components such as the embedded server and benchmark + suite. + To install using this package, see Section 2.5.3, "Installing + MySQL with the MSI Package." - 7. Configure the release and compile everything: -shell> ./configure --prefix=/usr/local/mysql -shell> make - When you run configure, you might want to specify other - options. Run ./configure --help for a list of options. Section - 2.10.2, "Typical configure Options," discusses some of the - more useful options. - If configure fails and you are going to send mail to a MySQL - mailing list to ask for assistance, please include any lines - from config.log that you think can help solve the problem. - Also include the last couple of lines of output from - configure. To file a bug report, please use the instructions - in Section 1.6, "How to Report Bugs or Problems." - If the compile fails, see Section 2.10.4, "Dealing with - Problems Compiling MySQL," for help. + * Without installer --- this package has a file name similar to + mysql-noinstall-5.1.41-win32.zip and contains all the files + found in the Complete install package, with the exception of + the MySQL Server Instance Config Wizard. This package does not + include an automated installer, and must be manually installed + and configured. + + The Essentials package is recommended for most users. Both the + Essentials and Complete distributions are available as an .msi + file for use with the Windows Installer. The Noinstall + distribution is packaged as Zip archives. To use Zip archives, you + must have a tool that can unpack .zip files. + + When using the MSI installers you can automate the installation + process. For more information, see Section 2.5.3.2, "Installing + MySQL Automatically using MSI." To automate the creation of a + MySQL instance, see Section 2.5.4.13, "Creating an Instance from + the Command Line." - 8. Install the distribution: -shell> make install - You might need to run this command as root. - If you want to set up an option file, use one of those present - in the support-files directory as a template. For example: -shell> cp support-files/my-medium.cnf /etc/my.cnf - You might need to run this command as root. - If you want to configure support for InnoDB tables, you should - edit the /etc/my.cnf file, remove the # character before the - option lines that start with innodb_..., and modify the option - values to be what you want. See Section 4.2.3.3, "Using Option - Files," and Section 13.6.2, "InnoDB Configuration." + Your choice of install package affects the installation process + you must follow. If you choose to install either the Essentials or + Complete install packages, see Section 2.5.3, "Installing MySQL + with the MSI Package." If you choose to install MySQL from the + Noinstall archive, see Section 2.5.5, "Installing MySQL from a + noinstall Zip Archive." - 9. Change location into the installation directory: -shell> cd /usr/local/mysql - 10. If you ran the make install command as root, the installed - files will be owned by root. Ensure that the installation is - accessible to mysql by executing the following commands as - root in the installation directory: -shell> chown -R mysql . -shell> chgrp -R mysql . - The first command changes the owner attribute of the files to - the mysql user. The second changes the group attribute to the - mysql group. - 11. If you have not installed MySQL before, you must create the - MySQL data directory and initialize the grant tables: -shell> bin/mysql_install_db --user=mysql - If you run the command as root, include the --user option as - shown. If you run the command while logged in as mysql, you - can omit the --user option. - The command should create the data directory and its contents - with mysql as the owner. - After using mysql_install_db to create the grant tables for - MySQL, you must restart the server manually. The mysqld_safe - command to do this is shown in a later step. - 12. Most of the MySQL installation can be owned by root if you - like. The exception is that the data directory must be owned - by mysql. To accomplish this, run the following commands as - root in the installation directory: -shell> chown -R root . -shell> chown -R mysql var - 13. If you want MySQL to start automatically when you boot your - machine, you can copy support-files/mysql.server to the - location where your system has its startup files. More - information can be found in the support-files/mysql.server - script itself; see also Section 2.11.2.2, "Starting and - Stopping MySQL Automatically." - 14. You can set up new accounts using the bin/mysql_setpermission - script if you install the DBI and DBD::mysql Perl modules. See - Section 4.6.14, "mysql_setpermission --- Interactively Set - Permissions in Grant Tables." For Perl module installation - instructions, see Section 2.15, "Perl Installation Notes." +2.5.3. Installing MySQL with the MSI Package + + The MSI package are designed to install and configure MySQL in + such a way that you can immediately get started using MySQL. + + The MySQL Installation Wizard and MySQL Config Wizard are + available in the Essentials and Complete install packages. They + are recommended for most standard MySQL installations. Exceptions + include users who need to install multiple instances of MySQL on a + single server host and advanced users who want complete control of + server configuration. - After everything has been installed, you should test your - distribution. To start the MySQL server, use the following - command: -shell> /usr/local/mysql/bin/mysqld_safe --user=mysql & + * For information on installing using the GUI MSI installer + process, see Section 2.5.3, "Installing MySQL with the MSI + Package." - If you run the command as root, you should use the --user option - as shown. The value of the option is the name of the login account - that you created in the first step to use for running the server. - If you run the command while logged in as that user, you can omit - the --user option. + * For information on installing using the command line using the + MSI package, see Section 2.5.3.2, "Installing MySQL + Automatically using MSI." + + * If you have previously installed MySQL using the MSI package + and want to remove MySQL, see Section 2.5.3.3, "Removing MySQL + Installed from the MSI Package." - If the command fails immediately and prints mysqld ended, you can - find some information in the host_name.err file in the data - directory. + The workflow sequence for using the installer is shown in the + figure below: - More information about mysqld_safe is given in Section 4.3.2, - "mysqld_safe --- MySQL Server Startup Script." + Figure 2.3. Installation Workflow for Windows using MSI Installer + Installation Workflow for Windows using MSI Installer Note - The accounts that are listed in the MySQL grant tables initially - have no passwords. After starting the server, you should set up - passwords for them using the instructions in Section 2.11, - "Post-Installation Setup and Testing." + Microsoft Windows XP and later include a firewall which + specifically blocks ports. If you plan on using MySQL through a + network port then you should open and create an exception for this + port before performing the installation. To check and if necessary + add an exception to the firewall settings: -2.10.2. Typical configure Options + 1. First ensure that you are logged in as an Administrator or a + user with Administrator privileges. - The configure script gives you a great deal of control over how - you configure a MySQL source distribution. Typically you do this - using options on the configure command line. You can also affect - configure using certain environment variables. See Section 2.14, - "Environment Variables." For a full list of options supported by - configure, run this command: -shell> ./configure --help + 2. Go to the Control Panel, and double click the Windows Firewall + icon. - A list of the available configure options is provided in the table - below. + 3. Choose the Allow a program through Windows Firewall option and + click the Add port button. - Table 2.1. Build (configure) Reference - Formats Description Default Introduced Removed - --bindir=DIR User executables EPREFIX/bin - --build=BUILD Configure for building on BUILD guessed - --cache-file=FILE Cache test results in FILE disabled - -C Alias for `--cache-file=config.cache' - --config-cache - --datadir=DIR Read-only architecture-independent data PREFIX/share + 4. Enter MySQL into the Name text box and 3306 (or the port of + your choice) into the Port number text box. - --disable-FEATURE Do not include FEATURE - --disable-dependency-tracking Disable dependency tracking - --disable-grant-options Disable GRANT options - --disable-largefile Omit support for large files - --disable-libtool-lock Disable libtool lock - --disable-thread-safe-client Compile the client without threads - 5.1.7 - --enable-FEATURE Enable FEATURE - --enable-assembler Use assembler versions of some string functions - if available - --enable-dependency-tracking Do not reject slow dependency - extractors - --enable-fast-install Optimize for fast installation yes - --enable-local-infile Enable LOAD DATA LOCAL INFILE disabled - --enable-shared Build shared libraries yes - --enable-static Build static libraries yes - --enable-thread-safe-client Compile the client with threads - --exec-prefix=EPREFIX Install architecture-dependent files in - EPREFIX - -h Display this help and exit - --help - --help=short Display options specific to this package - --help=recursive Display the short help of all the included - packages - --host=HOST Cross-compile to build programs to run on HOST - --includedir=DIR C header files PREFIX/include - --infodir=DIR Info documentation PREFIX/info - --libdir=DIR Object code libraries EPREFIX/lib - --libexecdir=DIR Program executables EPREFIX/libexec - --localstatedir=DIR Modifiable single-machine data PREFIX/var - --mandir=DIR man documentation PREFIX/man - -n Do not create output files - --no-create - --oldincludedir=DIR C header files for non-gcc /usr/include - --prefix=PREFIX Install architecture-independent files in PREFIX + 5. Also ensure that the TCP protocol radio button is selected. - --program-prefix=PREFIX Prepend PREFIX to installed program names + 6. If you wish, you can also limit access to the MySQL server by + choosing the Change scope button. - --program-suffix=SUFFIX Append SUFFIX to installed program names + 7. Confirm your choices by clicking the OK button. - --program-transform-name=PROGRAM run sed PROGRAM on installed - program names - -q Do not print `checking...' messages - --quiet - --sbindir=DIR System admin executables EPREFIX/sbin - --sharedstatedir=DIR Modifiable architecture-independent data - PREFIX/com - --srcdir=DIR Find the sources in DIR configure directory or .. - --sysconfdir=DIR Read-only single-machine data PREFIX/etc - --target=TARGET Configure for building compilers for TARGET - -V Display version information and exit - --version - --with-PACKAGE Use PACKAGE - --with-archive-storage-engine Enable the Archive Storage Engine no + Additionally, when running the MySQL Installation Wizard on + Windows Vista, ensure that you are logged in as a user with + administrative rights. - --with-atomic-ops Implement atomic operations using pthread - rwlocks or atomic CPU instructions for multi-processor 5.1.12 - --with-berkeley-db Use BerkeleyDB located in DIR no - --with-berkeley-db-includes Find Berkeley DB headers in DIR - --with-berkeley-db-libs Find Berkeley DB libraries in DIR - --with-big-tables Support tables with more than 4 G rows even on - 32 bit platforms - --with-blackhole-storage-engine Enable the Blackhole Storage - Engine no - --with-charset Default character set - --with-client-ldflags Extra linking arguments for clients - --with-collation Default collation - --with-comment Comment about compilation environment - --with-csv-storage-engine Enable the CSV Storage Engine yes - --with-darwin-mwcc Use Metrowerks CodeWarrior wrappers on OS - X/Darwin - --with-debug Add debug code 5.1.7 - --with-debug=full Add debug code (adds memory checker, very slow) +Note - --with-embedded-privilege-control Build parts to check user's - privileges (only affects embedded library) - --with-embedded-server Build the embedded server - --with-error-inject Enable error injection in MySQL Server - 5.1.11 - --with-example-storage-engine Enable the Example Storage Engine no + When using Windows Vista, you may want to disable User Account + Control (UAC) before performing the installation. If you do not do + so, then MySQL may be identified as a security risk, which will + mean that you need to enable MySQL. You can disable the security + checking by following these instructions: - --with-extra-charsets Use charsets in addition to default - --with-fast-mutexes Compile with fast mutexes enabled 5.1.5 - --with-federated-storage-engine Enable federated storage engine no - 5.1.3 5.1.9 - --with-gnu-ld Assume the C compiler uses GNU ld no - --with-innodb Enable innobase storage engine no 5.1.3 5.1.9 - --with-lib-ccflags Extra CC options for libraries - --with-libwrap=DIR Compile in libwrap (tcp_wrappers) support - --with-low-memory Try to use less memory to compile to avoid - memory limitations - --with-machine-type Set the machine type, like "powerpc" - --with-max-indexes=N Sets the maximum number of indexes per table - 64 - --with-mysqld-ldflags Extra linking arguments for mysqld - --with-mysqld-libs Extra libraries to link with for mysqld - --with-mysqld-user What user the mysqld daemon shall be run as + 1. Open Control Panel. - --with-mysqlmanager Build the mysqlmanager binary Build if server - is built - --with-named-curses-libs Use specified curses libraries - --with-named-thread-libs Use specified thread libraries - --with-ndb-ccflags Extra CC options for ndb compile - --with-ndb-docs Include the NDB Cluster ndbapi and mgmapi - documentation - --with-ndb-port Port for NDB Cluster management server - --with-ndb-port-base Port for NDB Cluster management server - --with-ndb-sci=DIR Provide MySQL with a custom location of sci - library - --with-ndb-test Include the NDB Cluster ndbapi test programs - --with-ndbcluster Include the NDB Cluster table handler no - --with-openssl=DIR Include the OpenSSL support - --with-openssl-includes Find OpenSSL headers in DIR - --with-openssl-libs Find OpenSSL libraries in DIR - --with-other-libc=DIR Link against libc and other standard - libraries installed in the specified nonstandard location - --with-pic Try to use only PIC/non-PIC objects Use both - --with-plugin-PLUGIN Forces the named plugin to be linked into - mysqld statically 5.1.11 - --with-plugins Plugins to include in mysqld none 5.1.11 - --with-pstack Use the pstack backtrace library - --with-pthread Force use of pthread library - --with-row-based-replication Include row-based replication 5.1.5 - 5.1.6 - --with-server-suffix Append value to the version string - --with-ssl=DIR Include SSL support 5.1.11 - --with-system-type Set the system type, like "sun-solaris10" - --with-tags Include additional configurations automatic - --with-tcp-port Which port to use for MySQL services 3306 - --with-unix-socket-path Where to put the unix-domain socket - --with-yassl Include the yaSSL support - --with-zlib-dir=no|bundled|DIR Provide MySQL with a custom - location of compression library - --without-PACKAGE Do not use PACKAGE - --without-bench Skip building of the benchmark suite - --without-debug Build a production version without debugging code + 2. Under the User Accounts and Family Safety, select Add or + remove user accounts. - --without-docs Skip building of the documentation - --without-extra-tools Skip building utilities in the tools - directory - --without-geometry Do not build geometry-related parts - --without-libedit Use system libedit instead of bundled copy - --without-man Skip building of the man pages - --without-ndb-binlog Disable ndb binlog 5.1.6 - --without-ndb-debug Disable special ndb debug features - --without-plugin-PLUGIN Exclude PLUGIN 5.1.11 - --without-query-cache Do not build query cache - --without-readline Use system readline instead of bundled copy + 3. Click on the Got to the main User Accounts page link. - --without-row-based-replication Don't include row-based - replication 5.1.7 5.1.14 - --without-server Only build the client - --without-uca Skip building of the national Unicode collations + 4. Click on Turn User Account Control on or off. You may be + prompted to provide permission to change this setting. Click + Continue. - Some of the configure options available are described here. For - options that may be of use if you have difficulties building - MySQL, see Section 2.10.4, "Dealing with Problems Compiling - MySQL." + 5. Deselect or unceck the checkbox next to Use User Account + Control (UAC) to help protect your computer. Click OK to save + the setting. - * To compile just the MySQL client libraries and client programs - and not the server, use the --without-server option: -shell> ./configure --without-server - If you have no C++ compiler, some client programs such as - mysql cannot be compiled because they require C++.. In this - case, you can remove the code in configure that tests for the - C++ compiler and then run ./configure with the - --without-server option. The compile step should still try to - build all clients, but you can ignore any warnings about files - such as mysql.cc. (If make stops, try make -k to tell it to - continue with the rest of the build even if errors occur.) + You will need to restart to complete the process. Click Restart + Now to reboot the machine and apply the changes. You can then + follow the instructions below for installing Windows. - * If you want to build the embedded MySQL library (libmysqld.a), - use the --with-embedded-server option. +2.5.3.1. Using the MySQL Installation Wizard - * If you don't want your log files and database directories - located under /usr/local/var, use a configure command - something like one of these: -shell> ./configure --prefix=/usr/local/mysql -shell> ./configure --prefix=/usr/local \ - --localstatedir=/usr/local/mysql/data - The first command changes the installation prefix so that - everything is installed under /usr/local/mysql rather than the - default of /usr/local. The second command preserves the - default installation prefix, but overrides the default - location for database directories (normally /usr/local/var) - and changes it to /usr/local/mysql/data. - You can also specify the installation directory and data - directory locations at server startup time by using the - --basedir and --datadir options. These can be given on the - command line or in an MySQL option file, although it is more - common to use an option file. See Section 4.2.3.3, "Using - Option Files." + MySQL Installation Wizard is an installer for the MySQL server + that uses the latest installer technologies for Microsoft Windows. + The MySQL Installation Wizard, in combination with the MySQL + Config Wizard, allows a user to install and configure a MySQL + server that is ready for use immediately after installation. - * If you are using Unix and you want the MySQL socket file - location to be somewhere other than the default location - (normally in the directory /tmp or /var/run), use a configure - command like this: -shell> ./configure \ - --with-unix-socket-path=/usr/local/mysql/tmp/mysql.sock - The socket file name must be an absolute path name. You can - also change the location of mysql.sock at server startup by - using a MySQL option file. See Section B.1.4.5, "How to - Protect or Change the MySQL Unix Socket File." + The MySQL Installation Wizard uses the standard Microsoft + Installer Engine (MSI) system is the standard installer for all + MySQL server distributions, version 4.1.5 and higher. Users of + previous versions of MySQL need to shut down and remove their + existing MySQL installations manually before installing MySQL with + the MySQL Installation Wizard. See Section 2.5.3.1.6, "Upgrading + MySQL with the Installation Wizard," for more information on + upgrading from a previous version. - * If you want to compile statically linked programs (for - example, to make a binary distribution, to get better - performance, or to work around problems with some Red Hat - Linux distributions), run configure like this: -shell> ./configure --with-client-ldflags=-all-static \ - --with-mysqld-ldflags=-all-static + If you are upgrading an installation from MySQL 5.1.31 or earlier + to MySQL 5.1.32 or later, read the notes provided in Section + 2.5.3.1.6, "Upgrading MySQL with the Installation Wizard." - * If you are using gcc and don't have libg++ or libstdc++ - installed, you can tell configure to use gcc as your C++ - compiler: -shell> CC=gcc CXX=gcc ./configure - When you use gcc as your C++ compiler, it does not attempt to - link in libg++ or libstdc++. This may be a good thing to do - even if you have those libraries installed. Some versions of - them have caused strange problems for MySQL users in the past. - The following list indicates some compilers and environment - variable settings that are commonly used with each one. + The Microsoft Windows Installer Engine was updated with the + release of Windows XP; those using a previous version of Windows + can reference this Microsoft Knowledge Base article + (http://support.microsoft.com/default.aspx?scid=kb;EN-US;292539) + for information on upgrading to the latest version of the Windows + Installer Engine. - + gcc 2.7.2: -CC=gcc CXX=gcc CXXFLAGS="-O3 -felide-constructors" + In addition, Microsoft has introduced the WiX (Windows Installer + XML) toolkit. This is the first highly acknowledged Open Source + project from Microsoft. We have switched to WiX because it is an + Open Source project and it allows us to handle the complete + Windows installation process in a flexible manner using scripts. - + gcc 2.95.2: -CFLAGS="-O3 -mpentiumpro" CXX=gcc CXXFLAGS="-O3 -mpentiumpro \ --felide-constructors -fno-exceptions -fno-rtti" + Improving the MySQL Installation Wizard depends on the support and + feedback of users like you. If you find that the MySQL + Installation Wizard is lacking some feature important to you, or + if you discover a bug, please report it in our bugs database using + the instructions given in Section 1.6, "How to Report Bugs or + Problems." - + pgcc 2.90.29 or newer: -CFLAGS="-O3 -mpentiumpro -mstack-align-double" CXX=gcc \ -CXXFLAGS="-O3 -mpentiumpro -mstack-align-double \ --felide-constructors -fno-exceptions -fno-rtti" - In most cases, you can get a reasonably optimized MySQL binary - by using the options from the preceding list and adding the - following options to the configure line: ---prefix=/usr/local/mysql --enable-assembler \ ---with-mysqld-ldflags=-all-static - The full configure line would, in other words, be something - like the following for all recent gcc versions: -CFLAGS="-O3 -mpentiumpro" CXX=gcc CXXFLAGS="-O3 -mpentiumpro \ --felide-constructors -fno-exceptions -fno-rtti" ./configure \ ---prefix=/usr/local/mysql --enable-assembler \ ---with-mysqld-ldflags=-all-static - The binaries we provide on the MySQL Web site at - http://dev.mysql.com/downloads/ are all compiled with full - optimization and should be perfect for most users. See Section - 2.1.2.4, "MySQL Binaries Compiled by Sun Microsystems, Inc.." - There are some configuration settings you can tweak to build - an even faster binary, but these are only for advanced users. - See Section 7.5.1, "How Compiling and Linking Affects the - Speed of MySQL." - If the build fails and produces errors about your compiler or - linker not being able to create the shared library - libmysqlclient.so.N (where N is a version number), you can - work around this problem by giving the --disable-shared option - to configure. In this case, configure does not build a shared - libmysqlclient.so.N library. +2.5.3.1.1. Downloading and Starting the MySQL Installation Wizard - * By default, MySQL uses the latin1 (cp1252 West European) - character set. To change the default set, use the - --with-charset option: -shell> ./configure --with-charset=CHARSET - CHARSET may be one of binary, armscii8, ascii, big5, cp1250, - cp1251, cp1256, cp1257, cp850, cp852, cp866, cp932, dec8, - eucjpms, euckr, gb2312, gbk, geostd8, greek, hebrew, hp8, - keybcs2, koi8r, koi8u, latin1, latin2, latin5, latin7, macce, - macroman, sjis, swe7, tis620, ucs2, ujis, utf8. See Section - 9.2, "The Character Set Used for Data and Sorting." - (Additional character sets might be available. Check the - output from ./configure --help for the current list.) - The default collation may also be specified. MySQL uses the - latin1_swedish_ci collation by default. To change this, use - the --with-collation option: -shell> ./configure --with-collation=COLLATION - To change both the character set and the collation, use both - the --with-charset and --with-collation options. The collation - must be a legal collation for the character set. (Use the SHOW - COLLATION statement to determine which collations are - available for each character set.) - With the configure option --with-extra-charsets=LIST, you can - define which additional character sets should be compiled into - the server. LIST is one of the following: + The MySQL installation packages can be downloaded from + http://dev.mysql.com/downloads/. If the package you download is + contained within a Zip archive, you need to extract the archive + first. - + A list of character set names separated by spaces + The process for starting the wizard depends on the contents of the + installation package you download. If there is a setup.exe file + present, double-click it to start the installation process. If + there is an .msi file present, double-click it to start the + installation process. - + complex to include all character sets that can't be - dynamically loaded +2.5.3.1.2. Choosing an Install Type - + all to include all character sets into the binaries - Clients that want to convert characters between the server and - the client should use the SET NAMES statement. See Section - 5.1.5, "Session System Variables," and Section 9.1.4, - "Connection Character Sets and Collations." + There are three installation types available: Typical, Complete, + and Custom. - * To configure MySQL with debugging code, use the --with-debug - option: -shell> ./configure --with-debug - This causes a safe memory allocator to be included that can - find some errors and that provides output about what is - happening. See MySQL Internals: Porting - (http://forge.mysql.com/wiki/MySQL_Internals_Porting) - As of MySQL 5.1.12, using --with-debug to configure MySQL with - debugging support enables you to use the - --debug="d,parser_debug" option when you start the server. - This causes the Bison parser that is used to process SQL - statements to dump a parser trace to the server's standard - error output. Typically, this output is written to the error - log. + The Typical installation type installs the MySQL server, the mysql + command-line client, and the command-line utilities. The + command-line clients and utilities include mysqldump, myisamchk, + and several other tools to help you manage the MySQL server. - * If your client programs are using threads, you must compile a - thread-safe version of the MySQL client library with the - --enable-thread-safe-client configure option. This creates a - libmysqlclient_r library with which you should link your - threaded applications. See Section 21.10.17, "How to Make a - Threaded Client." + The Complete installation type installs all components included in + the installation package. The full installation package includes + components such as the embedded server library, the benchmark + suite, support scripts, and documentation. - * Some features require that the server be built with - compression library support, such as the COMPRESS() and - UNCOMPRESS() functions, and compression of the client/server - protocol. The --with-zlib-dir=no|bundled|DIR option provides - control for compression library support. The value no - explicitly disables compression support. bundled causes the - zlib library bundled in the MySQL sources to be used. A DIR - path name specifies where to find the compression library - sources. + The Custom installation type gives you complete control over which + packages you wish to install and the installation path that is + used. See Section 2.5.3.1.3, "The Custom Install Dialog," for more + information on performing a custom install. - * It is possible to build MySQL with large table support using - the --with-big-tables option. - This option causes the variables that store table row counts - to be declared as unsigned long long rather than unsigned - long. This enables tables to hold up to approximately - 1.844E+19 ((2^32)^2) rows rather than 2^32 (~4.295E+09) rows. - Previously it was necessary to pass -DBIG_TABLES to the - compiler manually in order to enable this feature. + If you choose the Typical or Complete installation types and click + the Next button, you advance to the confirmation screen to verify + your choices and begin the installation. If you choose the Custom + installation type and click the Next button, you advance to the + custom installation dialog, described in Section 2.5.3.1.3, "The + Custom Install Dialog." - * Run configure with the --disable-grant-options option to cause - the --bootstrap, --skip-grant-tables, and --init-file options - for mysqld to be disabled. For Windows, the configure.js - script recognizes the DISABLE_GRANT_OPTIONS flag, which has - the same effect. The capability is available as of MySQL - 5.1.15. +2.5.3.1.3. The Custom Install Dialog - * This option allows MySQL Community Server features to be - enabled. Additional options may be required for individual - features, such as --enable-profiling to enable statement - profiling. This option was added in MySQL 5.1.24. It is - enabled by default as of MySQL 5.1.28; to disable it, use - --disable-community-features. + If you wish to change the installation path or the specific + components that are installed by the MySQL Installation Wizard, + choose the Custom installation type. - * When given with --enable-community-features, the - --enable-profiling option enables the statement profiling - capability exposed by the SHOW PROFILE and SHOW PROFILES - statements. (See Section 12.5.5.33, "SHOW PROFILES Syntax.") - This option was added in MySQL 5.1.24. It is enabled by - default as of MySQL 5.1.28; to disable it, use - --disable-profiling. + A tree view on the left side of the custom install dialog lists + all available components. Components that are not installed have a + red X icon; components that are installed have a gray icon. To + change whether a component is installed, click on that component's + icon and choose a new option from the drop-down list that appears. - * See Section 2.13, "Operating System-Specific Notes," for - options that pertain to particular operating systems. + You can change the default installation path by clicking the + Change... button to the right of the displayed installation path. - * See Section 5.5.7.2, "Using SSL Connections," for options that - pertain to configuring MySQL to support secure (encrypted) - connections. + After choosing your installation components and installation path, + click the Next button to advance to the confirmation dialog. - * Several configure options apply to plugin selection and - building: ---with-plugins=PLUGIN[,PLUGIN]... ---with-plugins=GROUP ---with-plugin-PLUGIN ---without-plugin-PLUGIN - PLUGIN is an individual plugin name such as csv or archive. - As shorthand, GROUP is a configuration group name such as none - (select no plugins) or all (select all plugins). - You can build a plugin as static (compiled into the server) or - dynamic (built as a dynamic library that must be installed - using the INSTALL PLUGIN statement before it can be used). - Some plugins might not support static or dynamic build. - configure --help shows the following information pertaining to - plugins: +2.5.3.1.4. The Confirmation Dialog - + The plugin-related options + Once you choose an installation type and optionally choose your + installation components, you advance to the confirmation dialog. + Your installation type and installation path are displayed for you + to review. - + The names of all available plugins + To install MySQL if you are satisfied with your settings, click + the Install button. To change your settings, click the Back + button. To exit the MySQL Installation Wizard without installing + MySQL, click the Cancel button. - + For each plugin, a description of its purpose, which - build types it supports (static or dynamic), and which - plugin groups it is a part of. - --with-plugins can take a list of one or more plugin names - separated by commas, or a plugin group name. The named plugins - are configured to be built as static plugins. - --with-plugin-PLUGIN configures the given plugin to be built - as a static plugin. - --without-plugin-PLUGIN disables the given plugin from being - built. - If a plugin is named both with a --with and --without option, - the result is undefined. - For any plugin that is not explicitly selected or disabled, it - is selected to be built dynamically if it supports dynamic - build, and not built if it does not support dynamic build. - (Thus, in the case that no plugin options are given, all - plugins that support dynamic build are selected to be built as - dynamic plugins. Plugins that do not support dynamic build are - not built.) + After installation is complete, you have the option of registering + with the MySQL web site. Registration gives you access to post in + the MySQL forums at forums.mysql.com (http://forums.mysql.com) + along with the ability to report bugs at bugs.mysql.com + (http://bugs.mysql.com) and to subscribe to our newsletter. The + final screen of the installer provides a summary of the + installation and gives you the option to launch the MySQL Config + Wizard, which you can use to create a configuration file, install + the MySQL service, and configure security settings. -2.10.3. Installing from the Development Source Tree +2.5.3.1.5. Changes Made by MySQL Installation Wizard -Caution + Once you click the Install button, the MySQL Installation Wizard + begins the installation process and makes certain changes to your + system which are described in the sections that follow. - You should read this section only if you are interested in helping - us test our new code. If you just want to get MySQL up and running - on your system, you should use a standard release distribution - (either a binary or source distribution). + Changes to the Registry - To obtain the most recent development source tree, you first need - to download and install Bazaar. You can obtain Bazaar from the - Bazaar VCS Website (http://bazaar-vcs.org) Bazaar is supported by - any platform that supports Python, and is therefore compatible - with any Linux, Unix, Windows or Mac OS X host. Instructions for - downloading and installing Bazaar on the different platforms are - available on the Bazaar website. + The MySQL Installation Wizard creates one Windows registry key in + a typical install situation, located in + HKEY_LOCAL_MACHINE\SOFTWARE\MySQL AB. - All MySQL projects are hosted on Launchpad - (http://launchpad.net/) MySQL projects, including MySQL server, - MySQL Workbench and others are available from the Sun/MySQL - Engineering (http://launchpad.net/~mysql) page. For the - repositories related only to MySQL server, see the MySQL Server - (http://launchpad.net/mysql-server) page. + The MySQL Installation Wizard creates a key named after the major + version of the server that is being installed, such as MySQL + Server 5.1. It contains two string values, Location and Version. + The Location string contains the path to the installation + directory. In a default installation it contains C:\Program + Files\MySQL\MySQL Server 5.1\. The Version string contains the + release number. For example, for an installation of MySQL Server + 5.1.41, the key contains a value of 5.1.41. - To build under Unix/Linux, you must have the following tools - installed: + These registry keys are used to help external tools identify the + installed location of the MySQL server, preventing a complete scan + of the hard-disk to determine the installation path of the MySQL + server. The registry keys are not required to run the server, and + if you install MySQL using the noinstall Zip archive, the registry + keys are not created. + + Changes to the Start Menu + + The MySQL Installation Wizard creates a new entry in the Windows + Start menu under a common MySQL menu heading named after the major + version of MySQL that you have installed. For example, if you + install MySQL 5.1, the MySQL Installation Wizard creates a MySQL + Server 5.1 section in the Start menu. - * GNU make, available from http://www.gnu.org/software/make/. - Although some platforms come with their own make - implementations, it is highly recommended that you use GNU - make. It may already be available on your system as gmake. + The following entries are created within the new Start menu + section: - * autoconf 2.58 (or newer), available from - http://www.gnu.org/software/autoconf/. + * MySQL Command Line Client: This is a shortcut to the mysql + command-line client and is configured to connect as the root + user. The shortcut prompts for a root user password when you + connect. - * automake 1.8.1, available from - http://www.gnu.org/software/automake/. + * MySQL Server Instance Config Wizard: This is a shortcut to the + MySQL Config Wizard. Use this shortcut to configure a newly + installed server, or to reconfigure an existing server. - * libtool 1.5, available from - http://www.gnu.org/software/libtool/. + * MySQL Documentation: This is a link to the MySQL server + documentation that is stored locally in the MySQL server + installation directory. This option is not available when the + MySQL server is installed using the Essentials installation + package. - * m4, available from http://www.gnu.org/software/m4/. + Changes to the File System - * bison, available from http://www.gnu.org/software/bison/. You - should use the latest version of bison where possible. Version - 1.75 and version 2.1 are known to work. There have been - reported problems with bison 1.875. If you experience - problems, upgrade to a later, rather than earlier, version. - Versions of bison older than 1.75 may report this error: -sql_yacc.yy:#####: fatal error: maximum table size (32767) exceeded - The maximum table size is not actually exceeded; the error is - caused by bugs in older versions of bison. + The MySQL Installation Wizard by default installs the MySQL 5.1 + server to C:\Program Files\MySQL\MySQL Server 5.1, where Program + Files is the default location for applications in your system, and + 5.1 is the major version of your MySQL server. This is the + recommended location for the MySQL server, replacing the former + default location C:\mysql. - To build under Windows you will need a copy of Microsoft Visual - C++ 2005 Express Edition, Visual Studio .Net 2003 (7.1), or Visual - Studio 2005 (8.0) compiler system. + By default, all MySQL applications are stored in a common + directory at C:\Program Files\MySQL, where Program Files is the + default location for applications in your Windows installation. A + typical MySQL installation on a developer machine might look like + this: +C:\Program Files\MySQL\MySQL Server 5.1 +C:\Program Files\MySQL\MySQL Workbench 5.1 OSS - Once you have the necessary tools installed, you first need to - create a local branch of the MySQL source code on your machine: + This approach makes it easier to manage and maintain all MySQL + applications installed on a particular system. - 1. To obtain a copy of the MySQL source code, you must create a - new Bazaar branch. If you do not already have a Bazaar - repository directory set up, you need to initialize a new - directory: -shell> mkdir mysql-server -shell> bzr init-repo --trees mysql-server + In MySQL 5.1.23 and earlier, the default location for the data + files used by MySQL is located within the corresponding MySQL + Server installation directory. For MySQL 5.1.24 and later, the + default location of the data directory is the AppData directory + configured for the user that installed the MySQL application. - 2. Once you have an initialized directory, you can branch from - the public MySQL server repositories. To create a branch of a - specific version: -shell> cd mysql-server -shell> bzr branch lp:mysql-server/5.1 mysql-5.1 +2.5.3.1.6. Upgrading MySQL with the Installation Wizard - 3. The initial download will take some time to complete, - depending on the speed of your connection. Please be patient. - Once you have downloaded the first tree, additional trees - should take significantly less time to download. + The MySQL Installation Wizard can perform server upgrades + automatically using the upgrade capabilities of MSI. That means + you do not need to remove a previous installation manually before + installing a new release. The installer automatically shuts down + and removes the previous MySQL service before installing the new + version. - 4. When building from the Bazaar branch, you may want to create a - copy of your active branch so that you can make configuration - and other changes without affecting the original branch - contents. You can achieve this by branching from the original - branch: -shell> bzr branch mysql-5.1 mysql-5.1-build + Automatic upgrades are available only when upgrading between + installations that have the same major and minor version numbers. + For example, you can upgrade automatically from MySQL 5.1.34 to + MySQL 5.1.37, but not from MySQL 5.0 to MySQL 5.1. - Once you have the local branch, you can start to build MySQL - server from the source code. On Windows, the build process is - different from Unix/Linux. To continue building MySQL on Windows, - see Section 2.10.6, "Installing MySQL from Source on Windows." - - On Unix/Linux you need to use the autoconf system to create the - configure script so that you can configure the build environment - before building. - - 1. The following example shows the typical commands required to - configure a source tree. The first cd command changes location - into the top-level directory of the tree; replace mysql-5.1 - with the appropriate directory name. + In MySQL 5.1.32 and later, the EXE version of the MSI installer + packages were removed. When upgrading an existing MySQL + installation from the old EXE based installer to the MSI based + installer, please keep the following notes in mind: + + * The MSI installer will not identify an existing installation + that was installed using the old EXE installer. This means + that the installer will not stop the existing server, or + detect that the existing password is required before + installing the new version. To work around this: + + 1. Stop the current server manually using net stop or + mysqladmin shutdown. + + 2. Remove the existing installation manually by using the + Add/Remove Programs control panel. This will keep the + existing configuration and data files, as these are not + removed automatically. + + 3. Install the new version of MySQL using the MSI installer. + When running the installation, skip updating the security + by deselecting the checkbox on the security screen. + + 4. Complete the installation, and then start the server + again. You should be able to login with your existing + user and password credentials. + + * You can only upgrade the version and release using the MSI + installer. For example, you can upgrade an open source + installation with an open source installer. You cannot upgrade + an open source installation using the enterprise installer. + + See Section 2.5.7, "Upgrading MySQL on Windows." + +2.5.3.2. Installing MySQL Automatically using MSI + + The Microsoft Installer (MSI) supports a both a quiet and a + passive mode that can be used to install MySQL automatically + without requireing intervention. You can use this either in + scripts to automatically install MySQL or through a terminal + connection such as Telnet where you do not have access to the + standard Windows user interface. The MSI packages can also be used + in combination with Microsoft's Group Policy system (part of + Windows Server 2003 and Windows Server 2008) to install MySQL + across multiple machines. + + To install MySQL from one of the MSI packages automatically from + the command line (or within a script), you need to use the + msiexec.exe tool. For example, to perform a quiet installation + (which shows no dialog boxes or progress): +shell> msiexec /i /quiet mysql-5.1.39.msi + + The /i indicates that you want to perform an installation. The + /quiet option indicates that you want no interactive elements. + + To provide a dialog box showing the progress during installation, + and the dialog boxes providing information on the installation and + registration of MySQL, use /passive mode instead of /quiet: +shell> msiexec /i /passive mysql-5.1.39.msi + + Regardless of the mode of the installation, installing the package + in this manner performs a 'Typical' installation, and installs the + default components into the standard location. + + You can also use this method to uninstall MySQL by using the + /uninstall or /x options: +shell> msiexec /x /quiet mysql-5.1.39.msi + + To install MySQL and configure a MySQL instance from the command + line, see Section 2.5.4.13, "Creating an Instance from the Command + Line." + + For information on using MSI packages to install software + automatically using Group Policy, see How to use Group Policy to + remotely install software in Windows Server 2003 + (http://support.microsoft.com/kb/816102) -Note - For MySQL 5.1.12 and earlier, you must separately configure - the INNODB storage engine. You can do this by running the - following command from the main source directory: -shell> (cd storage/innobase; autoreconf --force --install) -shell> cd mysql-5.1 -shell> autoreconf --force --install -shell> ./configure # Add your favorite options here -shell> make - Or you can use BUILD/autorun.sh as a shortcut for the - following sequence of commands: -shell> aclocal; autoheader -shell> libtoolize --automake --force -shell> automake --force --add-missing; autoconf - The command line that changes directory into the - storage/innobase directory is used to configure the InnoDB - storage engine. You can omit this lines if you do not require - InnoDB support. +2.5.3.3. Removing MySQL Installed from the MSI Package -Note - Beginning with MySQL 5.1, code specific to storage engines has - been moved under a storage directory. For example, InnoDB code - is now found in storage/innobase and NDBCLUSTER code is in - storage/ndb. - If you get some strange errors during this stage, verify that - you have the correct version of the libtool installed. - A collection of our standard configuration scripts is located - in the BUILD/ subdirectory. For example, you may find it more - convenient to use the BUILD/compile-pentium-debug script than - the preceding set of shell commands. To compile on a different - architecture, modify the script by removing flags that are - Pentium-specific, or use another script that may be more - appropriate. These scripts are provided on an "as-is" basis. - They are not officially maintained and their contents may - change from release to release. + To uninstall a MySQL where you have used the MSI packages, you + must use the Add/Remove Programs tool within Control Panel. To do + this: - 2. When the build is done, run make install. Be careful with this - on a production machine; the command may overwrite your live - release installation. If you have another installation of - MySQL, run ./configure with different values for the --prefix, - --with-tcp-port, and --with-unix-socket-path options than - those used for your production server. + 1. Right click on the start menu and choose Control Panel. - 3. Play hard with your new installation and try to make the new - features crash. Start by running make test. See Section - 22.1.2, "MySQL Test Suite." + 2. If the Control Panel is set to category mode (you will see + Pick a category at the top of the Control Panel window), + double click on Add or Remove Programs. If the Control is set + to classic mode, doubgle click on the Add or Remove Programs + icon. + + 3. Find MySQL in the list of installed software. MySQL Server is + installed against major version numbers (MySQL 5.0, MySQL 5.1, + etc.). Select the version that you want to remove and click + Remove. + + 4. You will be prompted to confirm the removal. Click Yes to + remove MySQL. + + When MySQL is removed using this method, only the installed + components are removed. Any database information (including the + tables and data), import or export files, log files, and binary + logs produced during execution are kept in their configured + location. - 4. If you have gotten to the make stage, but the distribution - does not compile, please enter the problem into our bugs - database using the instructions given in Section 1.6, "How to - Report Bugs or Problems." If you have installed the latest - versions of the required GNU tools, and they crash trying to - process our configuration files, please report that also. - However, if you execute aclocal and get a command not found - error or a similar problem, do not report it. Instead, make - sure that all the necessary tools are installed and that your - PATH variable is set correctly so that your shell can find - them. +2.5.4. MySQL Server Instance Config Wizard - 5. After initially copying the repository with bzr to obtain the - source tree, you should use pull option to periodically update - your local copy. To do this any time after you have set up the - repository, use this command: -shell> bzr pull + The MySQL Server Instance Config Wizard helps automate the process + of configuring your server. It creates a custom MySQL + configuration file (my.ini or my.cnf) by asking you a series of + questions and then applying your responses to a template to + generate the configuration file that is tuned to your + installation. - 6. You can examine the changeset comments for the tree by using - the log option to bzr: -shell> bzr log - You can also browse changesets, comments, and source code - online. To browse this information for MySQL 5.1, go to - http://launchpad.net/mysql-server/. - If you see diffs or code that you have a question about, do - not hesitate to send email to the MySQL internals mailing - list. See Section 1.5.1, "MySQL Mailing Lists." Also, if you - think you have a better idea on how to do something, send an - email message to the list with a patch. + The complete and essential MSI installation packages include the + MySQL Server Instance Config Wizard in the MySQL 5.1 server. The + MySQL Server Instance Config Wizard is only available for Windows. + + The workflow sequence for using the MySQL Server Instance Config + Wizard is shown in the figure below: + + Figure 2.4. MySQL Server Instance Config Wizard Workflow + MySQL Server Instance Config Wizard Workflow + +2.5.4.1. Starting the MySQL Server Instance Config Wizard + + The MySQL Server Instance Config Wizard is normally started as + part of the installation process. You should only need to run the + MySQL Server Instance Config Wizard again when you need to change + the configuration parameters of your server. -2.10.4. Dealing with Problems Compiling MySQL + If you chose not to open a port prior to installing MySQL on + Windows Vista, you can choose to use the MySQL Server Instance + Config Wizard after installation. However, you must open a port in + the Windows Firewall. To do this see the instructions given in + Section 2.5.3.1.1, "Downloading and Starting the MySQL + Installation Wizard." Rather than opening a port, you also have + the option of adding MySQL as a program that bypasses the Windows + Firewall. One or the other option is sufficient --- you need not + do both. Additionally, when running the MySQL Server Config Wizard + on Windows Vista ensure that you are logged in as a user with + administrative rights. + MySQL Server Instance Config Wizard - All MySQL programs compile cleanly for us with no warnings on - Solaris or Linux using gcc. On other systems, warnings may occur - due to differences in system include files. See Section 2.10.5, - "MIT-pthreads Notes," for warnings that may occur when using - MIT-pthreads. For other problems, check the following list. + You can launch the MySQL Config Wizard by clicking the MySQL + Server Instance Config Wizard entry in the MySQL section of the + Windows Start menu. - The solution to many problems involves reconfiguring. If you do - need to reconfigure, take note of the following: + Alternatively, you can navigate to the bin directory of your MySQL + installation and launch the MySQLInstanceConfig.exe file directly. - * If configure is run after it has previously been run, it may - use information that was gathered during its previous - invocation. This information is stored in config.cache. When - configure starts up, it looks for that file and reads its - contents if it exists, on the assumption that the information - is still correct. That assumption is invalid when you - reconfigure. + The MySQL Server Instance Config Wizard places the my.ini file in + the installation directory for the MySQL server. This helps + associate configuration files with particular server instances. - * Each time you run configure, you must run make again to - recompile. However, you may want to remove old object files - from previous builds first because they were compiled using - different configuration options. + To ensure that the MySQL server knows where to look for the my.ini + file, an argument similar to this is passed to the MySQL server as + part of the service installation: +--defaults-file="C:\Program Files\MySQL\MySQL Server 5.1\my.ini" - To prevent old configuration information or object files from - being used, run these commands before re-running configure: -shell> rm config.cache -shell> make clean + Here, C:\Program Files\MySQL\MySQL Server 5.1 is replaced with the + installation path to the MySQL Server. The --defaults-file option + instructs the MySQL server to read the specified file for + configuration options when it starts. - Alternatively, you can run make distclean. + Apart from making changes to the my.ini file by running the MySQL + Server Instance Config Wizard again, you can modify it by opening + it with a text editor and making any necessary changes. You can + also modify the server configuration with the MySQL Administrator + (http://www.mysql.com/products/administrator/) utility. For more + information about server configuration, see Section 5.1.2, "Server + Command Options." - The following list describes some of the problems when compiling - MySQL that have been found to occur most often: + MySQL clients and utilities such as the mysql and mysqldump + command-line clients are not able to locate the my.ini file + located in the server installation directory. To configure the + client and utility applications, create a new my.ini file in the + Windows installation directory (for example, C:\WINDOWS). - * If you get errors such as the ones shown here when compiling - sql_yacc.cc, you probably have run out of memory or swap - space: -Internal compiler error: program cc1plus got fatal signal 11 -Out of virtual memory -Virtual memory exhausted - The problem is that gcc requires a huge amount of memory to - compile sql_yacc.cc with inline functions. Try running - configure with the --with-low-memory option: -shell> ./configure --with-low-memory - This option causes -fno-inline to be added to the compile line - if you are using gcc and -O0 if you are using something else. - You should try the --with-low-memory option even if you have - so much memory and swap space that you think you can't - possibly have run out. This problem has been observed to occur - even on systems with generous hardware configurations, and the - --with-low-memory option usually fixes it. + Under Windows Server 2003, Windows Server 2000, Windows XP, and + Windows Vista MySQL Server Instance Config Wizard will configure + MySQL to work as a Windows service. To start and stop MySQL you + use the Services application that is supplied as part of the + Windows Administrator Tools. + +2.5.4.2. Choosing a Maintenance Option + + If the MySQL Server Instance Config Wizard detects an existing + configuration file, you have the option of either reconfiguring + your existing server, or removing the server instance by deleting + the configuration file and stopping and removing the MySQL + service. - * By default, configure picks c++ as the compiler name and GNU - c++ links with -lg++. If you are using gcc, that behavior can - cause problems during configuration such as this: -configure: error: installation or configuration problem: -C++ compiler cannot create executables. - You might also observe problems during compilation related to - g++, libg++, or libstdc++. - One cause of these problems is that you may not have g++, or - you may have g++ but not libg++, or libstdc++. Take a look at - the config.log file. It should contain the exact reason why - your C++ compiler didn't work. To work around these problems, - you can use gcc as your C++ compiler. Try setting the - environment variable CXX to "gcc -O3". For example: -shell> CXX="gcc -O3" ./configure - This works because gcc compiles C++ source files as well as - g++ does, but does not link in libg++ or libstdc++ by default. - Another way to fix these problems is to install g++, libg++, - and libstdc++. However, do not use libg++ or libstdc++ with - MySQL because this only increases the binary size of mysqld - without providing any benefits. Some versions of these - libraries have also caused strange problems for MySQL users in - the past. + To reconfigure an existing server, choose the Re-configure + Instance option and click the Next button. Any existing + configuration file is not overwritten, but renamed (within the + same directory) using a timestamp (Windows) or sequential number + (Linux). To remove the existing server instance, choose the Remove + Instance option and click the Next button. - * If your compile fails with errors such as any of the - following, you must upgrade your version of make to GNU make: -making all in mit-pthreads -make: Fatal error in reader: Makefile, line 18: -Badly formed macro assignment - Or: -make: file `Makefile' line 18: Must be a separator (: - Or: -pthread.h: No such file or directory - Solaris and FreeBSD are known to have troublesome make - programs. - GNU make 3.75 is known to work. + If you choose the Remove Instance option, you advance to a + confirmation window. Click the Execute button. The MySQL Server + Config Wizard stops and removes the MySQL service, and then + deletes the configuration file. The server installation and its + data folder are not removed. - * If you want to define flags to be used by your C or C++ - compilers, do so by adding the flags to the CFLAGS and - CXXFLAGS environment variables. You can also specify the - compiler names this way using CC and CXX. For example: -shell> CC=gcc -shell> CFLAGS=-O3 -shell> CXX=gcc -shell> CXXFLAGS=-O3 -shell> export CC CFLAGS CXX CXXFLAGS - See Section 2.1.2.4, "MySQL Binaries Compiled by Sun - Microsystems, Inc.," for a list of flag definitions that have - been found to be useful on various systems. + If you choose the Re-configure Instance option, you advance to the + Configuration Type dialog where you can choose the type of + installation that you wish to configure. - * If you get errors such as those shown here when compiling - mysqld, configure did not correctly detect the type of the - last argument to accept(), getsockname(), or getpeername(): -cxx: Error: mysqld.cc, line 645: In this statement, the referenced - type of the pointer value ''length'' is ''unsigned long'', - which is not compatible with ''int''. -new_sock = accept(sock, (struct sockaddr *)&cAddr, &length); - To fix this, edit the config.h file (which is generated by - configure). Look for these lines: -/* Define as the base type of the last arg to accept */ -#define SOCKET_SIZE_TYPE XXX - Change XXX to size_t or int, depending on your operating - system. (You must do this each time you run configure because - configure regenerates config.h.) +2.5.4.3. Choosing a Configuration Type - * The sql_yacc.cc file is generated from sql_yacc.yy. Normally, - the build process does not need to create sql_yacc.cc because - MySQL comes with a pre-generated copy. However, if you do need - to re-create it, you might encounter this error: -"sql_yacc.yy", line xxx fatal: default action causes potential... - This is a sign that your version of yacc is deficient. You - probably need to install bison (the GNU version of yacc) and - use that instead. + When you start the MySQL Server Instance Config Wizard for a new + MySQL installation, or choose the Re-configure Instance option for + an existing installation, you advance to the Configuration Type + dialog. + MySQL Server Instance Config Wizard: Configuration Type - * On Debian Linux 3.0, you need to install gawk instead of the - default mawk. + There are two configuration types available: Detailed + Configuration and Standard Configuration. The Standard + Configuration option is intended for new users who want to get + started with MySQL quickly without having to make many decisions + about server configuration. The Detailed Configuration option is + intended for advanced users who want more fine-grained control + over server configuration. - * If you need to debug mysqld or a MySQL client, run configure - with the --with-debug option, and then recompile and link your - clients with the new client library. See MySQL Internals: - Porting (http://forge.mysql.com/wiki/MySQL_Internals_Porting) + If you are new to MySQL and need a server configured as a + single-user developer machine, the Standard Configuration should + suit your needs. Choosing the Standard Configuration option causes + the MySQL Config Wizard to set all configuration options + automatically with the exception of Service Options and Security + Options. - * If you get a compilation error on Linux (for example, SuSE - Linux 8.1 or Red Hat Linux 7.3) similar to the following one, - you probably do not have g++ installed: -libmysql.c:1329: warning: passing arg 5 of `gethostbyname_r' from -incompatible pointer type -libmysql.c:1329: too few arguments to function `gethostbyname_r' -libmysql.c:1329: warning: assignment makes pointer from integer -without a cast -make[2]: *** [libmysql.lo] Error 1 - By default, the configure script attempts to determine the - correct number of arguments by using g++ (the GNU C++ - compiler). This test yields incorrect results if g++ is not - installed. There are two ways to work around this problem: + The Standard Configuration sets options that may be incompatible + with systems where there are existing MySQL installations. If you + have an existing MySQL installation on your system in addition to + the installation you wish to configure, the Detailed Configuration + option is recommended. - + Make sure that the GNU C++ g++ is installed. On some - Linux distributions, the required package is called gpp; - on others, it is named gcc-c++. + To complete the Standard Configuration, please refer to the + sections on Service Options and Security Options in Section + 2.5.4.10, "The Service Options Dialog," and Section 2.5.4.11, "The + Security Options Dialog," respectively. - + Use gcc as your C++ compiler by setting the CXX - environment variable to gcc: -export CXX="gcc" - You must run configure again after making either of those - changes. +2.5.4.4. The Server Type Dialog -2.10.5. MIT-pthreads Notes + There are three different server types available to choose from. + The server type that you choose affects the decisions that the + MySQL Server Instance Config Wizard makes with regard to memory, + disk, and processor usage. + MySQL Server Instance Config Wizard: Server Type - This section describes some of the issues involved in using - MIT-pthreads. + * Developer Machine: Choose this option for a typical desktop + workstation where MySQL is intended only for personal use. It + is assumed that many other desktop applications are running. + The MySQL server is configured to use minimal system + resources. - On Linux, you should not use MIT-pthreads. Use the installed - LinuxThreads implementation instead. See Section 2.13.1, "Linux - Notes." + * Server Machine: Choose this option for a server machine where + the MySQL server is running alongside other server + applications such as FTP, email, and Web servers. The MySQL + server is configured to use a moderate portion of the system + resources. - If your system does not provide native thread support, you should - build MySQL using the MIT-pthreads package. This includes older - FreeBSD systems, SunOS 4.x, Solaris 2.4 and earlier, and some - others. See Section 2.1.1, "Operating Systems Supported by MySQL - Community Server." + * Dedicated MySQL Server Machine: Choose this option for a + server machine that is intended to run only the MySQL server. + It is assumed that no other applications are running. The + MySQL server is configured to use all available system + resources. - MIT-pthreads is not part of the MySQL 5.1 source distribution. If - you require this package, you need to download it separately from - http://dev.mysql.com/Downloads/Contrib/pthreads-1_60_beta6-mysql.t - ar.gz +Note - After downloading, extract this source archive into the top level - of the MySQL source directory. It creates a new subdirectory named - mit-pthreads. + By selecting one of the preconfigured configurations, the values + and settings of various options in your my.cnf or my.ini will be + altered accordingly. The default values and options as described + in the reference manual may therefore be different to the options + and values that were created during the execution of the Config + Wizard. - * On most systems, you can force MIT-pthreads to be used by - running configure with the --with-mit-threads option: -shell> ./configure --with-mit-threads - Building in a nonsource directory is not supported when using - MIT-pthreads because we want to minimize our changes to this - code. +2.5.4.5. The Database Usage Dialog - * The checks that determine whether to use MIT-pthreads occur - only during the part of the configuration process that deals - with the server code. If you have configured the distribution - using --without-server to build only the client code, clients - do not know whether MIT-pthreads is being used and use Unix - socket file connections by default. Because Unix socket files - do not work under MIT-pthreads on some platforms, this means - you need to use -h or --host with a value other than localhost - when you run client programs. + The Database Usage dialog allows you to indicate the storage + engines that you expect to use when creating MySQL tables. The + option you choose determines whether the InnoDB storage engine is + available and what percentage of the server resources are + available to InnoDB. + MySQL Server Instance Config Wizard: Usage Dialog - * When MySQL is compiled using MIT-pthreads, system locking is - disabled by default for performance reasons. You can tell the - server to use system locking with the --external-locking - option. This is needed only if you want to be able to run two - MySQL servers against the same data files, but that is not - recommended, anyway. + * Multifunctional Database: This option enables both the InnoDB + and MyISAM storage engines and divides resources evenly + between the two. This option is recommended for users who use + both storage engines on a regular basis. - * Sometimes the pthread bind() command fails to bind to a socket - without any error message (at least on Solaris). The result is - that all connections to the server fail. For example: -shell> mysqladmin version -mysqladmin: connect to server at '' failed; -error: 'Can't connect to mysql server on localhost (146)' - The solution to this problem is to kill the mysqld server and - restart it. This has happened to us only when we have forcibly - stopped the server and restarted it immediately. + * Transactional Database Only: This option enables both the + InnoDB and MyISAM storage engines, but dedicates most server + resources to the InnoDB storage engine. This option is + recommended for users who use InnoDB almost exclusively and + make only minimal use of MyISAM. - * With MIT-pthreads, the sleep() system call isn't interruptible - with SIGINT (break). This is noticeable only when you run - mysqladmin --sleep. You must wait for the sleep() call to - terminate before the interrupt is served and the process - stops. + * Non-Transactional Database Only: This option disables the + InnoDB storage engine completely and dedicates all server + resources to the MyISAM storage engine. This option is + recommended for users who do not use InnoDB. - * When linking, you might receive warning messages like these - (at least on Solaris); they can be ignored: -ld: warning: symbol `_iob' has differing sizes: - (file /my/local/pthreads/lib/libpthread.a(findfp.o) value=0x4; -file /usr/lib/libc.so value=0x140); - /my/local/pthreads/lib/libpthread.a(findfp.o) definition taken -ld: warning: symbol `__iob' has differing sizes: - (file /my/local/pthreads/lib/libpthread.a(findfp.o) value=0x4; -file /usr/lib/libc.so value=0x140); - /my/local/pthreads/lib/libpthread.a(findfp.o) definition taken + The Config Wizard uses a template to generate the server + configuration file. The Database Usage dialog sets one of the + following option strings: +Multifunctional Database: MIXED +Transactional Database Only: INNODB +Non-Transactional Database Only: MYISAM - * Some other warnings also can be ignored: -implicit declaration of function `int strtoll(...)' -implicit declaration of function `int strtoul(...)' + When these options are processed through the default template + (my-template.ini) the result is: +Multifunctional Database: +default-storage-engine=InnoDB +_myisam_pct=50 - * We have not been able to make readline work with MIT-pthreads. - (This is not necessary, but may be of interest to some.) +Transactional Database Only: +default-storage-engine=InnoDB +_myisam_pct=5 -2.10.6. Installing MySQL from Source on Windows +Non-Transactional Database Only: +default-storage-engine=MyISAM +_myisam_pct=100 +skip-innodb - These instructions describe how to build binaries from source for - MySQL 5.1 on Windows. Instructions are provided for building - binaries from a standard source distribution or from the Bazaar - tree that contains the latest development source. + The _myisam_pct value is used to calculate the percentage of + resources dedicated to MyISAM. The remaining resources are + allocated to InnoDB. -Note +2.5.4.6. The InnoDB Tablespace Dialog - The instructions here are strictly for users who want to test - MySQL on Microsoft Windows from the latest source distribution or - from the Bazaar tree. For production use, we do not advise using a - MySQL server built by yourself from source. Normally, it is best - to use precompiled binary distributions of MySQL that are built - specifically for optimal performance on Windows by Sun - Microsystems, Inc. Instructions for installing binary - distributions are available in Section 2.3, "Installing MySQL on - Windows." + Some users may want to locate the InnoDB tablespace files in a + different location than the MySQL server data directory. Placing + the tablespace files in a separate location can be desirable if + your system has a higher capacity or higher performance storage + device available, such as a RAID storage system. + MySQL Server Instance Config Wizard: InnoDB Data Tablespace - To build MySQL on Windows from source, you must satisfy the - following system, compiler, and resource requirements: + To change the default location for the InnoDB tablespace files, + choose a new drive from the drop-down list of drive letters and + choose a new path from the drop-down list of paths. To create a + custom path, click the ... button. - * Windows 2000, Windows XP, or newer version. - Windows Vista is supported when using Visual Studio 2005 - provided you have installed the following updates: + If you are modifying the configuration of an existing server, you + must click the Modify button before you change the path. In this + situation you must move the existing tablespace files to the new + location manually before starting the server. - + Microsoft Visual Studio 2005 Professional Edition - ENU - Service Pack 1 (KB926601) - (http://support.microsoft.com/?kbid=926601) +2.5.4.7. The Concurrent Connections Dialog - + Security Update for Microsoft Visual Studio 2005 - Professional Edition - ENU (KB937061) - (http://support.microsoft.com/?kbid=937061) + To prevent the server from running out of resources, it is + important to limit the number of concurrent connections to the + MySQL server that can be established. The Concurrent Connections + dialog allows you to choose the expected usage of your server, and + sets the limit for concurrent connections accordingly. It is also + possible to set the concurrent connection limit manually. + MySQL Server Instance Config Wizard: Connections - + Update for Microsoft Visual Studio 2005 Professional - Edition - ENU (KB932232) - (http://support.microsoft.com/?kbid=932232) + * Decision Support (DSS)/OLAP: Choose this option if your server + does not require a large number of concurrent connections. The + maximum number of connections is set at 100, with an average + of 20 concurrent connections assumed. - * CMake, which can be downloaded from http://www.cmake.org. - After installing, modify your path to include the cmake - binary. + * Online Transaction Processing (OLTP): Choose this option if + your server requires a large number of concurrent connections. + The maximum number of connections is set at 500. - * Microsoft Visual C++ 2005 Express Edition, Visual Studio .Net - 2003 (7.1), or Visual Studio 2005 (8.0) compiler system. + * Manual Setting: Choose this option to set the maximum number + of concurrent connections to the server manually. Choose the + number of concurrent connections from the drop-down box + provided, or enter the maximum number of connections into the + drop-down box if the number you desire is not listed. - * If you are using Visual C++ 2005 Express Edition, you must - also install an appropriate Platform SDK. More information and - links to downloads for various Windows platforms is available - from - http://www.microsoft.com/downloads/details.aspx?familyid=0baf2 - b35-c656-4969-ace8-e4c0c0716adb. +2.5.4.8. The Networking and Strict Mode Options Dialog - * If you are compiling from a Bazaar tree or making changes to - the parser, you need bison for Windows, which can be - downloaded from - http://gnuwin32.sourceforge.net/packages/bison.htm. Download - the package labeled "Complete package, excluding sources". - After installing the package, modify your path to include the - bison binary and ensure that this binary is accessible from - Visual Studio. + Use the Networking Options dialog to enable or disable TCP/IP + networking and to configure the port number that is used to + connect to the MySQL server. + MySQL Server Instance Config Wizard: Network Configuration - * Cygwin might be necessary if you want to run the test script - or package the compiled binaries and support files into a Zip - archive. (Cygwin is needed only to test or package the - distribution, not to build it.) Cygwin is available from - http://cygwin.com. + TCP/IP networking is enabled by default. To disable TCP/IP + networking, uncheck the box next to the Enable TCP/IP Networking + option. - * 3GB to 5GB of disk space. + Port 3306 is used by default. To change the port used to access + MySQL, choose a new port number from the drop-down box or type a + new port number directly into the drop-down box. If the port + number you choose is in use, you are prompted to confirm your + choice of port number. - The exact system requirements can be found here: - http://msdn.microsoft.com/vstudio/Previous/2003/sysreqs/default.as - px and - http://msdn.microsoft.com/vstudio/products/sysreqs/default.aspx + Set the Server SQL Mode to either enable or disable strict mode. + Enabling strict mode (default) makes MySQL behave more like other + database management systems. If you run applications that rely on + MySQL's old "forgiving" behavior, make sure to either adapt those + applications or to disable strict mode. For more information about + strict mode, see Section 5.1.8, "Server SQL Modes." - You also need a MySQL source distribution for Windows, which can - be obtained two ways: +2.5.4.9. The Character Set Dialog - * Obtain a source distribution packaged by Sun Microsystems, - Inc. These are available from http://dev.mysql.com/downloads/. + The MySQL server supports multiple character sets and it is + possible to set a default server character set that is applied to + all tables, columns, and databases unless overridden. Use the + Character Set dialog to change the default character set of the + MySQL server. + MySQL Server Instance Config Wizard: Character Set - * Package a source distribution yourself from the latest Bazaar - developer source tree. For instructions on pulling the latest - source files, see Section 2.10.3, "Installing from the - Development Source Tree." + * Standard Character Set: Choose this option if you want to use + latin1 as the default server character set. latin1 is used for + English and many Western European languages. - If you find something not working as expected, or you have - suggestions about ways to improve the current build process on - Windows, please send a message to the win32 mailing list. See - Section 1.5.1, "MySQL Mailing Lists." + * Best Support For Multilingualism: Choose this option if you + want to use utf8 as the default server character set. This is + a Unicode character set that can store characters from many + different languages. -2.10.6.1. Building MySQL from Source Using CMake and Visual Studio + * Manual Selected Default Character Set / Collation: Choose this + option if you want to pick the server's default character set + manually. Choose the desired character set from the provided + drop-down list. - You can build MySQL on Windows by using a combination of cmake and - Microsoft Visual Studio .NET 2003 (7.1), Microsoft Visual Studio - 2005 (8.0) or Microsoft Visual C++ 2005 Express Edition. You must - have the appropriate Microsoft Platform SDK installed. +2.5.4.10. The Service Options Dialog -Note + On Windows platforms, the MySQL server can be installed as a + Windows service. When installed this way, the MySQL server can be + started automatically during system startup, and even restarted + automatically by Windows in the event of a service failure. - To compile from the source code on Windows you must use the - standard source distribution (for example, mysql-5.0.45.tar.gz). - You build from the same distribution as used to build MySQL on - Unix, Linux and other platforms. Do not use the Windows Source - distributions as they do not contain the necessary configuration - script and other files. + The MySQL Server Instance Config Wizard installs the MySQL server + as a service by default, using the service name MySQL. If you do + not wish to install the service, uncheck the box next to the + Install As Windows Service option. You can change the service name + by picking a new service name from the drop-down box provided or + by entering a new service name into the drop-down box. - Follow this procedure to build MySQL: +Note - 1. If you are installing from a packaged source distribution, - create a work directory (for example, C:\workdir), and unpack - the source distribution there using WinZip or another Windows - tool that can read .zip files. This directory is the work - directory in the following instructions. + Service names can include any legal character except forward (/) + or backward (\) slashes, and must be less than 256 characters + long. - 2. Using a command shell, navigate to the work directory and run - the following command: -C:\workdir>win\configure.js options - If you have associated the .js file extension with an - application such as a text editor, then you may need to use - the following command to force configure.js to be executed as - a script: -C:\workdir>cscript win\configure.js options - These options are available for configure.js: +Warning - + WITH_INNOBASE_STORAGE_ENGINE: Enable the InnoDB storage - engine. + If you are installing multiple versions of MySQL onto the same + machine, you must choose a different service name for each version + that you install. If you do not choose a different service for + each installed version then the service manager information will + be inconsistent and this will cause problems when you try to + uninstall a previous version. - + WITH_PARTITION_STORAGE_ENGINE: Enable user-defined - partitioning. + If you have already installed multiple versions using the same + service name, you must manually edit the contents of the + HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services parameters + within the Windows registry to update the association of the + service name with the correct server version. - + WITH_ARCHIVE_STORAGE_ENGINE: Enable the ARCHIVE storage - engine. + Typically, when installing multiple versions you create a service + name based on the version information. For example, you might + install MySQL 5.x as mysql5, or specific versions such as MySQL + 5.1.30 as mysql50130. - + WITH_BLACKHOLE_STORAGE_ENGINE: Enable the BLACKHOLE - storage engine. + To install the MySQL server as a service but not have it started + automatically at startup, uncheck the box next to the Launch the + MySQL Server Automatically option. - + WITH_EXAMPLE_STORAGE_ENGINE: Enable the EXAMPLE storage - engine. +2.5.4.11. The Security Options Dialog - + WITH_FEDERATED_STORAGE_ENGINE: Enable the FEDERATED - storage engine. + It is strongly recommended that you set a root password for your + MySQL server, and the MySQL Server Instance Config Wizard requires + by default that you do so. If you do not wish to set a root + password, uncheck the box next to the Modify Security Settings + option. + MySQL Server Instance Config Wizard: Security - + WITH_NDBCLUSTER_STORAGE_ENGINE (experimental): Enable the - NDBCLUSTER storage engine in the MySQL server; cause - binaries for the MySQL Cluster management and data node, - management client, and other programs to be built. - This option is supported only in MySQL Cluster NDB 7.0 - (NDBCLUSTER storage engine versions 6.4.0 and later) - using the MySQL Cluster sources. It cannot be used to - enable clustering support in other MySQL source trees or - distributions. + To set the root password, enter the desired password into both the + New root password and Confirm boxes. If you are reconfiguring an + existing server, you need to enter the existing root password into + the Current root password box. - + MYSQL_SERVER_SUFFIX=suffix: Server suffix, default none. + To allow root logins from across the network, check the box next + to the Enable root access from remote machines option. This + decreases the security of your root account. - + COMPILATION_COMMENT=comment: Server comment, default - "Source distribution". + To create an anonymous user account, check the box next to the + Create An Anonymous Account option. Creating an anonymous account + can decrease server security and cause login and permission + difficulties. For this reason, it is not recommended. - + MYSQL_TCP_PORT=port: Server port, default 3306. +2.5.4.12. The Confirmation Dialog - + DISABLE_GRANT_OPTIONS: Disables the --bootstrap, - --skip-grant-tables, and --init-file options for mysqld. - This option is available as of MySQL 5.1.15. - For example (type the command on one line): -C:\workdir>win\configure.js WITH_INNOBASE_STORAGE_ENGINE - WITH_PARTITION_STORAGE_ENGINE MYSQL_SERVER_SUFFIX=-pro + The final dialog in the MySQL Server Instance Config Wizard is the + Confirmation Dialog. To start the configuration process, click the + Execute button. To return to a previous dialog, click the Back + button. To exit the MySQL Server Instance Config Wizard without + configuring the server, click the Cancel button. + MySQL Server Instance Config Wizard: Confirmation - 3. From the work directory, execute the win\build-vs8.bat or - win\build-vs71.bat file, depending on the version of Visual - Studio you have installed. The script invokes CMake, which - generates the mysql.sln solution file. - You can also use win\build-vs8_x64.bat to build the 64-bit - version of MySQL. However, you cannot build the 64-bit version - with Visual Studio Express Edition. You must use Visual Studio - 2005 (8.0) or higher. + After you click the Execute button, the MySQL Server Instance + Config Wizard performs a series of tasks and displays the progress + onscreen as the tasks are performed. - 4. From the work directory, open the generated mysql.sln file - with Visual Studio and select the proper configuration using - the Configuration menu. The menu provides Debug, Release, - RelwithDebInfo, MinRelInfo options. Then select Solution > - Build to build the solution. - Remember the configuration that you use in this step. It is - important later when you run the test script because that - script needs to know which configuration you used. + The MySQL Server Instance Config Wizard first determines + configuration file options based on your choices using a template + prepared by MySQL developers and engineers. This template is named + my-template.ini and is located in your server installation + directory. - 5. Test the server. The server built using the preceding - instructions expects that the MySQL base directory and data - directory are C:\mysql and C:\mysql\data by default. If you - want to test your server using the source tree root directory - and its data directory as the base directory and data - directory, you need to tell the server their path names. You - can either do this on the command line with the --basedir and - --datadir options, or by placing appropriate options in an - option file. (See Section 4.2.3.3, "Using Option Files.") If - you have an existing data directory elsewhere that you want to - use, you can specify its path name instead. - When the server is running in standalone fashion or as a - service based on your configuration, try to connect to it from - the mysql interactive command-line utility. - You can also run the standard test script, mysql-test-run.pl. - This script is written in Perl, so you'll need either Cygwin - or ActiveState Perl to run it. You may also need to install - the modules required by the script. To run the test script, - change location into the mysql-test directory under the work - directory, set the MTR_VS_CONFIG environment variable to the - configuration you selected earlier (or use the --vs-config - option), and invoke mysql-test-run.pl. For example (using - Cygwin and the bash shell): -shell> cd mysql-test -shell> export MTR_VS_CONFIG=debug -shell> ./mysql-test-run.pl --force --timer -shell> ./mysql-test-run.pl --force --timer --ps-protocol + The MySQL Config Wizard then writes these options to the + corresponding configuration file. - When you are satisfied that the programs you have built are - working correctly, stop the server. Now you can install the - distribution. One way to do this is to use the make_win_bin_dist - script in the scripts directory of the MySQL source distribution - (see Section 4.4.2, "make_win_bin_dist --- Package MySQL - Distribution as ZIP Archive"). This is a shell script, so you must - have Cygwin installed if you want to use it. It creates a Zip - archive of the built executables and support files that you can - unpack in the location at which you want to install MySQL. + If you chose to create a service for the MySQL server, the MySQL + Server Instance Config Wizard creates and starts the service. If + you are reconfiguring an existing service, the MySQL Server + Instance Config Wizard restarts the service to apply your + configuration changes. + + If you chose to set a root password, the MySQL Config Wizard + connects to the server, sets your new root password, and applies + any other security settings you may have selected. + + After the MySQL Server Instance Config Wizard has completed its + tasks, it displays a summary. Click the Finish button to exit the + MySQL Server Config Wizard. + +2.5.4.13. Creating an Instance from the Command Line + + In addition to using the GUI interface to the MySQL Server + Instance Config Wizard, you can also create instances + automatically from the command line. + + To use the MySQL Server Instance Config Wizard on the command + line, you need to use the MySQLInstanceConfig.exe command that is + installed with MySQL in the bin directory within the installation + directory. MySQLInstanceConfig.exe takes a number of command-line + arguments the set the properties that would normally be selected + through the GUI interface, and then creates a new configuration + file (my.ini) by combining these selections with a template + configuration file to produce the working configuration file. + + The main command line options are provided in the table below. + Some of the options are required, while some options are optional. + + Table 2.4. MySQL Server Instance Config Wizard Command Line + Options + Option Description + Required Parameters + -nPRODUCTNAME The name of the instance when installed + -pPATH Path of the base directory for installation. This is + equivalent to the directory when using the basedir configuration + parameter + -vVERSION The version tag to use for this installation + Action to Perform + -i Install an instance + -r Remove an instance + -s Stop an existing instance + -q Perform the operation quietly + -lFILENAME Sae the installation progress in a logfile + Config File to Use + -tFILENAME Path to the template config file that will be used to + generate the installed configuration file + -cFILENAME Path to a config file to be generated + + The -t and -c options work together to set the configuration + parameters for a new instance. The -t option specifies the + template configuration file to use as the basic configuration, + which are then merged with the configuration parameters generated + by the MySQL Server Instance Config Wizard into the configuration + file specified by the -c option. + + A sample template file, my-template.ini is provided in the + toplevel MySQL installation directory. The file contains elements + are replaced automatically by the MySQL Server Instance Config + Wizard during configuration. + + If you specify a configuration file that already exists, the + existing configuration file will be saved in the file with the + original, with the date and time added. For example, the mysql.ini + will be copied to mysql 2009-10-27 1646.ini.bak. + + The parameters that you can specify on the command line are listed + in the table below. + + Table 2.5. MySQL Server Instance Config Wizard Parameters + Parameter Description + ServiceName=$ Specify the name of the service to be created + AddBinToPath={yes | no} Specifies whether to add the binary + directory of MySQL to the standard PATH environment variable + ServerType={DEVELOPMENT | SERVER | DEDICATED} Specify the server + type. For more information, see Section 2.5.4.4, "The Server Type + Dialog" + DatabaseType={MIXED | INNODB | MYISAM} Specify the default + database type. For more information, see Section 2.5.4.5, "The + Database Usage Dialog" + ConnectionUsage={DSS | OLTP} Specify the type of connection + support, this automates the setting for the number of concurrent + connections (see the ConnectionCount parameter). For more + information, see Section 2.5.4.7, "The Concurrent Connections + Dialog" + ConnectionCount=# Specify the number of concurrent connections to + support. For more information, see Section 2.5.4.4, "The Server + Type Dialog" + SkipNetworking={yes | no} Specify whether network support should + be supported. Specifying yes disables network access altogether + Port=# Specify the network port number to use for network + connections. For more information, see Section 2.5.4.8, "The + Networking and Strict Mode Options Dialog" + StrictMode={yes | no} Specify whether to use the strict SQL mode. + For more information, see Section 2.5.4.8, "The Networking and + Strict Mode Options Dialog" + Charset=$ Specify the default character set. For more information, + see Section 2.5.4.9, "The Character Set Dialog" + RootPassword=$ Specify the root password + RootCurrentPassword=$ Specify the current root password then + stopping and/or reconfiguring an existing service - It is also possible to install MySQL by copying directories and - files directly: +Note - 1. Create the directories where you want to install MySQL. For - example, to install into C:\mysql, use these commands: -C:\> mkdir C:\mysql -C:\> mkdir C:\mysql\bin -C:\> mkdir C:\mysql\data -C:\> mkdir C:\mysql\share -C:\> mkdir C:\mysql\scripts - If you want to compile other clients and link them to MySQL, - you should also create several additional directories: -C:\> mkdir C:\mysql\include -C:\> mkdir C:\mysql\lib -C:\> mkdir C:\mysql\lib\debug -C:\> mkdir C:\mysql\lib\opt - If you want to benchmark MySQL, create this directory: -C:\> mkdir C:\mysql\sql-bench - Benchmarking requires Perl support. See Section 2.15, "Perl - Installation Notes." + When specifying options on the command line, you can enclose the + entire command-line option and the value you are specifying using + double quotes. This enables you to use spaces in the options. For + example, "-cC:\mysql.ini". + + The following command installs a MySQL Server 5.1 instance from + the directory C:\Program Files\MySQL\MySQL Server 5.1 using the + service name MySQL51 and setting the root password to 1234. +shell> MySQLInstanceConfig.exe -i -q "-lC:\mysql_install_log.txt" » + "-nMySQL Server 5.1" "-pC:\Program Files\MySQL\MySQL Server 5.1" - +v5.1.39 » + "-tmy-template.ini" "-cC:\mytest.ini" ServerType=DEVELOPMENT Datab +aseType=MIXED » + ConnectionUsage=DSS Port=3311 ServiceName=MySQL51 RootPassword=123 +4 + + In the above example, a log file will be generated in + mysql_install_log.txt containing the information about the + instance creation process. The log file generated by the above + example is shown below: +Welcome to the MySQL Server Instance Configuration Wizard 1.0.16.0 +Date: 2009-10-27 17:07:21 + +Installing service ... + +Product Name: MySQL Server 5.1 +Version: 5.1.39 +Installation Path: C:\Program Files\MySQL\MySQL Server 5.1\ + +Creating configuration file C:\mytest.ini using template my-template. +ini. +Options: +DEVELOPMENT +MIXED +DSS +STRICTMODE + +Variables: +port: 3311 +default-character-set: latin1 +basedir: "C:/Program Files/MySQL/MySQL Server 5.1/" +datadir: "C:/Program Files/MySQL/MySQL Server 5.1/Data/" + + +Creating Windows service entry. +Service name: "MySQL51" +Parameters: "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" -- +defaults-file="C:\mytest.ini" MySQL51. +Windows service MySQL51 installed. + + When using the command-line, the return values in the following + table indicate an error performing the specified option. + + Table 2.6. Return Value from MySQL Server Instance Config Wizard + Value Description + 2 Configuration template file cannot be found + 3 The Windows service entry cannot be created + 4 Could not connect to the Service Control Manager + 5 The MySQL service cannot be started + 6 The MySQL service cannot be stopped + 7 The security settings cannot be applied + 8 The configuration file cannot be written + 9 The Windows service entry cannot be removed + + You can perform an installation of MySQL automatically using the + MSI packe. For more information, see Section 2.5.3.2, "Installing + MySQL Automatically using MSI." - 2. From the work directory, copy into the C:\mysql directory the - following directories: -C:\> cd \workdir -C:\workdir> copy client_release\*.exe C:\mysql\bin -C:\workdir> copy client_debug\mysqld.exe C:\mysql\bin\mysqld-debug.ex -e -C:\workdir> xcopy scripts\*.* C:\mysql\scripts /E -C:\workdir> xcopy share\*.* C:\mysql\share /E - If you want to compile other clients and link them to MySQL, - you should also copy several libraries and header files: -C:\workdir> copy lib_debug\mysqlclient.lib C:\mysql\lib\debug -C:\workdir> copy lib_debug\libmysql.* C:\mysql\lib\debug -C:\workdir> copy lib_debug\zlib.* C:\mysql\lib\debug -C:\workdir> copy lib_release\mysqlclient.lib C:\mysql\lib\opt -C:\workdir> copy lib_release\libmysql.* C:\mysql\lib\opt -C:\workdir> copy lib_release\zlib.* C:\mysql\lib\opt -C:\workdir> copy include\*.h C:\mysql\include -C:\workdir> copy libmysql\libmysql.def C:\mysql\include - If you want to benchmark MySQL, you should also do this: -C:\workdir> xcopy sql-bench\*.* C:\mysql\bench /E +2.5.5. Installing MySQL from a noinstall Zip Archive - After installation, set up and start the server in the same way as - for binary Windows distributions. See Section 2.3, "Installing - MySQL on Windows." + Users who are installing from the noinstall package can use the + instructions in this section to manually install MySQL. The + process for installing MySQL from a Zip archive is as follows: -2.10.7. Compiling MySQL Clients on Windows + 1. Extract the archive to the desired install directory - In your source files, you should include my_global.h before - mysql.h: -#include <my_global.h> -#include <mysql.h> + 2. Create an option file + + 3. Choose a MySQL server type - my_global.h includes any other files needed for Windows - compatibility (such as windows.h) if you compile your program on - Windows. + 4. Start the MySQL server - You can either link your code with the dynamic libmysql.lib - library, which is just a wrapper to load in libmysql.dll on - demand, or link with the static mysqlclient.lib library. + 5. Secure the default user accounts - The MySQL client libraries are compiled as threaded libraries, so - you should also compile your code to be multi-threaded. + This process is described in the sections that follow. -2.11. Post-Installation Setup and Testing +2.5.5.1. Extracting the Install Archive - After installing MySQL, there are some issues that you should - address. For example, on Unix, you should initialize the data - directory and create the MySQL grant tables. On all platforms, an - important security concern is that the initial accounts in the - grant tables have no passwords. You should assign passwords to - prevent unauthorized access to the MySQL server. Optionally, you - can create time zone tables to enable recognition of named time - zones. + To install MySQL manually, do the following: - The following sections include post-installation procedures that - are specific to Windows systems and to Unix systems. Another - section, Section 2.11.2.3, "Starting and Troubleshooting the MySQL - Server," applies to all platforms; it describes what to do if you - have trouble getting the server to start. Section 2.11.3, - "Securing the Initial MySQL Accounts," also applies to all - platforms. You should follow its instructions to make sure that - you have properly protected your MySQL accounts by assigning - passwords to them. + 1. If you are upgrading from a previous version please refer to + Section 2.5.7, "Upgrading MySQL on Windows," before beginning + the upgrade process. - When you are ready to create additional user accounts, you can - find information on the MySQL access control system and account - management in Section 5.4, "The MySQL Access Privilege System," - and Section 5.5, "MySQL User Account Management." + 2. Make sure that you are logged in as a user with administrator + privileges. -2.11.1. Windows Post-Installation Procedures + 3. Choose an installation location. Traditionally, the MySQL + server is installed in C:\mysql. The MySQL Installation Wizard + installs MySQL under C:\Program Files\MySQL. If you do not + install MySQL at C:\mysql, you must specify the path to the + install directory during startup or in an option file. See + Section 2.5.5.2, "Creating an Option File." - On Windows, the data directory and the grant tables do not have to - be created. MySQL Windows distributions include the grant tables - with a set of preinitialized accounts in the mysql database under - the data directory. It is unnecessary to run the mysql_install_db - script that is used on Unix. Regarding passwords, if you installed - MySQL using the Windows Installation Wizard, you may have already - assigned passwords to the accounts. (See Section 2.3.3, "Using the - MySQL Installation Wizard.") Otherwise, use the - password-assignment procedure given in Section 2.11.3, "Securing - the Initial MySQL Accounts." + 4. Extract the install archive to the chosen installation + location using your preferred Zip archive tool. Some tools may + extract the archive to a folder within your chosen + installation location. If this occurs, you can move the + contents of the subfolder into the chosen installation + location. - Before setting up passwords, you might want to try running some - client programs to make sure that you can connect to the server - and that it is operating properly. Make sure that the server is - running (see Section 2.3.9, "Starting the Server for the First - Time"), and then issue the following commands to verify that you - can retrieve information from the server. The output should be - similar to what is shown here: -C:\> C:\mysql\bin\mysqlshow -+--------------------+ -| Databases | -+--------------------+ -| information_schema | -| mysql | -| test | -+--------------------+ +2.5.5.2. Creating an Option File -C:\> C:\mysql\bin\mysqlshow mysql -Database: mysql -+---------------------------+ -| Tables | -+---------------------------+ -| columns_priv | -| db | -| event | -| func | -| general_log | -| help_category | -| help_keyword | -| help_relation | -| help_topic | -| host | -| plugin | -| proc | -| procs_priv | -| servers | -| slow_log | -| tables_priv | -| time_zone | -| time_zone_leap_second | -| time_zone_name | -| time_zone_transition | -| time_zone_transition_type | -| user | -+---------------------------+ + If you need to specify startup options when you run the server, + you can indicate them on the command line or place them in an + option file. For options that are used every time the server + starts, you may find it most convenient to use an option file to + specify your MySQL configuration. This is particularly true under + the following circumstances: + * The installation or data directory locations are different + from the default locations (C:\Program Files\MySQL\MySQL + Server 5.1 and C:\Program Files\MySQL\MySQL Server 5.1\data). -C:\> C:\mysql\bin\mysql -e "SELECT Host,Db,User FROM db" mysql -+------+-------+------+ -| host | db | user | -+------+-------+------+ -| % | test% | | -+------+-------+------+ + * You need to tune the server settings, such as memory, cache, + or InnoDB configuration information. - You may need to specify a different directory from the one shown; - if you used the Windows Installation Wizard, then the default - directory is C:\Program Files\MySQL\MySQL Server 5.1, and the - mysql and mysqlshow client programs are in C:\Program - Files\MySQL\MySQL Server 5.1\bin. See Section 2.3.3, "Using the - MySQL Installation Wizard," for more information. + When the MySQL server starts on Windows, it looks for option files + in several locations, such as the Windows directory, C:\, and the + MySQL installation directory (for the full list of locations, see + Section 4.2.3.3, "Using Option Files"). The Windows directory + typically is named something like C:\WINDOWS. You can determine + its exact location from the value of the WINDIR environment + variable using the following command: +C:\> echo %WINDIR% - If you have already secured the initial MySQL accounts, you may - need to use the -u and -p options to supply a user name and - password to the mysqlshow and mysql client programs; otherwise the - programs may fail with an error, or you may not be able to view - all databases. For example, if you have assigned the password - "secretpass" to the MySQL root account, then you can invoke - mysqlshow and mysql as shown here: -C:\> C:\mysql\bin\mysqlshow -uroot -psecretpass -+--------------------+ -| Databases | -+--------------------+ -| information_schema | -| mysql | -| test | -+--------------------+ + MySQL looks for options in each location first in the my.ini file, + and then in the my.cnf file. However, to avoid confusion, it is + best if you use only one file. If your PC uses a boot loader where + C: is not the boot drive, your only option is to use the my.ini + file. Whichever option file you use, it must be a plain text file. -C:\> C:\mysql\bin\mysqlshow -uroot -psecretpass mysql -Database: mysql -+---------------------------+ -| Tables | -+---------------------------+ -| columns_priv | -| db | -| event | -| func | -| general_log | -| help_category | -| help_keyword | -| help_relation | -| help_topic | -| host | -| plugin | -| proc | -| procs_priv | -| servers | -| slow_log | -| tables_priv | -| time_zone | -| time_zone_leap_second | -| time_zone_name | -| time_zone_transition | -| time_zone_transition_type | -| user | -+---------------------------+ + You can also make use of the example option files included with + your MySQL distribution; see Section 4.2.3.3.2, "Preconfigured + Option Files." + An option file can be created and modified with any text editor, + such as Notepad. For example, if MySQL is installed in E:\mysql + and the data directory is in E:\mydata\data, you can create an + option file containing a [mysqld] section to specify values for + the basedir and datadir options: +[mysqld] +# set basedir to your installation path +basedir=E:/mysql +# set datadir to the location of your data directory +datadir=E:/mydata/data -C:\> C:\mysql\bin\mysql -uroot -psecretpass -e "SELECT Host,Db,User F -ROM db" mysql -+------+-------+------+ -| host | db | user | -+------+-------+------+ -| % | test% | | -+------+-------+------+ + Note that Windows path names are specified in option files using + (forward) slashes rather than backslashes. If you do use + backslashes, double them: +[mysqld] +# set basedir to your installation path +basedir=E:\\mysql +# set datadir to the location of your data directory +datadir=E:\\mydata\\data - For more information about these programs, see Section 4.5.6, - "mysqlshow --- Display Database, Table, and Column Information," - and Section 4.5.1, "mysql --- The MySQL Command-Line Tool." + The rules for use of backslash in option file values are given in + Section 4.2.3.3, "Using Option Files." - If you are running a version of Windows that supports services and - you want the MySQL server to run automatically when Windows - starts, see Section 2.3.11, "Starting MySQL as a Windows Service." + MySQL Enterprise For expert advice on the start-up options + appropriate to your circumstances, subscribe to the MySQL + Enterprise Monitor. For more information, see + http://www.mysql.com/products/enterprise/advisors.html. -2.11.2. Unix Post-Installation Procedures + In MySQL 5.1.23 and earlier, the MySQL installer places the data + directory directly under the directory where you install MySQL. On + MySQL 5.1.24 and later, the data directory is located within the + AppData directory for the user running MySQL. - After installing MySQL on Unix, you need to initialize the grant - tables, start the server, and make sure that the server works - satisfactorily. You may also wish to arrange for the server to be - started and stopped automatically when your system starts and - stops. You should also assign passwords to the accounts in the - grant tables. + If you would like to use a data directory in a different location, + you should copy the entire contents of the data directory to the + new location. For example, if you want to use E:\mydata as the + data directory instead, you must do two things: - On Unix, the grant tables are set up by the mysql_install_db - program. For some installation methods, this program is run for - you automatically: + 1. Move the entire data directory and all of its contents from + the default location (for example C:\Program Files\MySQL\MySQL + Server 5.1\data) to E:\mydata. - * If you install MySQL on Linux using RPM distributions, the - server RPM runs mysql_install_db. + 2. Use a --datadir option to specify the new data directory + location each time you start the server. - * If you install MySQL on Mac OS X using a PKG distribution, the - installer runs mysql_install_db. +2.5.5.3. Selecting a MySQL Server Type - Otherwise, you'll need to run mysql_install_db yourself. + The following table shows the available servers for Windows in + MySQL 5.1.20 and earlier. + Binary Description + mysqld-nt Optimized binary with named-pipe support + mysqld Optimized binary without named-pipe support + mysqld-debug Like mysqld-nt, but compiled with full debugging and + automatic memory allocation checking - The following procedure describes how to initialize the grant - tables (if that has not previously been done) and then start the - server. It also suggests some commands that you can use to test - whether the server is accessible and working properly. For - information about starting and stopping the server automatically, - see Section 2.11.2.2, "Starting and Stopping MySQL Automatically." + The following table shows the available servers for Windows in + MySQL 5.1.21 and later. + Binary Description + mysqld Optimized binary with named-pipe support + mysqld-debug Like mysqld, but compiled with full debugging and + automatic memory allocation checking - After you complete the procedure and have the server running, you - should assign passwords to the accounts created by - mysql_install_db. Instructions for doing so are given in Section - 2.11.3, "Securing the Initial MySQL Accounts." + All of the preceding binaries are optimized for modern Intel + processors, but should work on any Intel i386-class or higher + processor. - In the examples shown here, the server runs under the user ID of - the mysql login account. This assumes that such an account exists. - Either create the account if it does not exist, or substitute the - name of a different existing login account that you plan to use - for running the server. + Each of the servers in a distribution support the same set of + storage engines. The SHOW ENGINES statement displays which engines + a given server supports. - 1. Change location into the top-level directory of your MySQL - installation, represented here by BASEDIR: -shell> cd BASEDIR - BASEDIR is likely to be something like /usr/local/mysql or - /usr/local. The following steps assume that you are located in - this directory. + All Windows MySQL 5.1 servers have support for symbolic linking of + database directories. - 2. If necessary, run the mysql_install_db program to set up the - initial MySQL grant tables containing the privileges that - determine how users are allowed to connect to the server. - You'll need to do this if you used a distribution type for - which the installation procedure doesn't run the program for - you. - Typically, mysql_install_db needs to be run only the first - time you install MySQL, so you can skip this step if you are - upgrading an existing installation, However, mysql_install_db - does not overwrite any existing privilege tables, so it should - be safe to run in any circumstances. - To initialize the grant tables, use one of the following - commands, depending on whether mysql_install_db is located in - the bin or scripts directory: -shell> bin/mysql_install_db --user=mysql -shell> scripts/mysql_install_db --user=mysql - It might be necessary to specify other options such as - --basedir or --datadir if mysql_install_db does not use the - correct locations for the installation directory or data - directory. For example: -shell> bin/mysql_install_db --user=mysql \ - --basedir=/opt/mysql/mysql \ - --datadir=/opt/mysql/mysql/data - The mysql_install_db script creates the server's data - directory. Under the data directory, it creates directories - for the mysql database that holds all database privileges and - the test database that you can use to test MySQL. The script - also creates privilege table entries for root and - anonymous-user accounts. The accounts have no passwords - initially. A description of their initial privileges is given - in Section 2.11.3, "Securing the Initial MySQL Accounts." - Briefly, these privileges allow the MySQL root user to do - anything, and allow anybody to create or use databases with a - name of test or starting with test_. - It is important to make sure that the database directories and - files are owned by the mysql login account so that the server - has read and write access to them when you run it later. To - ensure this, the --user option should be used as shown if you - run mysql_install_db as root. Otherwise, you should execute - the script while logged in as mysql, in which case you can - omit the --user option from the command. - mysql_install_db creates several tables in the mysql database, - including user, db, host, tables_priv, columns_priv, func, and - others. See Section 5.4, "The MySQL Access Privilege System," - for a complete listing and description of these tables. - If you don't want to have the test database, you can remove it - with mysqladmin -u root drop test after starting the server. - If you have trouble with mysql_install_db at this point, see - Section 2.11.2.1, "Problems Running mysql_install_db." + MySQL supports TCP/IP on all Windows platforms. MySQL servers on + Windows support named pipes as indicated in the following list. + However, the default is to use TCP/IP regardless of platform. + (Named pipes are slower than TCP/IP in many Windows + configurations.) - 3. Start the MySQL server: -shell> bin/mysqld_safe --user=mysql & - It is important that the MySQL server be run using an - unprivileged (non-root) login account. To ensure this, the - --user option should be used as shown if you run mysqld_safe - as system root. Otherwise, you should execute the script while - logged in to the system as mysql, in which case you can omit - the --user option from the command. - Further instructions for running MySQL as an unprivileged user - are given in Section 5.3.5, "How to Run MySQL as a Normal - User." - If you neglected to create the grant tables before proceeding - to this step, the following message appears in the error log - file when you start the server: -mysqld: Can't find file: 'host.frm' - If you have other problems starting the server, see Section - 2.11.2.3, "Starting and Troubleshooting the MySQL Server." + Use of named pipes is subject to these conditions: - 4. Use mysqladmin to verify that the server is running. The - following commands provide simple tests to check whether the - server is up and responding to connections: -shell> bin/mysqladmin version -shell> bin/mysqladmin variables - The output from mysqladmin version varies slightly depending - on your platform and version of MySQL, but should be similar - to that shown here: -shell> bin/mysqladmin version -mysqladmin Ver 14.12 Distrib 5.1.39, for pc-linux-gnu on i686 -... + * Named pipes are enabled only if you start the server with the + --enable-named-pipe option. It is necessary to use this option + explicitly because some users have experienced problems with + shutting down the MySQL server when named pipes were used. -Server version 5.1.39 -Protocol version 10 -Connection Localhost via UNIX socket -UNIX socket /var/lib/mysql/mysql.sock -Uptime: 14 days 5 hours 5 min 21 sec + * For MySQL 5.1.20 and earlier, named-pipe connections are + allowed only by the mysqld-nt and mysqld-debug servers. For + MySQL 5.1.21 and later, the mysqld and mysqld-debug servers + both contain support for named-pipe connections. -Threads: 1 Questions: 366 Slow queries: 0 -Opens: 0 Flush tables: 1 Open tables: 19 -Queries per second avg: 0.000 - To see what else you can do with mysqladmin, invoke it with - the --help option. +Note - 5. Verify that you can shut down the server: -shell> bin/mysqladmin -u root shutdown + Most of the examples in this manual use mysqld as the server name. + If you choose to use a different server, such as mysqld-nt or + mysqld-debug, make the appropriate substitutions in the commands + that are shown in the examples. - 6. Verify that you can start the server again. Do this by using - mysqld_safe or by invoking mysqld directly. For example: -shell> bin/mysqld_safe --user=mysql --log & - If mysqld_safe fails, see Section 2.11.2.3, "Starting and - Troubleshooting the MySQL Server." +2.5.5.4. Starting the Server for the First Time - 7. Run some simple tests to verify that you can retrieve - information from the server. The output should be similar to - what is shown here: -shell> bin/mysqlshow -+-----------+ -| Databases | -+-----------+ -| mysql | -| test | -+-----------+ + This section gives a general overview of starting the MySQL + server. The following sections provide more specific information + for starting the MySQL server from the command line or as a + Windows service. -shell> bin/mysqlshow mysql -Database: mysql -+---------------------------+ -| Tables | -+---------------------------+ -| columns_priv | -| db | -| func | -| help_category | -| help_keyword | -| help_relation | -| help_topic | -| host | -| proc | -| procs_priv | -| tables_priv | -| time_zone | -| time_zone_leap_second | -| time_zone_name | -| time_zone_transition | -| time_zone_transition_type | -| user | -+---------------------------+ + The information here applies primarily if you installed MySQL + using the Noinstall version, or if you wish to configure and test + MySQL manually rather than with the GUI tools. -shell> bin/mysql -e "SELECT Host,Db,User FROM db" mysql -+------+--------+------+ -| host | db | user | -+------+--------+------+ -| % | test | | -| % | test_% | | -+------+--------+------+ + The examples in these sections assume that MySQL is installed + under the default location of C:\Program Files\MySQL\MySQL Server + 5.1. Adjust the path names shown in the examples if you have MySQL + installed in a different location. - 8. There is a benchmark suite in the sql-bench directory (under - the MySQL installation directory) that you can use to compare - how MySQL performs on different platforms. The benchmark suite - is written in Perl. It requires the Perl DBI module that - provides a database-independent interface to the various - databases, and some other additional Perl modules: -DBI -DBD::mysql -Data::Dumper -Data::ShowTable - These modules can be obtained from CPAN - (http://www.cpan.org/) See also Section 2.15.1, "Installing - Perl on Unix." - The sql-bench/Results directory contains the results from many - runs against different databases and platforms. To run all - tests, execute these commands: -shell> cd sql-bench -shell> perl run-all-tests - If you don't have the sql-bench directory, you probably - installed MySQL using RPM files other than the source RPM. - (The source RPM includes the sql-bench benchmark directory.) - In this case, you must first install the benchmark suite - before you can use it. There are separate benchmark RPM files - named mysql-bench-VERSION.i386.rpm that contain benchmark code - and data. - If you have a source distribution, there are also tests in its - tests subdirectory that you can run. For example, to run - auto_increment.tst, execute this command from the top-level - directory of your source distribution: -shell> mysql -vvf test < ./tests/auto_increment.tst - The expected result of the test can be found in the - ./tests/auto_increment.res file. + Clients have two options. They can use TCP/IP, or they can use a + named pipe if the server supports named-pipe connections. - 9. At this point, you should have the server running. However, - none of the initial MySQL accounts have a password, so you - should assign passwords using the instructions found in - Section 2.11.3, "Securing the Initial MySQL Accounts." + MySQL for Windows also supports shared-memory connections if the + server is started with the --shared-memory option. Clients can + connect through shared memory by using the --protocol=MEMORY + option. - The MySQL 5.1 installation procedure creates time zone tables in - the mysql database. However, you must populate the tables manually - using the instructions in Section 9.7, "MySQL Server Time Zone - Support." + For information about which server binary to run, see Section + 2.5.5.3, "Selecting a MySQL Server Type." -2.11.2.1. Problems Running mysql_install_db + Testing is best done from a command prompt in a console window (or + "DOS window"). In this way you can have the server display status + messages in the window where they are easy to see. If something is + wrong with your configuration, these messages make it easier for + you to identify and fix any problems. - The purpose of the mysql_install_db script is to generate new - MySQL privilege tables. It does not overwrite existing MySQL - privilege tables, and it does not affect any other data. + To start the server, enter this command: +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" --console - If you want to re-create your privilege tables, first stop the - mysqld server if it is running. Then rename the mysql directory - under the data directory to save it, and then run - mysql_install_db. Suppose that your current directory is the MySQL - installation directory and that mysql_install_db is located in the - bin directory and the data directory is named data. To rename the - mysql database and re-run mysql_install_db, use these commands. -shell> mv data/mysql data/mysql.old -shell> bin/mysql_install_db --user=mysql + For a server that includes InnoDB support, you should see the + messages similar to those following as it starts (the path names + and sizes may differ): +InnoDB: The first specified datafile c:\ibdata\ibdata1 did not exist: +InnoDB: a new database to be created! +InnoDB: Setting file c:\ibdata\ibdata1 size to 209715200 +InnoDB: Database physically writes the file full: wait... +InnoDB: Log file c:\iblogs\ib_logfile0 did not exist: new to be creat +ed +InnoDB: Setting log file c:\iblogs\ib_logfile0 size to 31457280 +InnoDB: Log file c:\iblogs\ib_logfile1 did not exist: new to be creat +ed +InnoDB: Setting log file c:\iblogs\ib_logfile1 size to 31457280 +InnoDB: Log file c:\iblogs\ib_logfile2 did not exist: new to be creat +ed +InnoDB: Setting log file c:\iblogs\ib_logfile2 size to 31457280 +InnoDB: Doublewrite buffer not found: creating new +InnoDB: Doublewrite buffer created +InnoDB: creating foreign key constraint system tables +InnoDB: foreign key constraint system tables created +011024 10:58:25 InnoDB: Started - When you run mysql_install_db, you might encounter the following - problems: + When the server finishes its startup sequence, you should see + something like this, which indicates that the server is ready to + service client connections: +mysqld: ready for connections +Version: '5.1.41' socket: '' port: 3306 - * mysql_install_db fails to install the grant tables - You may find that mysql_install_db fails to install the grant - tables and terminates after displaying the following messages: -Starting mysqld daemon with databases from XXXXXX -mysqld ended - In this case, you should examine the error log file very - carefully. The log should be located in the directory XXXXXX - named by the error message and should indicate why mysqld - didn't start. If you do not understand what happened, include - the log when you post a bug report. See Section 1.6, "How to - Report Bugs or Problems." + The server continues to write to the console any further + diagnostic output it produces. You can open a new console window + in which to run client programs. - * There is a mysqld process running - This indicates that the server is running, in which case the - grant tables have probably been created already. If so, there - is no need to run mysql_install_db at all because it needs to - be run only once (when you install MySQL the first time). + If you omit the --console option, the server writes diagnostic + output to the error log in the data directory (C:\Program + Files\MySQL\MySQL Server 5.1\data by default). The error log is + the file with the .err extension. - * Installing a second mysqld server does not work when one - server is running - This can happen when you have an existing MySQL installation, - but want to put a new installation in a different location. - For example, you might have a production installation, but you - want to create a second installation for testing purposes. - Generally the problem that occurs when you try to run a second - server is that it tries to use a network interface that is in - use by the first server. In this case, you should see one of - the following error messages: -Can't start server: Bind on TCP/IP port: -Address already in use -Can't start server: Bind on unix socket... - For instructions on setting up multiple servers, see Section - 5.6, "Running Multiple MySQL Servers on the Same Machine." +Note - * You do not have write access to the /tmp directory - If you do not have write access to create temporary files or a - Unix socket file in the default location (the /tmp directory), - an error occurs when you run mysql_install_db or the mysqld - server. - You can specify different locations for the temporary - directory and Unix socket file by executing these commands - prior to starting mysql_install_db or mysqld, where - some_tmp_dir is the full path name to some directory for which - you have write permission: -shell> TMPDIR=/some_tmp_dir/ -shell> MYSQL_UNIX_PORT=/some_tmp_dir/mysql.sock -shell> export TMPDIR MYSQL_UNIX_PORT - Then you should be able to run mysql_install_db and start the - server with these commands: -shell> bin/mysql_install_db --user=mysql -shell> bin/mysqld_safe --user=mysql & - If mysql_install_db is located in the scripts directory, - modify the first command to scripts/mysql_install_db. - See Section B.1.4.5, "How to Protect or Change the MySQL Unix - Socket File," and Section 2.14, "Environment Variables." + The accounts that are listed in the MySQL grant tables initially + have no passwords. After starting the server, you should set up + passwords for them using the instructions in Section 2.13, + "Post-Installation Setup and Testing." - There are some alternatives to running the mysql_install_db script - provided in the MySQL distribution: +2.5.5.5. Starting MySQL from the Windows Command Line - * If you want the initial privileges to be different from the - standard defaults, you can modify mysql_install_db before you - run it. However, it is preferable to use GRANT and REVOKE to - change the privileges after the grant tables have been set up. - In other words, you can run mysql_install_db, and then use - mysql -u root mysql to connect to the server as the MySQL root - user so that you can issue the necessary GRANT and REVOKE - statements. - If you want to install MySQL on several machines with the same - privileges, you can put the GRANT and REVOKE statements in a - file and execute the file as a script using mysql after - running mysql_install_db. For example: -shell> bin/mysql_install_db --user=mysql -shell> bin/mysql -u root < your_script_file - By doing this, you can avoid having to issue the statements - manually on each machine. + The MySQL server can be started manually from the command line. + This can be done on any version of Windows. - * It is possible to re-create the grant tables completely after - they have previously been created. You might want to do this - if you're just learning how to use GRANT and REVOKE and have - made so many modifications after running mysql_install_db that - you want to wipe out the tables and start over. - To re-create the grant tables, remove all the .frm, .MYI, and - .MYD files in the mysql database directory. Then run the - mysql_install_db script again. + To start the mysqld server from the command line, you should start + a console window (or "DOS window") and enter this command: +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" - * You can start mysqld manually using the --skip-grant-tables - option and add the privilege information yourself using mysql: -shell> bin/mysqld_safe --user=mysql --skip-grant-tables & -shell> bin/mysql mysql - From mysql, manually execute the SQL commands contained in - mysql_install_db. Make sure that you run mysqladmin - flush-privileges or mysqladmin reload afterward to tell the - server to reload the grant tables. - Note that by not using mysql_install_db, you not only have to - populate the grant tables manually, you also have to create - them first. + The path to mysqld may vary depending on the install location of + MySQL on your system. -2.11.2.2. Starting and Stopping MySQL Automatically + You can stop the MySQL server by executing this command: +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqladmin" -u root + shutdown - Generally, you start the mysqld server in one of these ways: +Note - * Invoke mysqld directly. This works on any platform. + If the MySQL root user account has a password, you need to invoke + mysqladmin with the -p option and supply the password when + prompted. - * Run the MySQL server as a Windows service. The service can be - set to start the server automatically when Windows starts, or - as a manual service that you start on request. For - instructions, see Section 2.3.11, "Starting MySQL as a Windows - Service." + This command invokes the MySQL administrative utility mysqladmin + to connect to the server and tell it to shut down. The command + connects as the MySQL root user, which is the default + administrative account in the MySQL grant system. Note that users + in the MySQL grant system are wholly independent from any login + users under Windows. - * Invoke mysqld_safe, which tries to determine the proper - options for mysqld and then runs it with those options. This - script is used on Unix and Unix-like systems. See Section - 4.3.2, "mysqld_safe --- MySQL Server Startup Script." + If mysqld doesn't start, check the error log to see whether the + server wrote any messages there to indicate the cause of the + problem. The error log is located in the C:\Program + Files\MySQL\MySQL Server 5.1\data directory. It is the file with a + suffix of .err. You can also try to start the server as mysqld + --console; in this case, you may get some useful information on + the screen that may help solve the problem. - * Invoke mysql.server. This script is used primarily at system - startup and shutdown on systems that use System V-style run - directories, where it usually is installed under the name - mysql. The mysql.server script starts the server by invoking - mysqld_safe. See Section 4.3.3, "mysql.server --- MySQL Server - Startup Script." + The last option is to start mysqld with the --standalone and + --debug options. In this case, mysqld writes a log file + C:\mysqld.trace that should contain the reason why mysqld doesn't + start. See MySQL Internals: Porting + (http://forge.mysql.com/wiki/MySQL_Internals_Porting) - * On Mac OS X, install a separate MySQL Startup Item package to - enable the automatic startup of MySQL on system startup. The - Startup Item starts the server by invoking mysql.server. See - Section 2.5, "Installing MySQL on Mac OS X," for details. + Use mysqld --verbose --help to display all the options that mysqld + supports. - The mysqld_safe and mysql.server scripts and the Mac OS X Startup - Item can be used to start the server manually, or automatically at - system startup time. mysql.server and the Startup Item also can be - used to stop the server. +2.5.5.6. Starting MySQL as a Windows Service - To start or stop the server manually using the mysql.server - script, invoke it with start or stop arguments: -shell> mysql.server start -shell> mysql.server stop + On Windows, the recommended way to run MySQL is to install it as a + Windows service, whereby MySQL starts and stops automatically when + Windows starts and stops. A MySQL server installed as a service + can also be controlled from the command line using NET commands, + or with the graphical Services utility. Generally, to install + MySQL as a Windows service you should be logged in using an + account that has administrator rights. - Before mysql.server starts the server, it changes location to the - MySQL installation directory, and then invokes mysqld_safe. If you - want the server to run as some specific user, add an appropriate - user option to the [mysqld] group of the /etc/my.cnf option file, - as shown later in this section. (It is possible that you will need - to edit mysql.server if you've installed a binary distribution of - MySQL in a nonstandard location. Modify it to change location into - the proper directory before it runs mysqld_safe. If you do this, - your modified version of mysql.server may be overwritten if you - upgrade MySQL in the future, so you should make a copy of your - edited version that you can reinstall.) + The Services utility (the Windows Service Control Manager) can be + found in the Windows Control Panel (under Administrative Tools on + Windows 2000, XP, Vista and Server 2003). To avoid conflicts, it + is advisable to close the Services utility while performing server + installation or removal operations from the command line. - mysql.server stop stops the server by sending a signal to it. You - can also stop the server manually by executing mysqladmin - shutdown. + Before installing MySQL as a Windows service, you should first + stop the current server if it is running by using the following + command: +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqladmin" + -u root shutdown - To start and stop MySQL automatically on your server, you need to - add start and stop commands to the appropriate places in your - /etc/rc* files. +Note - If you use the Linux server RPM package - (MySQL-server-VERSION.rpm), the mysql.server script is installed - in the /etc/init.d directory with the name mysql. You need not - install it manually. See Section 2.4, "Installing MySQL from RPM - Packages on Linux," for more information on the Linux RPM - packages. + If the MySQL root user account has a password, you need to invoke + mysqladmin with the -p option and supply the password when + prompted. - Some vendors provide RPM packages that install a startup script - under a different name such as mysqld. + This command invokes the MySQL administrative utility mysqladmin + to connect to the server and tell it to shut down. The command + connects as the MySQL root user, which is the default + administrative account in the MySQL grant system. Note that users + in the MySQL grant system are wholly independent from any login + users under Windows. - If you install MySQL from a source distribution or using a binary - distribution format that does not install mysql.server - automatically, you can install it manually. The script can be - found in the support-files directory under the MySQL installation - directory or in a MySQL source tree. + Install the server as a service using this command: +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" --install - To install mysql.server manually, copy it to the /etc/init.d - directory with the name mysql, and then make it executable. Do - this by changing location into the appropriate directory where - mysql.server is located and executing these commands: -shell> cp mysql.server /etc/init.d/mysql -shell> chmod +x /etc/init.d/mysql + The service-installation command does not start the server. + Instructions for that are given later in this section. - Older Red Hat systems use the /etc/rc.d/init.d directory rather - than /etc/init.d. Adjust the preceding commands accordingly. - Alternatively, first create /etc/init.d as a symbolic link that - points to /etc/rc.d/init.d: -shell> cd /etc -shell> ln -s rc.d/init.d . + To make it easier to invoke MySQL programs, you can add the path + name of the MySQL bin directory to your Windows system PATH + environment variable: - After installing the script, the commands needed to activate it to - run at system startup depend on your operating system. On Linux, - you can use chkconfig: -shell> chkconfig --add mysql + * On the Windows desktop, right-click on the My Computer icon, + and select Properties. - On some Linux systems, the following command also seems to be - necessary to fully enable the mysql script: -shell> chkconfig --level 345 mysql on + * Next select the Advanced tab from the System Properties menu + that appears, and click the Environment Variables button. - On FreeBSD, startup scripts generally should go in - /usr/local/etc/rc.d/. The rc(8) manual page states that scripts in - this directory are executed only if their basename matches the - *.sh shell file name pattern. Any other files or directories - present within the directory are silently ignored. In other words, - on FreeBSD, you should install the mysql.server script as - /usr/local/etc/rc.d/mysql.server.sh to enable automatic startup. + * Under System Variables, select Path, and then click the Edit + button. The Edit System Variable dialogue should appear. - As an alternative to the preceding setup, some operating systems - also use /etc/rc.local or /etc/init.d/boot.local to start - additional services on startup. To start up MySQL using this - method, you could append a command like the one following to the - appropriate startup file: -/bin/sh -c 'cd /usr/local/mysql; ./bin/mysqld_safe --user=mysql &' + * Place your cursor at the end of the text appearing in the + space marked Variable Value. (Use the End key to ensure that + your cursor is positioned at the very end of the text in this + space.) Then enter the complete path name of your MySQL bin + directory (for example, C:\Program Files\MySQL\MySQL Server + 5.1\bin), Note that there should be a semicolon separating + this path from any values present in this field. Dismiss this + dialogue, and each dialogue in turn, by clicking OK until all + of the dialogues that were opened have been dismissed. You + should now be able to invoke any MySQL executable program by + typing its name at the DOS prompt from any directory on the + system, without having to supply the path. This includes the + servers, the mysql client, and all MySQL command-line + utilities such as mysqladmin and mysqldump. + You should not add the MySQL bin directory to your Windows + PATH if you are running multiple MySQL servers on the same + machine. - For other systems, consult your operating system documentation to - see how to install startup scripts. +Warning - You can add options for mysql.server in a global /etc/my.cnf file. - A typical /etc/my.cnf file might look like this: -[mysqld] -datadir=/usr/local/mysql/var -socket=/var/tmp/mysql.sock -port=3306 -user=mysql + You must exercise great care when editing your system PATH by + hand; accidental deletion or modification of any portion of the + existing PATH value can leave you with a malfunctioning or even + unusable system. -[mysql.server] -basedir=/usr/local/mysql + The following additional arguments can be used in MySQL 5.1 when + installing the service: - The mysql.server script supports the following options: basedir, - datadir, and pid-file. If specified, they must be placed in an - option file, not on the command line. mysql.server supports only - start and stop as command-line arguments. + * You can specify a service name immediately following the + --install option. The default service name is MySQL. - The following table shows which option groups the server and each - startup script read from option files. - Script Option Groups - mysqld [mysqld], [server], [mysqld-major_version] - mysqld_safe [mysqld], [server], [mysqld_safe] - mysql.server [mysqld], [mysql.server], [server] + * If a service name is given, it can be followed by a single + option. By convention, this should be + --defaults-file=file_name to specify the name of an option + file from which the server should read options when it starts. + The use of a single option other than --defaults-file is + possible but discouraged. --defaults-file is more flexible + because it enables you to specify multiple startup options for + the server by placing them in the named option file. - [mysqld-major_version] means that groups with names like - [mysqld-5.0] and [mysqld-5.1] are read by servers having versions - 5.0.x, 5.1.x, and so forth. This feature can be used to specify - options that can be read only by servers within a given release - series. + * You can also specify a --local-service option following the + service name. This causes the server to run using the + LocalService Windows account that has limited system + privileges. This account is available only for Windows XP or + newer. If both --defaults-file and --local-service are given + following the service name, they can be in any order. - For backward compatibility, mysql.server also reads the - [mysql_server] group and mysqld_safe also reads the [safe_mysqld] - group. However, you should update your option files to use the - [mysql.server] and [mysqld_safe] groups instead when using MySQL - 5.1. + For a MySQL server that is installed as a Windows service, the + following rules determine the service name and option files that + the server uses: - See Section 4.2.3.3, "Using Option Files." + * If the service-installation command specifies no service name + or the default service name (MySQL) following the --install + option, the server uses the a service name of MySQL and reads + options from the [mysqld] group in the standard option files. -2.11.2.3. Starting and Troubleshooting the MySQL Server + * If the service-installation command specifies a service name + other than MySQL following the --install option, the server + uses that service name. It reads options from the [mysqld] + group and the group that has the same name as the service in + the standard option files. This allows you to use the [mysqld] + group for options that should be used by all MySQL services, + and an option group with the service name for use by the + server installed with that service name. - This section provides troubleshooting suggestions for problems - starting the server on Unix. If you are using Windows, see Section - 2.3.13, "Troubleshooting a MySQL Installation Under Windows." + * If the service-installation command specifies a + --defaults-file option after the service name, the server + reads options only from the [mysqld] group of the named file + and ignores the standard option files. - If you have problems starting the server, here are some things to - try: + As a more complex example, consider the following command: +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" + --install MySQL --defaults-file=C:\my-opts.cnf - * Check the error log to see why the server does not start. + Here, the default service name (MySQL) is given after the + --install option. If no --defaults-file option had been given, + this command would have the effect of causing the server to read + the [mysqld] group from the standard option files. However, + because the --defaults-file option is present, the server reads + options from the [mysqld] option group, and only from the named + file. - * Specify any special options needed by the storage engines you - are using. + You can also specify options as Start parameters in the Windows + Services utility before you start the MySQL service. - * Make sure that the server knows where to find the data - directory. + Once a MySQL server has been installed as a service, Windows + starts the service automatically whenever Windows starts. The + service also can be started immediately from the Services utility, + or by using a NET START MySQL command. The NET command is not case + sensitive. - * Make sure that the server can access the data directory. The - ownership and permissions of the data directory and its - contents must be set such that the server can read and modify - them. + When run as a service, mysqld has no access to a console window, + so no messages can be seen there. If mysqld does not start, check + the error log to see whether the server wrote any messages there + to indicate the cause of the problem. The error log is located in + the MySQL data directory (for example, C:\Program + Files\MySQL\MySQL Server 5.1\data). It is the file with a suffix + of .err. - * Verify that the network interfaces the server wants to use are - available. + When a MySQL server has been installed as a service, and the + service is running, Windows stops the service automatically when + Windows shuts down. The server also can be stopped manually by + using the Services utility, the NET STOP MySQL command, or the + mysqladmin shutdown command. - Some storage engines have options that control their behavior. You - can create a my.cnf file and specify startup options for the - engines that you plan to use. If you are going to use storage - engines that support transactional tables (InnoDB, NDB), be sure - that you have them configured the way you want before starting the - server: + You also have the choice of installing the server as a manual + service if you do not wish for the service to be started + automatically during the boot process. To do this, use the + --install-manual option rather than the --install option: +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" --install-m +anual - * If you are using InnoDB tables, see Section 13.6.2, "InnoDB - Configuration." + To remove a server that is installed as a service, first stop it + if it is running by executing NET STOP MySQL. Then use the + --remove option to remove it: +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqld" --remove - * If you are using MySQL Cluster, see Section 17.3, "MySQL - Cluster Configuration." + If mysqld is not running as a service, you can start it from the + command line. For instructions, see Section 2.5.5.5, "Starting + MySQL from the Windows Command Line." - MySQL Enterprise For expert advice on start-up options appropriate - to your circumstances, subscribe to The MySQL Enterprise Monitor. - For more information, see - http://www.mysql.com/products/enterprise/advisors.html. + Please see Section 2.5.6, "Troubleshooting a MySQL Installation + Under Windows," if you encounter difficulties during installation. - Storage engines will use default option values if you specify - none, but it is recommended that you review the available options - and specify explicit values for those for which the defaults are - not appropriate for your installation. +2.5.5.7. Testing The MySQL Installation - When the mysqld server starts, it changes location to the data - directory. This is where it expects to find databases and where it - expects to write log files. The server also writes the pid - (process ID) file in the data directory. + You can test whether the MySQL server is working by executing any + of the following commands: +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqlshow" +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqlshow" -u root +mysql +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqladmin" version + status proc +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysql" test - The data directory location is hardwired in when the server is - compiled. This is where the server looks for the data directory by - default. If the data directory is located somewhere else on your - system, the server will not work properly. You can determine what - the default path settings are by invoking mysqld with the - --verbose and --help options. +Note - If the default locations don't match the MySQL installation layout - on your system, you can override them by specifying options to - mysqld or mysqld_safe on the command line or in an option file. + By default, mysqlshow will try to connect using the ODBC user. + This user is not created by default. You should specify a valid + user, or root with the right password to check the operation of + the server. - To specify the location of the data directory explicitly, use the - --datadir option. However, normally you can tell mysqld the - location of the base directory under which MySQL is installed and - it looks for the data directory there. You can do this with the - --basedir option. + If mysqld is slow to respond to TCP/IP connections from client + programs, there is probably a problem with your DNS. In this case, + start mysqld with the --skip-name-resolve option and use only + localhost and IP numbers in the Host column of the MySQL grant + tables. - To check the effect of specifying path options, invoke mysqld with - those options followed by the --verbose and --help options. For - example, if you change location into the directory where mysqld is - installed and then run the following command, it shows the effect - of starting the server with a base directory of /usr/local: -shell> ./mysqld --basedir=/usr/local --verbose --help + You can force a MySQL client to use a named-pipe connection rather + than TCP/IP by specifying the --pipe or --protocol=PIPE option, or + by specifying . (period) as the host name. Use the --socket option + to specify the name of the pipe if you do not want to use the + default pipe name. - You can specify other options such as --datadir as well, but - --verbose and --help must be the last options. + Note that if you have set a password for the root account, deleted + the anonymous account, or created a new user account, then you + must use the appropriate -u and -p options with the commands shown + above in order to connect with the MySQL Server. See Section + 4.2.2, "Connecting to the MySQL Server." - Once you determine the path settings you want, start the server - without --verbose and --help. + For more information about mysqlshow, see Section 4.5.6, + "mysqlshow --- Display Database, Table, and Column Information." - If mysqld is currently running, you can find out what path - settings it is using by executing this command: -shell> mysqladmin variables +2.5.6. Troubleshooting a MySQL Installation Under Windows - Or: -shell> mysqladmin -h host_name variables + When installing and running MySQL for the first time, you may + encounter certain errors that prevent the MySQL server from + starting. The purpose of this section is to help you diagnose and + correct some of these errors. - host_name is the name of the MySQL server host. + Your first resource when troubleshooting server issues is the + error log. The MySQL server uses the error log to record + information relevant to the error that prevents the server from + starting. The error log is located in the data directory specified + in your my.ini file. The default data directory location is + C:\Program Files\MySQL\MySQL Server 5.1\data. See Section 5.2.2, + "The Error Log." - If you get Errcode 13 (which means Permission denied) when - starting mysqld, this means that the privileges of the data - directory or its contents do not allow the server access. In this - case, you change the permissions for the involved files and - directories so that the server has the right to use them. You can - also start the server as root, but this raises security issues and - should be avoided. + Another source of information regarding possible errors is the + console messages displayed when the MySQL service is starting. Use + the NET START MySQL command from the command line after installing + mysqld as a service to see any error messages regarding the + starting of the MySQL server as a service. See Section 2.5.5.6, + "Starting MySQL as a Windows Service." + + The following examples show other common error messages you may + encounter when installing MySQL and starting the server for the + first time: - On Unix, change location into the data directory and check the - ownership of the data directory and its contents to make sure the - server has access. For example, if the data directory is - /usr/local/mysql/var, use this command: -shell> ls -la /usr/local/mysql/var + * If the MySQL server cannot find the mysql privileges database + or other critical files, you may see these messages: +System error 1067 has occurred. +Fatal error: Can't open privilege tables: Table 'mysql.host' doesn't +exist + These messages often occur when the MySQL base or data + directories are installed in different locations than the + default locations (C:\Program Files\MySQL\MySQL Server 5.1 and + C:\Program Files\MySQL\MySQL Server 5.1\data, respectively). + This situation may occur when MySQL is upgraded and installed + to a new location, but the configuration file is not updated + to reflect the new location. In addition, there may be old and + new configuration files that conflict. Be sure to delete or + rename any old configuration files when upgrading MySQL. + If you have installed MySQL to a directory other than + C:\Program Files\MySQL\MySQL Server 5.1, you need to ensure + that the MySQL server is aware of this through the use of a + configuration (my.ini) file. The my.ini file needs to be + located in your Windows directory, typically C:\WINDOWS. You + can determine its exact location from the value of the WINDIR + environment variable by issuing the following command from the + command prompt: +C:\> echo %WINDIR% + An option file can be created and modified with any text + editor, such as Notepad. For example, if MySQL is installed in + E:\mysql and the data directory is D:\MySQLdata, you can + create the option file and set up a [mysqld] section to + specify values for the basedir and datadir options: +[mysqld] +# set basedir to your installation path +basedir=E:/mysql +# set datadir to the location of your data directory +datadir=D:/MySQLdata + Note that Windows path names are specified in option files + using (forward) slashes rather than backslashes. If you do use + backslashes, double them: +[mysqld] +# set basedir to your installation path +basedir=C:\\Program Files\\MySQL\\MySQL Server 5.1 +# set datadir to the location of your data directory +datadir=D:\\MySQLdata + The rules for use of backslash in option file values are given + in Section 4.2.3.3, "Using Option Files." + If you change the datadir value in your MySQL configuration + file, you must move the contents of the existing MySQL data + directory before restarting the MySQL server. + See Section 2.5.5.2, "Creating an Option File." - If the data directory or its files or subdirectories are not owned - by the login account that you use for running the server, change - their ownership to that account. If the account is named mysql, - use these commands: -shell> chown -R mysql /usr/local/mysql/var -shell> chgrp -R mysql /usr/local/mysql/var + * If you reinstall or upgrade MySQL without first stopping and + removing the existing MySQL service and install MySQL using + the MySQL Config Wizard, you may see this error: +Error: Cannot create Windows service for MySql. Error: 0 + This occurs when the Config Wizard tries to install the + service and finds an existing service with the same name. + One solution to this problem is to choose a service name other + than mysql when using the configuration wizard. This allows + the new service to be installed correctly, but leaves the + outdated service in place. Although this is harmless, it is + best to remove old services that are no longer in use. + To permanently remove the old mysql service, execute the + following command as a user with administrative privileges, on + the command-line: +C:\> sc delete mysql +[SC] DeleteService SUCCESS + If the sc utility is not available for your version of + Windows, download the delsrv utility from + http://www.microsoft.com/windows2000/techinfo/reskit/tools/exi + sting/delsrv-o.asp and use the delsrv mysql syntax. - If the server fails to start up correctly, check the error log. - Log files are located in the data directory (typically C:\Program - Files\MySQL\MySQL Server 5.1\data on Windows, - /usr/local/mysql/data for a Unix binary distribution, and - /usr/local/var for a Unix source distribution). Look in the data - directory for files with names of the form host_name.err and - host_name.log, where host_name is the name of your server host. - Then examine the last few lines of these files. On Unix, you can - use tail to display them: -shell> tail host_name.err -shell> tail host_name.log +2.5.7. Upgrading MySQL on Windows - The error log should contain information that indicates why the - server couldn't start. + This section lists some of the steps you should take when + upgrading MySQL on Windows. - If either of the following errors occur, it means that some other - program (perhaps another mysqld server) is using the TCP/IP port - or Unix socket file that mysqld is trying to use: -Can't start server: Bind on TCP/IP port: Address already in use -Can't start server: Bind on unix socket... + 1. Review Section 2.4.1, "Upgrading MySQL," for additional + information on upgrading MySQL that is not specific to + Windows. - Use ps to determine whether you have another mysqld server - running. If so, shut down the server before starting mysqld again. - (If another server is running, and you really want to run multiple - servers, you can find information about how to do so in Section - 5.6, "Running Multiple MySQL Servers on the Same Machine.") + 2. You should always back up your current MySQL installation + before performing an upgrade. See Section 6.1, "Database + Backup Methods." - If no other server is running, try to execute the command telnet - your_host_name tcp_ip_port_number. (The default MySQL port number - is 3306.) Then press Enter a couple of times. If you don't get an - error message like telnet: Unable to connect to remote host: - Connection refused, some other program is using the TCP/IP port - that mysqld is trying to use. You'll need to track down what - program this is and disable it, or else tell mysqld to listen to a - different port with the --port option. In this case, you'll also - need to specify the port number for client programs when - connecting to the server via TCP/IP. + 3. Download the latest Windows distribution of MySQL from + http://dev.mysql.com/downloads/. - Another reason the port might be inaccessible is that you have a - firewall running that blocks connections to it. If so, modify the - firewall settings to allow access to the port. + 4. Before upgrading MySQL, you must stop the server. If the + server is installed as a service, stop the service with the + following command from the command prompt: +C:\> NET STOP MySQL + If you are not running the MySQL server as a service, use the + following command to stop it: +C:\> "C:\Program Files\MySQL\MySQL Server 5.1\bin\mysqladmin" -u root + shutdown - If the server starts but you can't connect to it, you should make - sure that you have an entry in /etc/hosts that looks like this: -127.0.0.1 localhost +Note + If the MySQL root user account has a password, you need to + invoke mysqladmin with the -p option and supply the password + when prompted. - This problem occurs only on systems that do not have a working - thread library and for which MySQL must be configured to use - MIT-pthreads. + 5. When upgrading to MySQL 5.1 from a version previous to 4.1.5, + or when upgrading from a version of MySQL installed from a Zip + archive to a version of MySQL installed with the MySQL + Installation Wizard, you must manually remove the previous + installation and MySQL service (if the server is installed as + a service). + To remove the MySQL service, use the following command: +C:\> C:\mysql\bin\mysqld --remove + If you do not remove the existing service, the MySQL + Installation Wizard may fail to properly install the new MySQL + service. - If you cannot get mysqld to start, you can try to make a trace - file to find the problem by using the --debug option. See MySQL - Internals: Porting - (http://forge.mysql.com/wiki/MySQL_Internals_Porting) + 6. When upgrading from MySQL 5.1.23 to MySQL 5.1.24, the change + in the default location of the data directory from a directory + within the MySQL installation to the AppData folder means that + you must manually copy the data files from your old + installation to the new location. -2.11.3. Securing the Initial MySQL Accounts + 7. If you are using the MySQL Installation Wizard, start the + wizard as described in Section 2.5.3.1, "Using the MySQL + Installation Wizard." - Part of the MySQL installation process is to set up the mysql - database that contains the grant tables: + 8. If you are installing MySQL from a Zip archive, extract the + archive. You may either overwrite your existing MySQL + installation (usually located at C:\mysql), or install it into + a different directory, such as C:\mysql5. Overwriting the + existing installation is recommended. - * Windows distributions contain preinitialized grant tables that - are installed automatically. + 9. If you were running MySQL as a Windows service and you had to + remove the service earlier in this procedure, reinstall the + service. (See Section 2.5.5.6, "Starting MySQL as a Windows + Service.") + 10. Restart the server. For example, use NET START MySQL if you + run MySQL as a service, or invoke mysqld directly otherwise. + 11. If you encounter errors, see Section 2.5.6, "Troubleshooting a + MySQL Installation Under Windows." - * On Unix, the grant tables are populated by the - mysql_install_db program. Some installation methods run this - program for you. Others require that you execute it manually. - For details, see Section 2.11.2, "Unix Post-Installation - Procedures." +2.5.8. Windows Post-Installation Procedures - The grant tables define the initial MySQL user accounts and their - access privileges. These accounts are set up as follows: + On Windows, the data directory and the grant tables do not have to + be created. MySQL Windows distributions include the grant tables + with a set of preinitialized accounts in the mysql database under + the data directory. It is unnecessary to run the mysql_install_db + script that is used on Unix. Regarding passwords, if you installed + MySQL using the Windows Installation Wizard, you may have already + assigned passwords to the accounts. (See Section 2.5.3.1, "Using + the MySQL Installation Wizard.") Otherwise, use the + password-assignment procedure given in Section 2.13.2, "Securing + the Initial MySQL Accounts." - * Accounts with the user name root are created. These are - superuser accounts that can do anything. The initial root - account passwords are empty, so anyone can connect to the - MySQL server as root --- without a password --- and be granted - all privileges. + Before setting up passwords, you might want to try running some + client programs to make sure that you can connect to the server + and that it is operating properly. Make sure that the server is + running (see Section 2.5.5.4, "Starting the Server for the First + Time"), and then issue the following commands to verify that you + can retrieve information from the server. The output should be + similar to what is shown here: +C:\> C:\mysql\bin\mysqlshow ++--------------------+ +| Databases | ++--------------------+ +| information_schema | +| mysql | +| test | ++--------------------+ - + On Windows, one root account is created; this account - allows connecting from the local host only. The Windows - installer will optionally create an account allowing for - connections from any host only if the user selects the - Enable root access from remote machines option during - installation. +Note - + On Unix, both root accounts are for connections from the - local host. Connections must be made from the local host - by specifying a host name of localhost for one of the - accounts, or the actual host name or IP number for the - other. + The above may not work if the correct user does not exist. If you + installed using the MSI packages and used the MySQL Server + Instance Config Wizard, then the root will haqve been created + automatically with the password you supplied. In this case, you + should use the -u and -p options where you will be prompted for + the password. - * Two anonymous-user accounts are created, each with an empty - user name. The anonymous accounts have no password, so anyone - can use them to connect to the MySQL server. +Note - + On Windows, one anonymous account is for connections from - the local host. It has no global privileges. (Before - MySQL 5.1.16, it has all global privileges, just like the - root accounts.) The other is for connections from any - host and has all privileges for the test database and for - other databases with names that start with test. + The list of installed databases may vary, but will always include + the minimum of mysql and information_schema. In most cases, the + test database will also be installed automatically. + + If you specify the name of the database, then a list of the tables + within a given database will be displayed: +C:\> C:\mysql\bin\mysqlshow mysql +Database: mysql ++---------------------------+ +| Tables | ++---------------------------+ +| columns_priv | +| db | +| event | +| func | +| general_log | +| help_category | +| help_keyword | +| help_relation | +| help_topic | +| host | +| plugin | +| proc | +| procs_priv | +| servers | +| slow_log | +| tables_priv | +| time_zone | +| time_zone_leap_second | +| time_zone_name | +| time_zone_transition | +| time_zone_transition_type | +| user | ++---------------------------+ - + On Unix, both anonymous accounts are for connections from - the local host. Connections must be made from the local - host by specifying a host name of localhost for one of - the accounts, or the actual host name or IP number for - the other. These accounts have all privileges for the - test database and for other databases with names that - start with test_. - As noted, none of the initial accounts have passwords. This means - that your MySQL installation is unprotected until you do something - about it: +C:\> C:\mysql\bin\mysql -e "SELECT Host,Db,User FROM db" mysql ++------+-------+------+ +| host | db | user | ++------+-------+------+ +| % | test% | | ++------+-------+------+ - * If you want to prevent clients from connecting as anonymous - users without a password, you should either assign a password - to each anonymous account or else remove the accounts. + You may need to specify a different directory from the one shown; + if you used the Windows Installation Wizard, then the default + directory is C:\Program Files\MySQL\MySQL Server 5.1, and the + mysql and mysqlshow client programs are in C:\Program + Files\MySQL\MySQL Server 5.1\bin. See Section 2.5.3.1, "Using the + MySQL Installation Wizard," for more information. - * You should assign a password to each MySQL root account. + If you have already secured the initial MySQL accounts, you may + need to use the -u and -p options to supply a user name and + password to the mysqlshow and mysql client programs; otherwise the + programs may fail with an error, or you may not be able to view + all databases. For example, if you have assigned the password + "secretpass" to the MySQL root account, then you can invoke + mysqlshow and mysql as shown here: +C:\> C:\mysql\bin\mysqlshow -uroot -psecretpass ++--------------------+ +| Databases | ++--------------------+ +| information_schema | +| mysql | +| test | ++--------------------+ - The following instructions describe how to set up passwords for - the initial MySQL accounts, first for the anonymous accounts and - then for the root accounts. Replace "newpwd" in the examples with - the actual password that you want to use. The instructions also - cover how to remove the anonymous accounts, should you prefer not - to allow anonymous access at all. +C:\> C:\mysql\bin\mysqlshow -uroot -psecretpass mysql +Database: mysql ++---------------------------+ +| Tables | ++---------------------------+ +| columns_priv | +| db | +| event | +| func | +| general_log | +| help_category | +| help_keyword | +| help_relation | +| help_topic | +| host | +| plugin | +| proc | +| procs_priv | +| servers | +| slow_log | +| tables_priv | +| time_zone | +| time_zone_leap_second | +| time_zone_name | +| time_zone_transition | +| time_zone_transition_type | +| user | ++---------------------------+ - You might want to defer setting the passwords until later, so that - you don't need to specify them while you perform additional setup - or testing. However, be sure to set them before using your - installation for production purposes. - Anonymous Account Password Assignment +C:\> C:\mysql\bin\mysql -uroot -psecretpass -e "SELECT Host,Db,User F +ROM db" mysql ++------+-------+------+ +| host | db | user | ++------+-------+------+ +| % | test% | | ++------+-------+------+ - To assign passwords to the anonymous accounts, connect to the - server as root and then use either SET PASSWORD or UPDATE. In - either case, be sure to encrypt the password using the PASSWORD() - function. + For more information about these programs, see Section 4.5.6, + "mysqlshow --- Display Database, Table, and Column Information," + and Section 4.5.1, "mysql --- The MySQL Command-Line Tool." - To use SET PASSWORD on Windows, do this: -shell> mysql -u root -mysql> SET PASSWORD FOR ''@'localhost' = PASSWORD('newpwd'); -mysql> SET PASSWORD FOR ''@'%' = PASSWORD('newpwd'); + If you are running a version of Windows that supports services and + you want the MySQL server to run automatically when Windows + starts, see Section 2.5.5.6, "Starting MySQL as a Windows + Service." - To use SET PASSWORD on Unix, do this: -shell> mysql -u root -mysql> SET PASSWORD FOR ''@'localhost' = PASSWORD('newpwd'); -mysql> SET PASSWORD FOR ''@'host_name' = PASSWORD('newpwd'); +2.5.9. MySQL on Windows Compared to MySQL on Unix - In the second SET PASSWORD statement, replace host_name with the - name of the server host. This is the name that is specified in the - Host column of the non-localhost record for root in the user - table. If you don't know what host name this is, issue the - following statement before using SET PASSWORD: -mysql> SELECT Host, User FROM mysql.user; + MySQL for Windows has proven itself to be very stable. The Windows + version of MySQL has the same features as the corresponding Unix + version, with the following exceptions: - Look for the record that has root in the User column and something - other than localhost in the Host column. Then use that Host value - in the second SET PASSWORD statement. + * Limited number of ports + Windows systems have about 4,000 ports available for client + connections, and after a connection on a port closes, it takes + two to four minutes before the port can be reused. In + situations where clients connect to and disconnect from the + server at a high rate, it is possible for all available ports + to be used up before closed ports become available again. If + this happens, the MySQL server appears to be unresponsive even + though it is running. Note that ports may be used by other + applications running on the machine as well, in which case the + number of ports available to MySQL is lower. + For more information about this problem, see + http://support.microsoft.com/default.aspx?scid=kb;en-us;196271 + . - Anonymous Account Removal + * Concurrent reads + MySQL depends on the pread() and pwrite() system calls to be + able to mix INSERT and SELECT. Currently, we use mutexes to + emulate pread() and pwrite(). We intend to replace the file + level interface with a virtual interface in the future so that + we can use the readfile()/writefile() interface to get more + speed. The current implementation limits the number of open + files that MySQL 5.1 can use to 2,048, which means that you + cannot run as many concurrent threads on Windows as on Unix. - If you prefer to remove the anonymous accounts instead, do so as - follows: -shell> mysql -u root -mysql> DROP USER ''; + * Blocking read + MySQL uses a blocking read for each connection. That has the + following implications if named-pipe connections are enabled: - The DROP statement applies both to Windows and to Unix. On - Windows, if you want to remove only the anonymous account that has - the same privileges as root, do this instead: -shell> mysql -u root -mysql> DROP USER ''@'localhost'; + + A connection is not disconnected automatically after + eight hours, as happens with the Unix version of MySQL. - That account allows anonymous access but has full privileges, so - removing it improves security. + + If a connection hangs, it is not possible to break it + without killing MySQL. - root Account Password Assignment + + mysqladmin kill does not work on a sleeping connection. - You can assign passwords to the root accounts in several ways. The - following discussion demonstrates three methods: + + mysqladmin shutdown cannot abort as long as there are + sleeping connections. + We plan to fix this problem in the future. - * Use the SET PASSWORD statement + * ALTER TABLE + While you are executing an ALTER TABLE statement, the table is + locked from being used by other threads. This has to do with + the fact that on Windows, you can't delete a file that is in + use by another thread. In the future, we may find some way to + work around this problem. - * Use the mysqladmin command-line client program + * DATA DIRECTORY and INDEX DIRECTORY + The DATA DIRECTORY and INDEX DIRECTORY options for CREATE + TABLE are ignored on Windows, because Windows doesn't support + symbolic links. These options also are ignored on systems that + have a nonfunctional realpath() call. - * Use the UPDATE statement + * DROP DATABASE + You cannot drop a database that is in use by another thread. - To assign passwords using SET PASSWORD, connect to the server as - root and issue SET PASSWORD statements. Be sure to encrypt the - password using the PASSWORD() function. + * Case-insensitive names + File names are not case sensitive on Windows, so MySQL + database and table names are also not case sensitive on + Windows. The only restriction is that database and table names + must be specified using the same case throughout a given + statement. See Section 8.2.2, "Identifier Case Sensitivity." - For Windows, do this: -shell> mysql -u root -mysql> SET PASSWORD FOR 'root'@'localhost' = PASSWORD('newpwd'); -mysql> SET PASSWORD FOR 'root'@'%' = PASSWORD('newpwd'); + * Directory and file names + On Windows, MySQL Server supports only directory and file + names that are compatible with the current ANSI code pages. + For example, the following Japanese directory name will not + work in the Western locale (code page 1252): +datadir="C:/维基百科关于中文维基百科" + The same limitation applies to directory and file names + referred to in SQL statements, such as the data file path name + in LOAD DATA INFILE. - For Unix, do this: -shell> mysql -u root -mysql> SET PASSWORD FOR 'root'@'localhost' = PASSWORD('newpwd'); -mysql> SET PASSWORD FOR 'root'@'host_name' = PASSWORD('newpwd'); + * The "\" path name separator character + Path name components in Windows are separated by the "\" + character, which is also the escape character in MySQL. If you + are using LOAD DATA INFILE or SELECT ... INTO OUTFILE, use + Unix-style file names with "/" characters: +mysql> LOAD DATA INFILE 'C:/tmp/skr.txt' INTO TABLE skr; +mysql> SELECT * INTO OUTFILE 'C:/tmp/skr.txt' FROM skr; + Alternatively, you must double the "\" character: +mysql> LOAD DATA INFILE 'C:\\tmp\\skr.txt' INTO TABLE skr; +mysql> SELECT * INTO OUTFILE 'C:\\tmp\\skr.txt' FROM skr; - In the second SET PASSWORD statement, replace host_name with the - name of the server host. This is the same host name that you used - when you assigned the anonymous account passwords. + * Problems with pipes + Pipes do not work reliably from the Windows command-line + prompt. If the pipe includes the character ^Z / CHAR(24), + Windows thinks that it has encountered end-of-file and aborts + the program. + This is mainly a problem when you try to apply a binary log as + follows: +C:\> mysqlbinlog binary_log_file | mysql --user=root + If you have a problem applying the log and suspect that it is + because of a ^Z / CHAR(24) character, you can use the + following workaround: +C:\> mysqlbinlog binary_log_file --result-file=/tmp/bin.sql +C:\> mysql --user=root --execute "source /tmp/bin.sql" + The latter command also can be used to reliably read in any + SQL file that may contain binary data. - If the user table contains an account with User and Host values of - 'root' and '127.0.0.1', use an additional SET PASSWORD statement - to set that account's password: -mysql> SET PASSWORD FOR 'root'@'127.0.0.1' = PASSWORD('newpwd'); + * Access denied for user error + If MySQL cannot resolve your host name properly, you may get + the following error when you attempt to run a MySQL client + program to connect to a server running on the same machine: +Access denied for user 'some_user'@'unknown' +to database 'mysql' + To fix this problem, you should create a file named + \windows\hosts containing the following information: +127.0.0.1 localhost - To assign passwords to the root accounts using mysqladmin, execute - the following commands: -shell> mysqladmin -u root password "newpwd" -shell> mysqladmin -u root -h host_name password "newpwd" + Here are some open issues for anyone who might want to help us + improve MySQL on Windows: + + * Add macros to use the faster thread-safe increment/decrement + methods provided by Windows. + +2.5.10. Installing MySQL from Source on Windows - These commands apply both to Windows and to Unix. In the second - command, replace host_name with the name of the server host. The - double quotes around the password are not always necessary, but - you should use them if the password contains spaces or other - characters that are special to your command interpreter. + These instructions describe how to build binaries from source for + MySQL 5.1 on Windows. Instructions are provided for building + binaries from a standard source distribution or from the Bazaar + tree that contains the latest development source. - The mysqladmin method of setting the root account passwords does - not set the password for the 'root'@'127.0.0.1' account. To do so, - use SET PASSWORD as shown earlier. +Note - You can also use UPDATE to modify the user table directly. The - following UPDATE statement assigns a password to all root - accounts: -shell> mysql -u root -mysql> UPDATE mysql.user SET Password = PASSWORD('newpwd') - -> WHERE User = 'root'; -mysql> FLUSH PRIVILEGES; + The instructions here are strictly for users who want to test + MySQL on Microsoft Windows from the latest source distribution or + from the Bazaar tree. For production use, we do not advise using a + MySQL server built by yourself from source. Normally, it is best + to use precompiled binary distributions of MySQL that are built + specifically for optimal performance on Windows by Sun + Microsystems, Inc. Instructions for installing binary + distributions are available in Section 2.5, "Installing MySQL on + Windows." - The UPDATE statement applies both to Windows and to Unix. + To build MySQL on Windows from source, you must satisfy the + following system, compiler, and resource requirements: - After the passwords have been set, you must supply the appropriate - password whenever you connect to the server. For example, if you - want to use mysqladmin to shut down the server, you can do so - using this command: -shell> mysqladmin -u root -p shutdown -Enter password: (enter root password here) + * Windows 2000, Windows XP, or newer version. + Windows Vista is supported when using Visual Studio 2005 + provided you have installed the following updates: -Note + + Microsoft Visual Studio 2005 Professional Edition - ENU + Service Pack 1 (KB926601) + (http://support.microsoft.com/?kbid=926601) - If you forget your root password after setting it up, Section - B.1.4.1, "How to Reset the Root Password," covers the procedure - for resetting it. + + Security Update for Microsoft Visual Studio 2005 + Professional Edition - ENU (KB937061) + (http://support.microsoft.com/?kbid=937061) - To set up additional accounts, you can use the GRANT statement. - For instructions, see Section 5.5.2, "Adding User Accounts." + + Update for Microsoft Visual Studio 2005 Professional + Edition - ENU (KB932232) + (http://support.microsoft.com/?kbid=932232) -2.12. Upgrading or Downgrading MySQL + * CMake, which can be downloaded from http://www.cmake.org. + After installing, modify your path to include the cmake + binary. -2.12.1. Upgrading MySQL + * Microsoft Visual C++ 2005 Express Edition, Visual Studio .Net + 2003 (7.1), or Visual Studio 2005 (8.0) compiler system. - As a general rule, to upgrade from one release series to another, - you should go to the next series rather than skipping a series. To - upgrade from a release series previous to MySQL 5.0, upgrade to - each successive release series in turn until you have reached - MySQL 5.0, and then proceed with the upgrade to MySQL 5.1. For - example, if you currently are running MySQL 4.0 and wish to - upgrade to a newer series, upgrade to MySQL 4.1 first before - upgrading to 5.0, and so forth. For information on upgrading to - MySQL 5.0, see the MySQL 5.0 Reference Manual; for earlier - releases, see the MySQL 3.23, 4.0, 4.1 Reference Manual. + * If you are using Visual C++ 2005 Express Edition, you must + also install an appropriate Platform SDK. More information and + links to downloads for various Windows platforms is available + from + http://www.microsoft.com/downloads/details.aspx?familyid=0baf2 + b35-c656-4969-ace8-e4c0c0716adb. - To upgrade from MySQL 5.0 to 5.1, use the items in the following - checklist as a guide: + * If you are compiling from a Bazaar tree or making changes to + the parser, you need bison for Windows, which can be + downloaded from + http://gnuwin32.sourceforge.net/packages/bison.htm. Download + the package labeled "Complete package, excluding sources". + After installing the package, modify your path to include the + bison binary and ensure that this binary is accessible from + Visual Studio. - * Before any upgrade, back up your databases, including the - mysql database that contains the grant tables. See Section - 6.1, "Database Backups." + * Cygwin might be necessary if you want to run the test script + or package the compiled binaries and support files into a Zip + archive. (Cygwin is needed only to test or package the + distribution, not to build it.) Cygwin is available from + http://cygwin.com. - * Read all the notes in Section 2.12.1.1, "Upgrading from MySQL - 5.0 to 5.1." These notes enable you to identify upgrade issues - that apply to your current MySQL installation. Some - incompatibilities discussed in that section require your - attention before upgrading. Others should be dealt with after - upgrading. + * 3GB to 5GB of disk space. - * Read Appendix C, "MySQL Change History" as well, which - provides information about features that are new in MySQL 5.1 - or differ from those found in MySQL 5.0. + The exact system requirements for Visual Studio can be found here: + http://msdn.microsoft.com/vstudio/Previous/2003/sysreqs/default.as + px and + http://msdn.microsoft.com/vstudio/products/sysreqs/default.aspx - * After you upgrade to a new version of MySQL, run mysql_upgrade - (see Section 4.4.8, "mysql_upgrade --- Check Tables for MySQL - Upgrade"). This program checks your tables, and attempts to - repair them if necessary. It also updates your grant tables to - make sure that they have the current structure so that you can - take advantage of any new capabilities. (Some releases of - MySQL introduce changes to the structure of the grant tables - to add new privileges or features.) + You also need a MySQL source distribution for Windows, which can + be obtained two ways: - * If you are running MySQL Server on Windows, see Section - 2.3.14, "Upgrading MySQL on Windows." + * Obtain a source distribution packaged by Sun Microsystems, + Inc. These are available from http://dev.mysql.com/downloads/. - * If you are using replication, see Section 16.3.3, "Upgrading a - Replication Setup," for information on upgrading your - replication setup. + * Package a source distribution yourself from the latest Bazaar + developer source tree. For instructions on pulling the latest + source files, see Section 2.3.3, "Installing from the + Development Source Tree." - * If you are upgrading an installation originally produced by - installing multiple RPM packages, it is best to upgrade all - the packages, not just some. For example, if you previously - installed the server and client RPMs, do not upgrade just the - server RPM. + If you find something not working as expected, or you have + suggestions about ways to improve the current build process on + Windows, please send a message to the win32 mailing list. See + Section 1.5.1, "MySQL Mailing Lists." - * As of MySQL 5.1.9, the mysqld-max server is included in binary - distributions. There is no separate MySQL-Max distribution. As - of MySQL 5.1.12, there is no mysqld-max server at all in - binary distributions. They contain a server that includes the - features previously included in mysqld-max. +2.5.10.1. Building MySQL from Source Using CMake and Visual Studio - * If you have created a user-defined function (UDF) with a given - name and upgrade MySQL to a version that implements a new - built-in function with the same name, the UDF becomes - inaccessible. To correct this, use DROP FUNCTION to drop the - UDF, and then use CREATE FUNCTION to re-create the UDF with a - different nonconflicting name. The same is true if the new - version of MySQL implements a built-in function with the same - name as an existing stored function. See Section 8.2.4, - "Function Name Parsing and Resolution," for the rules - describing how the server interprets references to different - kinds of functions. + You can build MySQL on Windows by using a combination of cmake and + Microsoft Visual Studio .NET 2003 (7.1), Microsoft Visual Studio + 2005 (8.0) or Microsoft Visual C++ 2005 Express Edition. You must + have the appropriate Microsoft Platform SDK installed. - You can always move the MySQL format files and data files between - different versions on systems with the same architecture as long - as you stay within versions for the same release series of MySQL. +Note - If you are cautious about using new versions, you can always - rename your old mysqld before installing a newer one. For example, - if you are using MySQL 5.0.13 and want to upgrade to 5.1.10, - rename your current server from mysqld to mysqld-5.0.13. If your - new mysqld then does something unexpected, you can simply shut it - down and restart with your old mysqld. + To compile from the source code on Windows you must use the + standard source distribution (for example, mysql-5.1.41.tar.gz). + You build from the same distribution as used to build MySQL on + Unix, Linux and other platforms. Do not use the Windows Source + distributions as they do not contain the necessary configuration + script and other files. - If, after an upgrade, you experience problems with recompiled - client programs, such as Commands out of sync or unexpected core - dumps, you probably have used old header or library files when - compiling your programs. In this case, you should check the date - for your mysql.h file and libmysqlclient.a library to verify that - they are from the new MySQL distribution. If not, recompile your - programs with the new headers and libraries. + Follow this procedure to build MySQL: - If problems occur, such as that the new mysqld server does not - start or that you cannot connect without a password, verify that - you do not have an old my.cnf file from your previous - installation. You can check this with the --print-defaults option - (for example, mysqld --print-defaults). If this command displays - anything other than the program name, you have an active my.cnf - file that affects server or client operation. + 1. If you are installing from a packaged source distribution, + create a work directory (for example, C:\workdir), and unpack + the source distribution there using WinZip or another Windows + tool that can read .zip files. This directory is the work + directory in the following instructions. - If your MySQL installation contains a large amount of data that - might take a long time to convert after an in-place upgrade, you - might find it useful to create a "dummy" database instance for - assessing what conversions might be needed and the work involved - to perform them. Make a copy of your MySQL instance that contains - a full copy of the mysql database, plus all other databases - without data. Run your upgrade procedure on this dummy instance to - see what actions might be needed so that you can better evaluate - the work involved when performing actual data conversion on your - original database instance. + 2. Using a command shell, navigate to the work directory and run + the following command: +C:\workdir>win\configure.js options + If you have associated the .js file extension with an + application such as a text editor, then you may need to use + the following command to force configure.js to be executed as + a script: +C:\workdir>cscript win\configure.js options + These options are available for configure.js: - It is a good idea to rebuild and reinstall the Perl DBD::mysql - module whenever you install a new release of MySQL. The same - applies to other MySQL interfaces as well, such as PHP mysql - extensions and the Python MySQLdb module. + + WITH_INNOBASE_STORAGE_ENGINE: Enable the InnoDB storage + engine. -2.12.1.1. Upgrading from MySQL 5.0 to 5.1 + + WITH_PARTITION_STORAGE_ENGINE: Enable user-defined + partitioning. - After upgrading a 5.0 installation to 5.0.10 or above, it is - necessary to upgrade your grant tables. Otherwise, creating stored - procedures and functions might not work. To perform this upgrade, - run mysql_upgrade. + + WITH_ARCHIVE_STORAGE_ENGINE: Enable the ARCHIVE storage + engine. -Note + + WITH_BLACKHOLE_STORAGE_ENGINE: Enable the BLACKHOLE + storage engine. - It is good practice to back up your data before installing any new - version of software. Although MySQL works very hard to ensure a - high level of quality, you should protect your data by making a - backup. + + WITH_EXAMPLE_STORAGE_ENGINE: Enable the EXAMPLE storage + engine. - To upgrade to 5.1 from any previous version, MySQL recommends that - you dump your tables with mysqldump before upgrading and reload - the dump file after upgrading. + + WITH_FEDERATED_STORAGE_ENGINE: Enable the FEDERATED + storage engine. - In general, you should do the following when upgrading from MySQL - 5.0 to 5.1: + + WITH_NDBCLUSTER_STORAGE_ENGINE (experimental): Enable the + NDBCLUSTER storage engine in the MySQL server; cause + binaries for the MySQL Cluster management and data node, + management client, and other programs to be built. + This option is supported only in MySQL Cluster NDB 7.0 + (NDBCLUSTER storage engine versions 6.4.0 and later) + using the MySQL Cluster sources. It cannot be used to + enable clustering support in other MySQL source trees or + distributions. - * Read all the items in the following sections to see whether - any of them might affect your applications: + + MYSQL_SERVER_SUFFIX=suffix: Server suffix, default none. - + Section 2.12.1, "Upgrading MySQL," has general update - information. + + COMPILATION_COMMENT=comment: Server comment, default + "Source distribution". - + The items in the change lists found later in this section - enable you to identify upgrade issues that apply to your - current MySQL installation. + + MYSQL_TCP_PORT=port: Server port, default 3306. + + + DISABLE_GRANT_OPTIONS: Disables the --bootstrap, + --skip-grant-tables, and --init-file options for mysqld. + This option is available as of MySQL 5.1.15. + For example (type the command on one line): +C:\workdir>win\configure.js WITH_INNOBASE_STORAGE_ENGINE + WITH_PARTITION_STORAGE_ENGINE MYSQL_SERVER_SUFFIX=-pro - + The MySQL 5.1 change history describes significant new - features you can use in 5.1 or that differ from those - found in MySQL 5.0. Some of these changes may result in - incompatibilities. See Section C.1, "Changes in Release - 5.1.x (Production)." + 3. From the work directory, execute the win\build-vs8.bat or + win\build-vs71.bat file, depending on the version of Visual + Studio you have installed. The script invokes CMake, which + generates the mysql.sln solution file. + You can also use win\build-vs8_x64.bat to build the 64-bit + version of MySQL. However, you cannot build the 64-bit version + with Visual Studio Express Edition. You must use Visual Studio + 2005 (8.0) or higher. - * Note particularly any changes that are marked Known issue or - Incompatible change. These incompatibilities with earlier - versions of MySQL may require your attention before you - upgrade. - Our aim is to avoid these changes, but occasionally they are - necessary to correct problems that would be worse than an - incompatibility between releases. If any upgrade issue - applicable to your installation involves an incompatibility - that requires special handling, follow the instructions given - in the incompatibility description. Often this will involve a - dump and reload, or use of a statement such as CHECK TABLE or - REPAIR TABLE. - For dump and reload instructions, see Section 2.12.4, - "Rebuilding or Repairing Tables or Indexes." Any procedure - that involves REPAIR TABLE with the USE_FRM option must be - done before upgrading. Use of this statement with a version of - MySQL different from the one used to create the table (that - is, using it after upgrading) may damage the table. See - Section 12.5.2.6, "REPAIR TABLE Syntax." + 4. From the work directory, open the generated mysql.sln file + with Visual Studio and select the proper configuration using + the Configuration menu. The menu provides Debug, Release, + RelwithDebInfo, MinRelInfo options. Then select Solution > + Build to build the solution. + Remember the configuration that you use in this step. It is + important later when you run the test script because that + script needs to know which configuration you used. - * After you upgrade to a new version of MySQL, run mysql_upgrade - (see Section 4.4.8, "mysql_upgrade --- Check Tables for MySQL - Upgrade"). This program checks your tables, and attempts to - repair them if necessary. It also updates your grant tables to - make sure that they have the current structure so that you can - take advantage of any new capabilities. (Some releases of - MySQL introduce changes to the structure of the grant tables - to add new privileges or features.) + 5. Test the server. The server built using the preceding + instructions expects that the MySQL base directory and data + directory are C:\mysql and C:\mysql\data by default. If you + want to test your server using the source tree root directory + and its data directory as the base directory and data + directory, you need to tell the server their path names. You + can either do this on the command line with the --basedir and + --datadir options, or by placing appropriate options in an + option file. (See Section 4.2.3.3, "Using Option Files.") If + you have an existing data directory elsewhere that you want to + use, you can specify its path name instead. + When the server is running in standalone fashion or as a + service based on your configuration, try to connect to it from + the mysql interactive command-line utility. + You can also run the standard test script, mysql-test-run.pl. + This script is written in Perl, so you'll need either Cygwin + or ActiveState Perl to run it. You may also need to install + the modules required by the script. To run the test script, + change location into the mysql-test directory under the work + directory, set the MTR_VS_CONFIG environment variable to the + configuration you selected earlier (or use the --vs-config + option), and invoke mysql-test-run.pl. For example (using + Cygwin and the bash shell): +shell> cd mysql-test +shell> export MTR_VS_CONFIG=debug +shell> ./mysql-test-run.pl --force --timer +shell> ./mysql-test-run.pl --force --timer --ps-protocol - * Check Section 2.12.3, "Checking Whether Table Indexes Must Be - Rebuilt," to see whether changes to character sets or - collations were made that affect your table indexes. If so, - you will need to rebuild the affected indexes using the - instructions in Section 2.12.4, "Rebuilding or Repairing - Tables or Indexes." + When you are satisfied that the programs you have built are + working correctly, stop the server. Now you can install the + distribution. One way to do this is to use the make_win_bin_dist + script in the scripts directory of the MySQL source distribution + (see Section 4.4.2, "make_win_bin_dist --- Package MySQL + Distribution as ZIP Archive"). This is a shell script, so you must + have Cygwin installed if you want to use it. It creates a Zip + archive of the built executables and support files that you can + unpack in the location at which you want to install MySQL. - * If you are running MySQL Server on Windows, see Section - 2.3.14, "Upgrading MySQL on Windows." + It is also possible to install MySQL by copying directories and + files directly: - * If you are using replication, see Section 16.3.3, "Upgrading a - Replication Setup," for information on upgrading your - replication setup. + 1. Create the directories where you want to install MySQL. For + example, to install into C:\mysql, use these commands: +C:\> mkdir C:\mysql +C:\> mkdir C:\mysql\bin +C:\> mkdir C:\mysql\data +C:\> mkdir C:\mysql\share +C:\> mkdir C:\mysql\scripts + If you want to compile other clients and link them to MySQL, + you should also create several additional directories: +C:\> mkdir C:\mysql\include +C:\> mkdir C:\mysql\lib +C:\> mkdir C:\mysql\lib\debug +C:\> mkdir C:\mysql\lib\opt + If you want to benchmark MySQL, create this directory: +C:\> mkdir C:\mysql\sql-bench + Benchmarking requires Perl support. See Section 2.15, "Perl + Installation Notes." - If your MySQL installation contains a large amount of data that - might take a long time to convert after an in-place upgrade, you - might find it useful to create a "dummy" database instance for - assessing what conversions might be needed and the work involved - to perform them. Make a copy of your MySQL instance that contains - a full copy of the mysql database, plus all other databases - without data. Run your upgrade procedure on this dummy instance to - see what actions might be needed so that you can better evaluate - the work involved when performing actual data conversion on your - original database instance. + 2. From the work directory, copy into the C:\mysql directory the + following directories: +C:\> cd \workdir +C:\workdir> copy client_release\*.exe C:\mysql\bin +C:\workdir> copy client_debug\mysqld.exe C:\mysql\bin\mysqld-debug.ex +e +C:\workdir> xcopy scripts\*.* C:\mysql\scripts /E +C:\workdir> xcopy share\*.* C:\mysql\share /E + If you want to compile other clients and link them to MySQL, + you should also copy several libraries and header files: +C:\workdir> copy lib_debug\mysqlclient.lib C:\mysql\lib\debug +C:\workdir> copy lib_debug\libmysql.* C:\mysql\lib\debug +C:\workdir> copy lib_debug\zlib.* C:\mysql\lib\debug +C:\workdir> copy lib_release\mysqlclient.lib C:\mysql\lib\opt +C:\workdir> copy lib_release\libmysql.* C:\mysql\lib\opt +C:\workdir> copy lib_release\zlib.* C:\mysql\lib\opt +C:\workdir> copy include\*.h C:\mysql\include +C:\workdir> copy libmysql\libmysql.def C:\mysql\include + If you want to benchmark MySQL, you should also do this: +C:\workdir> xcopy sql-bench\*.* C:\mysql\bench /E - MySQL Enterprise MySQL Enterprise subscribers will find more - information about upgrading in the Knowledge Base articles found - at Upgrading - (https://kb.mysql.com/search.php?cat=search&category=41) Access - to the MySQL Knowledge Base collection of articles is one of the - advantages of subscribing to MySQL Enterprise. For more - information, see - http://www.mysql.com/products/enterprise/advisors.html. + After installation, set up and start the server in the same way as + for binary Windows distributions. See Section 2.5, "Installing + MySQL on Windows." - The following lists describe changes that may affect applications - and that you should watch out for when upgrading to MySQL 5.1. +2.5.11. Compiling MySQL Clients on Windows - Configuration Changes: + In your source files, you should include my_global.h before + mysql.h: +#include <my_global.h> +#include <mysql.h> - * Before MySQL 5.1.11, to build MySQL from source with SSL - support enabled, you would invoke configure with either the - --with-openssl or --with-yassl option. In MySQL 5.1.11, those - options both have been replaced by the --with-ssl option. By - default, --with-ssl causes the bundled yaSSL library to be - used. To select OpenSSL instead, give the option as - --with-ssl=path, where path is the directory where the OpenSSL - header files and libraries are located. + my_global.h includes any other files needed for Windows + compatibility (such as windows.h) if you compile your program on + Windows. - Server Changes: + You can either link your code with the dynamic libmysql.lib + library, which is just a wrapper to load in libmysql.dll on + demand, or link with the static mysqlclient.lib library. - * Known issue: Dumps performed by using mysqldump to generate a - dump file before the upgrade and reloading the file after - upgrading are subject to the following problem: - Before MySQL 5.0.40, mysqldump displays SPATIAL index - definitions using prefix lengths for the indexed columns. - These prefix lengths are accepted in MySQL 5.0, but not as of - MySQL 5.1. If you use mysqldump from versions of MySQL older - than 5.0.40, any table containing SPATIAL indexes will cause - an error when the dump file is reloaded into MySQL 5.1 or - higher. - For example, a table definition might look like this when - dumped in MySQL 5.0: -CREATE TABLE `t` ( - `g` geometry NOT NULL, - SPATIAL KEY `g` (`g`(32)) -) ENGINE=MyISAM DEFAULT CHARSET=latin1 - The SPATIAL index definition will not be accepted in MySQL - 5.1. To work around this, edit the dump file to remove the - prefix: -CREATE TABLE `t` ( - `g` geometry NOT NULL, - SPATIAL KEY `g` (`g`) -) ENGINE=MyISAM DEFAULT CHARSET=latin1 - Dump files can be large, so it may be preferable to dump table - definitions and data separately to make it easier to edit the - definitions: -shell> mysqldump --no-data other_args > definitions.sql -shell> mysqldump --no-create-info other_args > data.sql - Then edit definitions.sql before reloading definitions.sql and - data.sql, in that order. - If you upgrade to a version of MySQL 5.0 higher than 5.0.40 - before upgrading to MySQL 5.1, this problem does not occur. + The MySQL client libraries are compiled as threaded libraries, so + you should also compile your code to be multi-threaded. - * Known issue: Before MySQL 5.1.30, the CHECK TABLE ... FOR - UPGRADE statement did not check for incompatible collation - changes made in MySQL 5.1.24. (This also affects mysqlcheck - and mysql_upgrade, which cause that statement to be executed.) - Prior to the fix made in 5.1.30, a binary upgrade (performed - without dumping tables with mysqldump before the upgrade and - reloading the dump file after the upgrade) would corrupt - tables. After the fix, CHECK TABLE ... FOR UPGRADE properly - detects the problem and warns about tables that need repair. - However, the fix is not backward compatible and can result in - a downgrading problem under these circumstances: +2.6. Installing MySQL on Linux - 1. Perform a binary upgrade to a version of MySQL that - includes the fix. + The following sections covers the installation of Linux using + RPMs. For information on using a generic binary package using tar, + see Section 2.2, "Installing MySQL from Generic Binaries on + Unix/Linux." For information on installing from source, see + Section 2.3, "MySQL Installation Using a Source Distribution." - 2. Run CHECK TABLE ... FOR UPGRADE (or mysqlcheck or - mysql_upgrade) to upgrade tables. + mysql.server can be found in the support-files directory under the + MySQL installation directory or in a MySQL source tree. You can + install it as /etc/init.d/mysql for automatic MySQL startup and + shutdown. See Section 2.13.1.2, "Starting and Stopping MySQL + Automatically." - 3. Perform a binary downgrade to a version of MySQL that - does not include the fix. - The solution is to dump tables with mysqldump before the - downgrade and reload the dump file after the downgrade. - Alternatively, drop and recreate affected indexes. +2.6.1. Installing MySQL from RPM Packages on Linux - * Known issue: MySQL introduces encoding for table names that - have non-ASCII characters (see Section 8.2.3, "Mapping of - Identifiers to File Names"). After a binary upgrade from MySQL - 5.0 to 5.1 or higher, the server recognizes names that have - non-ASCII characters and adds a #mysql50# prefix to them. - As of MySQL 5.1.31, mysql_upgrade encodes these names by - executing the following command: -mysqlcheck --all-databases --check-upgrade --fix-db-names --fix-table --names - Prior to MySQL 5.1.31, mysql_upgrade does not execute this - command, so you should execute it manually if you have - database or table names that contain nonalphanumeric - characters. - Prior to MySQL 5.1.23, the mysqlcheck command does not perform - the name encoding for views. To work around this problem, drop - each affected view and recreate it. - mysqlcheck cannot fix names that contain literal instances of - the @ character that is used for encoding special characters. - If you have databases or tables that contain this character, - use mysqldump to dump them before upgrading to MySQL 5.1, and - then reload the dump file after upgrading. + The recommended way to install MySQL on RPM-based Linux + distributions is by using the RPM packages. The RPMs that we + provide to the community should work on all versions of Linux that + support RPM packages and use glibc 2.3. To obtain RPM packages, + see Section 2.1.3, "How to Get MySQL." - * Known issue: When upgrading from MySQL 5.0 to versions of 5.1 - prior to 5.1.23, running mysqlcheck (or mysql_upgrade, which - runs mysqlcheck) to upgrade tables fails for names that must - be written as quoted identifiers. To work around this problem, - rename each affected table to a name that does not require - quoting: -RENAME TABLE `tab``le_a` TO table_a; -RENAME TABLE `table b` TO table_b; - After renaming the tables, run the mysql_upgrade program. Then - rename the tables back to their original names: -RENAME TABLE table_a TO `tab``le_a`; -RENAME TABLE table_b TO `table b`; + For non-RPM Linux distributions, you can install MySQL using a + .tar.gz package. See Section 2.2, "Installing MySQL from Generic + Binaries on Unix/Linux." - * Known issue: In connection with view creation, the server - created arc directories inside database directories and - maintained useless copies of .frm files there. Creation and - renaming procedures of those copies as well as creation of arc - directories has been discontinued in MySQL 5.1.29. - This change does cause a problem when downgrading to older - server versions which manifests itself under these - circumstances: + We do provide some platform-specific RPMs; the difference between + a platform-specific RPM and a generic RPM is that a + platform-specific RPM is built on the targeted platform and is + linked dynamically whereas a generic RPM is linked statically with + LinuxThreads. - 1. Create a view v_orig in MySQL 5.1.29 or higher. +Note - 2. Rename the view to v_new and then back to v_orig. + RPM distributions of MySQL often are provided by other vendors. Be + aware that they may differ in features and capabilities from those + built by us, and that the instructions in this manual do not + necessarily apply to installing them. The vendor's instructions + should be consulted instead. - 3. Downgrade to an older 5.1.x server and run mysql_upgrade. + In most cases, you need to install only the MySQL-server and + MySQL-client packages to get a functional MySQL installation. The + other packages are not required for a standard installation. - 4. Try to rename v_orig to v_new again. This operation - fails. - As a workaround to avoid this problem, use either of these - approaches: + RPMs for MySQL Cluster. Beginning with MySQL 5.1.24, standard + MySQL server RPMs built by MySQL no longer provide support for the + NDBCLUSTER storage engine. MySQL Cluster users wanting to upgrade + MySQL 5.1.23 or earlier installations from RPMs built by MySQL + should upgrade to MySQL Cluster NDB 6.2 or MySQL Cluster NDB 6.3; + RPMs that should work with most Linux distributions are available + for both of these release series. - + Dump your data using mysqldump before downgrading and - reload the dump file after downgrading. +Important - + Instead of renaming a view after the downgrade, drop it - and recreate it. + When upgrading a MySQL Cluster RPM installation, you must upgrade + all installed RPMs, including the Server and Client RPMs. - * Incompatible change: Character set or collation changes were - made in MySQL 5.1.21, 5.1.23, and 5.1.24 that may require - table indexes to be rebuilt. For details, see Section 2.12.3, - "Checking Whether Table Indexes Must Be Rebuilt." + For more information about installing MySQL Cluster from RPMs, see + Section 17.2.1, "MySQL Cluster Multi-Computer Installation." - * Incompatible change: In MySQL 5.1.36, options for loading - plugins such as pluggable storage engines were changed from - boolean to tristate format. The implementations overlap, but - if you previously used options of the form --plugin_name=0 or - --plugin_name=1, you should instead use --plugin_name=OFF or - --plugin_name=ON, respectively. For details, see Section - 5.1.3, "Server Options for Loading Plugins." + For upgrades, if your installation was originally produced by + installing multiple RPM packages, it is best to upgrade all the + packages, not just some. For example, if you previously installed + the server and client RPMs, do not upgrade just the server RPM. - * Incompatible change: From MySQL 5.1.24 to 5.1.31, the UPDATE - statement was changed such that assigning NULL to a NOT NULL - column caused an error even when strict SQL mode was not - enabled. The original behavior before MySQL 5.1.24 was that - such assignments caused an error only in strict SQL mode, and - otherwise set the column to the implicit default value for the - column data type and generated a warning. (For information - about implicit default values, see Section 10.1.4, "Data Type - Default Values.") - The change caused compatibility problems for applications that - relied on the original behavior. It also caused replication - problems between servers that had the original behavior and - those that did not, for applications that assigned NULL to NOT - NULL columns in UPDATE statements without strict SQL mode - enabled. The change was reverted in MySQL 5.1.32 so that - UPDATE again had the original behavior. Problems can still - occur if you replicate between servers that have the modified - UPDATE behavior and those that do not. + The RPM packages shown in the following list are available. The + names shown here use a suffix of .glibc23.i386.rpm, but particular + packages can have different suffixes, described later. - * Incompatible change: As of MySQL 5.1.29, the default binary - logging mode has been changed from MIXED to STATEMENT for - compatibility with MySQL 5.0. + * MySQL-server-VERSION.glibc23.i386.rpm + The MySQL server. You need this unless you only want to + connect to a MySQL server running on another machine. - * Incompatible change: In MySQL 5.1.25, a change was made to the - way that the server handles prepared statements. This affects - prepared statements processed at the SQL level (using the - PREPARE statement) and those processed using the binary - client-server protocol (using the mysql_stmt_prepare() C API - function). - Previously, changes to metadata of tables or views referred to - in a prepared statement could cause a server crash when the - statement was next executed, or perhaps an error at execute - time with a crash occurring later. For example, this could - happen after dropping a table and recreating it with a - different definition. - Now metadata changes to tables or views referred to by - prepared statements are detected and cause automatic - repreparation of the statement when it is next executed. - Metadata changes occur for DDL statements such as those that - create, drop, alter, rename, or truncate tables, or that - analyze, optimize, or repair tables. Repreparation also occurs - after referenced tables or views are flushed from the table - definition cache, either implicitly to make room for new - entries in the cache, or explicitly due to FLUSH TABLES. - Repreparation is automatic, but to the extent that it occurs, - performance of prepared statements is diminished. - Table content changes (for example, with INSERT or UPDATE) do - not cause repreparation, nor do SELECT statements. - An incompatibility with previous versions of MySQL is that a - prepared statement may now return a different set of columns - or different column types from one execution to the next. For - example, if the prepared statement is SELECT * FROM t1, - altering t1 to contain a different number of columns causes - the next execution to return a number of columns different - from the previous execution. - Older versions of the client library cannot handle this change - in behavior. For applications that use prepared statements - with the new server, an upgrade to the new client library is - strongly recommended. - Along with this change to statement repreparation, the default - value of the table_definition_cache system variable has been - increased from 128 to 256. The purpose of this increase is to - lessen the chance that prepared statements will need - repreparation due to referred-to tables/views having been - flushed from the cache to make room for new entries. - A new status variable, Com_stmt_reprepare, has been introduced - to track the number of repreparations. + * MySQL-client-VERSION.glibc23.i386.rpm + The standard MySQL client programs. You probably always want + to install this package. - * Incompatible change: As of MySQL 5.1.23, within a stored - routine, it is no longer allowable to declare a cursor for a - SHOW or DESCRIBE statement. This happened to work in some - instances, but is no longer supported. In many cases, a - workaround for this change is to use the cursor with a SELECT - query to read from an INFORMATION_SCHEMA table that produces - the same information as the SHOW statement. + * MySQL-devel-VERSION.glibc23.i386.rpm + The libraries and include files that are needed if you want to + compile other MySQL clients, such as the Perl modules. - * Incompatible change: SHOW CREATE VIEW displays view - definitions using an AS alias_name clause for each column. If - a column is created from an expression, the default alias is - the expression text, which can be quite long. As of MySQL - 5.1.23, aliases for column names in CREATE VIEW statements are - checked against the maximum column length of 64 characters - (not the maximum alias length of 256 characters). As a result, - views created from the output of SHOW CREATE VIEW fail if any - column alias exceeds 64 characters. This can cause problems - for replication or loading dump files. For additional - information and workarounds, see Section D.4, "Restrictions on - Views." + * MySQL-debuginfo-VERSION.glibc23.i386.rpm + This package contains debugging information. debuginfo RPMs + are never needed to use MySQL software; this is true both for + the server and for client programs. However, they contain + additional information that might be needed by a debugger to + analyze a crash. - * Incompatible change: MySQL 5.1 implements support for a plugin - API that allows the loading and unloading of components at - runtime, without restarting the server. Section 22.2, "The - MySQL Plugin Interface." The plugin API requires the - mysql.plugin table. After upgrading from an older version of - MySQL, you should run the mysql_upgrade command to create this - table. See Section 4.4.8, "mysql_upgrade --- Check Tables for - MySQL Upgrade." - Plugins are installed in the directory named by the plugin_dir - system variable. This variable also controls the location from - which the server loads user-defined functions (UDFs), which is - a change from earlier versions of MySQL. That is, all UDF - library files now must be installed in the plugin directory. - When upgrading from an older version of MySQL, you must - migrate your UDF files to the plugin directory. + * MySQL-shared-VERSION.glibc23.i386.rpm + This package contains the shared libraries + (libmysqlclient.so*) that certain languages and applications + need to dynamically load and use MySQL. It contains + single-threaded and thread-safe libraries. If you install this + package, do not install the MySQL-shared-compat package. - * Incompatible change: The table_cache system variable has been - renamed to table_open_cache. Any scripts that refer to - table_cache must be updated to use the new name. + * MySQL-shared-compat-VERSION.glibc23.i386.rpm + This package includes the shared libraries for MySQL 3.23, + 4.0, and so on, up to the current release. It contains + single-threaded and thread-safe libraries. Install this + package instead of MySQL-shared if you have applications + installed that are dynamically linked against older versions + of MySQL but you want to upgrade to the current version + without breaking the library dependencies. - * Incompatible change: Several issues were identified for stored - programs (stored procedures and functions, triggers, and - events) and views containing non-ASCII symbols. These issues - involved conversion errors due to incomplete character set - information when translating these objects to and from stored - format. - To address these problems, the representation for these - objects was changed in MySQL 5.1.21. However, the fixes affect - all stored programs and views. (For example, you will see - warnings about "no creation context.") To avoid warnings from - the server about the use of old definitions from any release - prior to 5.1.21, you should dump stored programs and views - with mysqldump after upgrading to 5.1.21 or higher, and then - reload them to recreate them with new definitions. Invoke - mysqldump with a --default-character-set option that names the - non-ASCII character set that was used for the definitions when - the objects were originally defined. + * MySQL-shared-compat-advanced-gpl-VERSION.glibc23.i386.rpm, + MySQL-shared-compat-advanced-VERSION.glibc23.i386.rpm + These are like the MySQL-shared-compat package, but are for + the "MySQL Enterprise Server - Advanced Edition" products. + Install these packages rather than the normal + MySQL-shared-compat package if you want to included shared + client libraries for older MySQL versions. - * Incompatible change: As of MySQL 5.1.20, mysqld_safe supports - error logging to syslog on systems that support the logger - command. The new --syslog and --skip-syslog options can be - used instead of the --log-error option to control logging - behavior, as described in Section 4.3.2, "mysqld_safe --- - MySQL Server Startup Script." - In 5.1.21 and up, the default is --skip-syslog, which is - compatible with the default behavior of writing an error log - file for releases prior to 5.1.20. - In 5.1.20 only, the following conditions apply: 1) The default - is to use syslog, which is not compatible with releases prior - to 5.1.20. 2) Logging to syslog may fail to operate correctly - in some cases. For these reasons, avoid using MySQL 5.1.20. + * MySQL-embedded-VERSION.glibc23.i386.rpm + The embedded MySQL server library. - * Incompatible change: As of MySQL 5.1.15, InnoDB rolls back - only the last statement on a transaction timeout. A new - option, --innodb_rollback_on_timeout, causes InnoDB to abort - and roll back the entire transaction if a transaction timeout - occurs (the same behavior as in MySQL 4.1). + * MySQL-ndb-management-VERSION.glibc23.i386.rpm, + MySQL-ndb-storage-VERSION.glibc23.i386.rpm, + MySQL-ndb-tools-VERSION.glibc23.i386.rpm, + MySQL-ndb-extra-VERSION.glibc23.i386.rpm + Packages that contain additional files for MySQL Cluster + installations. + +Note + The MySQL-ndb-tools RPM requires a working installation of + perl. Prior to MySQL 5.1.18, the DBI and HTML::Template + packages were also required. See Section 2.15, "Perl + Installation Notes," and Section 17.4.21, "ndb_size.pl --- + NDBCLUSTER Size Requirement Estimator," for more information. + + * MySQL-test-VERSION.glibc23.i386.rpm + This package includes the MySQL test suite. + + * MySQL-VERSION.src.rpm + This contains the source code for all of the previous + packages. It can also be used to rebuild the RPMs on other + architectures (for example, Alpha or SPARC). - * Incompatible change: As of MySQL 5.1.15, the following - conditions apply to enabling the read_only system variable: + The suffix of RPM package names (following the VERSION value) has + the following syntax: +.PLATFORM.CPU.rpm - + If you attempt to enable read_only while you have any - explicit locks (acquired with LOCK TABLES or have a - pending transaction, an error will occur. + The PLATFORM and CPU values indicate the type of system for which + the package is built. PLATFORM indicates the platform and CPU + indicates the processor type or family. - + If other clients hold explicit table locks or have - pending transactions, the attempt to enable read_only - blocks until the locks are released and the transactions - end. While the attempt to enable read_only is pending, - requests by other clients for table locks or to begin - transactions also block until read_only has been set. + All packages are dynamically linked against glibc 2.3. The + PLATFORM value indicates whether the package is platform + independent or intended for a specific platform, as shown in the + following table. + glibc23 Platform independent, should run on any Linux distribution + that supports glibc 2.3 + rhel3, rhel4 Red Hat Enterprise Linux 3 or 4 + sles9, sles10 SuSE Linux Enterprise Server 9 or 10 - + read_only can be enabled while you hold a global read - lock (acquired with FLUSH TABLES WITH READ LOCK) because - that does not involve table locks. - Previously, the attempt to enable read_only would return - immediately even if explicit locks or transactions were - pending, so some data changes could occur for statements - executing in the server at the same time. + In MySQL 5.1, only glibc23 packages are available currently. - * Incompatible change: The number of function names affected by - IGNORE_SPACE was reduced significantly in MySQL 5.1.13, from - about 200 to about 30. (For details about IGNORE_SPACE, see - Section 8.2.4, "Function Name Parsing and Resolution.") This - change improves the consistency of parser operation. However, - it also introduces the possibility of incompatibility for old - SQL code that relies on the following conditions: + The CPU value indicates the processor type or family for which the + package is built. + i386 x86 processor, 386 and up + i586 x86 processor, Pentium and up + x86_64 64-bit x86 processor + ia64 Itanium (IA-64) processor - + IGNORE_SPACE is disabled. + To see all files in an RPM package (for example, a MySQL-server + RPM), run a command like this: +shell> rpm -qpl MySQL-server-VERSION.glibc23.i386.rpm - + The presence or absence of whitespace following a - function name is used to distinguish between a built-in - function and stored function that have the same name (for - example, PI() versus PI ()). - For functions that are no longer affected by IGNORE_SPACE as - of MySQL 5.1.13, that strategy no longer works. Either of the - following approaches can be used if you have code that is - subject to the preceding incompatibility: + To perform a standard minimal installation, install the server and + client RPMs: +shell> rpm -i MySQL-server-VERSION.glibc23.i386.rpm +shell> rpm -i MySQL-client-VERSION.glibc23.i386.rpm - + If a stored function has a name that conflicts with a - built-in function, refer to the stored function with a - schema name qualifier, regardless of whether whitespace - is present. For example, write schema_name.PI() or - schema_name.PI (). + To install only the client programs, install just the client RPM: +shell> rpm -i MySQL-client-VERSION.glibc23.i386.rpm - + Alternatively, rename the stored function to use a - nonconflicting name and change invocations of the - function to use the new name. + RPM provides a feature to verify the integrity and authenticity of + packages before installing them. If you would like to learn more + about this feature, see Section 2.1.4, "Verifying Package + Integrity Using MD5 Checksums or GnuPG." - * Incompatible change: For utf8 columns, the full-text parser - incorrectly considered several nonword punctuation and - whitespace characters as word characters, causing some - searches to return incorrect results. The fix involves a - change to the full-text parser in MySQL 5.1.12, so as of - 5.1.12, any tables that have FULLTEXT indexes on utf8 columns - must be repaired with REPAIR TABLE: -REPAIR TABLE tbl_name QUICK; + The server RPM places data under the /var/lib/mysql directory. The + RPM also creates a login account for a user named mysql (if one + does not exist) to use for running the MySQL server, and creates + the appropriate entries in /etc/init.d/ to start the server + automatically at boot time. (This means that if you have performed + a previous installation and have made changes to its startup + script, you may want to make a copy of the script so that you + don't lose it when you install a newer RPM.) See Section 2.13.1.2, + "Starting and Stopping MySQL Automatically," for more information + on how MySQL can be started automatically on system startup. - * Incompatible change: Storage engines can be pluggable at - runtime, so the distinction between disabled and invalid - storage engines no longer applies. As of MySQL 5.1.12, this - affects the NO_ENGINE_SUBSTITUTION SQL mode, as described in - Section 5.1.8, "Server SQL Modes." + If you want to install the MySQL RPM on older Linux distributions + that do not support initialization scripts in /etc/init.d + (directly or via a symlink), you should create a symbolic link + that points to the location where your initialization scripts + actually are installed. For example, if that location is + /etc/rc.d/init.d, use these commands before installing the RPM to + create /etc/init.d as a symbolic link that points there: +shell> cd /etc +shell> ln -s rc.d/init.d . - * Incompatible change: The structure of FULLTEXT indexes has - been changed in MySQL 5.1.6. After upgrading to MySQL 5.1.6 or - greater, any tables that have FULLTEXT indexes must be - repaired with REPAIR TABLE: -REPAIR TABLE tbl_name QUICK; + However, all current major Linux distributions should support the + new directory layout that uses /etc/init.d, because it is required + for LSB (Linux Standard Base) compliance. - * Incompatible change: In MySQL 5.1.6, when log tables were - implemented, the default log destination for the general query - and slow query log was TABLE. As of MySQL 5.1.21, this default - has been changed to FILE, which is compatible with MySQL 5.0, - but incompatible with earlier releases of MySQL 5.1. If you - are upgrading from MySQL 5.0 to 5.1.21 or higher, no logging - option changes should be necessary. However, if you are - upgrading from 5.1.6 through 5.1.20 to 5.1.21 or higher and - were using TABLE logging, use the --log-output=TABLE option - explicitly to preserve your server's table-logging behavior. + If the RPM files that you install include MySQL-server, the mysqld + server should be up and running after installation. You should be + able to start using MySQL. - * Incompatible change: For ENUM columns that had enumeration - values containing commas, the commas were mapped to 0xff - internally. However, this rendered the commas - indistinguishable from true 0xff characters in the values. - This no longer occurs. However, the fix requires that you dump - and reload any tables that have ENUM columns containing true - 0xff in their values: Dump the tables using mysqldump with the - current server before upgrading from a version of MySQL 5.1 - older than 5.1.15 to version 5.1.15 or newer. + If something goes wrong, you can find more information in the + binary installation section. See Section 2.2, "Installing MySQL + from Generic Binaries on Unix/Linux." - * As of MySQL 5.1.12, the lc_time_names system variable - specifies the locale that controls the language used to - display day and month names and abbreviations. This variable - affects the output from the DATE_FORMAT(), DAYNAME() and - MONTHNAME() functions. See Section 9.8, "MySQL Server Locale - Support." +Note - * As of MySQL 5.1.9, mysqld_safe no longer implicitly invokes - mysqld-max if it exists. Instead, it invokes mysqld unless a - --mysqld or --mysqld-version option is given to specify - another server explicitly. If you previously relied on the - implicit invocation of mysqld-max, you should use an - appropriate option now. As of MySQL 5.1.12, there is no longer - any separate mysqld-max server, so no change should be - necessary. + The accounts that are listed in the MySQL grant tables initially + have no passwords. After starting the server, you should set up + passwords for them using the instructions in Section 2.13, + "Post-Installation Setup and Testing." - SQL Changes: + During RPM installation, a user named mysql and a group named + mysql are created on the system. This is done using the useradd, + groupadd, and usermod commands. Those commands require appropriate + administrative privileges, which is ensured for locally managed + users and groups (as listed in the /etc/passwd and /etc/group + files) by the RPM installation process being run by root. - * Known issue: Prior to MySQL 5.1.17, the parser accepted - invalid code in SQL condition handlers, leading to server - crashes or unexpected execution behavior in stored programs. - Specifically, the parser allowed a condition handler to refer - to labels for blocks that enclose the handler declaration. - This was incorrect because block label scope does not include - the code for handlers declared within the labeled block. - As of 5.1.17, the parser rejects this invalid construct, but - if you perform a binary upgrade (without dumping and reloading - your databases), existing handlers that contain the construct - still are invalid and should be rewritten even if they appear - to function as you expect. - To find affected handlers, use mysqldump to dump all stored - procedures and functions, triggers, and events. Then attempt - to reload them into an upgraded server. Handlers that contain - illegal label references will be rejected. - For more information about condition handlers and writing them - to avoid invalid jumps, see Section 12.8.4.2, "DECLARE for - Handlers." + For nonlocal user management (LDAP, NIS, and so forth), the + administrative tools may require additional authentication (such + as a password), and will fail if the installing user does not + provide this authentication. Even if they fail, the RPM + installation will not abort but succeed, and this is intentional. + If they failed, some of the intended transfer of ownership may be + missing, and it is recommended that the system administrator then + manually ensures some appropriate user andgroup exists and + manually transfers ownership following the actions in the RPM spec + file. - * Incompatible change: The parser accepted statements that - contained /* ... */ that were not properly closed with */, - such as SELECT 1 /* + 2. As of MySQL 5.1.23, statements that - contain unclosed /*-comments now are rejected with a syntax - error. - This fix has the potential to cause incompatibilities. Because - of Bug#26302: http://bugs.mysql.com/26302, which caused the - trailing */ to be truncated from comments in views, stored - routines, triggers, and events, it is possible that objects of - those types may have been stored with definitions that now - will be rejected as syntactically invalid. Such objects should - be dropped and re-created so that their definitions do not - contain truncated comments. +2.7. Installing MySQL on Mac OS X - * Incompatible change: Multiple-table DELETE statements - containing ambiguous aliases could have unintended side - effects such as deleting rows from the wrong table. Example: -DELETE FROM t1 AS a2 USING t1 AS a1 INNER JOIN t2 AS a2; - As of MySQL 5.1.23, alias declarations can be declared only in - the table_references part. Elsewhere in the statement, alias - references are allowed but not alias declarations. Statements - containing aliases that are no longer allowed must be - rewritten. + MySQL for Mac OS X is available in a number of different forms: - * Incompatible change: As of MySQL 5.1.8, TYPE = engine_name is - still accepted as a synonym for the ENGINE = engine_name table - option but generates a warning. You should note that this - option is not available in MySQL 5.1.7, and is removed - altogether as of MySQL 6.0 and produces a syntax error. - TYPE has been deprecated since MySQL 4.0. + * Native Package Installer format, which uses the native Mac OS + X installer to walk you through the installation of MySQL. For + more information, see Section 2.7.1, "Installing MySQL Using + the Installation Package." You can use the package installer + with Mac OS X 10.3 and later, and available for both PowerPC + and Intel architectures, and both 32-bit and 64-bit + architectures. There is no Universal Binary available using + the package installation method. The user you use to perform + the installation must have administrator privileges. + + * Tar package format, which uses a file packaged using the Unix + tar and gzip commands. To use this method, you will need to + open a Terminal window. You do not need administrator + privileges using this method, as you can install the MySQL + server anywhere using this method. For more information on + using this method, you can use the generic instructions for + using a tarball, Section 2.2, "Installing MySQL from Generic + Binaries on Unix/Linux."You can use the package installer with + Mac OS X 10.3 and later, and available for both PowerPC and + Intel architectures, and both 32-bit and 64-bit architectures. + A Universal Binary, incorporating both Power PC and Intel + architectures and 32-bit and 64-bit binaries is available. + In addition to the core installation, the Package Installer + also includes Section 2.7.2, "Installing the MySQL Startup + Item" and Section 2.7.3, "Installing and Using the MySQL + Preference Pane," both of which simplify the management of + your installation. + + * Mac OS X server includes a version of MySQL as standard. If + you want to use a more recent version than that supplied with + the Mac OS X server release, you can make use of the package + or tar formats. For more information on using the MySQL + bundled with Mac OS X, see Section 2.7.4, "Using MySQL on Mac + OS X Server." - * Incompatible change: The namespace for triggers changed in - MySQL 5.0.10. Previously, trigger names had to be unique per - table. Now they must be unique within the schema (database). - An implication of this change is that DROP TRIGGER syntax now - uses a schema name instead of a table name (schema name is - optional and, if omitted, the current schema will be used). - When upgrading from a version of MySQL 5 older than 5.0.10 to - MySQL 5.0.10 or newer, you must drop all triggers and - re-create them or DROP TRIGGER will not work after the - upgrade. Here is a suggested procedure for doing this: + For additional information on using MySQL on Mac OS X, see Section + 2.7.5, "MySQL Installation on Mac OS X Notes." - 1. Upgrade to MySQL 5.0.10 or later to be able to access - trigger information in the INFORMATION_SCHEMA.TRIGGERS - table. (This should work even for pre-5.0.10 triggers.) +2.7.1. Installing MySQL Using the Installation Package - 2. Dump all trigger definitions using the following SELECT - statement: -SELECT CONCAT('CREATE TRIGGER ', t.TRIGGER_SCHEMA, '.', t.TRIGGER_NAM -E, - ' ', t.ACTION_TIMING, ' ', t.EVENT_MANIPULATION, ' ON ' -, - t.EVENT_OBJECT_SCHEMA, '.', t.EVENT_OBJECT_TABLE, - ' FOR EACH ROW ', t.ACTION_STATEMENT, '//' ) -INTO OUTFILE '/tmp/triggers.sql' -FROM INFORMATION_SCHEMA.TRIGGERS AS t; - The statement uses INTO OUTFILE, so you must have the - FILE privilege. The file will be created on the server - host. Use a different file name if you like. To be 100% - safe, inspect the trigger definitions in the triggers.sql - file, and perhaps make a backup of the file. + You can install MySQL on Mac OS X 10.3.x ("Panther") or newer + using a Mac OS X binary package in PKG format instead of the + binary tarball distribution. Please note that older versions of + Mac OS X (for example, 10.1.x or 10.2.x) are not supported by this + package. - 3. Stop the server and drop all triggers by removing all - .TRG files in your database directories. Change location - to your data directory and issue this command: -shell> rm */*.TRG + The package is located inside a disk image (.dmg) file that you + first need to mount by double-clicking its icon in the Finder. It + should then mount the image and display its contents. - 4. Start the server and re-create all triggers using the - triggers.sql file: -mysql> delimiter // ; -mysql> source /tmp/triggers.sql // +Note - 5. Check that all triggers were successfully created using - the SHOW TRIGGERS statement. + Before proceeding with the installation, be sure to shut down all + running MySQL server instances by either using the MySQL Manager + Application (on Mac OS X Server) or via mysqladmin shutdown on the + command line. - * Incompatible change: MySQL 5.1.6 introduces the TRIGGER - privilege. Previously, the SUPER privilege was needed to - create or drop triggers. Now those operations require the - TRIGGER privilege. This is a security improvement because you - no longer need to grant users the SUPER privilege to enable - them to create triggers. However, the requirement that the - account named in a trigger's DEFINER clause must have the - SUPER privilege has changed to a requirement for the TRIGGER - privilege. When upgrading from a previous version of MySQL 5.0 - or 5.1 to MySQL 5.1.6 or newer, be sure to update your grant - tables by running mysql_upgrade. This will assign the TRIGGER - privilege to all accounts that had the SUPER privilege. If you - fail to update the grant tables, triggers may fail when - activated. After updating the grant tables, you can revoke the - SUPER privilege from those accounts that no longer otherwise - require it. + When installing from the package version, you should also install + the MySQL Preference Pane, which will allow you to control the + startup and execution of your MySQL server from System + Preferences. For more information, see Section 2.7.3, "Installing + and Using the MySQL Preference Pane." + + When installing using the package installer, the files are + installed into a directory within /usr/local matching the name of + the installation version and platform. For example, the installer + file mysql-5.1.39-osx10.5-x86_64.pkg installs MySQL into + /usr/local/mysql-5.1.39-osx10.5-x86_64 . The installation layout + of the directory is as shown in the following table: + Directory Contents of Directory + bin Client programs and the mysqld server + data Log files, databases + docs Manual in Info format + include Include (header) files + lib Libraries + man Unix manual pages + mysql-test MySQL test suite + scripts Contains the mysql_install_db script + share/mysql Error message files + sql-bench Benchmarks + support-files Scripts and sample configuration files + /tmp/mysql.sock The location of the MySQL Unix socket + + During the package installer process, a symbolic link from + /usr/local/mysql to the version/platform specific directory + created during installation will be created automatically. + + 1. Download and open the MySQL package installer, which is + provided on a disk image (.dmg). Double-click to open the disk + image, which includes the main MySQL installation package, the + MySQLStartupItem.pkg installation package, and the + MySQL.prefPane. + + 2. Double-click on the MySQL installer package. It will be named + according to the version of MySQL you have downloaded. For + example, if you have downloaded MySQL 5.1.39, double-click + mysql-5.1.39-osx10.5-x86.pkg. + + 3. You will be presented with the openin installer dialog. Click + Continue to begihn installation. + MySQL Package Installer: Step 1 + + 4. A copy of the installation instructions and other important + information relevant to this installation are display. Click + Continue . + + 5. If you have downloaded the community version of MySQL, you + will be shown a copy of the relevent GNU General Public + License. Click Continue . + + 6. Select the drive you want to use to install the MySQL Startup + Item. The drive must have a valid, bootable, Mac OS X + operating system installed. Click Continue. + MySQL Package Installer: Step 4 + + 7. You will be asked to confirm the details of the installation, + including the space required for the installation. To change + the drive on which the startup item is installed you can click + either Go Back or Change Install Location.... To install the + startup item, click Install. - * Some keywords are reserved in MySQL 5.1 that were not reserved - in MySQL 5.0. See Section 8.3, "Reserved Words." + 8. Once the installation has been completed successfully, you + will be given an Install Succeeded message. - * The LOAD DATA FROM MASTER and LOAD TABLE FROM MASTER - statements are deprecated. See Section 12.6.2.2, "LOAD DATA - FROM MASTER Syntax," for recommended alternatives. + Once you have completed the basic installation, you must complete + the post-installation steps as specifed in Section 2.13, + "Post-Installation Setup and Testing." - * The INSTALL PLUGIN and UNINSTALL PLUGIN statements that are - used for the plugin API are new. So is the WITH PARSER clause - for FULLTEXT index creation that associates a parser plugin - with a full-text index. Section 22.2, "The MySQL Plugin - Interface." + For convenience, you may also want to install the Section 2.7.2, + "Installing the MySQL Startup Item" and Section 2.7.3, "Installing + and Using the MySQL Preference Pane." + +2.7.2. Installing the MySQL Startup Item + + The MySQL Installation Package includes a startup item that can be + used to automatically startup and shutdown MySQL during boot. + + To install the MySQL Startup Item: + + 1. Download and open the MySQL package installer, which is + provided on a disk image (.dmg). Double-click to open the disk + image, which includes the main MySQL installation package, the + MySQLStartupItem.pkg installation package, and the + MySQL.prefPane. + + 2. Double-click on the MySQLStartItem.pkg file to start the + installation process. + + 3. You will be presented with the Install MySQL Startup Item + dialog. + MySQL Startup Item Installer: Step 1 + Click Continue to continue the installation process. + + 4. A copy of the installation instructions and other important + information relevant to this installation are display. Click + Continue . + + 5. Select the drive you want to use to install the MySQL Startup + Item. The drive must have a valid, bootable, Mac OS X + operating system installed. Click Continue. + MySQL Startup Item Installer: Step 3 + + 6. You will be asked to confirm the details of the installation. + To change the drive on which the startup item is installed you + can click either Go Back or Change Install Location.... To + install the startup item, click Install. + + 7. Once the installation has been completed successfully, you + will be given an Install Succeeded message. + MySQL Startup Item Installer: Step 5 - C API Changes: + The Startup Item for MySQL is installed into + /Library/StartupItems/MySQLCOM. The Startup Item installation adds + a variable MYSQLCOM=-YES- to the system configuration file + /etc/hostconfig. If you want to disable the automatic startup of + MySQL, simply change this variable to MYSQLCOM=-NO-. - * Incompatible change: As of MySQL 5.1.7, the - mysql_stmt_attr_get() C API function returns a boolean rather - than an unsigned int for STMT_ATTR_UPDATE_MAX_LENGTH. - (Bug#16144: http://bugs.mysql.com/16144) + After the installation, you can start up MySQL by running the + following commands in a terminal window. You must have + administrator privileges to perform this task. -2.12.2. Downgrading MySQL + If you have installed the Startup Item, use this command to start + the server: +shell> sudo /Library/StartupItems/MySQLCOM/MySQLCOM start - This section describes what you should do to downgrade to an older - MySQL version in the unlikely case that the previous version - worked better than the new one. + You may be prompted for your password to complete the startup. - If you are downgrading within the same release series (for - example, from 5.0.13 to 5.0.12) the general rule is that you just - have to install the new binaries on top of the old ones. There is - no need to do anything with the databases. As always, however, it - is always a good idea to make a backup. + If you have installed the Startup Item, use this command to stop + the server: +shell> sudo /Library/StartupItems/MySQLCOM/MySQLCOM stop + + You may be prompted for your password to complete the shutdown. + +2.7.3. Installing and Using the MySQL Preference Pane + + The MySQL Package installer disk image also includes a custom + MySQL Preference Pane that enables you to start, stop and control + automated startup during boot of your MySQL installation. + + To install the MySQL Preference Pane: + + 1. Download and open the MySQL package installer package, which + is provided on a disk image (.dmg). Double-click to open the + disk image, which includes the main MySQL installation + package, the MySQLStartupItem.pkg installation package, and + the MySQL.prefPane. + + 2. Double click on MySQL.prefPane. The MySQL System Preferences + will open. + + 3. If this is the first time you have installed the preference + pane, you will be asked to confirm installation and whether + you want to install the preference pane for all users, or only + the current user. To install the preference pane for all users + you will need administrator privileges. If necessary, you will + be prompted for the username and password for a user with + administrator privileges. + + 4. If you already have the MySQL Preference Pane installed, you + will be asked to confirm whether you want to overwrite the + existing MySQL Preference Pane. - The following items form a checklist of things you should do - whenever you perform a downgrade: +Note - * Read the upgrading section for the release series from which - you are downgrading to be sure that it does not have any - features you really need. See Section 2.12.1, "Upgrading - MySQL." + The MySQL Preference Pane only starts and stops MySQL installation + installed from the MySQL package installation that have been + installed in the default location. + + Once the MySQL Preference Pane has been installed, you can control + your MySQL server instance using the preference pane. To use the + preference pane, open the System Preferences... from the Apple + menu. Select the MySQL preference pane by clicking on the MySQL + logo within the Other section of the preference panes list. + MySQL Preference Pane + + The MySQL Preference Pane shows the current status of the MySQL + server, showing stopped (in red) if the server is not running and + running (in green) if the server has already been started. The + preference pane will also show the current setting for whether the + MySQL server has been set to start up automatically. + + * To start MySQL using the preference pane: + Click Start MySQL Server. You may be prompted for the username + and password of a user with administrator privileges to start + the MySQL server. + + * To stop MySQL using the preference pane: + Click Stop MySQL Server. You may be prompted for the username + and password of a user with administrator privileges to + shutdown the MySQL server. + + * To automatically start the MySQL server when the system boots: + Check the checkbox next to Automatically Start MySQL Server on + Startup. + + * To disable the automatic starting of the MySQL server when the + system boots: + Uncheck the checkbox next to Automatically Start MySQL Server + on Startup. - * If there is a downgrading section for that version, you should - read that as well. + You can close the System Preferences... once you have completed + your settings. - * To see which new features were added between the version to - which you are downgrading and your current version, see the - change logs (Appendix C, "MySQL Change History"). +2.7.4. Using MySQL on Mac OS X Server - * Check Section 2.12.3, "Checking Whether Table Indexes Must Be - Rebuilt," to see whether changes to character sets or - collations were made between your current version of MySQL and - the version to which you are downgrading. If so and these - changes affect your table indexes, you will need to rebuild - the affected indexes using the instructions in Section 2.12.4, - "Rebuilding or Repairing Tables or Indexes." + If you are running Mac OS X Server, a version of MySQL should + already be installed. The following table shows the versions of + MySQL that ship with Mac OS X Server versions. + Mac OS X Server Version MySQL Version + 10.2-10.2.2 3.23.51 + 10.2.3-10.2.6 3.23.53 + 10.3 4.0.14 + 10.3.2 4.0.16 + 10.4.0 4.1.10a + 10.5.0 5.0.45 + 10.6.0 5.0.82 - In most cases, you can move the MySQL format files and data files - between different versions on the same architecture as long as you - stay within versions for the same release series of MySQL. + The installation layout of MySQL on Mac OS X Server is as shown in + the table below: + Directory Contents of Directory + /usr/bin Client programs + /var/mysql Log files, databases + /usr/libexec The mysqld server + /usr/share/man Unix manual pages + /usr/share/mysql/mysql-test MySQL test suite + /usr/share/mysql Contains the mysql_install_db script + /var/mysql/mysql.sock The location of the MySQL Unix socket - If you downgrade from one release series to another, there may be - incompatibilities in table storage formats. In this case, use - mysqldump to dump your tables before downgrading. After - downgrading, reload the dump file using mysql or mysqlimport to - re-create your tables. For examples, see Section 2.12.5, "Copying - MySQL Databases to Another Machine." +Note - A typical symptom of a downward-incompatible table format change - when you downgrade is that you cannot open tables. In that case, - use the following procedure: + The MySQL server bundled with Mac OS X Server does not include the + MySQL client libraries and header files required if you want to + access and use MySQL from a third-party driver, such as Perl DBI + or PHP. For more information on obtaining and installing MySQL + libraries, see Mac OS X Server version 10.5: MySQL libraries + available for download (http://support.apple.com/kb/TA25017) + Alternatively, you can ignore the bundled MySQL server and install + MySQL from the package or tarball installation. + + For more information on managing the bundled MySQL instance in Mac + OS X Server 10.5, see Mac OS X Server: Web Technologies + Administration For Version 10.5 Leopard + (http://images.apple.com/server/macosx/docs/Web_Technologies_Admin + _v10.5.pdf). For more information on managing the bundled MySQL + instance in Mac OS X Server 10.6, see Mac OS X Server: Web + Technologies Administration Version 10.6 Snow Leopard + (http://manuals.info.apple.com/en_US/WebTech_v10.6.pdf) + +2.7.5. MySQL Installation on Mac OS X Notes + + You should keep the following issues and notes in mind: + + * The default location for the MySQL Unix socket is different on + Mac OS X and Mac OS X Server depending on the installation + type you chose. The default locations by installation are as + follows: - 1. Stop the older MySQL server that you are downgrading to. + Package Installer from MySQL /tmp/mysql.sock + Tarball from MySQL /tmp/mysql.sock + MySQL Bundled with Mac OS X Server /var/mysql/mysql.sock + To prevent issues, you should either change the configuration + of the socket used within your application (for example, + changing php.ini), or you should configure the socket location + using a MySQL configuration file and the socket option. For + more information, see Section 5.1.2, "Server Command Options." + + * You may need (or want) to create a specific mysql user to own + the MySQL directory and data. On Mac OS X 10.4 and lower you + can do this by using the Netinfo Manager application, located + within the Utilities folder within the Applications folder. On + Mac OS X 10.5 and later you can do this through the Directory + Utility. From Mac OS X 10.5 and later (including Mac OS X + Server 10.5) the mysql should already exist. For use in single + user mode, an entry for _mysql (note the underscore prefix) + should already exist within the system /etc/passwd file. - 2. Restart the newer MySQL server you are downgrading from. + * Due to a bug in the Mac OS X package installer, you may see + this error message in the destination disk selection dialog: +You cannot install this software on this disk. (null) + If this error occurs, simply click the Go Back button once to + return to the previous screen. Then click Continue to advance + to the destination disk selection again, and you should be + able to choose the destination disk correctly. We have + reported this bug to Apple and it is investigating this + problem. + + * Because the MySQL package installer installs the MySQL + contents into a version and platform specific directory, you + can use this to upgrade and migrate your database between + versions. You will need to either copy the data directory from + the old version to the new version, or alternatively specify + an alternative datadir value to set location of the data + directory. - 3. Dump any tables that were inaccessible to the older server by - using mysqldump to create a dump file. + * You might want to add aliases to your shell's resource file to + make it easier to access commonly used programs such as mysql + and mysqladmin from the command line. The syntax for bash is: +alias mysql=/usr/local/mysql/bin/mysql +alias mysqladmin=/usr/local/mysql/bin/mysqladmin + For tcsh, use: +alias mysql /usr/local/mysql/bin/mysql +alias mysqladmin /usr/local/mysql/bin/mysqladmin + Even better, add /usr/local/mysql/bin to your PATH environment + variable. You can do this by modifying the appropriate startup + file for your shell. For more information, see Section 4.2.1, + "Invoking MySQL Programs." + + * After you have copied over the MySQL database files from the + previous installation and have successfully started the new + server, you should consider removing the old installation + files to save disk space. Additionally, you should also remove + older versions of the Package Receipt directories located in + /Library/Receipts/mysql-VERSION.pkg. - 4. Stop the newer MySQL server and restart the older one. +2.8. Installing MySQL on Solaris - 5. Reload the dump file into the older server. Your tables should - be accessible. + To obtain a binary MySQL distribution for Solaris in tarball or + PKG format, http://dev.mysql.com/downloads/mysql/5.1.html. - It might also be the case that the structure of the system tables - in the mysql database has changed and that downgrading introduces - some loss of functionality or requires some adjustments. Here are - some examples: + If you install MySQL using a binary tarball distribution on + Solaris, you may run into trouble even before you get the MySQL + distribution unpacked, as the Solaris tar cannot handle long file + names. This means that you may see errors when you try to unpack + MySQL. - * Trigger creation requires the TRIGGER privilege as of MySQL - 5.1. In MySQL 5.0, there is no TRIGGER privilege and SUPER is - required instead. If you downgrade from MySQL 5.1 to 5.0, you - will need to give the SUPER privilege to those accounts that - had the TRIGGER privilege in 5.1. + If this occurs, you must use GNU tar (gtar) to unpack the + distribution. - * Triggers were added in MySQL 5.0, so if you downgrade from 5.0 - to 4.1, you cannot use triggers at all. + You can install MySQL on Solaris using a binary package in PKG + format instead of the binary tarball distribution. Before + installing using the binary PKG format, you should create the + mysql user and group, for example: +groupadd mysql +useradd -g mysql mysql -2.12.2.1. Downgrading to MySQL 5.0 + Some basic PKG-handling commands follow: - When downgrading to MySQL 5.0 from MySQL 5.1 or a later version, - you should keep in mind the following issues relating to features - found in MySQL 5.1 and later, but not in MySQL 5.0: + * To add a package: +pkgadd -d package_name.pkg - * Partitioning. MySQL 5.0 does not support user-defined - partitioning. If a table was created as a partitioned table in - 5.1 (or if an table created in a previous version of MySQL was - altered to include partitions after an upgrade to 5.1), the - table is accessible after downgrade only if you do one of the - following: + * To remove a package: +pkgrm package_name - + Export the table using mysqldump and then drop it in - MySQL 5.1; import the table again following the downgrade - to MySQL 5.0. + * To get a full list of installed packages: +pkginfo - + Prior to the downgrade, remove the table's partitioning - using ALTER TABLE table_name REMOVE PARTITIONING. + * To get detailed information for a package: +pkginfo -l package_name - * Event Scheduler. MySQL 5.0 does not support scheduled events. - If your databases contain scheduled event definitions, you - should prevent them from being dumped when you use mysqldump - by using the --skip-events option. (See Section 4.5.4, - "mysqldump --- A Database Backup Program.") + * To list the files belonging to a package: +pkgchk -v package_name - * Stored routines. MySQL 5.1.21 added a number of new columns - to the mysql.proc table in which stored routine definitions - are stored. If you are downgrading from MySQL 5.1.21 or later - to MySQL 5.0, you cannot import the MySQL 5.1 routine - definitions into MySQL 5.0.46 or earlier using the dump of - mysql.proc created by mysqldump (such as when using the - --all-databases option). Instead, you should run mysqldump - --routines prior to performing the downgrade and run the - stored routines DDL statements following the downgrade. - See Bug#11986: http://bugs.mysql.com/11986, - Bug#30029: http://bugs.mysql.com/30029, and - Bug#30660: http://bugs.mysql.com/30660, for more information. + * To get packaging information for an arbitrary file: +pkgchk -l -p file_name - * Triggers. Trigger creation requires the TRIGGER privilege as - of MySQL 5.1. In MySQL 5.0, there is no TRIGGER privilege and - SUPER is required instead. If you downgrade from MySQL 5.1 to - 5.0, you will need to give the SUPER privilege to those - accounts that had the TRIGGER privilege in 5.1. +2.8.1. Solaris Notes -2.12.3. Checking Whether Table Indexes Must Be Rebuilt + For information about installing MySQL on Solaris using PKG + distributions, see Section 2.8, "Installing MySQL on Solaris." - A binary upgrade or downgrade is one that installs one version of - MySQL "in place" over an existing version, without dumping and - reloading tables: + On Solaris, you may run into trouble even before you get the MySQL + distribution unpacked, as the Solaris tar cannot handle long file + names. This means that you may see errors when you try to unpack + MySQL. - 1. Stop the server for the existing version if it is running. + If this occurs, you must use GNU tar (gtar) to unpack the + distribution. - 2. Install a different version of MySQL. This is an upgrade if - the new version is higher than the original version, a - downgrade if the version is lower. + If you have an UltraSPARC system, you can get 4% better + performance by adding -mcpu=v8 -Wa,-xarch=v8plusa to the CFLAGS + and CXXFLAGS environment variables. - 3. Start the server for the new version. + If you have Sun's Forte 5.0 (or newer) compiler, you can run + configure like this: +CC=cc CFLAGS="-Xa -fast -native -xstrconst -mt" \ +CXX=CC CXXFLAGS="-noex -mt" \ +./configure --prefix=/usr/local/mysql --enable-assembler - In many cases, the tables from the previous version of MySQL can - be used without change by the new version. However, sometimes - modifications are made to the handling of character sets or - collations that change the character sort order, which causes the - ordering of entries in any index that uses an affected character - set or collation to be incorrect. Such changes result in several - possible problems: + To create a 64-bit binary with Sun's Forte compiler, use the + following configuration options: +CC=cc CFLAGS="-Xa -fast -native -xstrconst -mt -xarch=v9" \ +CXX=CC CXXFLAGS="-noex -mt -xarch=v9" ASFLAGS="-xarch=v9" \ +./configure --prefix=/usr/local/mysql --enable-assembler - * Comparison results that differ from previous results + To create a 64-bit Solaris binary using gcc, add -m64 to CFLAGS + and CXXFLAGS and remove --enable-assembler from the configure + line. - * Inability to find some index values due to misordered index - entries + In the MySQL benchmarks, we obtained a 4% speed increase on + UltraSPARC when using Forte 5.0 in 32-bit mode, as compared to + using gcc 3.2 with the -mcpu flag. - * Misordered ORDER BY results + If you create a 64-bit mysqld binary, it is 4% slower than the + 32-bit binary, but can handle more threads and memory. - * Tables that CHECK TABLE reports as being in need of repair + When using Solaris 10 for x86_64, you should mount any file + systems on which you intend to store InnoDB files with the + forcedirectio option. (By default mounting is done without this + option.) Failing to do so will cause a significant drop in + performance when using the InnoDB storage engine on this platform. - The solution to these problems is to rebuild any indexes that use - an affected character set or collation, either by dropping and - re-creating the indexes, or by dumping and reloading the entire - table. For information about rebuilding indexes, see Section - 2.12.4, "Rebuilding or Repairing Tables or Indexes." + If you get a problem with fdatasync or sched_yield, you can fix + this by adding LIBS=-lrt to the configure line - To check whether a table has indexes that must be rebuilt, consult - the following list. It indicates which versions of MySQL - introduced character set or collation changes that require indexes - to be rebuilt. Each entry indicates the version in which the - change occurred and the character sets or collations that the - change affects. If the change is associated with a particular bug - report, the bug number is given. + Solaris does not provide static versions of all system libraries + (libpthreads and libdl), so you cannot compile MySQL with + --static. If you try to do so, you get one of the following + errors: +ld: fatal: library -ldl: not found +undefined reference to `dlopen' +cannot find -lrt + + If you link your own MySQL client programs, you may see the + following error at runtime: +ld.so.1: fatal: libmysqlclient.so.#: +open failed: No such file or directory + + This problem can be avoided by one of the following methods: + + * Link clients with the -Wl,r/full/path/to/libmysqlclient.so + flag rather than with -Lpath). - The list applies both for binary upgrades and downgrades. For - example, Bug#29461: http://bugs.mysql.com/29461 was fixed in MySQL - 5.0.48, so it applies to upgrades from versions older than 5.0.48 - to 5.0.48 or newer, and also to downgrades from 5.0.48 or newer to - versions older than 5.0.48. + * Copy libmysqclient.so to /usr/lib. - If you have tables with indexes that are affected, rebuild the - indexes using the instructions given in Section 2.12.4, - "Rebuilding or Repairing Tables or Indexes." + * Add the path name of the directory where libmysqlclient.so is + located to the LD_RUN_PATH environment variable before running + your client. - In many cases, you can use CHECK TABLE ... FOR UPGRADE to identify - tables for which index rebuilding is required. (It will report: - Table upgrade required. Please do "REPAIR TABLE `tbl_name`" or - dump/reload to fix it!) In these cases, you can also use - mysqlcheck --check-upgrade or mysql_upgrade, which execute CHECK - TABLE. However, the use of CHECK TABLE applies only after - upgrades, not downgrades. Also, CHECK TABLE is not applicable to - all storage engines. For details about which storage engines CHECK - TABLE supports, see Section 12.5.2.3, "CHECK TABLE Syntax." + If you have problems with configure trying to link with -lz when + you don't have zlib installed, you have two options: - Changes that cause index rebuilding to be necessary: + * If you want to be able to use the compressed communication + protocol, you need to get and install zlib from ftp.gnu.org. - * MySQL 5.0.48 (Bug#29461: http://bugs.mysql.com/29461) - Affects indexes for columns that use any of these character - sets: eucjpms, euc_kr, gb2312, latin7, macce, ujis - Affected tables can be detected by CHECK TABLE ... FOR UPGRADE - as of MySQL 5.1.29, 6.0.8 (see - Bug#39585: http://bugs.mysql.com/39585) + * Run configure with the --with-named-z-libs=no option when + building MySQL. - * MySQL 5.0.48 (Bug#27562: http://bugs.mysql.com/27562) - Affects indexes that use the ascii_general_ci collation for - columns that contain any of these characters: '`' GRAVE - ACCENT, '[' LEFT SQUARE BRACKET, '\' REVERSE SOLIDUS, ']' - RIGHT SQUARE BRACKET, '~' TILDE - Affected tables can be detected by CHECK TABLE ... FOR UPGRADE - as of MySQL 5.1.29, 6.0.8 (see - Bug#39585: http://bugs.mysql.com/39585) + If you are using gcc and have problems with loading user-defined + functions (UDFs) into MySQL, try adding -lgcc to the link line for + the UDF. - * MySQL 5.1.21 (Bug#29461: http://bugs.mysql.com/29461) - Affects indexes for columns that use any of these character - sets: eucjpms, euc_kr, gb2312, latin7, macce, ujis - Affected tables can be detected by CHECK TABLE ... FOR UPGRADE - as of MySQL 5.1.29, 6.0.8 (see - Bug#39585: http://bugs.mysql.com/39585) + If you would like MySQL to start automatically, you can copy + support-files/mysql.server to /etc/init.d and create a symbolic + link to it named /etc/rc3.d/S99mysql.server. - * MySQL 5.1.23 (Bug#27562: http://bugs.mysql.com/27562) - Affects indexes that use the ascii_general_ci collation for - columns that contain any of these characters: '`' GRAVE - ACCENT, '[' LEFT SQUARE BRACKET, '\' REVERSE SOLIDUS, ']' - RIGHT SQUARE BRACKET, '~' TILDE - Affected tables can be detected by CHECK TABLE ... FOR UPGRADE - as of MySQL 5.1.29, 6.0.8 (see - Bug#39585: http://bugs.mysql.com/39585) + If too many processes try to connect very rapidly to mysqld, you + should see this error in the MySQL log: +Error in accept: Protocol error - * MySQL 5.1.24 (Bug#27877: http://bugs.mysql.com/27877) - Affects indexes that use the utf8_general_ci or - ucs2_general_ci collation for columns that contain 'ß' LATIN - SMALL LETTER SHARP S (German). - Affected tables can be detected by CHECK TABLE ... FOR UPGRADE - as of MySQL 5.1.30, 6.0.8 (see - Bug#40053: http://bugs.mysql.com/40053) + You might try starting the server with the --back_log=50 option as + a workaround for this. (Use -O back_log=50 before MySQL 4.) - * * MySQL 6.0.1 (WL#3664) - Affects indexes that use the latin2_czech_cs collation. - Affected tables can be detected by CHECK TABLE ... FOR UPGRADE - as of MySQL 5.4.4, 6.0.9 (see - Bug#40054: http://bugs.mysql.com/40054) - MySQL 6.0.5 (Bug#33452: http://bugs.mysql.com/33452) - Affects indexes that use the latin2_czech_cs collation. - Affected tables can be detected by CHECK TABLE ... FOR UPGRADE - as of MySQL 5.4.4, 6.0.9 (see - Bug#40054: http://bugs.mysql.com/40054) + To configure the generation of core files on Solaris you should + use the coreadm command. Because of the security implications of + generating a core on a setuid() application, by default, Solaris + does not support core files on setuid() programs. However, you can + modify this behavior using coreadm. If you enable setuid() core + files for the current user, they will be generated using the mode + 600 and owned by the superuser. - * MySQL 6.0.5 (Bug#27877: http://bugs.mysql.com/27877) - Affects indexes that use the utf8_general_ci or - ucs2_general_ci collation for columns that contain 'ß' LATIN - SMALL LETTER SHARP S (German). - Affected tables can be detected by CHECK TABLE ... FOR UPGRADE - as of MySQL 6.0.8 (see - Bug#40053: http://bugs.mysql.com/40053) +2.9. Installing MySQL on i5/OS - * MySQL 6.0.6 (Bug#25420: http://bugs.mysql.com/25420) - Affects indexes for columns that use the following collations, - if the columns contain the indicated characters: - big5_chinese_ci: '~' TILDE or '`' GRAVE ACCENT; - cp866_general_ci: j LATIN SMALL LETTER J; gb2312_chinese_ci: - '~' TILDE; gbk_chinese_ci: '~' TILDE - Affected tables can be detected by CHECK TABLE ... FOR UPGRADE - as of MySQL 5.4.4, 6.0.9 (see - Bug#40054: http://bugs.mysql.com/40054) + The i5/OS POWER MySQL package was created in cooperation with IBM. + MySQL works within the Portable Application Solution Environment + (PASE) on the System i series of hardware and will also provide + database services for the Zend Core for i5/OS. -2.12.4. Rebuilding or Repairing Tables or Indexes + MySQL for i5/OS is provided both as a tar file and as a save file + (.savf) package that can be downloaded and installed directly + without any additional installation steps required. To install + MySQL using the tar file, see Section 2.2, "Installing MySQL from + Generic Binaries on Unix/Linux." - This section describes how to rebuild a table. This can be - necessitated by changes to MySQL such as how data types are - handled or changes to character set handling. For example, an - error in a collation might have been corrected, necessitating a - table rebuild to rebuild the indexes for character columns that - use the collation. It might also be that a table repair or upgrade - should be done as indicated by a table check operation such as - that performed by CHECK TABLE, mysqlcheck, or mysql_upgrade. + MySQL is only supported on i5/OS V5R4 or later releases. The i5/OS + PASE must be installed for MySQL to operate. You must be able to + login as a user in *SECOFR class. - Methods for rebuilding a table include dumping and reloading it, - or using ALTER TABLE or REPAIR TABLE. + You should the installation notes and tips for i5/OS before + starting installation. See i5/OS Installation Notes. + + Before Installation: Note - If you are rebuilding tables because a different version of MySQL - will not handle them after a binary upgrade or downgrade, you must - use the dump-and-reload method. Dump the tables before upgrading - or downgrading (using your original version of MySQL), and reload - the tables after upgrading or downgrading (after installing the - new version). + The installation package will use an existing configuration if you + have previously installed MySQL (which is identified by looking + for the file /etc/my.cnf). The values for the data directory + (DATADIR) and owner of the MySQL files (USRPRF) specified during + the installation will be ignored, and the values determined from + the /etc/my.cnf will be used instead. - If you use the dump-and-reload method of rebuilding tables only - for the purpose of rebuilding indexes, you can perform the dump - either before or after upgrading or downgrading. Reloading still - must be done afterward. + If you want to change these parameters during a new install, you + should temporarily rename /etc/my.cnf, install MySQL using the new + parameters you want to use, and then merge your previous + /etc/my.cnf configuration settings with the new /etc/my.cnf file + that is created during installation. - To re-create a table by dumping and reloading it, use mysqldump to - create a dump file and mysql to reload the file: -shell> mysqldump db_name t1 > dump.sql -shell> mysql db_name < dump.sql + * You must have a user profile with PASE with suitable + privileges. The user should be within the *SECOFR class, such + as the QSECOFR user ID. You can use the WRKUSRPRF command to + check your user profile. + + * For network connections to MySQL, you must have TCP/IP + enabled. You should also check the following: + + + Ensure that a name has defined for the system. Run the + Configure TCP/IP (CFGTCP) command and select option 12 + (Change TCP/IP domain information) to display this + setting. Make sure that a value is listed in the Host + name field. - To recreate all the tables in a single database, specify the - database name without any following table name: -shell> mysqldump db_name > dump.sql -shell> mysql db_name < dump.sql + + Make sure that the system has a loopback entry which + represents the localhost or 127.0.0.1. - To recreate all tables in all databases, use the --all-databases - option: -shell> mysqldump --all-databases > dump.sql -shell> mysql < dump.sql + + Ensure that the IP address of the IBM i machine is mapped + correctly to the host name. - To rebuild a table with ALTER TABLE, use a statement that - "changes" the table to use the storage engine that it already has. - For example, if t1 is a MyISAM table, use this statement: -mysql> ALTER TABLE t1 ENGINE = MyISAM; + To install MySQL on i5/OS, follow these steps: - If you are not sure which storage engine to specify in the ALTER - TABLE statement, use SHOW CREATE TABLE to display the table - definition. + 1. On the System i machine, create a save file that will be used + to receive the downloaded installation save file. The file + should be located within the General Purpose Library (QGPL): +CRTSAVF FILE(QGPL/MYSQLINST) TESXT('MySQL Save file') - If you must rebuild a table because a table checking operation - indicates that the table is corrupt or needs an upgrade, you can - use REPAIR TABLE if that statement supports the table's storage - engine. For example, to repair a MyISAM table, use this statement: -mysql> REPAIR TABLE t1; + 2. Download the MySQL installation save file in 32-bit + (mysql-5.1.39-i5os-power-32bit.savf) or 64-bit + (mysql-5.1.39-i5os-power-64bit.savf) from MySQL Downloads + (http://dev.mysql.com/downloads) - For storage engines such as InnoDB that REPAIR TABLE does not - support, use mysqldump to create a dump file and mysql to reload - the file, as described earlier. + 3. You need to FTP the downloaded .savf file directly into the + QGPL/MYSQLINST file on the System i server. You can do this + through FTP using the following steps after logging in to the + System i machine: +ftp> bin +ftp> cd qgpl +ftp> put mysql-5.1.39-i5os-power.savf mysqlinst - For specifics about which storage engines REPAIR TABLE supports, - see Section 12.5.2.6, "REPAIR TABLE Syntax." + 4. Log into the System i server using a user in the *SECOFR + class, such as the QSECOFR user ID. -2.12.5. Copying MySQL Databases to Another Machine + 5. You need to restore the installation library stored in the + .savf save file: +RSTLIB MYSQLINST DEV(*SAVF) SAVF(QGPL/MYSQLINST) MBROPT(*ALL) ALWOBJD +IF(*ALL) - You can copy the .frm, .MYI, and .MYD files for MyISAM tables - between different architectures that support the same - floating-point format. (MySQL takes care of any byte-swapping - issues.) See Section 13.5, "The MyISAM Storage Engine." +Note + You can ignore the security changes-type message at the bottom + of the installation panel. - In cases where you need to transfer databases between different - architectures, you can use mysqldump to create a file containing - SQL statements. You can then transfer the file to the other - machine and feed it as input to the mysql client. + 6. Once you have finished restoring the MYSQLINST library, check + that all the necessary objects for installation are on the + system by using the Display Library (DSPLIB) command: +DSPLIB LIB(MYSQLINST) - Use mysqldump --help to see what options are available. + 7. You need to execute the installation command, + MYSQLINST/INSMYSQL. You can specify three parameter settings + during installation: - The easiest (although not the fastest) way to move a database - between two machines is to run the following commands on the - machine on which the database is located: -shell> mysqladmin -h 'other_hostname' create db_name -shell> mysqldump db_name | mysql -h 'other_hostname' db_name + + DIR('/QOpenSys/usr/local/mysql') sets the installation + location for the MySQL files. The directory will be + created if it does not already exist. + + + DATADIR('/QOpenSys/usr/local/mysql/data') sets the + location of the directory that will be used to store the + database files and binary logs. The default setting is + /QOpenSys/usr/local/mysql/data. Note that if the + installer detects an existing installation (due to the + existence of /etc/my.cnf), then the existing setting will + be used instead of the default. - If you want to copy a database from a remote machine over a slow - network, you can use these commands: -shell> mysqladmin create db_name -shell> mysqldump -h 'other_hostname' --compress db_name | mysql db_na -me + + USRPRF(MYSQL) sets the user profile that will own the + files that are installed. The profile will be created if + it does not already exist. - You can also store the dump in a file, transfer the file to the - target machine, and then load the file into the database there. - For example, you can dump a database to a compressed file on the - source machine like this: -shell> mysqldump --quick db_name | gzip > db_name.gz +Note + You should choose an appropriate user for using the MySQL + server installation. The user will be used whenever you + need to do any administration on the MySQL server. + Once you have set the appropriate parameters, you can begin + the installation. + The installation copies all the necessary files into a + directory matching the DIR configuration value; sets the + ownership on those files, sets up the MySQL environment and + creates the MySQL configuration file (in /etc/my.cnf) + completing all the steps in a typical binary installation + process automatically. If this is a new installation of MySQL, + or if the installer detects that this is a new version + (because the /etc/my.cnf file does not exist), then the + initial core MySQL databases will also be created during + installation. + Once the installation has been completed, you will get a + notice advising you to set the password for the root user. For + more information, Section 2.13, "Post-Installation Setup and + Testing." - Transfer the file containing the database contents to the target - machine and run these commands there: -shell> mysqladmin create db_name -shell> gunzip < db_name.gz | mysql db_name + 8. Once the installation has completed, you can delete the + installation file: +DLTLIB LIB(MYSQLINST) - You can also use mysqldump and mysqlimport to transfer the - database. For large tables, this is much faster than simply using - mysqldump. In the following commands, DUMPDIR represents the full - path name of the directory you use to store the output from - mysqldump. + Upgrading an existing MySQL instance - First, create the directory for the output files and dump the - database: -shell> mkdir DUMPDIR -shell> mysqldump --tab=DUMPDIR db_name + You need to execute the upgrade command, MYSQLINST/UPGMYSQL. You + must specify 6 parameters to perform an upgrade: - Then transfer the files in the DUMPDIR directory to some - corresponding directory on the target machine and load the files - into MySQL there: -shell> mysqladmin create db_name # create database -shell> cat DUMPDIR/*.sql | mysql db_name # create tables in databas -e -shell> mysqlimport db_name DUMPDIR/*.txt # load data into tables + * DIR('/QOpenSys/usr/local/') --- sets the installation location + for the MySQL files. The directory will be created if it does + not already exist. This is the directory that the MySQL server + will be installed into, inside a directory with a name + matching the version and release. For example if installing + MySQL 5.1.39 with the DIR set to /QOpenSys/usr/local/ would + result in /QOpenSys/usr/local/mysql-5.1.39-i5os-power64 and a + symbolic link to this directory will be created in + /QOpenSys/usr/local/mysql. + + * DATADIR('/QOpenSys/mysql/data') --- sets the location of the + directory that will be upgraded. + + * USRPRF('MYSQL') --- sets the user profile that will own the + files that are installed. The profile will be created if it + does not already exist; if it is created as part of the + upgrade process, it will be disabled initially. You may wish + to enable this user profile so that it can be used to start + the MySQL server later. It is best practice to use the one + previously created during the first installation. + + * MYSQLUSR('root user') --- any user account in the current + MySQL server with SUPER privileges. + + * PASSWORD('root user password') --- the password for the above + account. This is necessary as the upgrade starts the MySQL + server to upgrade the tables and the password is need to be + able to shutdown the MySQL server. + + * CURINST('path to previous install') --- the full path to the + installation that is being upgraded. For example an + installation in /QOpenSys/usr/local/ will be + /QOpenSys/usr/local/msyql-5.1.30-i5os-power64. Failure to + specify this option may result in corruption of your existing + data files. + + For example: +MYSQLINST/UPGMYSQL DIR('/QOpenSys/usr/local/') DATADIR('/QOpenSys/mys +ql/data') » + USERPRF(MYSQL) MYSQLUSR('root') PASSWORD('root') CURINST('/QOpen +Sys/usr/local/mysql-5.1.30-i5os-power64') + + You should receive a Program Message indicating UPGRADE + SUCCESSFUL! upon completion or an error message if there is a + problem.You can view the upgrade programs progression and the + error in the text file upgrade.log in the installation directory. - Do not forget to copy the mysql database because that is where the - grant tables are stored. You might have to run commands as the - MySQL root user on the new machine until you have the mysql - database in place. + To start MySQL: - After you import the mysql database on the new machine, execute - mysqladmin flush-privileges so that the server reloads the grant - table information. + 1. Log into the System i server using the user profile create or + specified during installation. By default, this is MYSQL. -2.13. Operating System-Specific Notes +Note + You should start mysqld_safe using a user that in the PASE + environment has the id=0 (the equivalent of the standard Unix + root user). If you do not use a user with this ID then the + system will be unable to change the user when executing mysqld + as set using --user option. If this happens, mysqld may be + unable to read the files located within the MySQL data + directory and the execution will fail. -2.13.1. Linux Notes + 2. Enter the PASE environment using call qp2term. - This section discusses issues that have been found to occur on - Linux. The first few subsections describe general operating - system-related issues, problems that can occur when using binary - or source distributions, and post-installation issues. The - remaining subsections discuss problems that occur with Linux on - specific platforms. + 3. Start the MySQL server by changing to the installation + directory and running mysqld_safe, specifying the user name + used to install the server. The installer conveniently + installs a symbolic link to the installation directory + (mysql-5.0.42-i5os-power-32bit) as /opt/mysql/mysql: +> cd /opt/mysql/mysql +> bin/mysqld_safe --user=mysql & + You should see a message similar to the following: +Starting mysqld daemon with databases » + from /opt/mysql/mysql-enterprise-5.0.42-i5os-power-32bit/data - Note that most of these problems occur on older versions of Linux. - If you are running a recent version, you may see none of them. + If you are having problems starting MySQL server, see Section + 2.13.1.3, "Starting and Troubleshooting the MySQL Server." -2.13.1.1. Linux Operating System Notes + To stop MySQL: - MySQL needs at least Linux version 2.0. + 1. Log into the System i server using the user profile create or + specified during installation. By default, this is MYSQL. -Warning + 2. Enter the PASE environment using call qp2term. - We have seen some strange problems with Linux 2.2.14 and MySQL on - SMP systems. We also have reports from some MySQL users that they - have encountered serious stability problems using MySQL with - kernel 2.2.14. If you are using this kernel, you should upgrade to - 2.2.19 (or newer) or to a 2.4 kernel. If you have a multiple-CPU - box, you should seriously consider using 2.4 because it gives you - a significant speed boost. Your system should be more stable. - - When using LinuxThreads, you should see a minimum of three mysqld - processes running. These are in fact threads. There is one thread - for the LinuxThreads manager, one thread to handle connections, - and one thread to handle alarms and signals. - -2.13.1.2. Linux Binary Distribution Notes - - The Linux-Intel binary and RPM releases of MySQL are configured - for the highest possible speed. We are always trying to use the - fastest stable compiler available. - - The binary release is linked with -static, which means you do not - normally need to worry about which version of the system libraries - you have. You need not install LinuxThreads, either. A program - linked with -static is slightly larger than a dynamically linked - program, but also slightly faster (3-5%). However, one problem - with a statically linked program is that you can't use - user-defined functions (UDFs). If you are going to write or use - UDFs (this is something for C or C++ programmers only), you must - compile MySQL yourself using dynamic linking. - - A known issue with binary distributions is that on older Linux - systems that use libc (such as Red Hat 4.x or Slackware), you get - some (nonfatal) issues with host name resolution. If your system - uses libc rather than glibc2, you probably will encounter some - difficulties with host name resolution and getpwnam(). This - happens because glibc (unfortunately) depends on some external - libraries to implement host name resolution and getpwent(), even - when compiled with -static. These problems manifest themselves in - two ways: - - * You may see the following error message when you run - mysql_install_db: -Sorry, the host 'xxxx' could not be looked up - You can deal with this by executing mysql_install_db --force, - which does not execute the resolveip test in mysql_install_db. - The downside is that you cannot use host names in the grant - tables: except for localhost, you must use IP numbers instead. - If you are using an old version of MySQL that does not support - --force, you must manually remove the resolveip test in - mysql_install_db using a text editor. - - * You also may see the following error when you try to run - mysqld with the --user option: -getpwnam: No such file or directory - To work around this problem, start mysqld by using the su - command rather than by specifying the --user option. This - causes the system itself to change the user ID of the mysqld - process so that mysqld need not do so. - - Another solution, which solves both problems, is not to use a - binary distribution. Obtain a MySQL source distribution (in RPM or - tar.gz format) and install that instead. - - On some Linux 2.2 versions, you may get the error Resource - temporarily unavailable when clients make a great many new - connections to a mysqld server over TCP/IP. The problem is that - Linux has a delay between the time that you close a TCP/IP socket - and the time that the system actually frees it. There is room for - only a finite number of TCP/IP slots, so you encounter the - resource-unavailable error if clients attempt too many new TCP/IP - connections over a short period of time. For example, you may see - the error when you run the MySQL test-connect benchmark over - TCP/IP. - - We have inquired about this problem a few times on different Linux - mailing lists but have never been able to find a suitable - resolution. The only known "fix" is for clients to use persistent - connections, or, if you are running the database server and - clients on the same machine, to use Unix socket file connections - rather than TCP/IP connections. - -2.13.1.3. Linux Source Distribution Notes - - The following notes regarding glibc apply only to the situation - when you build MySQL yourself. If you are running Linux on an x86 - machine, in most cases it is much better for you to use our - binary. We link our binaries against the best patched version of - glibc we can find and with the best compiler options, in an - attempt to make it suitable for a high-load server. For a typical - user, even for setups with a lot of concurrent connections or - tables exceeding the 2GB limit, our binary is the best choice in - most cases. After reading the following text, if you are in doubt - about what to do, try our binary first to determine whether it - meets your needs. If you discover that it is not good enough, you - may want to try your own build. In that case, we would appreciate - a note about it so that we can build a better binary next time. - - MySQL uses LinuxThreads on Linux. If you are using an old Linux - version that doesn't have glibc2, you must install LinuxThreads - before trying to compile MySQL. You can obtain LinuxThreads from - http://dev.mysql.com/downloads/os-linux.html. - - Note that glibc versions before and including version 2.1.1 have a - fatal bug in pthread_mutex_timedwait() handling, which is used - when INSERT DELAYED statements are issued. Do not use INSERT - DELAYED before upgrading glibc. - - Note that Linux kernel and the LinuxThread library can by default - handle a maximum of 1,024 threads. If you plan to have more than - 1,000 concurrent connections, you need to make some changes to - LinuxThreads, as follows: - - * Increase PTHREAD_THREADS_MAX in - sysdeps/unix/sysv/linux/bits/local_lim.h to 4096 and decrease - STACK_SIZE in linuxthreads/internals.h to 256KB. The paths are - relative to the root of glibc. (Note that MySQL is not stable - with 600-1000 connections if STACK_SIZE is the default of - 2MB.) - - * Recompile LinuxThreads to produce a new libpthread.a library, - and relink MySQL against it. - - There is another issue that greatly hurts MySQL performance, - especially on SMP systems. The mutex implementation in - LinuxThreads in glibc 2.1 is very poor for programs with many - threads that hold the mutex only for a short time. This produces a - paradoxical result: If you link MySQL against an unmodified - LinuxThreads, removing processors from an SMP actually improves - MySQL performance in many cases. We have made a patch available - for glibc 2.1.3 to correct this behavior - (http://dev.mysql.com/Downloads/Linux/linuxthreads-2.1-patch) - - With glibc 2.2.2, MySQL uses the adaptive mutex, which is much - better than even the patched one in glibc 2.1.3. Be warned, - however, that under some conditions, the current mutex code in - glibc 2.2.2 overspins, which hurts MySQL performance. The - likelihood that this condition occurs can be reduced by re-nicing - the mysqld process to the highest priority. We have also been able - to correct the overspin behavior with a patch, available at - http://dev.mysql.com/Downloads/Linux/linuxthreads-2.2.2.patch. It - combines the correction of overspin, maximum number of threads, - and stack spacing all in one. You need to apply it in the - linuxthreads directory with patch -p0 - </tmp/linuxthreads-2.2.2.patch. We hope it is included in some - form in future releases of glibc 2.2. In any case, if you link - against glibc 2.2.2, you still need to correct STACK_SIZE and - PTHREAD_THREADS_MAX. We hope that the defaults is corrected to - some more acceptable values for high-load MySQL setup in the - future, so that the commands needed to produce your own build can - be reduced to ./configure; make; make install. - - If you use these patches to build a special static version of - libpthread.a, use it only for statically linking against MySQL. We - know that these patches are safe for MySQL and significantly - improve its performance, but we cannot say anything about their - effects on other applications. If you link other applications that - require LinuxThreads against the patched static version of the - library, or build a patched shared version and install it on your - system, you do so at your own risk. - - If you experience any strange problems during the installation of - MySQL, or with some common utilities hanging, it is very likely - that they are either library or compiler related. If this is the - case, using our binary resolves them. + 3. Stop the MySQL server by changing into the installation + directory and running mysqladmin, specifying the user name + used to install the server: +> cd /opt/mysql/mysql +> bin/mysqladmin -u root shutdown + If the session that you started and stopped MySQL are the + same, you may get the log output from mysqld: + STOPPING server from pid file » + /opt/mysql/mysql-enterprise-5.0.42-i5os-power-32bit/data/I5DBX.R +CHLAND.IBM.COM.pid + 070718 10:34:20 mysqld ended + If the sessions used to start and stop MySQL are different, + you will not receive any confirmation of the shutdown. - If you link your own MySQL client programs, you may see the - following error at runtime: -ld.so.1: fatal: libmysqlclient.so.#: -open failed: No such file or directory + Note and tips - This problem can be avoided by one of the following methods: + * A problem has been identified with the installation process on + DBCS systems. If you are having problems install MySQL on a + DBCS system, you need to change your job's coded character set + identifier (CSSID) to 37 (EBCDIC) before executing the install + command, INSMYSQL. To do this, determine your existing CSSID + (using DSPJOB and selecting option 2), execute CHGJOB + CSSID(37), run INSMYSQL to install MySQL and then execute + CHGJOB again with your original CSSID. - * Link clients with the -Wl,r/full/path/to/libmysqlclient.so - flag rather than with -Lpath). + * If you want to use the Perl scripts that are included with + MySQL, you need to download the iSeries Tools for Developers + (5799-PTL). See + http://www-03.ibm.com/servers/enable/site/porting/tools/. - * Copy libmysqclient.so to /usr/lib. +2.10. Installing MySQL on FreeBSD - * Add the path name of the directory where libmysqlclient.so is - located to the LD_RUN_PATH environment variable before running - your client. + This section provides information about using MySQL on variants of + FreeBSD Unix. - If you are using the Fujitsu compiler (fcc/FCC), you may have some - problems compiling MySQL because the Linux header files are very - gcc oriented. The following configure line should work with - fcc/FCC: -CC=fcc CFLAGS="-O -K fast -K lib -K omitfp -Kpreex -D_GNU_SOURCE \ - -DCONST=const -DNO_STRTOLL_PROTO" \ -CXX=FCC CXXFLAGS="-O -K fast -K lib \ - -K omitfp -K preex --no_exceptions --no_rtti -D_GNU_SOURCE \ - -DCONST=const -Dalloca=__builtin_alloca -DNO_STRTOLL_PROTO \ - '-D_EXTERN_INLINE=static __inline'" \ -./configure \ - --prefix=/usr/local/mysql --enable-assembler \ - --with-mysqld-ldflags=-all-static --disable-shared \ - --with-low-memory + The easiest (and preferred) way to install MySQL is to use the + mysql-server and mysql-client ports available at + http://www.freebsd.org/. Using these ports gives you the following + benefits: -2.13.1.4. Linux Post-Installation Notes + * A working MySQL with all optimizations enabled that are known + to work on your version of FreeBSD. - mysql.server can be found in the support-files directory under the - MySQL installation directory or in a MySQL source tree. You can - install it as /etc/init.d/mysql for automatic MySQL startup and - shutdown. See Section 2.11.2.2, "Starting and Stopping MySQL - Automatically." + * Automatic configuration and build. - If MySQL cannot open enough files or connections, it may be that - you have not configured Linux to handle enough files. + * Startup scripts installed in /usr/local/etc/rc.d. - In Linux 2.2 and onward, you can check the number of allocated - file handles as follows: -shell> cat /proc/sys/fs/file-max -shell> cat /proc/sys/fs/dquot-max -shell> cat /proc/sys/fs/super-max - - If you have more than 16MB of memory, you should add something - like the following to your init scripts (for example, - /etc/init.d/boot.local on SuSE Linux): -echo 65536 > /proc/sys/fs/file-max -echo 8192 > /proc/sys/fs/dquot-max -echo 1024 > /proc/sys/fs/super-max - - You can also run the echo commands from the command line as root, - but these settings are lost the next time your computer restarts. - - Alternatively, you can set these parameters on startup by using - the sysctl tool, which is used by many Linux distributions - (including SuSE Linux 8.0 and later). Put the following values - into a file named /etc/sysctl.conf: -# Increase some values for MySQL -fs.file-max = 65536 -fs.dquot-max = 8192 -fs.super-max = 1024 - - You should also add the following to /etc/my.cnf: -[mysqld_safe] -open-files-limit=8192 - - This should allow the server a limit of 8,192 for the combined - number of connections and open files. - - The STACK_SIZE constant in LinuxThreads controls the spacing of - thread stacks in the address space. It needs to be large enough so - that there is plenty of room for each individual thread stack, but - small enough to keep the stack of some threads from running into - the global mysqld data. Unfortunately, as we have experimentally - discovered, the Linux implementation of mmap() successfully unmaps - a mapped region if you ask it to map out an address currently in - use, zeroing out the data on the entire page instead of returning - an error. So, the safety of mysqld or any other threaded - application depends on the "gentlemanly" behavior of the code that - creates threads. The user must take measures to make sure that the - number of running threads at any given time is sufficiently low - for thread stacks to stay away from the global heap. With mysqld, - you should enforce this behavior by setting a reasonable value for - the max_connections variable. - - If you build MySQL yourself, you can patch LinuxThreads for better - stack use. See Section 2.13.1.3, "Linux Source Distribution - Notes." If you do not want to patch LinuxThreads, you should set - max_connections to a value no higher than 500. It should be even - less if you have a large key buffer, large heap tables, or some - other things that make mysqld allocate a lot of memory, or if you - are running a 2.2 kernel with a 2GB patch. If you are using our - binary or RPM version, you can safely set max_connections at 1500, - assuming no large key buffer or heap tables with lots of data. The - more you reduce STACK_SIZE in LinuxThreads the more threads you - can safely create. Values between 128KB and 256KB are recommended. - - If you use a lot of concurrent connections, you may suffer from a - "feature" in the 2.2 kernel that attempts to prevent fork bomb - attacks by penalizing a process for forking or cloning a child. - This causes MySQL not to scale well as you increase the number of - concurrent clients. On single-CPU systems, we have seen this - manifest as very slow thread creation; it may take a long time to - connect to MySQL (as long as one minute), and it may take just as - long to shut it down. On multiple-CPU systems, we have observed a - gradual drop in query speed as the number of clients increases. In - the process of trying to find a solution, we have received a - kernel patch from one of our users who claimed it helped for his - site. This patch is available at - http://dev.mysql.com/Downloads/Patches/linux-fork.patch. We have - done rather extensive testing of this patch on both development - and production systems. It has significantly improved MySQL - performance without causing any problems and is recommended for - users who still run high-load servers on 2.2 kernels. - - This issue has been fixed in the 2.4 kernel, so if you are not - satisfied with the current performance of your system, rather than - patching your 2.2 kernel, it might be easier to upgrade to 2.4. On - SMP systems, upgrading also gives you a nice SMP boost in addition - to fixing the fairness bug. - - We have tested MySQL on the 2.4 kernel on a two-CPU machine and - found MySQL scales much better. There was virtually no slowdown on - query throughput all the way up to 1,000 clients, and the MySQL - scaling factor (computed as the ratio of maximum throughput to the - throughput for one client) was 180%. We have observed similar - results on a four-CPU system: Virtually no slowdown as the number - of clients was increased up to 1,000, and a 300% scaling factor. - Based on these results, for a high-load SMP server using a 2.2 - kernel, it is definitely recommended to upgrade to the 2.4 kernel - at this point. - - We have discovered that it is essential to run the mysqld process - with the highest possible priority on the 2.4 kernel to achieve - maximum performance. This can be done by adding a renice -20 $$ - command to mysqld_safe. In our testing on a four-CPU machine, - increasing the priority resulted in a 60% throughput increase with - 400 clients. - - We are currently also trying to collect more information on how - well MySQL performs with a 2.4 kernel on four-way and eight-way - systems. If you have access such a system and have done some - benchmarks, please send an email message to benchmarks(a)mysql.com - with the results. We will review them for inclusion in the manual. - - If you see a dead mysqld server process with ps, this usually - means that you have found a bug in MySQL or you have a corrupted - table. See Section B.1.4.2, "What to Do If MySQL Keeps Crashing." - - To get a core dump on Linux if mysqld dies with a SIGSEGV signal, - you can start mysqld with the --core-file option. Note that you - also probably need to raise the core file size by adding ulimit -c - 1000000 to mysqld_safe or starting mysqld_safe with - --core-file-size=1000000. See Section 4.3.2, "mysqld_safe --- - MySQL Server Startup Script." - -2.13.1.5. Linux x86 Notes - - MySQL requires libc 5.4.12 or newer. It is known to work with libc - 5.4.46. glibc 2.0.6 and later should also work. There have been - some problems with the glibc RPMs from Red Hat, so if you have - problems, check whether there are any updates. The glibc 2.0.7-19 - and 2.0.7-29 RPMs are known to work. - - If you are using Red Hat 8.0 or a new glibc 2.2.x library, you may - see mysqld die in gethostbyaddr(). This happens because the new - glibc library requires a stack size greater than 128KB for this - call. To fix the problem, start mysqld with the - --thread-stack=192K option. (Use -O thread_stack=192K before MySQL - 4.) This stack size is the default on MySQL 4.0.10 and above, so - you should not see the problem. - - If you are using gcc 3.0 and above to compile MySQL, you must - install the libstdc++v3 library before compiling MySQL; if you - don't do this, you get an error about a missing __cxa_pure_virtual - symbol during linking. - - On some older Linux distributions, configure may produce an error - like this: -Syntax error in sched.h. Change _P to __P in the -/usr/include/sched.h file. -See the Installation chapter in the Reference Manual. - - Just do what the error message says. Add an extra underscore to - the _P macro name that has only one underscore, and then try - again. - - You may get some warnings when compiling. Those shown here can be - ignored: -mysqld.cc -o objs-thread/mysqld.o -mysqld.cc: In function `void init_signals()': -mysqld.cc:315: warning: assignment of negative value `-1' to -`long unsigned int' -mysqld.cc: In function `void * signal_hand(void *)': -mysqld.cc:346: warning: assignment of negative value `-1' to -`long unsigned int' - - If mysqld always dumps core when it starts, the problem may be - that you have an old /lib/libc.a. Try renaming it, and then remove - sql/mysqld and do a new make install and try again. This problem - has been reported on some Slackware installations. - - If you get the following error when linking mysqld, it means that - your libg++.a is not installed correctly: -/usr/lib/libc.a(putc.o): In function `_IO_putc': -putc.o(.text+0x0): multiple definition of `_IO_putc' - - You can avoid using libg++.a by running configure like this: -shell> CXX=gcc ./configure - -2.13.1.6. Linux SPARC Notes - - In some implementations, readdir_r() is broken. The symptom is - that the SHOW DATABASES statement always returns an empty set. - This can be fixed by removing HAVE_READDIR_R from config.h after - configuring and before compiling. - -2.13.1.7. Linux Alpha Notes - - We have tested MySQL 5.1 on Alpha with our benchmarks and test - suite, and it appears to work well. - - We currently build the MySQL binary packages on SuSE Linux 7.0 for - AXP, kernel 2.4.4-SMP, Compaq C compiler (V6.2-505) and Compaq C++ - compiler (V6.3-006) on a Compaq DS20 machine with an Alpha EV6 - processor. + * The ability to use pkg_info -L to see which files are + installed. - You can find the preceding compilers at - http://www.support.compaq.com/alpha-tools/. By using these - compilers rather than gcc, we get about 9-14% better MySQL - performance. - - For MySQL on Alpha, we use the -arch generic flag to our compile - options, which ensures that the binary runs on all Alpha - processors. We also compile statically to avoid library problems. - The configure command looks like this: -CC=ccc CFLAGS="-fast -arch generic" CXX=cxx \ -CXXFLAGS="-fast -arch generic -noexceptions -nortti" \ -./configure --prefix=/usr/local/mysql --disable-shared \ - --with-extra-charsets=complex --enable-thread-safe-client \ - --with-mysqld-ldflags=-non_shared --with-client-ldflags=-non_shar -ed + * The ability to use pkg_delete to remove MySQL if you no longer + want it on your machine. - Some known problems when running MySQL on Linux-Alpha: + The MySQL build process requires GNU make (gmake) to work. If GNU + make is not available, you must install it first before compiling + MySQL. - * Debugging threaded applications like MySQL does not work with - gdb 4.18. You should use gdb 5.1 instead. + The recommended way to compile and install MySQL on FreeBSD with + gcc (2.95.2 and up) is: +CC=gcc CFLAGS="-O2 -fno-strength-reduce" \ + CXX=gcc CXXFLAGS="-O2 -fno-rtti -fno-exceptions \ + -felide-constructors -fno-strength-reduce" \ + ./configure --prefix=/usr/local/mysql --enable-assembler +gmake +gmake install +cd /usr/local/mysql +bin/mysql_install_db --user=mysql +bin/mysqld_safe & - * If you try linking mysqld statically when using gcc, the - resulting image dumps core at startup time. In other words, do - not use --with-mysqld-ldflags=-all-static with gcc. - -2.13.1.8. Linux PowerPC Notes - - MySQL should work on MkLinux with the newest glibc package (tested - with glibc 2.0.7). - -2.13.1.9. Linux MIPS Notes - - To get MySQL to work on Qube2 (Linux Mips), you need the newest - glibc libraries. glibc-2.0.7-29C2 is known to work. You must also - use gcc 2.95.2 or newer). - -2.13.1.10. Linux IA-64 Notes - - To get MySQL to compile on Linux IA-64, we use the following - configure command for building with gcc 2.96: -CC=gcc \ -CFLAGS="-O3 -fno-omit-frame-pointer" \ -CXX=gcc \ -CXXFLAGS="-O3 -fno-omit-frame-pointer -felide-constructors \ - -fno-exceptions -fno-rtti" \ - ./configure --prefix=/usr/local/mysql \ - "--with-comment=Official MySQL binary" \ - --with-extra-charsets=complex - - On IA-64, the MySQL client binaries use shared libraries. This - means that if you install our binary distribution at a location - other than /usr/local/mysql, you need to add the path of the - directory where you have libmysqlclient.so installed either to the - /etc/ld.so.conf file or to the value of your LD_LIBRARY_PATH - environment variable. - - See Section B.1.3.1, "Problems Linking to the MySQL Client - Library." - -2.13.1.11. SELinux Notes - - RHEL4 comes with SELinux, which supports tighter access control - for processes. If SELinux is enabled (SELINUX in - /etc/selinux/config is set to enforcing, SELINUXTYPE is set to - either targeted or strict), you might encounter problems - installing Sun Microsystems, Inc. RPM packages. - - Red Hat has an update that solves this. It involves an update of - the "security policy" specification to handle the install - structure of the RPMs provided by Sun Microsystems, Inc. For - further information, see - https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=167551 and - http://rhn.redhat.com/errata/RHBA-2006-0049.html. - - The preceding discussion applies only to RHEL4. The patch is - unnecessary for RHEL5. - -2.13.2. Mac OS X Notes - - On Mac OS X, tar cannot handle long file names. If you need to - unpack a .tar.gz distribution, use gnutar instead. - -2.13.2.1. Mac OS X 10.x (Darwin) - - MySQL should work without major problems on Mac OS X 10.x - (Darwin). - - Known issues: - - * If you have problems with performance under heavy load, try - using the --skip-thread-priority option to mysqld. This runs - all threads with the same priority. On Mac OS X, this gives - better performance, at least until Apple fixes its thread - scheduler. - - * The connection times (wait_timeout, interactive_timeout and - net_read_timeout) values are not honored. - This is probably a signal handling problem in the thread - library where the signal doesn't break a pending read and we - hope that a future update to the thread libraries will fix - this. - - Our binary for Mac OS X is compiled on Darwin 6.3 with the - following configure line: -CC=gcc CFLAGS="-O3 -fno-omit-frame-pointer" CXX=gcc \ -CXXFLAGS="-O3 -fno-omit-frame-pointer -felide-constructors \ - -fno-exceptions -fno-rtti" \ - ./configure --prefix=/usr/local/mysql \ - --with-extra-charsets=complex --enable-thread-safe-client \ - --enable-local-infile --disable-shared - - See Section 2.5, "Installing MySQL on Mac OS X." - -2.13.2.2. Mac OS X Server 1.2 (Rhapsody) - - For current versions of Mac OS X Server, no operating system - changes are necessary before compiling MySQL. Compiling for the - Server platform is the same as for the client version of Mac OS X. + FreeBSD is known to have a very low default file handle limit. See + Section B.5.2.18, "'File' Not Found and Similar Errors." Start the + server by using the --open-files-limit option for mysqld_safe, or + raise the limits for the mysqld user in /etc/login.conf and + rebuild it with cap_mkdb /etc/login.conf. Also be sure that you + set the appropriate class for this user in the password file if + you are not using the default (use chpass mysqld-user-name). See + Section 4.3.2, "mysqld_safe --- MySQL Server Startup Script." - For older versions (Mac OS X Server 1.2, a.k.a. Rhapsody), you - must first install a pthread package before trying to configure - MySQL. + In current versions of FreeBSD (at least 4.x and greater), you may + increase the limit on the amount of memory available for a process + by adding the following entries to the /boot/loader.conf file and + rebooting the machine (these are not settings that can be changed + at run time with the sysctl command): +kern.maxdsiz="1073741824" # 1GB +kern.dfldsiz="1073741824" # 1GB +kern.maxssiz="134217728" # 128MB - See Section 2.5, "Installing MySQL on Mac OS X." + For older versions of FreeBSD, you must recompile your kernel to + change the maximum data segment size for a process. In this case, + you should look at the MAXDSIZ option in the LINT config file for + more information. -2.13.3. Solaris Notes + If you get problems with the current date in MySQL, setting the TZ + variable should help. See Section 2.14, "Environment Variables." - For information about installing MySQL on Solaris using PKG - distributions, see Section 2.6, "Installing MySQL on Solaris." +2.11. Installing MySQL on HP-UX - On Solaris, you may run into trouble even before you get the MySQL - distribution unpacked, as the Solaris tar cannot handle long file + If you install MySQL using a binary tarball distribution on HP-UX, + you may run into trouble even before you get the MySQL + distribution unpacked, as the HP-UX tar cannot handle long file names. This means that you may see errors when you try to unpack MySQL. If this occurs, you must use GNU tar (gtar) to unpack the distribution. - Sun native threads work only on Solaris 2.5 and higher. For - Solaris 2.4 and earlier, MySQL automatically uses MIT-pthreads. - See Section 2.10.5, "MIT-pthreads Notes." - - If you get the following error from configure, it means that you - have something wrong with your compiler installation: -checking for restartable system calls... configure: error can not -run test programs while cross compiling - - In this case, you should upgrade your compiler to a newer version. - You may also be able to solve this problem by inserting the - following row into the config.cache file: -ac_cv_sys_restartable_syscalls=${ac_cv_sys_restartable_syscalls='no'} - - If you are using Solaris on a SPARC, the recommended compiler is - gcc 2.95.2 or 3.2. You can find this at http://gcc.gnu.org/. Note - that gcc 2.8.1 does not work reliably on SPARC. - - The recommended configure line when using gcc 2.95.2 is: -CC=gcc CFLAGS="-O3" \ -CXX=gcc CXXFLAGS="-O3 -felide-constructors -fno-exceptions -fno-rtti" - \ -./configure --prefix=/usr/local/mysql --with-low-memory \ - --enable-assembler + Because of some critical bugs in the standard HP-UX libraries, you + should install the following patches before trying to run MySQL on + HP-UX 11.0: +PHKL_22840 Streams cumulative +PHNE_22397 ARPA cumulative - If you have an UltraSPARC system, you can get 4% better - performance by adding -mcpu=v8 -Wa,-xarch=v8plusa to the CFLAGS - and CXXFLAGS environment variables. + This solves the problem of getting EWOULDBLOCK from recv() and + EBADF from accept() in threaded applications. - If you have Sun's Forte 5.0 (or newer) compiler, you can run - configure like this: -CC=cc CFLAGS="-Xa -fast -native -xstrconst -mt" \ -CXX=CC CXXFLAGS="-noex -mt" \ -./configure --prefix=/usr/local/mysql --enable-assembler + If you are using gcc 2.95.1 on an unpatched HP-UX 11.x system, you + may get the following error: +In file included from /usr/include/unistd.h:11, + from ../include/global.h:125, + from mysql_priv.h:15, + from item.cc:19: +/usr/include/sys/unistd.h:184: declaration of C function ... +/usr/include/sys/pthread.h:440: previous declaration ... +In file included from item.h:306, + from mysql_priv.h:158, + from item.cc:19: - To create a 64-bit binary with Sun's Forte compiler, use the - following configuration options: -CC=cc CFLAGS="-Xa -fast -native -xstrconst -mt -xarch=v9" \ -CXX=CC CXXFLAGS="-noex -mt -xarch=v9" ASFLAGS="-xarch=v9" \ -./configure --prefix=/usr/local/mysql --enable-assembler + The problem is that HP-UX does not define pthreads_atfork() + consistently. It has conflicting prototypes in + /usr/include/sys/unistd.h:184 and /usr/include/sys/pthread.h:440. - To create a 64-bit Solaris binary using gcc, add -m64 to CFLAGS - and CXXFLAGS and remove --enable-assembler from the configure - line. + One solution is to copy /usr/include/sys/unistd.h into + mysql/include and edit unistd.h and change it to match the + definition in pthread.h. Look for this line: +extern int pthread_atfork(void (*prepare)(), void (*parent)(), + void (*child)()); - In the MySQL benchmarks, we obtained a 4% speed increase on - UltraSPARC when using Forte 5.0 in 32-bit mode, as compared to - using gcc 3.2 with the -mcpu flag. + Change it to look like this: +extern int pthread_atfork(void (*prepare)(void), void (*parent)(void) +, + void (*child)(void)); - If you create a 64-bit mysqld binary, it is 4% slower than the - 32-bit binary, but can handle more threads and memory. + After making the change, the following configure line should work: +CFLAGS="-fomit-frame-pointer -O3 -fpic" CXX=gcc \ +CXXFLAGS="-felide-constructors -fno-exceptions -fno-rtti -O3" \ +./configure --prefix=/usr/local/mysql --disable-shared - When using Solaris 10 for x86_64, you should mount any file - systems on which you intend to store InnoDB files with the - forcedirectio option. (By default mounting is done without this - option.) Failing to do so will cause a significant drop in - performance when using the InnoDB storage engine on this platform. + If you are using HP-UX compiler, you can use the following command + (which has been tested with cc B.11.11.04): +CC=cc CXX=aCC CFLAGS=+DD64 CXXFLAGS=+DD64 ./configure \ + --with-extra-character-set=complex - If you get a problem with fdatasync or sched_yield, you can fix - this by adding LIBS=-lrt to the configure line + You can ignore any errors of the following type: +aCC: warning 901: unknown option: `-3': use +help for online +documentation - For compilers older than WorkShop 5.3, you might have to edit the - configure script. Change this line: -#if !defined(__STDC__) || __STDC__ != 1 - - To this: -#if !defined(__STDC__) - - If you turn on __STDC__ with the -Xc option, the Sun compiler - can't compile with the Solaris pthread.h header file. This is a - Sun bug (broken compiler or broken include file). - - If mysqld issues the following error message when you run it, you - have tried to compile MySQL with the Sun compiler without enabling - the -mt multi-thread option: -libc internal error: _rmutex_unlock: rmutex not held - - Add -mt to CFLAGS and CXXFLAGS and recompile. - - If you are using the SFW version of gcc (which comes with Solaris - 8), you must add /opt/sfw/lib to the environment variable - LD_LIBRARY_PATH before running configure. - - If you are using the gcc available from sunfreeware.com, you may - have many problems. To avoid this, you should recompile gcc and - GNU binutils on the machine where you are running them. - - If you get the following error when compiling MySQL with gcc, it - means that your gcc is not configured for your version of Solaris: -shell> gcc -O3 -g -O2 -DDBUG_OFF -o thr_alarm ... -./thr_alarm.c: In function `signal_hand': -./thr_alarm.c:556: too many arguments to function `sigwait' - - The proper thing to do in this case is to get the newest version - of gcc and compile it with your current gcc compiler. At least for - Solaris 2.5, almost all binary versions of gcc have old, unusable - include files that break all programs that use threads, and - possibly other programs as well. + If you get the following error from configure, verify that you + don't have the path to the K&R compiler before the path to the + HP-UX C and C++ compiler: +checking for cc option to accept ANSI C... no +configure: error: MySQL requires an ANSI C compiler (and a C++ compil +er). +Try gcc. See the Installation chapter in the Reference Manual. - Solaris does not provide static versions of all system libraries - (libpthreads and libdl), so you cannot compile MySQL with - --static. If you try to do so, you get one of the following - errors: -ld: fatal: library -ldl: not found -undefined reference to `dlopen' -cannot find -lrt + Another reason for not being able to compile is that you didn't + define the +DD64 flags as just described. - If you link your own MySQL client programs, you may see the - following error at runtime: -ld.so.1: fatal: libmysqlclient.so.#: -open failed: No such file or directory + Another possibility for HP-UX 11 is to use the MySQL binaries + provided at http://dev.mysql.com/downloads/, which we have built + and tested ourselves. We have also received reports that the HP-UX + 10.20 binaries supplied by MySQL can be run successfully on HP-UX + 11. If you encounter problems, you should be sure to check your + HP-UX patch level. - This problem can be avoided by one of the following methods: +2.12. Installing MySQL on AIX - * Link clients with the -Wl,r/full/path/to/libmysqlclient.so - flag rather than with -Lpath). + Automatic detection of xlC is missing from Autoconf, so a number + of variables need to be set before running configure. The + following example uses the IBM compiler: +export CC="xlc_r -ma -O3 -qstrict -qoptimize=3 -qmaxmem=8192 " +export CXX="xlC_r -ma -O3 -qstrict -qoptimize=3 -qmaxmem=8192" +export CFLAGS="-I /usr/local/include" +export LDFLAGS="-L /usr/local/lib" +export CPPFLAGS=$CFLAGS +export CXXFLAGS=$CFLAGS - * Copy libmysqclient.so to /usr/lib. +./configure --prefix=/usr/local \ + --localstatedir=/var/mysql \ + --sbindir='/usr/local/bin' \ + --libexecdir='/usr/local/bin' \ + --enable-thread-safe-client \ + --enable-large-files - * Add the path name of the directory where libmysqlclient.so is - located to the LD_RUN_PATH environment variable before running - your client. + The preceding options are used to compile the MySQL distribution + that can be found at http://www-frec.bull.com/. - If you have problems with configure trying to link with -lz when - you don't have zlib installed, you have two options: + If you change the -O3 to -O2 in the preceding configure line, you + must also remove the -qstrict option. This is a limitation in the + IBM C compiler. - * If you want to be able to use the compressed communication - protocol, you need to get and install zlib from ftp.gnu.org. + If you are using gcc to compile MySQL, you must use the + -fno-exceptions flag, because the exception handling in gcc is not + thread-safe! There are also some known problems with IBM's + assembler that may cause it to generate bad code when used with + gcc. - * Run configure with the --with-named-z-libs=no option when - building MySQL. + Use the following configure line with gcc 2.95 on AIX: +CC="gcc -pipe -mcpu=power -Wa,-many" \ +CXX="gcc -pipe -mcpu=power -Wa,-many" \ +CXXFLAGS="-felide-constructors -fno-exceptions -fno-rtti" \ +./configure --prefix=/usr/local/mysql --with-low-memory - If you are using gcc and have problems with loading user-defined - functions (UDFs) into MySQL, try adding -lgcc to the link line for - the UDF. + The -Wa,-many option is necessary for the compile to be + successful. IBM is aware of this problem but is in no hurry to fix + it because of the workaround that is available. We don't know if + the -fno-exceptions is required with gcc 2.95, but because MySQL + doesn't use exceptions and the option generates faster code, you + should always use it with gcc. - If you would like MySQL to start automatically, you can copy - support-files/mysql.server to /etc/init.d and create a symbolic - link to it named /etc/rc3.d/S99mysql.server. + If you get a problem with assembler code, try changing the + -mcpu=xxx option to match your CPU. Typically power2, power, or + powerpc may need to be used. Alternatively, you might need to use + 604 or 604e. We are not positive but suspect that power would + likely be safe most of the time, even on a power2 machine. - If too many processes try to connect very rapidly to mysqld, you - should see this error in the MySQL log: -Error in accept: Protocol error + If you don't know what your CPU is, execute a uname -m command. It + produces a string that looks like 000514676700, with a format of + xxyyyyyymmss where xx and ss are always 00, yyyyyy is a unique + system ID and mm is the ID of the CPU Planar. A chart of these + values can be found at + http://www16.boulder.ibm.com/pseries/en_US/cmds/aixcmds5/uname.htm + . - You might try starting the server with the --back_log=50 option as - a workaround for this. (Use -O back_log=50 before MySQL 4.) + This gives you a machine type and a machine model you can use to + determine what type of CPU you have. - Solaris doesn't support core files for setuid() applications, so - you can't get a core file from mysqld if you are using the --user - option. + If you have problems with threads on AIX 5.3, you should upgrade + AIX 5.3 to technology level 7 (5300-07). + + If you have problems with signals (MySQL dies unexpectedly under + high load), you may have found an OS bug with threads and signals. + In this case, you can tell MySQL not to use signals by configuring + as follows: +CFLAGS=-DDONT_USE_THR_ALARM CXX=gcc \ +CXXFLAGS="-felide-constructors -fno-exceptions -fno-rtti \ +-DDONT_USE_THR_ALARM" \ +./configure --prefix=/usr/local/mysql --with-debug \ + --with-low-memory + + This doesn't affect the performance of MySQL, but has the side + effect that you can't kill clients that are "sleeping" on a + connection with mysqladmin kill or mysqladmin shutdown. Instead, + the client dies when it issues its next command. + + On some versions of AIX, linking with libbind.a makes + getservbyname() dump core. This is an AIX bug and should be + reported to IBM. + + For AIX 4.2.1 and gcc, you have to make the following changes. -2.13.3.1. Solaris 2.7/2.8 Notes + After configuring, edit config.h and include/my_config.h and + change the line that says this: +#define HAVE_SNPRINTF 1 - Normally, you can use a Solaris 2.6 binary on Solaris 2.7 and 2.8. - Most of the Solaris 2.6 issues also apply for Solaris 2.7 and 2.8. + to this: +#undef HAVE_SNPRINTF - MySQL should be able to detect new versions of Solaris - automatically and enable workarounds for the following problems. + And finally, in mysqld.cc, you need to add a prototype for + initgroups(). +#ifdef _AIX41 +extern "C" int initgroups(const char *,int); +#endif - Solaris 2.7 / 2.8 has some bugs in the include files. You may see - the following error when you use gcc: -/usr/include/widec.h:42: warning: `getwc' redefined -/usr/include/wchar.h:326: warning: this is the location of the previo -us -definition + For 32-bit binaries, if you need to allocate a lot of memory to + the mysqld process, it is not enough to just use ulimit -d + unlimited. You may also have to modify mysqld_safe to add a line + something like this: +export LDR_CNTRL='MAXDATA=0x80000000' - If this occurs, you can fix the problem by copying - /usr/include/widec.h to .../lib/gcc-lib/os/gcc-version/include and - changing line 41 from this: -#if !defined(lint) && !defined(__lint) + You can find more information about using a lot of memory at + http://publib16.boulder.ibm.com/pseries/en_US/aixprggd/genprogc/lr + g_prg_support.htm. - To this: -#if !defined(lint) && !defined(__lint) && !defined(getwc) + Users of AIX 4.3 should use gmake instead of the make utility + included with AIX. - Alternatively, you can edit /usr/include/widec.h directly. Either - way, after you make the fix, you should remove config.cache and - run configure again. + As of AIX 4.1, the C compiler has been unbundled from AIX as a + separate product. gcc 3.3.2 can be obtained here: + ftp://ftp.software.ibm.com/aix/freeSoftware/aixtoolbox/RPMS/ppc/gc + c/ - If you get the following errors when you run make, it is because - configure didn't detect the curses.h file (probably because of the - error in /usr/include/widec.h): -In file included from mysql.cc:50: -/usr/include/term.h:1060: syntax error before `,' -/usr/include/term.h:1081: syntax error before `;' + The steps for compiling MySQL on AIX with gcc 3.3.2 are similar to + those for using gcc 2.95 (in particular, the need to edit config.h + and my_config.h after running configure). However, before running + configure, you should also patch the curses.h file as follows: +/opt/freeware/lib/gcc-lib/powerpc-ibm-aix5.2.0.0/3.3.2/include/curses +.h.ORIG + Mon Dec 26 02:17:28 2005 +--- /opt/freeware/lib/gcc-lib/powerpc-ibm-aix5.2.0.0/3.3.2/include/cu +rses.h +Mon Dec 26 02:40:13 2005 +*************** +*** 2023,2029 **** - The solution to this problem is to do one of the following: - 1. Configure with CFLAGS=-DHAVE_CURSES_H CXXFLAGS=-DHAVE_CURSES_H - ./configure. + #endif /* _AIX32_CURSES */ +! #if defined(__USE_FIXED_PROTOTYPES__) || defined(__cplusplus) || de +fined +(__STRICT_ANSI__) + extern int delwin (WINDOW *); + extern int endwin (void); + extern int getcurx (WINDOW *); +--- 2023,2029 ---- - 2. Edit /usr/include/widec.h as indicated in the preceding - discussion and re-run configure. - 3. Remove the #define HAVE_TERM line from the config.h file and - run make again. + #endif /* _AIX32_CURSES */ +! #if 0 && (defined(__USE_FIXED_PROTOTYPES__) || defined(__cplusplus) +|| defined +(__STRICT_ANSI__)) + extern int delwin (WINDOW *); + extern int endwin (void); + extern int getcurx (WINDOW *); - If your linker cannot find -lz when linking client programs, the - problem is probably that your libz.so file is installed in - /usr/local/lib. You can fix this problem by one of the following - methods: +2.13. Post-Installation Setup and Testing - * Add /usr/local/lib to LD_LIBRARY_PATH. + After installing MySQL, there are some issues that you should + address. For example, on Unix, you should initialize the data + directory and create the MySQL grant tables. On all platforms, an + important security concern is that the initial accounts in the + grant tables have no passwords. You should assign passwords to + prevent unauthorized access to the MySQL server. Optionally, you + can create time zone tables to enable recognition of named time + zones. - * Add a link to libz.so from /lib. + The following sections include post-installation procedures that + are specific to Windows systems and to Unix systems. Another + section, Section 2.13.1.3, "Starting and Troubleshooting the MySQL + Server," applies to all platforms; it describes what to do if you + have trouble getting the server to start. Section 2.13.2, + "Securing the Initial MySQL Accounts," also applies to all + platforms. You should follow its instructions to make sure that + you have properly protected your MySQL accounts by assigning + passwords to them. - * If you are using Solaris 8, you can install the optional zlib - from your Solaris 8 CD distribution. + When you are ready to create additional user accounts, you can + find information on the MySQL access control system and account + management in Section 5.4, "The MySQL Access Privilege System," + and Section 5.5, "MySQL User Account Management." - * Run configure with the --with-named-z-libs=no option when - building MySQL. +2.13.1. Unix Post-Installation Procedures -2.13.3.2. Solaris x86 Notes + After installing MySQL on Unix, you need to initialize the grant + tables, start the server, and make sure that the server works + satisfactorily. You may also wish to arrange for the server to be + started and stopped automatically when your system starts and + stops. You should also assign passwords to the accounts in the + grant tables. - On Solaris 8 on x86, mysqld dumps core if you remove the debug - symbols using strip. + On Unix, the grant tables are set up by the mysql_install_db + program. For some installation methods, this program is run for + you automatically: - If you are using gcc on Solaris x86 and you experience problems - with core dumps under load, you should use the following configure - command: -CC=gcc CFLAGS="-O3 -fomit-frame-pointer -DHAVE_CURSES_H" \ -CXX=gcc \ -CXXFLAGS="-O3 -fomit-frame-pointer -felide-constructors \ - -fno-exceptions -fno-rtti -DHAVE_CURSES_H" \ -./configure --prefix=/usr/local/mysql + * If you install MySQL on Linux using RPM distributions, the + server RPM runs mysql_install_db. - This avoids problems with the libstdc++ library and with C++ - exceptions. + * If you install MySQL on Mac OS X using a PKG distribution, the + installer runs mysql_install_db. - If this doesn't help, you should compile a debug version and run - it with a trace file or under gdb. See MySQL Internals: Porting - (http://forge.mysql.com/wiki/MySQL_Internals_Porting) + Otherwise, you'll need to run mysql_install_db yourself. -2.13.4. BSD Notes + The following procedure describes how to initialize the grant + tables (if that has not previously been done) and then start the + server. It also suggests some commands that you can use to test + whether the server is accessible and working properly. For + information about starting and stopping the server automatically, + see Section 2.13.1.2, "Starting and Stopping MySQL Automatically." - This section provides information about using MySQL on variants of - BSD Unix. + After you complete the procedure and have the server running, you + should assign passwords to the accounts created by + mysql_install_db. Instructions for doing so are given in Section + 2.13.2, "Securing the Initial MySQL Accounts." -2.13.4.1. FreeBSD Notes + In the examples shown here, the server runs under the user ID of + the mysql login account. This assumes that such an account exists. + Either create the account if it does not exist, or substitute the + name of a different existing login account that you plan to use + for running the server. - FreeBSD 4.x or newer is recommended for running MySQL, because the - thread package is much more integrated. To get a secure and stable - system, you should use only FreeBSD kernels that are marked - -RELEASE. + 1. Change location into the top-level directory of your MySQL + installation, represented here by BASEDIR: +shell> cd BASEDIR + BASEDIR is likely to be something like /usr/local/mysql or + /usr/local. The following steps assume that you are located in + this directory. - The easiest (and preferred) way to install MySQL is to use the - mysql-server and mysql-client ports available at - http://www.freebsd.org/. Using these ports gives you the following - benefits: + 2. If necessary, run the mysql_install_db program to set up the + initial MySQL grant tables containing the privileges that + determine how users are allowed to connect to the server. + You'll need to do this if you used a distribution type for + which the installation procedure doesn't run the program for + you. + Typically, mysql_install_db needs to be run only the first + time you install MySQL, so you can skip this step if you are + upgrading an existing installation, However, mysql_install_db + does not overwrite any existing privilege tables, so it should + be safe to run in any circumstances. + To initialize the grant tables, use one of the following + commands, depending on whether mysql_install_db is located in + the bin or scripts directory: +shell> bin/mysql_install_db --user=mysql +shell> scripts/mysql_install_db --user=mysql + It might be necessary to specify other options such as + --basedir or --datadir if mysql_install_db does not use the + correct locations for the installation directory or data + directory. For example: +shell> bin/mysql_install_db --user=mysql \ + --basedir=/opt/mysql/mysql \ + --datadir=/opt/mysql/mysql/data + The mysql_install_db script creates the server's data + directory. Under the data directory, it creates directories + for the mysql database that holds all database privileges and + the test database that you can use to test MySQL. The script + also creates privilege table entries for root and + anonymous-user accounts. The accounts have no passwords + initially. A description of their initial privileges is given + in Section 2.13.2, "Securing the Initial MySQL Accounts." + Briefly, these privileges allow the MySQL root user to do + anything, and allow anybody to create or use databases with a + name of test or starting with test_. + It is important to make sure that the database directories and + files are owned by the mysql login account so that the server + has read and write access to them when you run it later. To + ensure this, the --user option should be used as shown if you + run mysql_install_db as root. Otherwise, you should execute + the script while logged in as mysql, in which case you can + omit the --user option from the command. + mysql_install_db creates several tables in the mysql database, + including user, db, host, tables_priv, columns_priv, func, and + others. See Section 5.4, "The MySQL Access Privilege System," + for a complete listing and description of these tables. + If you don't want to have the test database, you can remove it + with mysqladmin -u root drop test after starting the server. + If you have trouble with mysql_install_db at this point, see + Section 2.13.1.1, "Problems Running mysql_install_db." - * A working MySQL with all optimizations enabled that are known - to work on your version of FreeBSD. + 3. Start the MySQL server: +shell> bin/mysqld_safe --user=mysql & + It is important that the MySQL server be run using an + unprivileged (non-root) login account. To ensure this, the + --user option should be used as shown if you run mysqld_safe + as system root. Otherwise, you should execute the script while + logged in to the system as mysql, in which case you can omit + the --user option from the command. + Further instructions for running MySQL as an unprivileged user + are given in Section 5.3.5, "How to Run MySQL as a Normal + User." + If you neglected to create the grant tables before proceeding + to this step, the following message appears in the error log + file when you start the server: +mysqld: Can't find file: 'host.frm' + If you have other problems starting the server, see Section + 2.13.1.3, "Starting and Troubleshooting the MySQL Server." - * Automatic configuration and build. + 4. Use mysqladmin to verify that the server is running. The + following commands provide simple tests to check whether the + server is up and responding to connections: +shell> bin/mysqladmin version +shell> bin/mysqladmin variables + The output from mysqladmin version varies slightly depending + on your platform and version of MySQL, but should be similar + to that shown here: +shell> bin/mysqladmin version +mysqladmin Ver 14.12 Distrib 5.1.41, for pc-linux-gnu on i686 +... - * Startup scripts installed in /usr/local/etc/rc.d. +Server version 5.1.41 +Protocol version 10 +Connection Localhost via UNIX socket +UNIX socket /var/lib/mysql/mysql.sock +Uptime: 14 days 5 hours 5 min 21 sec - * The ability to use pkg_info -L to see which files are - installed. +Threads: 1 Questions: 366 Slow queries: 0 +Opens: 0 Flush tables: 1 Open tables: 19 +Queries per second avg: 0.000 + To see what else you can do with mysqladmin, invoke it with + the --help option. - * The ability to use pkg_delete to remove MySQL if you no longer - want it on your machine. + 5. Verify that you can shut down the server: +shell> bin/mysqladmin -u root shutdown - It is recommended you use MIT-pthreads on FreeBSD 2.x, and native - threads on FreeBSD 3 and up. It is possible to run with native - threads on some late 2.2.x versions, but you may encounter - problems shutting down mysqld. - - Unfortunately, certain function calls on FreeBSD are not yet fully - thread-safe. Most notably, this includes the gethostbyname() - function, which is used by MySQL to convert host names into IP - addresses. Under certain circumstances, the mysqld process - suddenly causes 100% CPU load and is unresponsive. If you - encounter this problem, try to start MySQL using the - --skip-name-resolve option. - - Alternatively, you can link MySQL on FreeBSD 4.x against the - LinuxThreads library, which avoids a few of the problems that the - native FreeBSD thread implementation has. For a very good - comparison of LinuxThreads versus native threads, see Jeremy - Zawodny's article FreeBSD or Linux for your MySQL Server? at - http://jeremy.zawodny.com/blog/archives/000697.html. - - Known problem when using LinuxThreads on FreeBSD is: - - * The connection times (wait_timeout, interactive_timeout and - net_read_timeout) values are not honored. The symptom is that - persistent connections can hang for a very long time without - getting closed down and that a 'kill' for a thread will not - take affect until the thread does it a new command - This is probably a signal handling problem in the thread - library where the signal doesn't break a pending read. This is - supposed to be fixed in FreeBSD 5.0 + 6. Verify that you can start the server again. Do this by using + mysqld_safe or by invoking mysqld directly. For example: +shell> bin/mysqld_safe --user=mysql --log & + If mysqld_safe fails, see Section 2.13.1.3, "Starting and + Troubleshooting the MySQL Server." - The MySQL build process requires GNU make (gmake) to work. If GNU - make is not available, you must install it first before compiling - MySQL. + 7. Run some simple tests to verify that you can retrieve + information from the server. The output should be similar to + what is shown here: +shell> bin/mysqlshow ++-----------+ +| Databases | ++-----------+ +| mysql | +| test | ++-----------+ - The recommended way to compile and install MySQL on FreeBSD with - gcc (2.95.2 and up) is: -CC=gcc CFLAGS="-O2 -fno-strength-reduce" \ - CXX=gcc CXXFLAGS="-O2 -fno-rtti -fno-exceptions \ - -felide-constructors -fno-strength-reduce" \ - ./configure --prefix=/usr/local/mysql --enable-assembler -gmake -gmake install -cd /usr/local/mysql -bin/mysql_install_db --user=mysql -bin/mysqld_safe & +shell> bin/mysqlshow mysql +Database: mysql ++---------------------------+ +| Tables | ++---------------------------+ +| columns_priv | +| db | +| func | +| help_category | +| help_keyword | +| help_relation | +| help_topic | +| host | +| proc | +| procs_priv | +| tables_priv | +| time_zone | +| time_zone_leap_second | +| time_zone_name | +| time_zone_transition | +| time_zone_transition_type | +| user | ++---------------------------+ - If you notice that configure uses MIT-pthreads, you should read - the MIT-pthreads notes. See Section 2.10.5, "MIT-pthreads Notes." +shell> bin/mysql -e "SELECT Host,Db,User FROM db" mysql ++------+--------+------+ +| host | db | user | ++------+--------+------+ +| % | test | | +| % | test_% | | ++------+--------+------+ - If you get an error from make install that it can't find - /usr/include/pthreads, configure didn't detect that you need - MIT-pthreads. To fix this problem, remove config.cache, and then - re-run configure with the --with-mit-threads option. - - Be sure that your name resolver setup is correct. Otherwise, you - may experience resolver delays or failures when connecting to - mysqld. Also make sure that the localhost entry in the /etc/hosts - file is correct. The file should start with a line similar to - this: -127.0.0.1 localhost localhost.your.domain + 8. There is a benchmark suite in the sql-bench directory (under + the MySQL installation directory) that you can use to compare + how MySQL performs on different platforms. The benchmark suite + is written in Perl. It requires the Perl DBI module that + provides a database-independent interface to the various + databases, and some other additional Perl modules: +DBI +DBD::mysql +Data::Dumper +Data::ShowTable + These modules can be obtained from CPAN + (http://www.cpan.org/) See also Section 2.15.1, "Installing + Perl on Unix." + The sql-bench/Results directory contains the results from many + runs against different databases and platforms. To run all + tests, execute these commands: +shell> cd sql-bench +shell> perl run-all-tests + If you don't have the sql-bench directory, you probably + installed MySQL using RPM files other than the source RPM. + (The source RPM includes the sql-bench benchmark directory.) + In this case, you must first install the benchmark suite + before you can use it. There are separate benchmark RPM files + named mysql-bench-VERSION.i386.rpm that contain benchmark code + and data. + If you have a source distribution, there are also tests in its + tests subdirectory that you can run. For example, to run + auto_increment.tst, execute this command from the top-level + directory of your source distribution: +shell> mysql -vvf test < ./tests/auto_increment.tst + The expected result of the test can be found in the + ./tests/auto_increment.res file. - FreeBSD is known to have a very low default file handle limit. See - Section B.1.2.18, "'File' Not Found and Similar Errors." Start the - server by using the --open-files-limit option for mysqld_safe, or - raise the limits for the mysqld user in /etc/login.conf and - rebuild it with cap_mkdb /etc/login.conf. Also be sure that you - set the appropriate class for this user in the password file if - you are not using the default (use chpass mysqld-user-name). See - Section 4.3.2, "mysqld_safe --- MySQL Server Startup Script." + 9. At this point, you should have the server running. However, + none of the initial MySQL accounts have a password, so you + should assign passwords using the instructions found in + Section 2.13.2, "Securing the Initial MySQL Accounts." - FreeBSD limits the size of a process to 512MB, even if you have - much more RAM available on the system. So you may get an error - such as this: -Out of memory (Needed 16391 bytes) + The MySQL 5.1 installation procedure creates time zone tables in + the mysql database. However, you must populate the tables manually + using the instructions in Section 9.7, "MySQL Server Time Zone + Support." - In current versions of FreeBSD (at least 4.x and greater), you may - increase this limit by adding the following entries to the - /boot/loader.conf file and rebooting the machine (these are not - settings that can be changed at run time with the sysctl command): -kern.maxdsiz="1073741824" # 1GB -kern.dfldsiz="1073741824" # 1GB -kern.maxssiz="134217728" # 128MB +2.13.1.1. Problems Running mysql_install_db - For older versions of FreeBSD, you must recompile your kernel to - change the maximum data segment size for a process. In this case, - you should look at the MAXDSIZ option in the LINT config file for - more information. + The purpose of the mysql_install_db script is to generate new + MySQL privilege tables. It does not overwrite existing MySQL + privilege tables, and it does not affect any other data. - If you get problems with the current date in MySQL, setting the TZ - variable should help. See Section 2.14, "Environment Variables." + If you want to re-create your privilege tables, first stop the + mysqld server if it is running. Then rename the mysql directory + under the data directory to save it, and then run + mysql_install_db. Suppose that your current directory is the MySQL + installation directory and that mysql_install_db is located in the + bin directory and the data directory is named data. To rename the + mysql database and re-run mysql_install_db, use these commands. +shell> mv data/mysql data/mysql.old +shell> bin/mysql_install_db --user=mysql -2.13.4.2. NetBSD Notes + When you run mysql_install_db, you might encounter the following + problems: - To compile on NetBSD, you need GNU make. Otherwise, the build - process fails when make tries to run lint on C++ files. + * mysql_install_db fails to install the grant tables + You may find that mysql_install_db fails to install the grant + tables and terminates after displaying the following messages: +Starting mysqld daemon with databases from XXXXXX +mysqld ended + In this case, you should examine the error log file very + carefully. The log should be located in the directory XXXXXX + named by the error message and should indicate why mysqld + didn't start. If you do not understand what happened, include + the log when you post a bug report. See Section 1.6, "How to + Report Bugs or Problems." -2.13.4.3. OpenBSD 2.5 Notes + * There is a mysqld process running + This indicates that the server is running, in which case the + grant tables have probably been created already. If so, there + is no need to run mysql_install_db at all because it needs to + be run only once (when you install MySQL the first time). - On OpenBSD 2.5, you can compile MySQL with native threads with the - following options: -CFLAGS=-pthread CXXFLAGS=-pthread ./configure --with-mit-threads=no + * Installing a second mysqld server does not work when one + server is running + This can happen when you have an existing MySQL installation, + but want to put a new installation in a different location. + For example, you might have a production installation, but you + want to create a second installation for testing purposes. + Generally the problem that occurs when you try to run a second + server is that it tries to use a network interface that is in + use by the first server. In this case, you should see one of + the following error messages: +Can't start server: Bind on TCP/IP port: +Address already in use +Can't start server: Bind on unix socket... + For instructions on setting up multiple servers, see Section + 5.6, "Running Multiple MySQL Servers on the Same Machine." -2.13.4.4. BSD/OS Version 2.x Notes + * You do not have write access to the /tmp directory + If you do not have write access to create temporary files or a + Unix socket file in the default location (the /tmp directory), + an error occurs when you run mysql_install_db or the mysqld + server. + You can specify different locations for the temporary + directory and Unix socket file by executing these commands + prior to starting mysql_install_db or mysqld, where + some_tmp_dir is the full path name to some directory for which + you have write permission: +shell> TMPDIR=/some_tmp_dir/ +shell> MYSQL_UNIX_PORT=/some_tmp_dir/mysql.sock +shell> export TMPDIR MYSQL_UNIX_PORT + Then you should be able to run mysql_install_db and start the + server with these commands: +shell> bin/mysql_install_db --user=mysql +shell> bin/mysqld_safe --user=mysql & + If mysql_install_db is located in the scripts directory, + modify the first command to scripts/mysql_install_db. + See Section B.5.4.5, "How to Protect or Change the MySQL Unix + Socket File," and Section 2.14, "Environment Variables." - If you get the following error when compiling MySQL, your ulimit - value for virtual memory is too low: -item_func.h: In method -`Item_func_ge::Item_func_ge(const Item_func_ge &)': -item_func.h:28: virtual memory exhausted -make[2]: *** [item_func.o] Error 1 + There are some alternatives to running the mysql_install_db script + provided in the MySQL distribution: - Try using ulimit -v 80000 and run make again. If this doesn't work - and you are using bash, try switching to csh or sh; some BSDI - users have reported problems with bash and ulimit. + * If you want the initial privileges to be different from the + standard defaults, you can modify mysql_install_db before you + run it. However, it is preferable to use GRANT and REVOKE to + change the privileges after the grant tables have been set up. + In other words, you can run mysql_install_db, and then use + mysql -u root mysql to connect to the server as the MySQL root + user so that you can issue the necessary GRANT and REVOKE + statements. + If you want to install MySQL on several machines with the same + privileges, you can put the GRANT and REVOKE statements in a + file and execute the file as a script using mysql after + running mysql_install_db. For example: +shell> bin/mysql_install_db --user=mysql +shell> bin/mysql -u root < your_script_file + By doing this, you can avoid having to issue the statements + manually on each machine. - If you are using gcc, you may also use have to use the - --with-low-memory flag for configure to be able to compile - sql_yacc.cc. + * It is possible to re-create the grant tables completely after + they have previously been created. You might want to do this + if you're just learning how to use GRANT and REVOKE and have + made so many modifications after running mysql_install_db that + you want to wipe out the tables and start over. + To re-create the grant tables, remove all the .frm, .MYI, and + .MYD files in the mysql database directory. Then run the + mysql_install_db script again. - If you get problems with the current date in MySQL, setting the TZ - variable should help. See Section 2.14, "Environment Variables." + * You can start mysqld manually using the --skip-grant-tables + option and add the privilege information yourself using mysql: +shell> bin/mysqld_safe --user=mysql --skip-grant-tables & +shell> bin/mysql mysql + From mysql, manually execute the SQL commands contained in + mysql_install_db. Make sure that you run mysqladmin + flush-privileges or mysqladmin reload afterward to tell the + server to reload the grant tables. + Note that by not using mysql_install_db, you not only have to + populate the grant tables manually, you also have to create + them first. -2.13.4.5. BSD/OS Version 3.x Notes +2.13.1.2. Starting and Stopping MySQL Automatically - Upgrade to BSD/OS 3.1. If that is not possible, install BSDIpatch - M300-038. + Generally, you start the mysqld server in one of these ways: - Use the following command when configuring MySQL: -env CXX=shlicc++ CC=shlicc2 \ -./configure \ - --prefix=/usr/local/mysql \ - --localstatedir=/var/mysql \ - --without-perl \ - --with-unix-socket-path=/var/mysql/mysql.sock - - The following is also known to work: -env CC=gcc CXX=gcc CXXFLAGS=-O3 \ -./configure \ - --prefix=/usr/local/mysql \ - --with-unix-socket-path=/var/mysql/mysql.sock - - You can change the directory locations if you wish, or just use - the defaults by not specifying any locations. - - If you have problems with performance under heavy load, try using - the --skip-thread-priority option to mysqld. This runs all threads - with the same priority. On BSDI 3.1, this gives better - performance, at least until BSDI fixes its thread scheduler. - - If you get the error virtual memory exhausted while compiling, you - should try using ulimit -v 80000 and running make again. If this - doesn't work and you are using bash, try switching to csh or sh; - some BSDI users have reported problems with bash and ulimit. - -2.13.4.6. BSD/OS Version 4.x Notes - - BSDI 4.x has some thread-related bugs. If you want to use MySQL on - this, you should install all thread-related patches. At least - M400-023 should be installed. - - On some BSDI 4.x systems, you may get problems with shared - libraries. The symptom is that you can't execute any client - programs, for example, mysqladmin. In this case, you need to - reconfigure not to use shared libraries with the --disable-shared - option to configure. - - Some customers have had problems on BSDI 4.0.1 that the mysqld - binary after a while can't open tables. This occurs because some - library/system-related bug causes mysqld to change current - directory without having asked for that to happen. - - The fix is to either upgrade MySQL to at least version 3.23.34 or, - after running configure, remove the line #define HAVE_REALPATH - from config.h before running make. - - Note that this means that you can't symbolically link a database - directories to another database directory or symbolic link a table - to another database on BSDI. (Making a symbolic link to another - disk is okay). + * Invoke mysqld directly. This works on any platform. -2.13.5. Other Unix Notes + * Run the MySQL server as a Windows service. The service can be + set to start the server automatically when Windows starts, or + as a manual service that you start on request. For + instructions, see Section 2.5.5.6, "Starting MySQL as a + Windows Service." -2.13.5.1. HP-UX Version 10.20 Notes + * Invoke mysqld_safe, which tries to determine the proper + options for mysqld and then runs it with those options. This + script is used on Unix and Unix-like systems. See Section + 4.3.2, "mysqld_safe --- MySQL Server Startup Script." - If you install MySQL using a binary tarball distribution on HP-UX, - you may run into trouble even before you get the MySQL - distribution unpacked, as the HP-UX tar cannot handle long file - names. This means that you may see errors when you try to unpack - MySQL. + * Invoke mysql.server. This script is used primarily at system + startup and shutdown on systems that use System V-style run + directories, where it usually is installed under the name + mysql. The mysql.server script starts the server by invoking + mysqld_safe. See Section 4.3.3, "mysql.server --- MySQL Server + Startup Script." - If this occurs, you must use GNU tar (gtar) to unpack the - distribution. + * On Mac OS X, install a separate MySQL Startup Item package to + enable the automatic startup of MySQL on system startup. The + Startup Item starts the server by invoking mysql.server. See + Section 2.7, "Installing MySQL on Mac OS X," for details. - There are a couple of small problems when compiling MySQL on - HP-UX. Use gcc instead of the HP-UX native compiler, because gcc - produces better code. - - Use gcc 2.95 on HP-UX. Don't use high optimization flags (such as - -O6) because they may not be safe on HP-UX. - - The following configure line should work with gcc 2.95: -CFLAGS="-I/opt/dce/include -fpic" \ -CXXFLAGS="-I/opt/dce/include -felide-constructors -fno-exceptions \ --fno-rtti" \ -CXX=gcc \ -./configure --with-pthread \ - --with-named-thread-libs='-ldce' \ - --prefix=/usr/local/mysql --disable-shared - - The following configure line should work with gcc 3.1: -CFLAGS="-DHPUX -I/opt/dce/include -O3 -fPIC" CXX=gcc \ -CXXFLAGS="-DHPUX -I/opt/dce/include -felide-constructors \ - -fno-exceptions -fno-rtti -O3 -fPIC" \ -./configure --prefix=/usr/local/mysql \ - --with-extra-charsets=complex --enable-thread-safe-client \ - --enable-local-infile --with-pthread \ - --with-named-thread-libs=-ldce --with-lib-ccflags=-fPIC - --disable-shared + The mysqld_safe and mysql.server scripts and the Mac OS X Startup + Item can be used to start the server manually, or automatically at + system startup time. mysql.server and the Startup Item also can be + used to stop the server. -2.13.5.2. HP-UX Version 11.x Notes + To start or stop the server manually using the mysql.server + script, invoke it with start or stop arguments: +shell> mysql.server start +shell> mysql.server stop - If you install MySQL using a binary tarball distribution on HP-UX, - you may run into trouble even before you get the MySQL - distribution unpacked, as the HP-UX tar cannot handle long file - names. This means that you may see errors when you try to unpack - MySQL. + Before mysql.server starts the server, it changes location to the + MySQL installation directory, and then invokes mysqld_safe. If you + want the server to run as some specific user, add an appropriate + user option to the [mysqld] group of the /etc/my.cnf option file, + as shown later in this section. (It is possible that you will need + to edit mysql.server if you've installed a binary distribution of + MySQL in a nonstandard location. Modify it to change location into + the proper directory before it runs mysqld_safe. If you do this, + your modified version of mysql.server may be overwritten if you + upgrade MySQL in the future, so you should make a copy of your + edited version that you can reinstall.) - If this occurs, you must use GNU tar (gtar) to unpack the - distribution. + mysql.server stop stops the server by sending a signal to it. You + can also stop the server manually by executing mysqladmin + shutdown. - Because of some critical bugs in the standard HP-UX libraries, you - should install the following patches before trying to run MySQL on - HP-UX 11.0: -PHKL_22840 Streams cumulative -PHNE_22397 ARPA cumulative + To start and stop MySQL automatically on your server, you need to + add start and stop commands to the appropriate places in your + /etc/rc* files. - This solves the problem of getting EWOULDBLOCK from recv() and - EBADF from accept() in threaded applications. + If you use the Linux server RPM package + (MySQL-server-VERSION.rpm), the mysql.server script is installed + in the /etc/init.d directory with the name mysql. You need not + install it manually. See Section 2.6.1, "Installing MySQL from RPM + Packages on Linux," for more information on the Linux RPM + packages. - If you are using gcc 2.95.1 on an unpatched HP-UX 11.x system, you - may get the following error: -In file included from /usr/include/unistd.h:11, - from ../include/global.h:125, - from mysql_priv.h:15, - from item.cc:19: -/usr/include/sys/unistd.h:184: declaration of C function ... -/usr/include/sys/pthread.h:440: previous declaration ... -In file included from item.h:306, - from mysql_priv.h:158, - from item.cc:19: + Some vendors provide RPM packages that install a startup script + under a different name such as mysqld. - The problem is that HP-UX does not define pthreads_atfork() - consistently. It has conflicting prototypes in - /usr/include/sys/unistd.h:184 and /usr/include/sys/pthread.h:440. + If you install MySQL from a source distribution or using a binary + distribution format that does not install mysql.server + automatically, you can install it manually. The script can be + found in the support-files directory under the MySQL installation + directory or in a MySQL source tree. - One solution is to copy /usr/include/sys/unistd.h into - mysql/include and edit unistd.h and change it to match the - definition in pthread.h. Look for this line: -extern int pthread_atfork(void (*prepare)(), void (*parent)(), - void (*child)()); + To install mysql.server manually, copy it to the /etc/init.d + directory with the name mysql, and then make it executable. Do + this by changing location into the appropriate directory where + mysql.server is located and executing these commands: +shell> cp mysql.server /etc/init.d/mysql +shell> chmod +x /etc/init.d/mysql - Change it to look like this: -extern int pthread_atfork(void (*prepare)(void), void (*parent)(void) -, - void (*child)(void)); + Older Red Hat systems use the /etc/rc.d/init.d directory rather + than /etc/init.d. Adjust the preceding commands accordingly. + Alternatively, first create /etc/init.d as a symbolic link that + points to /etc/rc.d/init.d: +shell> cd /etc +shell> ln -s rc.d/init.d . - After making the change, the following configure line should work: -CFLAGS="-fomit-frame-pointer -O3 -fpic" CXX=gcc \ -CXXFLAGS="-felide-constructors -fno-exceptions -fno-rtti -O3" \ -./configure --prefix=/usr/local/mysql --disable-shared + After installing the script, the commands needed to activate it to + run at system startup depend on your operating system. On Linux, + you can use chkconfig: +shell> chkconfig --add mysql - If you are using HP-UX compiler, you can use the following command - (which has been tested with cc B.11.11.04): -CC=cc CXX=aCC CFLAGS=+DD64 CXXFLAGS=+DD64 ./configure \ - --with-extra-character-set=complex + On some Linux systems, the following command also seems to be + necessary to fully enable the mysql script: +shell> chkconfig --level 345 mysql on - You can ignore any errors of the following type: -aCC: warning 901: unknown option: `-3': use +help for online -documentation + On FreeBSD, startup scripts generally should go in + /usr/local/etc/rc.d/. The rc(8) manual page states that scripts in + this directory are executed only if their basename matches the + *.sh shell file name pattern. Any other files or directories + present within the directory are silently ignored. In other words, + on FreeBSD, you should install the mysql.server script as + /usr/local/etc/rc.d/mysql.server.sh to enable automatic startup. - If you get the following error from configure, verify that you - don't have the path to the K&R compiler before the path to the - HP-UX C and C++ compiler: -checking for cc option to accept ANSI C... no -configure: error: MySQL requires an ANSI C compiler (and a C++ compil -er). -Try gcc. See the Installation chapter in the Reference Manual. + As an alternative to the preceding setup, some operating systems + also use /etc/rc.local or /etc/init.d/boot.local to start + additional services on startup. To start up MySQL using this + method, you could append a command like the one following to the + appropriate startup file: +/bin/sh -c 'cd /usr/local/mysql; ./bin/mysqld_safe --user=mysql &' - Another reason for not being able to compile is that you didn't - define the +DD64 flags as just described. + For other systems, consult your operating system documentation to + see how to install startup scripts. - Another possibility for HP-UX 11 is to use the MySQL binaries - provided at http://dev.mysql.com/downloads/, which we have built - and tested ourselves. We have also received reports that the HP-UX - 10.20 binaries supplied by MySQL can be run successfully on HP-UX - 11. If you encounter problems, you should be sure to check your - HP-UX patch level. + You can add options for mysql.server in a global /etc/my.cnf file. + A typical /etc/my.cnf file might look like this: +[mysqld] +datadir=/usr/local/mysql/var +socket=/var/tmp/mysql.sock +port=3306 +user=mysql -2.13.5.3. IBM-AIX notes +[mysql.server] +basedir=/usr/local/mysql - Automatic detection of xlC is missing from Autoconf, so a number - of variables need to be set before running configure. The - following example uses the IBM compiler: -export CC="xlc_r -ma -O3 -qstrict -qoptimize=3 -qmaxmem=8192 " -export CXX="xlC_r -ma -O3 -qstrict -qoptimize=3 -qmaxmem=8192" -export CFLAGS="-I /usr/local/include" -export LDFLAGS="-L /usr/local/lib" -export CPPFLAGS=$CFLAGS -export CXXFLAGS=$CFLAGS + The mysql.server script supports the following options: basedir, + datadir, and pid-file. If specified, they must be placed in an + option file, not on the command line. mysql.server supports only + start and stop as command-line arguments. -./configure --prefix=/usr/local \ - --localstatedir=/var/mysql \ - --sbindir='/usr/local/bin' \ - --libexecdir='/usr/local/bin' \ - --enable-thread-safe-client \ - --enable-large-files + The following table shows which option groups the server and each + startup script read from option files. + Script Option Groups + mysqld [mysqld], [server], [mysqld-major_version] + mysqld_safe [mysqld], [server], [mysqld_safe] + mysql.server [mysqld], [mysql.server], [server] - The preceding options are used to compile the MySQL distribution - that can be found at http://www-frec.bull.com/. + [mysqld-major_version] means that groups with names like + [mysqld-5.0] and [mysqld-5.1] are read by servers having versions + 5.0.x, 5.1.x, and so forth. This feature can be used to specify + options that can be read only by servers within a given release + series. - If you change the -O3 to -O2 in the preceding configure line, you - must also remove the -qstrict option. This is a limitation in the - IBM C compiler. + For backward compatibility, mysql.server also reads the + [mysql_server] group and mysqld_safe also reads the [safe_mysqld] + group. However, you should update your option files to use the + [mysql.server] and [mysqld_safe] groups instead when using MySQL + 5.1. - If you are using gcc to compile MySQL, you must use the - -fno-exceptions flag, because the exception handling in gcc is not - thread-safe! There are also some known problems with IBM's - assembler that may cause it to generate bad code when used with - gcc. + See Section 4.2.3.3, "Using Option Files." - Use the following configure line with gcc 2.95 on AIX: -CC="gcc -pipe -mcpu=power -Wa,-many" \ -CXX="gcc -pipe -mcpu=power -Wa,-many" \ -CXXFLAGS="-felide-constructors -fno-exceptions -fno-rtti" \ -./configure --prefix=/usr/local/mysql --with-low-memory +2.13.1.3. Starting and Troubleshooting the MySQL Server - The -Wa,-many option is necessary for the compile to be - successful. IBM is aware of this problem but is in no hurry to fix - it because of the workaround that is available. We don't know if - the -fno-exceptions is required with gcc 2.95, but because MySQL - doesn't use exceptions and the option generates faster code, you - should always use it with gcc. + This section provides troubleshooting suggestions for problems + starting the server on Unix. If you are using Windows, see Section + 2.5.6, "Troubleshooting a MySQL Installation Under Windows." - If you get a problem with assembler code, try changing the - -mcpu=xxx option to match your CPU. Typically power2, power, or - powerpc may need to be used. Alternatively, you might need to use - 604 or 604e. We are not positive but suspect that power would - likely be safe most of the time, even on a power2 machine. + If you have problems starting the server, here are some things to + try: - If you don't know what your CPU is, execute a uname -m command. It - produces a string that looks like 000514676700, with a format of - xxyyyyyymmss where xx and ss are always 00, yyyyyy is a unique - system ID and mm is the ID of the CPU Planar. A chart of these - values can be found at - http://www16.boulder.ibm.com/pseries/en_US/cmds/aixcmds5/uname.htm - . + * Check the error log to see why the server does not start. - This gives you a machine type and a machine model you can use to - determine what type of CPU you have. + * Specify any special options needed by the storage engines you + are using. - If you have problems with signals (MySQL dies unexpectedly under - high load), you may have found an OS bug with threads and signals. - In this case, you can tell MySQL not to use signals by configuring - as follows: -CFLAGS=-DDONT_USE_THR_ALARM CXX=gcc \ -CXXFLAGS="-felide-constructors -fno-exceptions -fno-rtti \ --DDONT_USE_THR_ALARM" \ -./configure --prefix=/usr/local/mysql --with-debug \ - --with-low-memory + * Make sure that the server knows where to find the data + directory. - This doesn't affect the performance of MySQL, but has the side - effect that you can't kill clients that are "sleeping" on a - connection with mysqladmin kill or mysqladmin shutdown. Instead, - the client dies when it issues its next command. + * Make sure that the server can access the data directory. The + ownership and permissions of the data directory and its + contents must be set such that the server can read and modify + them. - On some versions of AIX, linking with libbind.a makes - getservbyname() dump core. This is an AIX bug and should be - reported to IBM. + * Verify that the network interfaces the server wants to use are + available. - For AIX 4.2.1 and gcc, you have to make the following changes. + Some storage engines have options that control their behavior. You + can create a my.cnf file and specify startup options for the + engines that you plan to use. If you are going to use storage + engines that support transactional tables (InnoDB, NDB), be sure + that you have them configured the way you want before starting the + server: - After configuring, edit config.h and include/my_config.h and - change the line that says this: -#define HAVE_SNPRINTF 1 + * If you are using InnoDB tables, see Section 13.6.2, "InnoDB + Configuration." - to this: -#undef HAVE_SNPRINTF + * If you are using MySQL Cluster, see Section 17.3, "MySQL + Cluster Configuration." - And finally, in mysqld.cc, you need to add a prototype for - initgroups(). -#ifdef _AIX41 -extern "C" int initgroups(const char *,int); -#endif + MySQL Enterprise For expert advice on start-up options appropriate + to your circumstances, subscribe to The MySQL Enterprise Monitor. + For more information, see + http://www.mysql.com/products/enterprise/advisors.html. - For 32-bit binaries, if you need to allocate a lot of memory to - the mysqld process, it is not enough to just use ulimit -d - unlimited. You may also have to modify mysqld_safe to add a line - something like this: -export LDR_CNTRL='MAXDATA=0x80000000' + Storage engines will use default option values if you specify + none, but it is recommended that you review the available options + and specify explicit values for those for which the defaults are + not appropriate for your installation. - You can find more information about using a lot of memory at - http://publib16.boulder.ibm.com/pseries/en_US/aixprggd/genprogc/lr - g_prg_support.htm. + When the mysqld server starts, it changes location to the data + directory. This is where it expects to find databases and where it + expects to write log files. The server also writes the pid + (process ID) file in the data directory. - Users of AIX 4.3 should use gmake instead of the make utility - included with AIX. + The data directory location is hardwired in when the server is + compiled. This is where the server looks for the data directory by + default. If the data directory is located somewhere else on your + system, the server will not work properly. You can determine what + the default path settings are by invoking mysqld with the + --verbose and --help options. - As of AIX 4.1, the C compiler has been unbundled from AIX as a - separate product. gcc 3.3.2 can be obtained here: - ftp://ftp.software.ibm.com/aix/freeSoftware/aixtoolbox/RPMS/ppc/gc - c/ + If the default locations don't match the MySQL installation layout + on your system, you can override them by specifying options to + mysqld or mysqld_safe on the command line or in an option file. - The steps for compiling MySQL on AIX with gcc 3.3.2 are similar to - those for using gcc 2.95 (in particular, the need to edit config.h - and my_config.h after running configure). However, before running - configure, you should also patch the curses.h file as follows: -/opt/freeware/lib/gcc-lib/powerpc-ibm-aix5.2.0.0/3.3.2/include/curses -.h.ORIG - Mon Dec 26 02:17:28 2005 ---- /opt/freeware/lib/gcc-lib/powerpc-ibm-aix5.2.0.0/3.3.2/include/cu -rses.h -Mon Dec 26 02:40:13 2005 -*************** -*** 2023,2029 **** + To specify the location of the data directory explicitly, use the + --datadir option. However, normally you can tell mysqld the + location of the base directory under which MySQL is installed and + it looks for the data directory there. You can do this with the + --basedir option. + + To check the effect of specifying path options, invoke mysqld with + those options followed by the --verbose and --help options. For + example, if you change location into the directory where mysqld is + installed and then run the following command, it shows the effect + of starting the server with a base directory of /usr/local: +shell> ./mysqld --basedir=/usr/local --verbose --help + You can specify other options such as --datadir as well, but + --verbose and --help must be the last options. - #endif /* _AIX32_CURSES */ -! #if defined(__USE_FIXED_PROTOTYPES__) || defined(__cplusplus) || de -fined -(__STRICT_ANSI__) - extern int delwin (WINDOW *); - extern int endwin (void); - extern int getcurx (WINDOW *); ---- 2023,2029 ---- + Once you determine the path settings you want, start the server + without --verbose and --help. + If mysqld is currently running, you can find out what path + settings it is using by executing this command: +shell> mysqladmin variables - #endif /* _AIX32_CURSES */ -! #if 0 && (defined(__USE_FIXED_PROTOTYPES__) || defined(__cplusplus) -|| defined -(__STRICT_ANSI__)) - extern int delwin (WINDOW *); - extern int endwin (void); - extern int getcurx (WINDOW *); + Or: +shell> mysqladmin -h host_name variables -2.13.5.4. SunOS 4 Notes + host_name is the name of the MySQL server host. - On SunOS 4, MIT-pthreads is needed to compile MySQL. This in turn - means you need GNU make. + If you get Errcode 13 (which means Permission denied) when + starting mysqld, this means that the privileges of the data + directory or its contents do not allow the server access. In this + case, you change the permissions for the involved files and + directories so that the server has the right to use them. You can + also start the server as root, but this raises security issues and + should be avoided. - Some SunOS 4 systems have problems with dynamic libraries and - libtool. You can use the following configure line to avoid this - problem: -./configure --disable-shared --with-mysqld-ldflags=-all-static - - When compiling readline, you may get warnings about duplicate - defines. These can be ignored. - - When compiling mysqld, there are some implicit declaration of - function warnings. These can be ignored. - -2.13.5.5. Alpha-DEC-UNIX Notes (Tru64) - - If you are using egcs 1.1.2 on Digital Unix, you should upgrade to - gcc 2.95.2, because egcs on DEC has some serious bugs! - - When compiling threaded programs under Digital Unix, the - documentation recommends using the -pthread option for cc and cxx - and the -lmach -lexc libraries (in addition to -lpthread). You - should run configure something like this: -CC="cc -pthread" CXX="cxx -pthread -O" \ -./configure --with-named-thread-libs="-lpthread -lmach -lexc -lc" - - When compiling mysqld, you may see a couple of warnings like this: -mysqld.cc: In function void handle_connections()': -mysqld.cc:626: passing long unsigned int *' as argument 3 of -accept(int,sockadddr *, int *)' - - You can safely ignore these warnings. They occur because configure - can detect only errors, not warnings. - - If you start the server directly from the command line, you may - have problems with it dying when you log out. (When you log out, - your outstanding processes receive a SIGHUP signal.) If so, try - starting the server like this: -nohup mysqld [options] & - - nohup causes the command following it to ignore any SIGHUP signal - sent from the terminal. Alternatively, start the server by running - mysqld_safe, which invokes mysqld using nohup for you. See Section - 4.3.2, "mysqld_safe --- MySQL Server Startup Script." - - If you get a problem when compiling mysys/get_opt.c, just remove - the #define _NO_PROTO line from the start of that file. - - If you are using Compaq's CC compiler, the following configure - line should work: -CC="cc -pthread" -CFLAGS="-O4 -ansi_alias -ansi_args -fast -inline speed \ - -speculate all -arch host" -CXX="cxx -pthread" -CXXFLAGS="-O4 -ansi_alias -ansi_args -fast -inline speed \ - -speculate all -arch host -noexceptions -nortti" -export CC CFLAGS CXX CXXFLAGS -./configure \ - --prefix=/usr/local/mysql \ - --with-low-memory \ - --enable-large-files \ - --enable-shared=yes \ - --with-named-thread-libs="-lpthread -lmach -lexc -lc" -gnumake - - If you get a problem with libtool when compiling with shared - libraries as just shown, when linking mysql, you should be able to - get around this by issuing these commands: -cd mysql -/bin/sh ../libtool --mode=link cxx -pthread -O3 -DDBUG_OFF \ - -O4 -ansi_alias -ansi_args -fast -inline speed \ - -speculate all \ -arch host -DUNDEF_HAVE_GETHOSTBYNAME_R \ - -o mysql mysql.o readline.o sql_string.o completion_hash.o \ - ../readline/libreadline.a -lcurses \ - ../libmysql/.libs/libmysqlclient.so -lm -cd .. -gnumake -gnumake install -scripts/mysql_install_db - -2.13.5.6. Alpha-DEC-OSF/1 Notes - - If you have problems compiling and have DEC CC and gcc installed, - try running configure like this: -CC=cc CFLAGS=-O CXX=gcc CXXFLAGS=-O3 \ -./configure --prefix=/usr/local/mysql - - If you get problems with the c_asm.h file, you can create and use - a 'dummy' c_asm.h file with: -touch include/c_asm.h -CC=gcc CFLAGS=-I./include \ -CXX=gcc CXXFLAGS=-O3 \ -./configure --prefix=/usr/local/mysql - - Note that the following problems with the ld program can be fixed - by downloading the latest DEC (Compaq) patch kit from: - http://ftp.support.compaq.com/public/unix/. - - On OSF/1 V4.0D and compiler "DEC C V5.6-071 on Digital Unix V4.0 - (Rev. 878)," the compiler had some strange behavior (undefined asm - symbols). /bin/ld also appears to be broken (problems with _exit - undefined errors occurring while linking mysqld). On this system, - we have managed to compile MySQL with the following configure - line, after replacing /bin/ld with the version from OSF 4.0C: -CC=gcc CXX=gcc CXXFLAGS=-O3 ./configure --prefix=/usr/local/mysql - - With the Digital compiler "C++ V6.1-029," the following should - work: -CC=cc -pthread -CFLAGS=-O4 -ansi_alias -ansi_args -fast -inline speed \ - -speculate all -arch host -CXX=cxx -pthread -CXXFLAGS=-O4 -ansi_alias -ansi_args -fast -inline speed \ - -speculate all -arch host -noexceptions -nortti -export CC CFLAGS CXX CXXFLAGS -./configure --prefix=/usr/mysql/mysql \ - --with-mysqld-ldflags=-all-static --disable-shared \ - --with-named-thread-libs="-lmach -lexc -lc" - - In some versions of OSF/1, the alloca() function is broken. Fix - this by removing the line in config.h that defines 'HAVE_ALLOCA'. - - The alloca() function also may have an incorrect prototype in - /usr/include/alloca.h. This warning resulting from this can be - ignored. + On Unix, change location into the data directory and check the + ownership of the data directory and its contents to make sure the + server has access. For example, if the data directory is + /usr/local/mysql/var, use this command: +shell> ls -la /usr/local/mysql/var - configure uses the following thread libraries automatically: - --with-named-thread-libs="-lpthread -lmach -lexc -lc". + If the data directory or its files or subdirectories are not owned + by the login account that you use for running the server, change + their ownership to that account. If the account is named mysql, + use these commands: +shell> chown -R mysql /usr/local/mysql/var +shell> chgrp -R mysql /usr/local/mysql/var - When using gcc, you can also try running configure like this: -CFLAGS=-D_PTHREAD_USE_D4 CXX=gcc CXXFLAGS=-O3 ./configure ... + If it possible that even with correct ownership, MySQL may fail to + start up if there is other security software running on your + system that manages application access to various parts of the + file system. In this case, you may need to reconfigure that + software to enable mysqld to access the directories it uses during + normal operation. - If you have problems with signals (MySQL dies unexpectedly under - high load), you may have found an OS bug with threads and signals. - In this case, you can tell MySQL not to use signals by configuring - with: -CFLAGS=-DDONT_USE_THR_ALARM \ -CXXFLAGS=-DDONT_USE_THR_ALARM \ -./configure ... + If the server fails to start up correctly, check the error log. + Log files are located in the data directory (typically C:\Program + Files\MySQL\MySQL Server 5.1\data on Windows, + /usr/local/mysql/data for a Unix binary distribution, and + /usr/local/var for a Unix source distribution). Look in the data + directory for files with names of the form host_name.err and + host_name.log, where host_name is the name of your server host. + Then examine the last few lines of these files. On Unix, you can + use tail to display them: +shell> tail host_name.err +shell> tail host_name.log - This does not affect the performance of MySQL, but has the side - effect that you can't kill clients that are "sleeping" on a - connection with mysqladmin kill or mysqladmin shutdown. Instead, - the client dies when it issues its next command. + The error log should contain information that indicates why the + server couldn't start. + + If either of the following errors occur, it means that some other + program (perhaps another mysqld server) is using the TCP/IP port + or Unix socket file that mysqld is trying to use: +Can't start server: Bind on TCP/IP port: Address already in use +Can't start server: Bind on unix socket... - With gcc 2.95.2, you may encounter the following compile error: -sql_acl.cc:1456: Internal compiler error in `scan_region', -at except.c:2566 -Please submit a full bug report. - - To fix this, you should change to the sql directory and do a - cut-and-paste of the last gcc line, but change -O3 to -O0 (or add - -O0 immediately after gcc if you don't have any -O option on your - compile line). After this is done, you can just change back to the - top-level directory and run make again. - -2.13.5.7. SGI Irix Notes - - As of MySQL 5.0, we don't provide binaries for Irix any more. - - If you are using Irix 6.5.3 or newer, mysqld is able to create - threads only if you run it as a user that has CAP_SCHED_MGT - privileges (such as root) or give the mysqld server this privilege - with the following shell command: -chcap "CAP_SCHED_MGT+epi" /opt/mysql/libexec/mysqld - - You may have to undefine some symbols in config.h after running - configure and before compiling. - - In some Irix implementations, the alloca() function is broken. If - the mysqld server dies on some SELECT statements, remove the lines - from config.h that define HAVE_ALLOC and HAVE_ALLOCA_H. If - mysqladmin create doesn't work, remove the line from config.h that - defines HAVE_READDIR_R. You may have to remove the HAVE_TERM_H - line as well. - - SGI recommends that you install all the patches on this page as a - set: - http://support.sgi.com/surfzone/patches/patchset/6.2_indigo.rps.ht - ml - - At the very minimum, you should install the latest kernel rollup, - the latest rld rollup, and the latest libc rollup. - - You definitely need all the POSIX patches on this page, for - pthreads support: - - http://support.sgi.com/surfzone/patches/patchset/6.2_posix.rps.htm - l - - If you get the something like the following error when compiling - mysql.cc: -"/usr/include/curses.h", line 82: error(1084): -invalid combination of type - - Type the following in the top-level directory of your MySQL source - tree: -extra/replace bool curses_bool < /usr/include/curses.h > include/curs -es.h -make - - There have also been reports of scheduling problems. If only one - thread is running, performance is slow. Avoid this by starting - another client. This may lead to a two-to-tenfold increase in - execution speed thereafter for the other thread. This is a poorly - understood problem with Irix threads; you may have to improvise to - find solutions until this can be fixed. + Use ps to determine whether you have another mysqld server + running. If so, shut down the server before starting mysqld again. + (If another server is running, and you really want to run multiple + servers, you can find information about how to do so in Section + 5.6, "Running Multiple MySQL Servers on the Same Machine.") - If you are compiling with gcc, you can use the following configure - command: -CC=gcc CXX=gcc CXXFLAGS=-O3 \ -./configure --prefix=/usr/local/mysql --enable-thread-safe-client \ - --with-named-thread-libs=-lpthread - - On Irix 6.5.11 with native Irix C and C++ compilers ver. 7.3.1.2, - the following is reported to work -CC=cc CXX=CC CFLAGS='-O3 -n32 -TARG:platform=IP22 -I/usr/local/includ -e \ --L/usr/local/lib' CXXFLAGS='-O3 -n32 -TARG:platform=IP22 \ --I/usr/local/include -L/usr/local/lib' \ -./configure --prefix=/usr/local/mysql --with-innodb \ - --with-libwrap=/usr/local \ - --with-named-curses-libs=/usr/local/lib/libncurses.a - -2.13.5.8. SCO UNIX and OpenServer 5.0.x Notes - - The current port is tested only on sco3.2v5.0.5, sco3.2v5.0.6, and - sco3.2v5.0.7 systems. There has also been progress on a port to - sco3.2v4.2. Open Server 5.0.8 (Legend) has native threads and - allows files greater than 2GB. The current maximum file size is - 2GB. - - We have been able to compile MySQL with the following configure - command on OpenServer with gcc 2.95.3. -CC=gcc CFLAGS="-D_FILE_OFFSET_BITS=64 -O3" \ -CXX=gcc CXXFLAGS="-D_FILE_OFFSET_BITS=64 -O3" \ -./configure --prefix=/usr/local/mysql \ - --enable-thread-safe-client --with-innodb \ - --with-openssl --with-vio --with-extra-charsets=complex - - gcc is available at - ftp://ftp.sco.com/pub/openserver5/opensrc/gnutools-5.0.7Kj. - - This development system requires the OpenServer Execution - Environment Supplement oss646B on OpenServer 5.0.6 and oss656B and - The OpenSource libraries found in gwxlibs. All OpenSource tools - are in the opensrc directory. They are available at - ftp://ftp.sco.com/pub/openserver5/opensrc/. - - Use the latest production release of MySQL. - - SCO provides operating system patches at - ftp://ftp.sco.com/pub/openserver5 for OpenServer 5.0.[0-6] and - ftp://ftp.sco.com/pub/openserverv5/507 for OpenServer 5.0.7. - - SCO provides information about security fixes at - ftp://ftp.sco.com/pub/security/OpenServer for OpenServer 5.0.x. - - The maximum file size on an OpenServer 5.0.x system is 2GB. - - The total memory which can be allocated for streams buffers, - clists, and lock records cannot exceed 60MB on OpenServer 5.0.x. - - Streams buffers are allocated in units of 4096 byte pages, clists - are 70 bytes each, and lock records are 64 bytes each, so: -(NSTRPAGES x 4096) + (NCLIST x 70) + (MAX_FLCKREC x 64) <= 62914560 - - Follow this procedure to configure the Database Services option. - If you are unsure whether an application requires this, see the - documentation provided with the application. - - 1. Log in as root. - - 2. Enable the SUDS driver by editing the /etc/conf/sdevice.d/suds - file. Change the N in the second field to a Y. - - 3. Use mkdev aio or the Hardware/Kernel Manager to enable support - for asynchronous I/O and relink the kernel. To allow users to - lock down memory for use with this type of I/O, update the - aiomemlock(F) file. This file should be updated to include the - names of users that can use AIO and the maximum amounts of - memory they can lock down. - - 4. Many applications use setuid binaries so that you need to - specify only a single user. See the documentation provided - with the application to determine whether this is the case for - your application. + If no other server is running, try to execute the command telnet + your_host_name tcp_ip_port_number. (The default MySQL port number + is 3306.) Then press Enter a couple of times. If you don't get an + error message like telnet: Unable to connect to remote host: + Connection refused, some other program is using the TCP/IP port + that mysqld is trying to use. You'll need to track down what + program this is and disable it, or else tell mysqld to listen to a + different port with the --port option. In this case, you'll also + need to specify the port number for client programs when + connecting to the server via TCP/IP. - After you complete this process, reboot the system to create a new - kernel incorporating these changes. + Another reason the port might be inaccessible is that you have a + firewall running that blocks connections to it. If so, modify the + firewall settings to allow access to the port. - By default, the entries in /etc/conf/cf.d/mtune are set as - follows: -Value Default Min Max ------ ------- --- --- -NBUF 0 24 450000 -NHBUF 0 32 524288 -NMPBUF 0 12 512 -MAX_INODE 0 100 64000 -MAX_FILE 0 100 64000 -CTBUFSIZE 128 0 256 -MAX_PROC 0 50 16000 -MAX_REGION 0 500 160000 -NCLIST 170 120 16640 -MAXUP 100 15 16000 -NOFILES 110 60 11000 -NHINODE 128 64 8192 -NAUTOUP 10 0 60 -NGROUPS 8 0 128 -BDFLUSHR 30 1 300 -MAX_FLCKREC 0 50 16000 -PUTBUFSZ 8000 2000 20000 -MAXSLICE 100 25 100 -ULIMIT 4194303 2048 4194303 -* Streams Parameters -NSTREAM 64 1 32768 -NSTRPUSH 9 9 9 -NMUXLINK 192 1 4096 -STRMSGSZ 16384 4096 524288 -STRCTLSZ 1024 1024 1024 -STRMAXBLK 524288 4096 524288 -NSTRPAGES 500 0 8000 -STRSPLITFRAC 80 50 100 -NLOG 3 3 3 -NUMSP 64 1 256 -NUMTIM 16 1 8192 -NUMTRW 16 1 8192 -* Semaphore Parameters -SEMMAP 10 10 8192 -SEMMNI 10 10 8192 -SEMMNS 60 60 8192 -SEMMNU 30 10 8192 -SEMMSL 25 25 150 -SEMOPM 10 10 1024 -SEMUME 10 10 25 -SEMVMX 32767 32767 32767 -SEMAEM 16384 16384 16384 -* Shared Memory Parameters -SHMMAX 524288 131072 2147483647 -SHMMIN 1 1 1 -SHMMNI 100 100 2000 -FILE 0 100 64000 -NMOUNT 0 4 256 -NPROC 0 50 16000 -NREGION 0 500 160000 - - Set these values as follows: - - * NOFILES should be 4096 or 2048. - - * MAXUP should be 2048. - - To make changes to the kernel, use the idtune name parameter - command. idtune modifies the /etc/conf/cf.d/stune file for you. - For example, to change SEMMS to 200, execute this command as root: -# /etc/conf/bin/idtune SEMMNS 200 - - Then rebuild and reboot the kernel by issuing this command: -# /etc/conf/bin/idbuild -B && init 6 - - To tune the system, the proper parameter values to use depend on - the number of users accessing the application or database and size - the of the database (that is, the used buffer pool). The following - kernel parameters can be set with idtune: - - * SHMMAX (recommended setting: 128MB) and SHMSEG (recommended - setting: 15). These parameters have an influence on the MySQL - database engine to create user buffer pools. - - * NOFILES and MAXUP should be set to at least 2048. - - * MAXPROC should be set to at least 3000/4000 (depends on number - of users) or more. - - * The following formulas are recommended to calculate values for - SEMMSL, SEMMNS, and SEMMNU: -SEMMSL = 13 - 13 is what has been found to be the best for both Progress and - MySQL. -SEMMNS = SEMMSL x number of db servers to be run on the system - Set SEMMNS to the value of SEMMSL multiplied by the number of - database servers (maximum) that you are running on the system - at one time. -SEMMNU = SEMMNS - Set the value of SEMMNU to equal the value of SEMMNS. You - could probably set this to 75% of SEMMNS, but this is a - conservative estimate. - - You need to at least install the SCO OpenServer Linker and - Application Development Libraries or the OpenServer Development - System to use gcc. You cannot use the GCC Dev system without - installing one of these. - - You should get the FSU Pthreads package and install it first. This - can be found at - http://moss.csc.ncsu.edu/~mueller/ftp/pub/PART/pthreads.tar.gz. - You can also get a precompiled package from - ftp://ftp.zenez.com/pub/zenez/prgms/FSU-threads-3.14.tar.gz. - - FSU Pthreads can be compiled with SCO Unix 4.2 with tcpip, or - using OpenServer 3.0 or Open Desktop 3.0 (OS 3.0 ODT 3.0) with the - SCO Development System installed using a good port of GCC 2.5.x. - For ODT or OS 3.0, you need a good port of GCC 2.5.x. There are a - lot of problems without a good port. The port for this product - requires the SCO Unix Development system. Without it, you are - missing the libraries and the linker that is needed. You also need - SCO-3.2v4.2-includes.tar.gz. This file contains the changes to the - SCO Development include files that are needed to get MySQL to - build. You need to replace the existing system include files with - these modified header files. They can be obtained from - ftp://ftp.zenez.com/pub/zenez/prgms/SCO-3.2v4.2-includes.tar.gz. - - To build FSU Pthreads on your system, all you should need to do is - run GNU make. The Makefile in FSU-threads-3.14.tar.gz is set up to - make FSU-threads. - - You can run ./configure in the threads/src directory and select - the SCO OpenServer option. This command copies Makefile.SCO5 to - Makefile. Then run make. + If the server starts but you can't connect to it, you should make + sure that you have an entry in /etc/hosts that looks like this: +127.0.0.1 localhost - To install in the default /usr/include directory, log in as root, - and then cd to the thread/src directory and run make install. + This problem occurs only on systems that do not have a working + thread library and for which MySQL must be configured to use + MIT-pthreads. - Remember that you must use GNU make to build MySQL. + If you cannot get mysqld to start, you can try to make a trace + file to find the problem by using the --debug option. See MySQL + Internals: Porting + (http://forge.mysql.com/wiki/MySQL_Internals_Porting) -Note +2.13.2. Securing the Initial MySQL Accounts - If you don't start mysqld_safe as root, you should get only the - default 110 open files per process. mysqld writes a note about - this in the log file. + Part of the MySQL installation process is to set up the mysql + database that contains the grant tables: + + * Windows distributions contain preinitialized grant tables that + are installed automatically. - With SCO 3.2V4.2, you should use FSU Pthreads version 3.14 or - newer. The following configure command should work: -CFLAGS="-D_XOPEN_XPG4" CXX=gcc CXXFLAGS="-D_XOPEN_XPG4" \ -./configure \ - --prefix=/usr/local/mysql \ - --with-named-thread-libs="-lgthreads -lsocket -lgen -lgthreads" \ - --with-named-curses-libs="-lcurses" + * On Unix, the grant tables are populated by the + mysql_install_db program. Some installation methods run this + program for you. Others require that you execute it manually. + For details, see Section 2.13.1, "Unix Post-Installation + Procedures." - You may have problems with some include files. In this case, you - can find new SCO-specific include files at - ftp://ftp.zenez.com/pub/zenez/prgms/SCO-3.2v4.2-includes.tar.gz. + The grant tables define the initial MySQL user accounts and their + access privileges. These accounts are set up as follows: - You should unpack this file in the include directory of your MySQL - source tree. + * Accounts with the user name root are created. These are + superuser accounts that can do anything. The initial root + account passwords are empty, so anyone can connect to the + MySQL server as root --- without a password --- and be granted + all privileges. - SCO development notes: + + On Windows, one root account is created; this account + allows connecting from the local host only. The Windows + installer will optionally create an account allowing for + connections from any host only if the user selects the + Enable root access from remote machines option during + installation. - * MySQL should automatically detect FSU Pthreads and link mysqld - with -lgthreads -lsocket -lgthreads. + + On Unix, both root accounts are for connections from the + local host. Connections must be made from the local host + by specifying a host name of localhost for one of the + accounts, or the actual host name or IP number for the + other. - * The SCO development libraries are re-entrant in FSU Pthreads. - SCO claims that its library functions are re-entrant, so they - must be re-entrant with FSU Pthreads. FSU Pthreads on - OpenServer tries to use the SCO scheme to make re-entrant - libraries. + * Two anonymous-user accounts are created, each with an empty + user name. The anonymous accounts have no password, so anyone + can use them to connect to the MySQL server. - * FSU Pthreads (at least the version at ftp://ftp.zenez.com) - comes linked with GNU malloc. If you encounter problems with - memory usage, make sure that gmalloc.o is included in - libgthreads.a and libgthreads.so. + + On Windows, one anonymous account is for connections from + the local host. It has no global privileges. (Before + MySQL 5.1.16, it has all global privileges, just like the + root accounts.) The other is for connections from any + host and has all privileges for the test database and for + other databases with names that start with test. - * In FSU Pthreads, the following system calls are - pthreads-aware: read(), write(), getmsg(), connect(), - accept(), select(), and wait(). + + On Unix, both anonymous accounts are for connections from + the local host. Connections must be made from the local + host by specifying a host name of localhost for one of + the accounts, or the actual host name or IP number for + the other. These accounts have all privileges for the + test database and for other databases with names that + start with test_. - * The CSSA-2001-SCO.35.2 (the patch is listed in custom as - erg711905-dscr_remap security patch (version 2.0.0)) breaks - FSU threads and makes mysqld unstable. You have to remove this - one if you want to run mysqld on an OpenServer 5.0.6 machine. + As noted, none of the initial accounts have passwords. This means + that your MySQL installation is unprotected until you do something + about it: - * If you use SCO OpenServer 5, you may need to recompile FSU - pthreads with -DDRAFT7 in CFLAGS. Otherwise, InnoDB may hang - at a mysqld startup. + * If you want to prevent clients from connecting as anonymous + users without a password, you should either assign a password + to each anonymous account or else remove the accounts. - * SCO provides operating system patches at - ftp://ftp.sco.com/pub/openserver5 for OpenServer 5.0.x. + * You should assign a password to each MySQL root account. - * SCO provides security fixes and libsocket.so.2 at - ftp://ftp.sco.com/pub/security/OpenServer and - ftp://ftp.sco.com/pub/security/sse for OpenServer 5.0.x. + The following instructions describe how to set up passwords for + the initial MySQL accounts, first for the anonymous accounts and + then for the root accounts. Replace "newpwd" in the examples with + the actual password that you want to use. The instructions also + cover how to remove the anonymous accounts, should you prefer not + to allow anonymous access at all. - * Pre-OSR506 security fixes. Also, the telnetd fix at - ftp://stage.caldera.com/pub/security/openserver/ or - ftp://stage.caldera.com/pub/security/openserver/CSSA-2001-SCO. - 10/ as both libsocket.so.2 and libresolv.so.1 with - instructions for installing on pre-OSR506 systems. - It is probably a good idea to install these patches before - trying to compile/use MySQL. + You might want to defer setting the passwords until later, so that + you don't need to specify them while you perform additional setup + or testing. However, be sure to set them before using your + installation for production purposes. - Beginning with Legend/OpenServer 6.0.0, there are native threads - and no 2GB file size limit. + Anonymous Account Password Assignment -2.13.5.9. SCO OpenServer 6.0.x Notes + To assign passwords to the anonymous accounts, connect to the + server as root and then use either SET PASSWORD or UPDATE. In + either case, be sure to encrypt the password using the PASSWORD() + function. - OpenServer 6 includes these key improvements: + To use SET PASSWORD on Windows, do this: +shell> mysql -u root +mysql> SET PASSWORD FOR ''@'localhost' = PASSWORD('newpwd'); +mysql> SET PASSWORD FOR ''@'%' = PASSWORD('newpwd'); - * Larger file support up to 1 TB + To use SET PASSWORD on Unix, do this: +shell> mysql -u root +mysql> SET PASSWORD FOR ''@'localhost' = PASSWORD('newpwd'); +mysql> SET PASSWORD FOR ''@'host_name' = PASSWORD('newpwd'); - * Multiprocessor support increased from 4 to 32 processors + In the second SET PASSWORD statement, replace host_name with the + name of the server host. This is the name that is specified in the + Host column of the non-localhost record for root in the user + table. If you don't know what host name this is, issue the + following statement before using SET PASSWORD: +mysql> SELECT Host, User FROM mysql.user; - * Increased memory support up to 64GB + Look for the record that has root in the User column and something + other than localhost in the Host column. Then use that Host value + in the second SET PASSWORD statement. - * Extending the power of UnixWare into OpenServer 6 + Anonymous Account Removal - * Dramatic performance improvement + If you prefer to remove the anonymous accounts instead, do so as + follows: +shell> mysql -u root +mysql> DROP USER ''; - OpenServer 6.0.0 commands are organized as follows: + The DROP statement applies both to Windows and to Unix. On + Windows, if you want to remove only the anonymous account that has + the same privileges as root, do this instead: +shell> mysql -u root +mysql> DROP USER ''@'localhost'; - * /bin is for commands that behave exactly the same as on - OpenServer 5.0.x. + That account allows anonymous access but has full privileges, so + removing it improves security. - * /u95/bin is for commands that have better standards - conformance, for example Large File System (LFS) support. + root Account Password Assignment - * /udk/bin is for commands that behave the same as on UnixWare - 7.1.4. The default is for the LFS support. + You can assign passwords to the root accounts in several ways. The + following discussion demonstrates three methods: - The following is a guide to setting PATH on OpenServer 6. If the - user wants the traditional OpenServer 5.0.x then PATH should be - /bin first. If the user wants LFS support, the path should be - /u95/bin:/bin. If the user wants UnixWare 7 support first, the - path would be /udk/bin:/u95/bin:/bin:. + * Use the SET PASSWORD statement - Use the latest production release of MySQL. Should you choose to - use an older release of MySQL on OpenServer 6.0.x, you must use a - version of MySQL at least as recent as 3.22.13 to get fixes for - some portability and OS problems. + * Use the mysqladmin command-line client program - MySQL distribution files with names of the following form are tar - archives of media are tar archives of media images suitable for - installation with the SCO Software Manager (/etc/custom) on SCO - OpenServer 6: -mysql-PRODUCT-5.1.39-sco-osr6-i686.VOLS.tar + * Use the UPDATE statement - A distribution where PRODUCT is pro-cert is the Commercially - licensed MySQL Pro Certified server. A distribution where PRODUCT - is pro-gpl-cert is the MySQL Pro Certified server licensed under - the terms of the General Public License (GPL). + To assign passwords using SET PASSWORD, connect to the server as + root and issue SET PASSWORD statements. Be sure to encrypt the + password using the PASSWORD() function. - Select whichever distribution you wish to install and, after - download, extract the tar archive into an empty directory. For - example: -shell> mkdir /tmp/mysql-pro -shell> cd /tmp/mysql-pro -shell> tar xf /tmp/mysql-pro-cert-5.1.39-sco-osr6-i686.VOLS.tar + For Windows, do this: +shell> mysql -u root +mysql> SET PASSWORD FOR 'root'@'localhost' = PASSWORD('newpwd'); +mysql> SET PASSWORD FOR 'root'@'%' = PASSWORD('newpwd'); - Prior to installation, back up your data in accordance with the - procedures outlined in Section 2.12.1, "Upgrading MySQL." + For Unix, do this: +shell> mysql -u root +mysql> SET PASSWORD FOR 'root'@'localhost' = PASSWORD('newpwd'); +mysql> SET PASSWORD FOR 'root'@'host_name' = PASSWORD('newpwd'); - Remove any previously installed pkgadd version of MySQL: -shell> pkginfo mysql 2>&1 > /dev/null && pkgrm mysql + In the second SET PASSWORD statement, replace host_name with the + name of the server host. This is the same host name that you used + when you assigned the anonymous account passwords. - Install MySQL Pro from media images using the SCO Software - Manager: -shell> /etc/custom -p SCO:MySQL -i -z /tmp/mysql-pro + If the user table contains an account with User and Host values of + 'root' and '127.0.0.1', use an additional SET PASSWORD statement + to set that account's password: +mysql> SET PASSWORD FOR 'root'@'127.0.0.1' = PASSWORD('newpwd'); - Alternatively, the SCO Software Manager can be displayed - graphically by clicking on the Software Manager icon on the - desktop, selecting Software -> Install New, selecting the host, - selecting Media Images for the Media Device, and entering - /tmp/mysql-pro as the Image Directory. + To assign passwords to the root accounts using mysqladmin, execute + the following commands: +shell> mysqladmin -u root password "newpwd" +shell> mysqladmin -u root -h host_name password "newpwd" - After installation, run mkdev mysql as the root user to configure - your newly installed MySQL Pro Certified server. + These commands apply both to Windows and to Unix. In the second + command, replace host_name with the name of the server host. The + double quotes around the password are not always necessary, but + you should use them if the password contains spaces or other + characters that are special to your command interpreter. -Note + The mysqladmin method of setting the root account passwords does + not set the password for the 'root'@'127.0.0.1' account. To do so, + use SET PASSWORD as shown earlier. - The installation procedure for VOLS packages does not create the - mysql user and group that the package uses by default. You should - either create the mysql user and group, or else select a different - user and group using an option in mkdev mysql. - - If you wish to configure your MySQL Pro server to interface with - the Apache Web server via PHP, download and install the PHP update - from SCO at - ftp://ftp.sco.com/pub/updates/OpenServer/SCOSA-2006.17/. - - We have been able to compile MySQL with the following configure - command on OpenServer 6.0.x: -CC=cc CFLAGS="-D_FILE_OFFSET_BITS=64 -O3" \ -CXX=CC CXXFLAGS="-D_FILE_OFFSET_BITS=64 -O3" \ -./configure --prefix=/usr/local/mysql \ - --enable-thread-safe-client \ - --with-extra-charsets=complex \ - --build=i686-unknown-sysv5SCO_SV6.0.0 - - If you use gcc, you must use gcc 2.95.3 or newer. -CC=gcc CXX=g++ ... ./configure ... - - SCO provides OpenServer 6 operating system patches at - ftp://ftp.sco.com/pub/openserver6. - - SCO provides information about security fixes at - ftp://ftp.sco.com/pub/security/OpenServer. - - By default, the maximum file size on a OpenServer 6.0.0 system is - 1TB. Some operating system utilities have a limitation of 2GB. The - maximum possible file size on UnixWare 7 is 1TB with VXFS or HTFS. + You can also use UPDATE to modify the user table directly. The + following UPDATE statement assigns a password to all root + accounts: +shell> mysql -u root +mysql> UPDATE mysql.user SET Password = PASSWORD('newpwd') + -> WHERE User = 'root'; +mysql> FLUSH PRIVILEGES; - OpenServer 6 can be configured for large file support (file sizes - greater than 2GB) by tuning the UNIX kernel. + The UPDATE statement applies both to Windows and to Unix. - By default, the entries in /etc/conf/cf.d/mtune are set as - follows: -Value Default Min Max ------ ------- --- --- -SVMMLIM 0x9000000 0x1000000 0x7FFFFFFF -HVMMLIM 0x9000000 0x1000000 0x7FFFFFFF - - To make changes to the kernel, use the idtune name parameter - command. idtune modifies the /etc/conf/cf.d/stune file for you. To - set the kernel values, execute the following commands as root: -# /etc/conf/bin/idtune SDATLIM 0x7FFFFFFF -# /etc/conf/bin/idtune HDATLIM 0x7FFFFFFF -# /etc/conf/bin/idtune SVMMLIM 0x7FFFFFFF -# /etc/conf/bin/idtune HVMMLIM 0x7FFFFFFF -# /etc/conf/bin/idtune SFNOLIM 2048 -# /etc/conf/bin/idtune HFNOLIM 2048 - - Then rebuild and reboot the kernel by issuing this command: -# /etc/conf/bin/idbuild -B && init 6 - - To tune the system, the proper parameter values to use depend on - the number of users accessing the application or database and size - the of the database (that is, the used buffer pool). The following - kernel parameters can be set with idtune: - - * SHMMAX (recommended setting: 128MB) and SHMSEG (recommended - setting: 15). These parameters have an influence on the MySQL - database engine to create user buffer pools. - - * SFNOLIM and HFNOLIM should be at maximum 2048. - - * NPROC should be set to at least 3000/4000 (depends on number - of users). - - * The following formulas are recommended to calculate values for - SEMMSL, SEMMNS, and SEMMNU: -SEMMSL = 13 - 13 is what has been found to be the best for both Progress and - MySQL. -SEMMNS = SEMMSL x number of db servers to be run on the system - Set SEMMNS to the value of SEMMSL multiplied by the number of - database servers (maximum) that you are running on the system - at one time. -SEMMNU = SEMMNS - Set the value of SEMMNU to equal the value of SEMMNS. You - could probably set this to 75% of SEMMNS, but this is a - conservative estimate. - -2.13.5.10. SCO UnixWare 7.1.x and OpenUNIX 8.0.0 Notes - - Use the latest production release of MySQL. Should you choose to - use an older release of MySQL on UnixWare 7.1.x, you must use a - version of MySQL at least as recent as 3.22.13 to get fixes for - some portability and OS problems. - - We have been able to compile MySQL with the following configure - command on UnixWare 7.1.x: -CC="cc" CFLAGS="-I/usr/local/include" \ -CXX="CC" CXXFLAGS="-I/usr/local/include" \ -./configure --prefix=/usr/local/mysql \ - --enable-thread-safe-client \ - --with-innodb --with-openssl --with-extra-charsets=complex - - If you want to use gcc, you must use gcc 2.95.3 or newer. -CC=gcc CXX=g++ ... ./configure ... - - SCO provides operating system patches at - ftp://ftp.sco.com/pub/unixware7 for UnixWare 7.1.1, - ftp://ftp.sco.com/pub/unixware7/713/ for UnixWare 7.1.3, - ftp://ftp.sco.com/pub/unixware7/714/ for UnixWare 7.1.4, and - ftp://ftp.sco.com/pub/openunix8 for OpenUNIX 8.0.0. - - SCO provides information about security fixes at - ftp://ftp.sco.com/pub/security/OpenUNIX for OpenUNIX and - ftp://ftp.sco.com/pub/security/UnixWare for UnixWare. - - The UnixWare 7 file size limit is 1 TB with VXFS. Some OS - utilities have a limitation of 2GB. - - On UnixWare 7.1.4 you do not need to do anything to get large file - support, but to enable large file support on prior versions of - UnixWare 7.1.x, run fsadm. -# fsadm -Fvxfs -o largefiles / -# fsadm / * Note -# ulimit unlimited -# /etc/conf/bin/idtune SFSZLIM 0x7FFFFFFF ** Note -# /etc/conf/bin/idtune HFSZLIM 0x7FFFFFFF ** Note -# /etc/conf/bin/idbuild -B + After the passwords have been set, you must supply the appropriate + password whenever you connect to the server. For example, if you + want to use mysqladmin to shut down the server, you can do so + using this command: +shell> mysqladmin -u root -p shutdown +Enter password: (enter root password here) -* This should report "largefiles". -** 0x7FFFFFFF represents infinity for these values. +Note - Reboot the system using shutdown. + If you forget your root password after setting it up, Section + B.5.4.1, "How to Reset the Root Password," covers the procedure + for resetting it. - By default, the entries in /etc/conf/cf.d/mtune are set as - follows: -Value Default Min Max ------ ------- --- --- -SVMMLIM 0x9000000 0x1000000 0x7FFFFFFF -HVMMLIM 0x9000000 0x1000000 0x7FFFFFFF - - To make changes to the kernel, use the idtune name parameter - command. idtune modifies the /etc/conf/cf.d/stune file for you. To - set the kernel values, execute the following commands as root: -# /etc/conf/bin/idtune SDATLIM 0x7FFFFFFF -# /etc/conf/bin/idtune HDATLIM 0x7FFFFFFF -# /etc/conf/bin/idtune SVMMLIM 0x7FFFFFFF -# /etc/conf/bin/idtune HVMMLIM 0x7FFFFFFF -# /etc/conf/bin/idtune SFNOLIM 2048 -# /etc/conf/bin/idtune HFNOLIM 2048 - - Then rebuild and reboot the kernel by issuing this command: -# /etc/conf/bin/idbuild -B && init 6 - - To tune the system, the proper parameter values to use depend on - the number of users accessing the application or database and size - the of the database (that is, the used buffer pool). The following - kernel parameters can be set with idtune: - - * SHMMAX (recommended setting: 128MB) and SHMSEG (recommended - setting: 15). These parameters have an influence on the MySQL - database engine to create user buffer pools. - - * SFNOLIM and HFNOLIM should be at maximum 2048. - - * NPROC should be set to at least 3000/4000 (depends on number - of users). - - * The following formulas are recommended to calculate values for - SEMMSL, SEMMNS, and SEMMNU: -SEMMSL = 13 - 13 is what has been found to be the best for both Progress and - MySQL. -SEMMNS = SEMMSL x number of db servers to be run on the system - Set SEMMNS to the value of SEMMSL multiplied by the number of - database servers (maximum) that you are running on the system - at one time. -SEMMNU = SEMMNS - Set the value of SEMMNU to equal the value of SEMMNS. You - could probably set this to 75% of SEMMNS, but this is a - conservative estimate. + To set up additional accounts, you can use the GRANT statement. + For instructions, see Section 5.5.2, "Adding User Accounts." 2.14. Environment Variables @@ -9003,7 +7980,7 @@ SEMMNU = SEMMNS PATH Used by the shell to find MySQL programs. TMPDIR The directory where temporary files are created. TZ This should be set to your local time zone. See Section - B.1.4.6, "Time Zone Problems." + B.5.4.6, "Time Zone Problems." UMASK The user-file creation mode when creating files. See note following table. UMASK_DIR The user-directory creation mode when creating @@ -9047,9 +8024,9 @@ SEMMNU = SEMMNS sections describe how to do this. Perl support for MySQL must be installed if you want to run the - MySQL benchmark scripts; see Section 7.1.4, "The MySQL Benchmark + MySQL benchmark scripts; see Section 7.1.3, "The MySQL Benchmark Suite." It is also required for the MySQL Cluster ndb_size.pl - utility; see Section 17.6.21, "ndb_size.pl --- NDBCLUSTER Size + utility; see Section 17.4.21, "ndb_size.pl --- NDBCLUSTER Size Requirement Estimator." 2.15.1. Installing Perl on Unix === modified file 'INSTALL-WIN-SOURCE' --- a/INSTALL-WIN-SOURCE 2009-09-16 12:03:18 +0000 +++ b/INSTALL-WIN-SOURCE 2009-12-01 07:24:05 +0000 @@ -1,5 +1,5 @@ -2.10.6. Installing MySQL from Source on Windows +2.5.10. Installing MySQL from Source on Windows These instructions describe how to build binaries from source for MySQL 5.1 on Windows. Instructions are provided for building @@ -15,7 +15,7 @@ Note to use precompiled binary distributions of MySQL that are built specifically for optimal performance on Windows by Sun Microsystems, Inc. Instructions for installing binary - distributions are available in Section 2.3, "Installing MySQL on + distributions are available in Section 2.5, "Installing MySQL on Windows." To build MySQL on Windows from source, you must satisfy the @@ -68,7 +68,7 @@ Note * 3GB to 5GB of disk space. - The exact system requirements can be found here: + The exact system requirements for Visual Studio can be found here: http://msdn.microsoft.com/vstudio/Previous/2003/sysreqs/default.as px and http://msdn.microsoft.com/vstudio/products/sysreqs/default.aspx @@ -81,7 +81,7 @@ Note * Package a source distribution yourself from the latest Bazaar developer source tree. For instructions on pulling the latest - source files, see Section 2.10.3, "Installing from the + source files, see Section 2.3.3, "Installing from the Development Source Tree." If you find something not working as expected, or you have @@ -89,7 +89,7 @@ Note Windows, please send a message to the win32 mailing list. See Section 1.5.1, "MySQL Mailing Lists." -2.10.6.1. Building MySQL from Source Using CMake and Visual Studio +2.5.10.1. Building MySQL from Source Using CMake and Visual Studio You can build MySQL on Windows by using a combination of cmake and Microsoft Visual Studio .NET 2003 (7.1), Microsoft Visual Studio @@ -99,7 +99,7 @@ Note Note To compile from the source code on Windows you must use the - standard source distribution (for example, mysql-5.0.45.tar.gz). + standard source distribution (for example, mysql-5.1.41.tar.gz). You build from the same distribution as used to build MySQL on Unix, Linux and other platforms. Do not use the Windows Source distributions as they do not contain the necessary configuration @@ -264,5 +264,5 @@ C:\workdir> copy libmysql\libmysql.def C C:\workdir> xcopy sql-bench\*.* C:\mysql\bench /E After installation, set up and start the server in the same way as - for binary Windows distributions. See Section 2.3, "Installing + for binary Windows distributions. See Section 2.5, "Installing MySQL on Windows." === modified file 'man/comp_err.1' --- a/man/comp_err.1 2009-09-16 12:03:18 +0000 +++ b/man/comp_err.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBcomp_err\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBCOMP_ERR\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBCOMP_ERR\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -91,7 +91,8 @@ Display a help message and exit\&. .\} .\" comp_err: charset option .\" charset option: comp_err -\fB\-\-charset=\fR\fB\fIpath\fR\fR\fB, \-C \fR\fB\fIpath\fR\fR +\fB\-\-charset=\fR\fB\fIpath\fR\fR, +\fB\-C \fR\fB\fIpath\fR\fR .sp The character set directory\&. The default is \&.\&./sql/share/charsets\&. @@ -107,7 +108,8 @@ The character set directory\&. The defau .\} .\" comp_err: debug option .\" debug option: comp_err -\fB\-\-debug=\fR\fB\fIdebug_options\fR\fR\fB, \-# \fR\fB\fIdebug_options\fR\fR +\fB\-\-debug=\fR\fB\fIdebug_options\fR\fR, +\fB\-# \fR\fB\fIdebug_options\fR\fR .sp Write a debugging log\&. A typical \fIdebug_options\fR @@ -142,7 +144,8 @@ Print some debugging information when th .\} .\" comp_err: header_file option .\" header_file option: comp_err -\fB\-\-header_file=\fR\fB\fIfile_name\fR\fR\fB, \-H \fR\fB\fIfile_name\fR\fR +\fB\-\-header_file=\fR\fB\fIfile_name\fR\fR, +\fB\-H \fR\fB\fIfile_name\fR\fR .sp The name of the error header file\&. The default is mysqld_error\&.h\&. @@ -158,7 +161,8 @@ mysqld_error\&.h\&. .\} .\" comp_err: in_file option .\" in_file option: comp_err -\fB\-\-in_file=\fR\fB\fIfile_name\fR\fR\fB, \-F \fR\fB\fIfile_name\fR\fR +\fB\-\-in_file=\fR\fB\fIfile_name\fR\fR, +\fB\-F \fR\fB\fIfile_name\fR\fR .sp The name of the input file\&. The default is \&.\&./sql/share/errmsg\&.txt\&. @@ -174,7 +178,8 @@ The name of the input file\&. The defaul .\} .\" comp_err: name_file option .\" name_file option: comp_err -\fB\-\-name_file=\fR\fB\fIfile_name\fR\fR\fB, \-N \fR\fB\fIfile_name\fR\fR +\fB\-\-name_file=\fR\fB\fIfile_name\fR\fR, +\fB\-N \fR\fB\fIfile_name\fR\fR .sp The name of the error name file\&. The default is mysqld_ername\&.h\&. @@ -190,7 +195,8 @@ mysqld_ername\&.h\&. .\} .\" comp_err: out_dir option .\" out_dir option: comp_err -\fB\-\-out_dir=\fR\fB\fIpath\fR\fR\fB, \-D \fR\fB\fIpath\fR\fR +\fB\-\-out_dir=\fR\fB\fIpath\fR\fR, +\fB\-D \fR\fB\fIpath\fR\fR .sp The name of the output base directory\&. The default is \&.\&./sql/share/\&. @@ -206,7 +212,8 @@ The name of the output base directory\&. .\} .\" comp_err: out_file option .\" out_file option: comp_err -\fB\-\-out_file=\fR\fB\fIfile_name\fR\fR\fB, \-O \fR\fB\fIfile_name\fR\fR +\fB\-\-out_file=\fR\fB\fIfile_name\fR\fR, +\fB\-O \fR\fB\fIfile_name\fR\fR .sp The name of the output file\&. The default is errmsg\&.sys\&. @@ -222,7 +229,8 @@ errmsg\&.sys\&. .\} .\" comp_err: statefile option .\" statefile option: comp_err -\fB\-\-statefile=\fR\fB\fIfile_name\fR\fR\fB, \-S \fR\fB\fIfile_name\fR\fR +\fB\-\-statefile=\fR\fB\fIfile_name\fR\fR, +\fB\-S \fR\fB\fIfile_name\fR\fR .sp The name for the SQLSTATE header file\&. The default is sql_state\&.h\&. === modified file 'man/innochecksum.1' --- a/man/innochecksum.1 2009-09-16 12:03:18 +0000 +++ b/man/innochecksum.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBinnochecksum\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBINNOCHECKSUM\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBINNOCHECKSUM\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -29,7 +29,20 @@ innochecksum \- offline InnoDB file chec \fBinnochecksum\fR prints checksums for InnoDB -files\&. +files\&. This tool reads an +InnoDB +tablespace file, calculates the checksum for each page, compares the calculated checksum to the stored checksum, and reports mismatches, which indicate damaged pages\&. It was originally developed to speed up verifying the integrity of tablespace files after power outages but can also be used after file copies\&. Because checksum mismatches will cause +InnoDB +to deliberately shut down a running server, it can be preferable to use this tool rather than waiting for a server in production usage to encounter the damaged pages\&. +.PP +\fBinnochecksum\fR +cannot be used on tablespace files that the server already has open\&. For such files, you should use +CHECK TABLE +to check tables within the tablespace\&. +.PP +If checksum mismatches are found, you would normally restore the tablespace from backup or start the server and attempt to use +\fBmysqldump\fR +to make a backup of the tables within the tablespace\&. .PP Invoke \fBinnochecksum\fR === modified file 'man/make_win_bin_dist.1' --- a/man/make_win_bin_dist.1 2009-09-16 12:03:18 +0000 +++ b/man/make_win_bin_dist.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmake_win_bin_dist\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMAKE_WIN_BIN_DIST" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMAKE_WIN_BIN_DIST" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/msql2mysql.1' --- a/man/msql2mysql.1 2009-09-16 12:03:18 +0000 +++ b/man/msql2mysql.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmsql2mysql\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMSQL2MYSQL\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMSQL2MYSQL\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/my_print_defaults.1' --- a/man/my_print_defaults.1 2009-09-16 12:03:18 +0000 +++ b/man/my_print_defaults.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmy_print_defaults\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMY_PRINT_DEFAULTS" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMY_PRINT_DEFAULTS" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -100,7 +100,8 @@ Read only the given option file\&. .\} .\" my_print_defaults: debug option .\" debug option: my_print_defaults -\fB\-\-debug=\fR\fB\fIdebug_options\fR\fR\fB, \-# \fR\fB\fIdebug_options\fR\fR +\fB\-\-debug=\fR\fB\fIdebug_options\fR\fR, +\fB\-# \fR\fB\fIdebug_options\fR\fR .sp Write a debugging log\&. A typical \fIdebug_options\fR === modified file 'man/myisam_ftdump.1' --- a/man/myisam_ftdump.1 2009-09-16 12:03:18 +0000 +++ b/man/myisam_ftdump.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmyisam_ftdump\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYISAM_FTDUMP\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYISAM_FTDUMP\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/myisamchk.1' --- a/man/myisamchk.1 2009-09-16 12:03:18 +0000 +++ b/man/myisamchk.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmyisamchk\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYISAMCHK\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYISAMCHK\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -275,7 +275,8 @@ Display a help message and exit\&. Optio .\} .\" myisamchk: debug option .\" debug option: myisamchk -\fB\-\-debug=\fR\fB\fIdebug_options\fR\fR\fB, \-# \fR\fB\fIdebug_options\fR\fR +\fB\-\-debug=\fR\fB\fIdebug_options\fR\fR, +\fB\-# \fR\fB\fIdebug_options\fR\fR .sp Write a debugging log\&. A typical \fIdebug_options\fR @@ -520,7 +521,7 @@ system variable\&. For more information, myisam_stats_method in Section\ \&5.1.4, \(lqServer System Variables\(rq, and -Section\ \&7.4.6, \(lqMyISAM Index Statistics Collection\(rq\&. For MySQL 5\&.1, +Section\ \&7.4.7, \(lqMyISAM Index Statistics Collection\(rq\&. For MySQL 5\&.1, stats_method was added in MySQL 5\&.0\&.14\&. For older versions, the statistics collection method is equivalent to nulls_equal\&. @@ -662,6 +663,9 @@ If you are using and have plenty of memory, setting the key_buffer_size variable to a large value helps the repair operation run faster\&. +.sp +For a description of the output format, see +the section called \(lqMYISAMCHK TABLE INFORMATION\(rq\&. .RE .sp .RS 4 @@ -782,7 +786,11 @@ server is using the table and you are ru .\" files: repairing .PP \fBmyisamchk\fR -supports the following options for table repair operations: +supports the following options for table repair operations (operations performed when an option such as +\fB\-\-recover\fR +or +\fB\-\-safe\-recover\fR +is given): .sp .RS 4 .ie n \{\ @@ -844,7 +852,8 @@ Correct the checksum information for the .\} .\" myisamchk: data-file-length option .\" data-file-length option: myisamchk -\fB\-\-data\-file\-length=\fR\fB\fIlen\fR\fR\fB, \-D \fR\fB\fIlen\fR\fR +\fB\-\-data\-file\-length=\fR\fB\fIlen\fR\fR, +\fB\-D \fR\fB\fIlen\fR\fR .sp The maximum length of the data file (when re\-creating data file when it is \(lqfull\(rq)\&. @@ -864,6 +873,9 @@ The maximum length of the data file (whe \fB\-e\fR .sp Do a repair that tries to recover every possible row from the data file\&. Normally, this also finds a lot of garbage rows\&. Do not use this option unless you are desperate\&. +.sp +For a description of the output format, see +the section called \(lqMYISAMCHK TABLE INFORMATION\(rq\&. .RE .sp .RS 4 @@ -1172,7 +1184,10 @@ Find the record that a block at the give \fB\-\-description\fR, \fB\-d\fR .sp -Print some descriptive information about the table\&. +Print some descriptive information about the table\&. Specifying the +\fB\-\-verbose\fR +option once or twice produces additional information\&. See +the section called \(lqMYISAMCHK TABLE INFORMATION\(rq\&. .RE .sp .RS 4 @@ -1243,47 +1258,187 @@ sorts and moves records, it just overwri \fBmyisamchk\fR must unpack key blocks first, then re\-create indexes and pack the key blocks again\&. (In this case, re\-creating indexes is faster than updating offsets for each index\&.) .RE -.SH "MYISAMCHK MEMORY USAGE" -.\" memory usage: myisamchk +.SH "MYISAMCHK TABLE INFORMATION" +.\" table description: myisamchk +.\" tables: information +.\" examples: myisamchk output +.\" myisamchk: example output .PP -Memory allocation is important when you run -\fBmyisamchk\fR\&. +To obtain a description of a +MyISAM +table or statistics about it, use the commands shown here\&. The output from these commands is explained later in this section\&. +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +\fBmyisamchk \-d \fR\fB\fItbl_name\fR\fR +.sp +Runs \fBmyisamchk\fR -uses no more memory than its memory\-related variables are set to\&. If you are going to use +in +\(lqdescribe mode\(rq +to produce a description of your table\&. If you start the MySQL server with external locking disabled, \fBmyisamchk\fR -on very large tables, you should first decide how much memory you want it to use\&. The default is to use only about 3MB to perform repairs\&. By using larger values, you can get +may report an error for a table that is updated while it runs\&. However, because \fBmyisamchk\fR -to operate faster\&. For example, if you have more than 32MB RAM, you could use options such as these (in addition to any other options you might specify): +does not change the table in describe mode, there is no risk of destroying data\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +\fBmyisamchk \-dv \fR\fB\fItbl_name\fR\fR +.sp +Adding +\fB\-v\fR +runs +\fBmyisamchk\fR +in verbose mode so that it produces more information about the table\&. Adding +\fB\-v\fR +a second time produces even more information\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +\fBmyisamchk \-eis \fR\fB\fItbl_name\fR\fR +.sp +Shows only the most important information from a table\&. This operation is slow because it must read the entire table\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +\fBmyisamchk \-eiv \fR\fB\fItbl_name\fR\fR +.sp +This is like +\fB\-eis\fR, but tells you what is being done\&. +.RE +.PP +The +\fItbl_name\fR +argument can be either the name of a +MyISAM +table or the name of its index file, as described in +\fBmyisamchk\fR(1)\&. Multiple +\fItbl_name\fR +arguments can be given\&. +.PP +Suppose that a table named +person +has the following structure\&. (The +MAX_ROWS +table option is included so that in the example output from +\fBmyisamchk\fR +shown later, some values are smaller and fit the output format more easily\&.) .sp .if n \{\ .RS 4 .\} .nf -shell> \fBmyisamchk \-\-sort_buffer_size=16M \-\-key_buffer_size=16M \e\fR - \fB\-\-read_buffer_size=1M \-\-write_buffer_size=1M \&.\&.\&.\fR +CREATE TABLE person +( + id INT NOT NULL AUTO_INCREMENT, + last_name VARCHAR(20) NOT NULL, + first_name VARCHAR(20) NOT NULL, + birth DATE, + death DATE, + PRIMARY KEY (id), + INDEX (last_name, first_name), + INDEX (birth) +) MAX_ROWS = 1000000; .fi .if n \{\ .RE .\} .PP -Using -\fB\-\-sort_buffer_size=16M\fR -should probably be enough for most cases\&. +Suppose also that the table has these data and index file sizes: +.sp +.if n \{\ +.RS 4 +.\} +.nf +\-rw\-rw\-\-\-\- 1 mysql mysql 9347072 Aug 19 11:47 person\&.MYD +\-rw\-rw\-\-\-\- 1 mysql mysql 6066176 Aug 19 11:47 person\&.MYI +.fi +.if n \{\ +.RE +.\} .PP -Be aware that -\fBmyisamchk\fR -uses temporary files in -TMPDIR\&. If -TMPDIR -points to a memory file system, you may easily get out of memory errors\&. If this happens, run -\fBmyisamchk\fR -with the -\fB\-\-tmpdir=\fR\fB\fIpath\fR\fR -option to specify some directory located on a file system that has more space\&. +Example of +\fBmyisamchk \-dvv\fR +output: +.sp +.if n \{\ +.RS 4 +.\} +.nf +MyISAM file: person +Record format: Packed +Character set: latin1_swedish_ci (8) +File\-version: 1 +Creation time: 2009\-08\-19 16:47:41 +Recover time: 2009\-08\-19 16:47:56 +Status: checked,analyzed,optimized keys +Auto increment key: 1 Last value: 306688 +Data records: 306688 Deleted blocks: 0 +Datafile parts: 306688 Deleted data: 0 +Datafile pointer (bytes): 4 Keyfile pointer (bytes): 3 +Datafile length: 9347072 Keyfile length: 6066176 +Max datafile length: 4294967294 Max keyfile length: 17179868159 +Recordlength: 54 +table description: +Key Start Len Index Type Rec/key Root Blocksize +1 2 4 unique long 1 99328 1024 +2 6 20 multip\&. varchar prefix 512 3563520 1024 + 27 20 varchar 512 +3 48 3 multip\&. uint24 NULL 306688 6065152 1024 +Field Start Length Nullpos Nullbit Type +1 1 1 +2 2 4 no zeros +3 6 21 varchar +4 27 21 varchar +5 48 3 1 1 no zeros +6 51 3 1 2 no zeros +.fi +.if n \{\ +.RE +.\} .PP -When repairing, +Explanations for the types of information \fBmyisamchk\fR -also needs a lot of disk space: +produces are given here\&. +\(lqKeyfile\(rq +refers to the index file\&. +\(lqRecord\(rq +and +\(lqrow\(rq +are synonymous, as are +\(lqfield\(rq +and +\(lqcolumn\&.\(rq +.PP +The initial part of the table description contains these values: .sp .RS 4 .ie n \{\ @@ -1293,9 +1448,11 @@ also needs a lot of disk space: .sp -1 .IP \(bu 2.3 .\} -Double the size of the data file (the original file and a copy)\&. This space is not needed if you do a repair with -\fB\-\-quick\fR; in this case, only the index file is re\-created\&. -\fIThis space must be available on the same file system as the original data file\fR, as the copy is created in the same directory as the original\&. +MyISAM file +.sp +Name of the +MyISAM +(index) file\&. .RE .sp .RS 4 @@ -1306,7 +1463,13 @@ Double the size of the data file (the or .sp -1 .IP \(bu 2.3 .\} -Space for the new index file that replaces the old one\&. The old index file is truncated at the start of the repair operation, so you usually ignore this space\&. This space must be available on the same file system as the original data file\&. +Record format +.sp +The format used to store table rows\&. The preceding examples use +Fixed length\&. Other possible values are +Compressed +and +Packed\&. .RE .sp .RS 4 @@ -1317,30 +1480,964 @@ Space for the new index file that replac .sp -1 .IP \(bu 2.3 .\} -When using -\fB\-\-recover\fR -or -\fB\-\-sort\-recover\fR -(but not when using -\fB\-\-safe\-recover\fR), you need space for a sort buffer\&. The following formula yields the amount of space required: +Chararacter set +.sp +The table default character set\&. +.RE .sp -.if n \{\ .RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c .\} -.nf -(\fIlargest_key\fR + \fIrow_pointer_length\fR) \(mu \fInumber_of_rows\fR \(mu 2 -.fi -.if n \{\ +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +File\-version +.sp +Version of +MyISAM +format\&. Currently always 1\&. .RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 .\} +Creation time .sp -You can check the length of the keys and the -row_pointer_length -with -\fBmyisamchk \-dv \fR\fB\fItbl_name\fR\fR\&. This space is allocated in the temporary directory (specified by -TMPDIR -or -\fB\-\-tmpdir=\fR\fB\fIpath\fR\fR)\&. +When the data file was created\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Recover time +.sp +When the index/data file was last reconstructed\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Status +.sp +Table status flags\&. Possible values are +crashed, +open, +changed, +analyzed, +optimized keys, and +sorted index pages\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Auto increment key, +Last value +.sp +The key number associated the table\'s +AUTO_INCREMENT +column, and the most recently generated value for this column\&. These fields do not appear if there is no such column\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Data records +.sp +The number of rows in the table\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Deleted blocks +.sp +How many deleted blocks still have reserved space\&. You can optimize your table to minimize this space\&. See +Section\ \&6.4.4, \(lqTable Optimization\(rq\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Datafile parts +.sp +For dynamic\-row format, this indicates how many data blocks there are\&. For an optimized table without fragmented rows, this is the same as +Data records\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Deleted data +.sp +How many bytes of unreclaimed deleted data there are\&. You can optimize your table to minimize this space\&. See +Section\ \&6.4.4, \(lqTable Optimization\(rq\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Datafile pointer +.sp +The size of the data file pointer, in bytes\&. It is usually 2, 3, 4, or 5 bytes\&. Most tables manage with 2 bytes, but this cannot be controlled from MySQL yet\&. For fixed tables, this is a row address\&. For dynamic tables, this is a byte address\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Keyfile pointer +.sp +The size of the index file pointer, in bytes\&. It is usually 1, 2, or 3 bytes\&. Most tables manage with 2 bytes, but this is calculated automatically by MySQL\&. It is always a block address\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Max datafile length +.sp +How long the table data file can become, in bytes\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Max keyfile length +.sp +How long the table index file can become, in bytes\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Recordlength +.sp +How much space each row takes, in bytes\&. +.RE +.PP +The +table description +part of the output includes a list of all keys in the table\&. For each key, +\fBmyisamchk\fR +displays some low\-level information: +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Key +.sp +This key\'s number\&. This value is shown only for the first column of the key\&. If this value is missing, the line corresponds to the second or later column of a multiple\-column key\&. For the table shown in the example, there are two +table description +lines for the second index\&. This indicates that it is a multiple\-part index with two parts\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Start +.sp +Where in the row this portion of the index starts\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Len +.sp +How long this portion of the index is\&. For packed numbers, this should always be the full length of the column\&. For strings, it may be shorter than the full length of the indexed column, because you can index a prefix of a string column\&. The total length of a multiple\-part key is the sum of the +Len +values for all key parts\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Index +.sp +Whether a key value can exist multiple times in the index\&. Possible values are +unique +or +multip\&. +(multiple)\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Type +.sp +What data type this portion of the index has\&. This is a +MyISAM +data type with the possible values +packed, +stripped, or +empty\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Root +.sp +Address of the root index block\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Blocksize +.sp +The size of each index block\&. By default this is 1024, but the value may be changed at compile time when MySQL is built from source\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Rec/key +.sp +This is a statistical value used by the optimizer\&. It tells how many rows there are per value for this index\&. A unique index always has a value of 1\&. This may be updated after a table is loaded (or greatly changed) with +\fBmyisamchk \-a\fR\&. If this is not updated at all, a default value of 30 is given\&. +.RE +.PP +The last part of the output provides information about each column: +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Field +.sp +The column number\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Start +.sp +The byte position of the column within table rows\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Length +.sp +The length of the column in bytes\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Nullpos, +Nullbit +.sp +For columns that can be +NULL, +MyISAM +stores +NULL +values as a flag in a byte\&. Depending on how many nullable columns there are, there can be one or more bytes used for this purpose\&. The +Nullpos +and +Nullbit +values, if nonempty, indicate which byte and bit contains that flag indicating whether the column is +NULL\&. +.sp +The position and number of bytes used to store +NULL +flags is shown in the line for field 1\&. This is why there are six +Field +lines for the +person +table even though it has only five columns\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Type +.sp +The data type\&. The value may contain any of the following descriptors: +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +constant +.sp +All rows have the same value\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +no endspace +.sp +Do not store endspace\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +no endspace, not_always +.sp +Do not store endspace and do not do endspace compression for all values\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +no endspace, no empty +.sp +Do not store endspace\&. Do not store empty values\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +table\-lookup +.sp +The column was converted to an +ENUM\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +zerofill(\fIN\fR) +.sp +The most significant +\fIN\fR +bytes in the value are always 0 and are not stored\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +no zeros +.sp +Do not store zeros\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +always zero +.sp +Zero values are stored using one bit\&. +.RE +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Huff tree +.sp +The number of the Huffman tree associated with the column\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Bits +.sp +The number of bits used in the Huffman tree\&. +.RE +.PP +The +Huff tree +and +Bits +fields are displayed if the table has been compressed with +\fBmyisampack\fR\&. See +\fBmyisampack\fR(1), for an example of this information\&. +.PP +Example of +\fBmyisamchk \-eiv\fR +output: +.sp +.if n \{\ +.RS 4 +.\} +.nf +Checking MyISAM file: person +Data records: 306688 Deleted blocks: 0 +\- check file\-size +\- check record delete\-chain +No recordlinks +\- check key delete\-chain +block_size 1024: +\- check index reference +\- check data record references index: 1 +Key: 1: Keyblocks used: 98% Packed: 0% Max levels: 3 +\- check data record references index: 2 +Key: 2: Keyblocks used: 99% Packed: 97% Max levels: 3 +\- check data record references index: 3 +Key: 3: Keyblocks used: 98% Packed: \-14% Max levels: 3 +Total: Keyblocks used: 98% Packed: 89% +\- check records and index references +\fI*** LOTS OF ROW NUMBERS DELETED ***\fR +Records: 306688 M\&.recordlength: 25 Packed: 83% +Recordspace used: 97% Empty space: 2% Blocks/Record: 1\&.00 +Record blocks: 306688 Delete blocks: 0 +Record data: 7934464 Deleted data: 0 +Lost space: 256512 Linkdata: 1156096 +User time 43\&.08, System time 1\&.68 +Maximum resident set size 0, Integral resident set size 0 +Non\-physical pagefaults 0, Physical pagefaults 0, Swaps 0 +Blocks in 0 out 7, Messages in 0 out 0, Signals 0 +Voluntary context switches 0, Involuntary context switches 0 +Maximum memory usage: 1046926 bytes (1023k) +.fi +.if n \{\ +.RE +.\} +.PP +\fBmyisamchk \-eiv\fR +output includes the following information: +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Data records +.sp +The number of rows in the table\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Deleted blocks +.sp +How many deleted blocks still have reserved space\&. You can optimize your table to minimize this space\&. See +Section\ \&6.4.4, \(lqTable Optimization\(rq\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Key +.sp +The key number\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Keyblocks used +.sp +What percentage of the keyblocks are used\&. When a table has just been reorganized with +\fBmyisamchk\fR, the values are very high (very near theoretical maximum)\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Packed +.sp +MySQL tries to pack key values that have a common suffix\&. This can only be used for indexes on +CHAR +and +VARCHAR +columns\&. For long indexed strings that have similar leftmost parts, this can significantly reduce the space used\&. In the preceding example, the second key is 40 bytes long and a 97% reduction in space is achieved\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Max levels +.sp +How deep the B\-tree for this key is\&. Large tables with long key values get high values\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Records +.sp +How many rows are in the table\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +M\&.recordlength +.sp +The average row length\&. This is the exact row length for tables with fixed\-length rows, because all rows have the same length\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Packed +.sp +MySQL strips spaces from the end of strings\&. The +Packed +value indicates the percentage of savings achieved by doing this\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Recordspace used +.sp +What percentage of the data file is used\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Empty space +.sp +What percentage of the data file is unused\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Blocks/Record +.sp +Average number of blocks per row (that is, how many links a fragmented row is composed of)\&. This is always 1\&.0 for fixed\-format tables\&. This value should stay as close to 1\&.0 as possible\&. If it gets too large, you can reorganize the table\&. See +Section\ \&6.4.4, \(lqTable Optimization\(rq\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Recordblocks +.sp +How many blocks (links) are used\&. For fixed\-format tables, this is the same as the number of rows\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Deleteblocks +.sp +How many blocks (links) are deleted\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Recorddata +.sp +How many bytes in the data file are used\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Deleted data +.sp +How many bytes in the data file are deleted (unused)\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Lost space +.sp +If a row is updated to a shorter length, some space is lost\&. This is the sum of all such losses, in bytes\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Linkdata +.sp +When the dynamic table format is used, row fragments are linked with pointers (4 to 7 bytes each)\&. +Linkdata +is the sum of the amount of storage used by all such pointers\&. +.RE +.SH "MYISAMCHK MEMORY USAGE" +.\" memory usage: myisamchk +.PP +Memory allocation is important when you run +\fBmyisamchk\fR\&. +\fBmyisamchk\fR +uses no more memory than its memory\-related variables are set to\&. If you are going to use +\fBmyisamchk\fR +on very large tables, you should first decide how much memory you want it to use\&. The default is to use only about 3MB to perform repairs\&. By using larger values, you can get +\fBmyisamchk\fR +to operate faster\&. For example, if you have more than 32MB RAM, you could use options such as these (in addition to any other options you might specify): +.sp +.if n \{\ +.RS 4 +.\} +.nf +shell> \fBmyisamchk \-\-sort_buffer_size=16M \e\fR + \fB\-\-key_buffer_size=16M \e\fR + \fB\-\-read_buffer_size=1M \e\fR + \fB\-\-write_buffer_size=1M \&.\&.\&.\fR +.fi +.if n \{\ +.RE +.\} +.PP +Using +\fB\-\-sort_buffer_size=16M\fR +should probably be enough for most cases\&. +.PP +Be aware that +\fBmyisamchk\fR +uses temporary files in +TMPDIR\&. If +TMPDIR +points to a memory file system, out of memory errors can easily occur\&. If this happens, run +\fBmyisamchk\fR +with the +\fB\-\-tmpdir=\fR\fB\fIpath\fR\fR +option to specify a directory located on a file system that has more space\&. +.PP +When performing repair operations, +\fBmyisamchk\fR +also needs a lot of disk space: +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Twice the size of the data file (the original file and a copy)\&. This space is not needed if you do a repair with +\fB\-\-quick\fR; in this case, only the index file is re\-created\&. +\fIThis space must be available on the same file system as the original data file\fR, as the copy is created in the same directory as the original\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +Space for the new index file that replaces the old one\&. The old index file is truncated at the start of the repair operation, so you usually ignore this space\&. This space must be available on the same file system as the original data file\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +When using +\fB\-\-recover\fR +or +\fB\-\-sort\-recover\fR +(but not when using +\fB\-\-safe\-recover\fR), you need space on disk for sorting\&. This space is allocated in the temporary directory (specified by +TMPDIR +or +\fB\-\-tmpdir=\fR\fB\fIpath\fR\fR)\&. The following formula yields the amount of space required: +.sp +.if n \{\ +.RS 4 +.\} +.nf +(\fIlargest_key\fR + \fIrow_pointer_length\fR) \(mu \fInumber_of_rows\fR \(mu 2 +.fi +.if n \{\ +.RE +.\} +.sp +You can check the length of the keys and the +\fIrow_pointer_length\fR +with +\fBmyisamchk \-dv \fR\fB\fItbl_name\fR\fR +(see +the section called \(lqMYISAMCHK TABLE INFORMATION\(rq)\&. The +\fIrow_pointer_length\fR +and +\fInumber_of_rows\fR +values are the +Datafile pointer +and +Data records +values in the table description\&. To determine the +\fIlargest_key\fR +value, check the +Key +lines in the table description\&. The +Len +column indicates the number of bytes for each key part\&. For a multiple\-column index, the key size is the sum of the +Len +values for all key parts\&. .RE .PP If you have a problem with disk space during repair, you can try === modified file 'man/myisamlog.1' --- a/man/myisamlog.1 2009-09-16 12:03:18 +0000 +++ b/man/myisamlog.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmyisamlog\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYISAMLOG\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYISAMLOG\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/myisampack.1' --- a/man/myisampack.1 2009-09-16 12:03:18 +0000 +++ b/man/myisampack.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmyisampack\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYISAMPACK\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYISAMPACK\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -652,9 +652,11 @@ After join The number of distinct Huffman trees left after joining trees to save some header space\&. .RE .PP -After a table has been compressed, +After a table has been compressed, the +Field +lines displayed by \fBmyisamchk \-dvv\fR -prints additional information about each column: +include additional information about each column: .sp .RS 4 .ie n \{\ === modified file 'man/mysql-stress-test.pl.1' --- a/man/mysql-stress-test.pl.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql-stress-test.pl.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql-stress-test.pl\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/08/2009 +.\" Date: 10/29/2009 .\" Manual: MySQL Database System .\" Source: MySQL .\" Language: English .\" -.TH "\FBMYSQL\-STRESS\-TE" "1" "08/08/2009" "MySQL" "MySQL Database System" +.TH "\FBMYSQL\-STRESS\-TE" "1" "10/29/2009" "MySQL" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysql-test-run.pl.1' --- a/man/mysql-test-run.pl.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql-test-run.pl.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql-test-run.pl\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/08/2009 +.\" Date: 10/29/2009 .\" Manual: MySQL Database System .\" Source: MySQL .\" Language: English .\" -.TH "\FBMYSQL\-TEST\-RUN\" "1" "08/08/2009" "MySQL" "MySQL Database System" +.TH "\FBMYSQL\-TEST\-RUN\" "1" "10/29/2009" "MySQL" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -700,6 +700,9 @@ Specify a file that contains a list of t code rather than [ fail ] if they fail\&. This option was added in MySQL 5\&.1\&.33/6\&.0\&.11\&. +.sp +For an example of a file that might be specified via this option, see +mysql\-test/collections/default\&.experimental\&. .RE .sp .RS 4 @@ -1132,6 +1135,26 @@ not to generate a timing file\&. .sp -1 .IP \(bu 2.3 .\} +.\" mysql-test-run.pl: parallel option +.\" parallel option: mysql-test-run.pl +\fB\-\-parallel={\fR\fB\fIN\fR\fR\fB|auto}\fR +.sp +Run tests using +\fIN\fR +parallel threads\&. By default, 1 thread is used\&. Use +\fB\-\-parallel=auto\fR +for auto\-setting of +\fIN\fR\&. This option was added in MySQL 5\&.1\&.36\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} .\" mysql-test-run.pl: ps-protocol option .\" ps-protocol option: mysql-test-run.pl \fB\-\-ps\-protocol\fR === modified file 'man/mysql.1' --- a/man/mysql.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -334,8 +334,18 @@ to display result set metadata\&. .sp Use \fIcharset_name\fR -as the default character set\&. See -Section\ \&9.2, \(lqThe Character Set Used for Data and Sorting\(rq\&. +as the default character set for the client and connection\&. +.sp +A common issue that can occur when the operating system uses +utf8 +or another multi\-byte character set is that output from the +\fBmysql\fR +client is formatted incorrectly, due to the fact that the MySQL client uses the +latin1 +character set by default\&. You can usually fix such issues by using this option to force the client to use the system character set instead\&. +.sp +See +Section\ \&9.2, \(lqThe Character Set Used for Data and Sorting\(rq, for more information\&. .RE .sp .RS 4 @@ -648,7 +658,7 @@ PAGER environment variable\&. Valid pagers are \fBless\fR, \fBmore\fR, -\fBcat [> filename]\fR, and so forth\&. This option works only on Unix\&. It does not work in batch mode\&. To disable paging, use +\fBcat [> filename]\fR, and so forth\&. This option works only on Unix and only in interactive mode\&. To disable paging, use \fB\-\-skip\-pager\fR\&. the section called \(lqMYSQL COMMANDS\(rq, discusses output paging further\&. .RE @@ -1026,7 +1036,7 @@ Display output in table format\&. This i .\" tee option: mysql \fB\-\-tee=\fR\fB\fIfile_name\fR\fR .sp -Append a copy of output to the given file\&. This option does not work in batch mode\&. +Append a copy of output to the given file\&. This option works only in interactive mode\&. the section called \(lqMYSQL COMMANDS\(rq, discusses tee files further\&. .RE .sp @@ -1523,7 +1533,7 @@ is set to something other than the defau \(lq;\(rq, instances of that character are sent to the server without interpretation\&. However, the server itself still interprets \(lq;\(rq as a statement delimiter and processes statements accordingly\&. This behavior on the server side comes into play for multiple\-statement execution (see -Section\ \&21.10.12, \(lqC API Support for Multiple Statement Execution\(rq), and for parsing the body of stored procedures and functions, triggers, and events (see +Section\ \&21.9.12, \(lqC API Support for Multiple Statement Execution\(rq), and for parsing the body of stored procedures and functions, triggers, and events (see Section\ \&19.1, \(lqDefining Stored Programs\(rq)\&. .RE .sp @@ -1680,7 +1690,7 @@ option when you invoke \fBmysql\fR checks the value of the PAGER -environment variable and sets the pager to that\&. +environment variable and sets the pager to that\&. Pager functionality works only in interactive mode\&. .sp Output paging can be enabled interactively with the \fBpager\fR @@ -1853,7 +1863,7 @@ By using the option when you invoke \fBmysql\fR, you can log statements and their output\&. All the data displayed on the screen is appended into a given file\&. This can be very useful for debugging purposes also\&. \fBmysql\fR -flushes results to the file after each statement, just before it prints its next prompt\&. +flushes results to the file after each statement, just before it prints its next prompt\&. Tee functionality works only in interactive mode\&. .sp You can enable this feature interactively with the \fBtee\fR @@ -2334,7 +2344,7 @@ prompt=(\e\eu@\e\eh) [\e\ed]>\e\e_ .sp In this example, note that the backslashes are doubled\&. If you set the prompt using the prompt -option in an option file, it is advisable to double the backslashes when using the special prompt options\&. There is some overlap in the set of allowable prompt options and the set of special escape sequences that are recognized in option files\&. (These sequences are listed in +option in an option file, it is advisable to double the backslashes when using the special prompt options\&. There is some overlap in the set of allowable prompt options and the set of special escape sequences that are recognized in option files\&. (The rules for escape sequences in option files are listed in Section\ \&4.2.3.3, \(lqUsing Option Files\(rq\&.) The overlap may cause you problems if you use single backslashes\&. For example, \es is interpreted as a space rather than as the current seconds value\&. The following example shows how to define a prompt within an option file to include the current time in @@ -2586,6 +2596,12 @@ SELECT \'<info_to_display>\' AS \' \'; The statement shown outputs <info_to_display>\&. .PP +You can also invoke +\fBmysql\fR +with the +\fB\-\-verbose\fR +option, which causes each statement to be displayed before the result that it produces\&. +.PP As of MySQL 5\&.1\&.23, \fBmysql\fR ignores Unicode byte order mark (BOM) characters at the beginning of input files\&. Previously, it read them and sent them to the server, resulting in a syntax error\&. Presence of a BOM does not cause @@ -2785,7 +2801,7 @@ client with the option\&. .PP For more information about auto\-reconnect and its effect on state information when a reconnection occurs, see -Section\ \&21.10.11, \(lqControlling Automatic Reconnection Behavior\(rq\&. +Section\ \&21.9.11, \(lqControlling Automatic Reconnection Behavior\(rq\&. .SH "COPYRIGHT" .br .PP === modified file 'man/mysql.server.1' --- a/man/mysql.server.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql.server.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql.server\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL\&.SERVER\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL\&.SERVER\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -40,7 +40,7 @@ script will be installed in the /etc/init\&.d directory with the name mysql\&. You need not install it manually\&. See -Section\ \&2.4, \(lqInstalling MySQL from RPM Packages on Linux\(rq, for more information on the Linux RPM packages\&. +Section\ \&2.6.1, \(lqInstalling MySQL from RPM Packages on Linux\(rq, for more information on the Linux RPM packages\&. .PP Some vendors provide RPM packages that install a startup script under a different name such as \fBmysqld\fR\&. @@ -48,7 +48,7 @@ Some vendors provide RPM packages that i If you install MySQL from a source distribution or using a binary distribution format that does not install \fBmysql\&.server\fR automatically, you can install it manually\&. Instructions are provided in -Section\ \&2.11.2.2, \(lqStarting and Stopping MySQL Automatically\(rq\&. +Section\ \&2.13.1.2, \(lqStarting and Stopping MySQL Automatically\(rq\&. .PP \fBmysql\&.server\fR reads options from the === modified file 'man/mysql_client_test.1' --- a/man/mysql_client_test.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_client_test.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_client_test\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/08/2009 +.\" Date: 10/29/2009 .\" Manual: MySQL Database System .\" Source: MySQL .\" Language: English .\" -.TH "\FBMYSQL_CLIENT_TEST" "1" "08/08/2009" "MySQL" "MySQL Database System" +.TH "\FBMYSQL_CLIENT_TEST" "1" "10/29/2009" "MySQL" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysql_config.1' --- a/man/mysql_config.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_config.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_config\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL_CONFIG\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL_CONFIG\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysql_convert_table_format.1' --- a/man/mysql_convert_table_format.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_convert_table_format.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_convert_table_format\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL_CONVERT_TAB" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL_CONVERT_TAB" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysql_find_rows.1' --- a/man/mysql_find_rows.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_find_rows.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_find_rows\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL_FIND_ROWS\F" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL_FIND_ROWS\F" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysql_fix_extensions.1' --- a/man/mysql_fix_extensions.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_fix_extensions.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_fix_extensions\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL_FIX_EXTENSI" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL_FIX_EXTENSI" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysql_fix_privilege_tables.1' --- a/man/mysql_fix_privilege_tables.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_fix_privilege_tables.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_fix_privilege_tables\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL_FIX_PRIVILE" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL_FIX_PRIVILE" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysql_install_db.1' --- a/man/mysql_install_db.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_install_db.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_install_db\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL_INSTALL_DB\" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL_INSTALL_DB\" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -81,7 +81,7 @@ with the and \fB\-\-skip\-grant\-tables\fR options (see -Section\ \&2.10.2, \(lqTypical configure Options\(rq)\&. If MySQL was configured with the +Section\ \&2.3.2, \(lqTypical configure Options\(rq)\&. If MySQL was configured with the \fB\-\-disable\-grant\-options\fR option, \fB\-\-bootstrap\fR === modified file 'man/mysql_secure_installation.1' --- a/man/mysql_secure_installation.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_secure_installation.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_secure_installation\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL_SECURE_INST" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL_SECURE_INST" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysql_setpermission.1' --- a/man/mysql_setpermission.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_setpermission.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_setpermission\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL_SETPERMISSI" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL_SETPERMISSI" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysql_tzinfo_to_sql.1' --- a/man/mysql_tzinfo_to_sql.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_tzinfo_to_sql.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_tzinfo_to_sql\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL_TZINFO_TO_S" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL_TZINFO_TO_S" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysql_upgrade.1' --- a/man/mysql_upgrade.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_upgrade.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_upgrade\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL_UPGRADE\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL_UPGRADE\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -41,7 +41,7 @@ script, which should no longer be used\& If a table is found to have a possible incompatibility, \fBmysql_upgrade\fR performs a table check\&. If any problems are found, a table repair is attempted\&. If the table cannot be repaired, see -Section\ \&2.12.4, \(lqRebuilding or Repairing Tables or Indexes\(rq +Section\ \&2.4.4, \(lqRebuilding or Repairing Tables or Indexes\(rq for manual table repair strategies\&. .if n \{\ .sp @@ -59,11 +59,11 @@ for manual table repair strategies\&. You should always back up your current MySQL installation \fIbefore\fR performing an upgrade\&. See -Section\ \&6.1, \(lqDatabase Backups\(rq\&. +Section\ \&6.1, \(lqDatabase Backup Methods\(rq\&. .PP Some upgrade incompatibilities may require special handling before you upgrade your MySQL installation and run \fBmysql_upgrade\fR\&. See -Section\ \&2.12.1, \(lqUpgrading MySQL\(rq, for instructions on determining whether any such incompatibilities apply to your installation and how to handle them\&. +Section\ \&2.4.1, \(lqUpgrading MySQL\(rq, for instructions on determining whether any such incompatibilities apply to your installation and how to handle them\&. .sp .5v .RE .PP @@ -189,7 +189,7 @@ If you install MySQL from RPM packages o \fBmysql_upgrade\fR is included in the server RPM but requires the client RPM because the latter includes \fBmysqlcheck\fR\&. (See -Section\ \&2.4, \(lqInstalling MySQL from RPM Packages on Linux\(rq\&.) +Section\ \&2.6.1, \(lqInstalling MySQL from RPM Packages on Linux\(rq\&.) .PP In MySQL 5\&.1\&.7, \fBmysql_upgrade \fR @@ -352,6 +352,26 @@ root\&. .sp Verbose mode\&. Print more information about what the program does\&. .RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} +.\" mysql_upgrade: write-binlog option +.\" write-binlog option: mysql_upgrade +\fB\-\-write\-binlog\fR +.sp +Cause binary logging to be enabled while +\fBmysql_upgrade\fR +runs\&. This is the default behavior; to disable binary logging during the upgrade, use the inverse of this option (that is, start the program with +\fB\-\-skip\-write\-binlog\fR)\&. +.sp +This option was introduced in MySQL 5\&.1\&.40\&. +.RE .SH "COPYRIGHT" .br .PP === modified file 'man/mysql_waitpid.1' --- a/man/mysql_waitpid.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_waitpid.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_waitpid\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL_WAITPID\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL_WAITPID\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysql_zap.1' --- a/man/mysql_zap.1 2009-09-16 12:03:18 +0000 +++ b/man/mysql_zap.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysql_zap\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQL_ZAP\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQL_ZAP\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysqlaccess.1' --- a/man/mysqlaccess.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqlaccess.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqlaccess\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLACCESS\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLACCESS\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysqladmin.1' --- a/man/mysqladmin.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqladmin.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqladmin\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLADMIN\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLADMIN\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysqlbinlog.1' --- a/man/mysqlbinlog.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqlbinlog.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqlbinlog\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLBINLOG\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLBINLOG\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -855,7 +855,7 @@ You can pipe the output of into the \fBmysql\fR client to execute the statements contained in the binary log\&. This is used to recover from a crash when you have an old backup (see -Section\ \&6.1, \(lqDatabase Backups\(rq)\&. For example: +Section\ \&6.1, \(lqDatabase Backup Methods\(rq)\&. For example: .sp .if n \{\ .RS 4 === modified file 'man/mysqlbug.1' --- a/man/mysqlbug.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqlbug.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqlbug\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLBUG\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLBUG\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysqlcheck.1' --- a/man/mysqlcheck.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqlcheck.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqlcheck\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLCHECK\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLCHECK\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -34,7 +34,14 @@ The \fBmysqlcheck\fR client performs table maintenance: It checks, repairs, optimizes, or analyzes tables\&. .PP -Each table is locked and therefore unavailable to other sessions while it is being processed\&. Table maintenance operations can be time\-consuming, particularly for large tables\&. If you use the +Each table is locked and therefore unavailable to other sessions while it is being processed, although for check operations, the table is locked with a +READ +lock only (see +Section\ \&12.4.5, \(lqLOCK TABLES and UNLOCK TABLES Syntax\(rq, for more information about +READ +and +WRITE +locks)\&. Table maintenance operations can be time\-consuming, particularly for large tables\&. If you use the \fB\-\-databases\fR or \fB\-\-all\-databases\fR @@ -94,7 +101,7 @@ note : The storage engine for the ta If \fBmysqlcheck\fR is unable to repair a table, see -Section\ \&2.12.4, \(lqRebuilding or Repairing Tables or Indexes\(rq +Section\ \&2.4.4, \(lqRebuilding or Repairing Tables or Indexes\(rq for manual table repair strategies\&. This will be the case, for example, for InnoDB tables, which can be checked with === modified file 'man/mysqld.8' --- a/man/mysqld.8 2009-09-16 12:03:18 +0000 +++ b/man/mysqld.8 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqld\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLD\FR" "8" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLD\FR" "8" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysqld_multi.1' --- a/man/mysqld_multi.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqld_multi.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqld_multi\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLD_MULTI\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLD_MULTI\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -51,7 +51,7 @@ to specify which servers you want to sta [mysqld] group used for starting \fBmysqld\fR\&. (See, for example, -Section\ \&2.11.2.2, \(lqStarting and Stopping MySQL Automatically\(rq\&.) However, when using multiple servers, it is necessary that each one use its own value for options such as the Unix socket file and TCP/IP port number\&. For more information on which options must be unique per server in a multiple\-server environment, see +Section\ \&2.13.1.2, \(lqStarting and Stopping MySQL Automatically\(rq\&.) However, when using multiple servers, it is necessary that each one use its own value for options such as the Unix socket file and TCP/IP port number\&. For more information on which options must be unique per server in a multiple\-server environment, see Section\ \&5.6, \(lqRunning Multiple MySQL Servers on the Same Machine\(rq\&. .PP To invoke @@ -183,12 +183,6 @@ Otherwise, option files in the standard option, if one is given\&. (If the option is given multiple times, the last value is used\&.) .RE .PP -Option files read are searched for -[mysqld_multi] -and -[mysqld\fIN\fR] -option groups\&. -.PP Before MySQL 5\&.1\&.18, the preceding options are not recognized\&. Files in the standard locations are read, and any file named by the \fB\-\-config\-file=\fR\fB\fIfile_name\fR\fR option, if one is given\&. A file named by @@ -199,6 +193,39 @@ option groups, not the [mysqld_multi] group\&. .PP +Option files read are searched for +[mysqld_multi] +and +[mysqld\fIN\fR] +option groups\&. The +[mysqld_multi] +group can be used for options to +\fBmysqld_multi\fR +itself\&. +[mysqld\fIN\fR] +groups can be used for options passed to specific +\fBmysqld\fR +instances\&. +.PP +As of MySQL 5\&.1\&.35, the +[mysqld] +or +[mysqld_safe] +groups can be used for common options read by all instances of +\fBmysqld\fR +or +\fBmysqld_safe\fR\&. You can specify a +\fB\-\-defaults\-file=\fR\fB\fIfile_name\fR\fR +option to use a different configuration file for that instance, in which case the +[mysqld] +or +[mysqld_safe] +groups from that file will be used for that instance\&. Before MySQL 5\&.1\&.35, some versions of +\fBmysqld_multi\fR +pass the +\fB\-\-no\-defaults\fR +options to instances, so these techniques are inapplicable\&. +.PP \fBmysqld_multi\fR supports the following options: .sp === modified file 'man/mysqld_safe.1' --- a/man/mysqld_safe.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqld_safe.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqld_safe\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLD_SAFE\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLD_SAFE\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysqldump.1' --- a/man/mysqldump.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqldump.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqldump\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLDUMP\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLDUMP\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -575,7 +575,7 @@ uses utf8, and earlier versions use latin1\&. .sp -This option has no effect for output data files produced by using the +Prior to MySQL 5\&.1\&.38, this option has no effect for output data files produced by using the \fB\-\-tab\fR option\&. See the description for that option\&. .RE @@ -1232,7 +1232,8 @@ statement for the table\&. \fB\-\-no\-set\-names\fR .sp This option is deprecated\&. Use -\fB\-\-skip\-set\-charset\fR\&. instead\&. +\fB\-\-skip\-set\-charset\fR +instead\&. .RE .sp .RS 4 @@ -1657,11 +1658,11 @@ and \fB\-\-lines\-terminated\-by\fR options\&. .sp -Column values are dumped using the -binary -character set and the +As of MySQL 5\&.1\&.38, column values are written converted to the character set specified by the \fB\-\-default\-character\-set\fR -option is ignored\&. In effect, there is no character set conversion\&. If a table contains columns in several character sets, the output data file will as well and you may not be able to reload the file correctly\&. +option\&. Prior to 5\&.1\&.38 or if no such option is present, values are dumped using the +binary +character set\&. In effect, there is no character set conversion\&. If a table contains columns in several character sets, the output data file will as well and you may not be able to reload the file correctly\&. .if n \{\ .sp .\} @@ -2110,7 +2111,7 @@ InnoDB storage engine\&. .PP For more information on making backups, see -Section\ \&6.1, \(lqDatabase Backups\(rq, and +Section\ \&6.1, \(lqDatabase Backup Methods\(rq, and Section\ \&6.2, \(lqExample Backup and Recovery Strategy\(rq\&. .\" mysqldump: views .\" mysqldump: problems === modified file 'man/mysqldumpslow.1' --- a/man/mysqldumpslow.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqldumpslow.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqldumpslow\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLDUMPSLOW\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLDUMPSLOW\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysqlhotcopy.1' --- a/man/mysqlhotcopy.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqlhotcopy.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqlhotcopy\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLHOTCOPY\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLHOTCOPY\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysqlimport.1' --- a/man/mysqlimport.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqlimport.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqlimport\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLIMPORT\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLIMPORT\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysqlmanager.8' --- a/man/mysqlmanager.8 2009-09-16 12:03:18 +0000 +++ b/man/mysqlmanager.8 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqlmanager\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLMANAGER\FR" "8" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLMANAGER\FR" "8" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -274,7 +274,8 @@ Drop all users from the password file\&. .\} .\" mysqlmanager: debug option .\" debug option: mysqlmanager -\fB\-\-debug=\fR\fB\fIdebug_options\fR\fR\fB, \-# \fR\fB\fIdebug_options\fR\fR +\fB\-\-debug=\fR\fB\fIdebug_options\fR\fR, +\fB\-# \fR\fB\fIdebug_options\fR\fR .sp Write a debugging log\&. A typical \fIdebug_options\fR === modified file 'man/mysqlshow.1' --- a/man/mysqlshow.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqlshow.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqlshow\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLSHOW\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLSHOW\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysqlslap.1' --- a/man/mysqlslap.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqlslap.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqlslap\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBMYSQLSLAP\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBMYSQLSLAP\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/mysqltest.1' --- a/man/mysqltest.1 2009-09-16 12:03:18 +0000 +++ b/man/mysqltest.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBmysqltest\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/08/2009 +.\" Date: 10/29/2009 .\" Manual: MySQL Database System .\" Source: MySQL .\" Language: English .\" -.TH "\FBMYSQLTEST\FR" "1" "08/08/2009" "MySQL" "MySQL Database System" +.TH "\FBMYSQLTEST\FR" "1" "10/29/2009" "MySQL" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/ndbd.8' --- a/man/ndbd.8 2009-09-16 12:03:18 +0000 +++ b/man/ndbd.8 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBndbd\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBNDBD\FR" "8" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBNDBD\FR" "8" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -41,8 +41,9 @@ processes cooperate in handling data\&. .\" command options (MySQL Cluster): ndbd .\" MySQL Cluster: ndbd process .PP -The following list describes command options specific to the MySQL Cluster data node program -\fBndbd\fR\&. +The following table includes command options specific to the MySQL Cluster data node program +\fBndbd\fR\&. Additional descriptions follow the table\&. For options common to all MySQL Cluster programs, see +Section\ \&17.4.2, \(lqOptions Common to MySQL Cluster Programs\(rq\&. .if n \{\ .sp .\} @@ -68,7 +69,7 @@ wherever the latter occurs in this secti For options common to all NDBCLUSTER programs, see -Section\ \&17.6.2, \(lqOptions Common to MySQL Cluster Programs\(rq\&. +Section\ \&17.4.2, \(lqOptions Common to MySQL Cluster Programs\(rq\&. .sp .RS 4 .ie n \{\ @@ -81,9 +82,11 @@ Section\ \&17.6.2, \(lqOptions Common to \fB\-\-bind\-address\fR .TS allbox tab(:); -l l -l l -l l. +l l s +l l s +l l s +^ l l +^ l l. T{ \fBVersion Introduced\fR T}:T{ @@ -95,28 +98,16 @@ T}:T{ \-\-bind\-address=name T} T{ -\fBPermitted Values \fR +\ \& T}:T{ -[\fInested\ table\fR]* +\fBPermitted Values \fR T} -.TE -.sp 1 -.sp -.it 1 an-trap -.nr an-no-space-flag 1 -.nr an-break-flag 1 -.br -.B *[nested\ table] -.TS -allbox tab(:); -l l -l l. -T{ +:T{ \fBType\fR T}:T{ string T} -T{ +:T{ \fBDefault\fR T}:T{ T} @@ -141,36 +132,26 @@ This option was added in MySQL 5\&.1\&.1 \fB\-d\fR .TS allbox tab(:); -l l -l l. +l l s +l l s +^ l l +^ l l. T{ \fBCommand Line Format\fR T}:T{ \-\-daemon T} T{ -\fBPermitted Values \fR +\ \& T}:T{ -[\fInested\ table\fR]* +\fBPermitted Values \fR T} -.TE -.sp 1 -.sp -.it 1 an-trap -.nr an-no-space-flag 1 -.nr an-break-flag 1 -.br -.B *[nested\ table] -.TS -allbox tab(:); -l l -l l. -T{ +:T{ \fBType\fR T}:T{ boolean T} -T{ +:T{ \fBDefault\fR T}:T{ TRUE @@ -182,6 +163,12 @@ Instructs to execute as a daemon process\&. This is the default behavior\&. \fB\-\-nodaemon\fR can be used to prevent the process from running as a daemon\&. +.sp +This option has no effect when running +\fBndbd\fR +or +\fBndbmtd\fR +on Windows platforms\&. .RE .sp .RS 4 @@ -197,36 +184,26 @@ can be used to prevent the process from \fB\-\-initial\fR .TS allbox tab(:); -l l -l l. +l l s +l l s +^ l l +^ l l. T{ \fBCommand Line Format\fR T}:T{ \-\-initial T} T{ -\fBPermitted Values \fR +\ \& T}:T{ -[\fInested\ table\fR]* +\fBPermitted Values \fR T} -.TE -.sp 1 -.sp -.it 1 an-trap -.nr an-no-space-flag 1 -.nr an-break-flag 1 -.br -.B *[nested\ table] -.TS -allbox tab(:); -l l -l l. -T{ +:T{ \fBType\fR T}:T{ boolean T} -T{ +:T{ \fBDefault\fR T}:T{ FALSE @@ -321,9 +298,11 @@ Backup files that have already been crea .IP \(bu 2.3 .\} MySQL Cluster Disk Data files (see -Section\ \&17.10, \(lqMySQL Cluster Disk Data Tables\(rq)\&. +Section\ \&17.5.9, \(lqMySQL Cluster Disk Data Tables\(rq)\&. .RE .RS 4 +.sp +This option also has no effect on recovery of data by a data node that is just starting (or restarting) from data nodes that are already running\&. This recovery of data occurs automatically, and requires no user intervention in a MySQL Cluster that is running normally\&. .sp .5v .RE It is permissible to use this option when starting the cluster for the very first time (that is, before any data node files have been created); however, it is @@ -344,9 +323,11 @@ necessary to do so\&. \fB\-\-initial\-start\fR .TS allbox tab(:); -l l -l l -l l. +l l s +l l s +l l s +^ l l +^ l l. T{ \fBVersion Introduced\fR T}:T{ @@ -358,28 +339,16 @@ T}:T{ \-\-initial\-start T} T{ -\fBPermitted Values \fR +\ \& T}:T{ -[\fInested\ table\fR]* +\fBPermitted Values \fR T} -.TE -.sp 1 -.sp -.it 1 an-trap -.nr an-no-space-flag 1 -.nr an-break-flag 1 -.br -.B *[nested\ table] -.TS -allbox tab(:); -l l -l l. -T{ +:T{ \fBType\fR T}:T{ boolean T} -T{ +:T{ \fBDefault\fR T}:T{ FALSE @@ -435,9 +404,11 @@ Prior to MySQL 5\&.1\&.19, it was not po \fB\-\-nowait\-nodes=\fR\fB\fInode_id_1\fR\fR\fB[, \fR\fB\fInode_id_2\fR\fR\fB[, \&.\&.\&.]]\fR .TS allbox tab(:); -l l -l l -l l. +l l s +l l s +l l s +^ l l +^ l l. T{ \fBVersion Introduced\fR T}:T{ @@ -449,28 +420,16 @@ T}:T{ \-\-nowait\-nodes=list T} T{ -\fBPermitted Values \fR +\ \& T}:T{ -[\fInested\ table\fR]* +\fBPermitted Values \fR T} -.TE -.sp 1 -.sp -.it 1 an-trap -.nr an-no-space-flag 1 -.nr an-break-flag 1 -.br -.B *[nested\ table] -.TS -allbox tab(:); -l l -l l. -T{ +:T{ \fBType\fR T}:T{ string T} -T{ +:T{ \fBDefault\fR T}:T{ T} @@ -508,36 +467,26 @@ This option was added in MySQL 5\&.1\&.9 \fB\-\-nodaemon\fR .TS allbox tab(:); -l l -l l. +l l s +l l s +^ l l +^ l l. T{ \fBCommand Line Format\fR T}:T{ \-\-nodaemon T} T{ -\fBPermitted Values \fR +\ \& T}:T{ -[\fInested\ table\fR]* +\fBPermitted Values \fR T} -.TE -.sp 1 -.sp -.it 1 an-trap -.nr an-no-space-flag 1 -.nr an-break-flag 1 -.br -.B *[nested\ table] -.TS -allbox tab(:); -l l -l l. -T{ +:T{ \fBType\fR T}:T{ boolean T} -T{ +:T{ \fBDefault\fR T}:T{ FALSE @@ -549,6 +498,12 @@ Instructs not to start as a daemon process\&. This is useful when \fBndbd\fR is being debugged and you want output to be redirected to the screen\&. +.sp +As of MySQL Cluster NDB 7\&.0\&.8, the default behavior for +\fBndbd\fR +and +\fBndbmtd\fR +on Windows is to run in the foreground, making this option unnecessary on Windows platforms\&. (\m[blue]\fBBug#45588\fR\m[]\&\s-2\u[2]\d\s+2) .RE .sp .RS 4 @@ -567,36 +522,26 @@ is being debugged and you want output to \fB\-n\fR .TS allbox tab(:); -l l -l l. +l l s +l l s +^ l l +^ l l. T{ \fBCommand Line Format\fR T}:T{ \-\-nostart T} T{ -\fBPermitted Values \fR +\ \& T}:T{ -[\fInested\ table\fR]* +\fBPermitted Values \fR T} -.TE -.sp 1 -.sp -.it 1 an-trap -.nr an-no-space-flag 1 -.nr an-break-flag 1 -.br -.B *[nested\ table] -.TS -allbox tab(:); -l l -l l. -T{ +:T{ \fBType\fR T}:T{ boolean T} -T{ +:T{ \fBDefault\fR T}:T{ FALSE @@ -610,7 +555,7 @@ not to start automatically\&. When this connects to the management server, obtains configuration data from it, and initializes communication objects\&. However, it does not actually start the execution engine until specifically requested to do so by the management server\&. This can be accomplished by issuing the proper START command in the management client (see -Section\ \&17.7.2, \(lqCommands in the MySQL Cluster Management Client\(rq)\&. +Section\ \&17.5.2, \(lqCommands in the MySQL Cluster Management Client\(rq)\&. .RE .\" MySQL Cluster: log files .\" log files (MySQL Cluster) @@ -668,7 +613,7 @@ TraceFile: ndb_2_trace\&.log\&.2 Listings of possible \fBndbd\fR exit codes and messages generated when a data node process shuts down prematurely can be found in -\m[blue]\fBndbd Error Messages\fR\m[]\&\s-2\u[2]\d\s+2\&. +\m[blue]\fBndbd Error Messages\fR\m[]\&\s-2\u[3]\d\s+2\&. .if n \{\ .sp .\} @@ -784,7 +729,7 @@ shell> \fBndbd \-\-connect\-string="node .\} .PP See -Section\ \&17.3.4.3, \(lqThe MySQL Cluster Connectstring\(rq, for additional information about this issue\&. +Section\ \&17.3.2.3, \(lqThe MySQL Cluster Connectstring\(rq, for additional information about this issue\&. \fBndbd\fR(8), describes other options for \fBndbd\fR\&. .PP @@ -810,7 +755,7 @@ process can consume up to 2 CPUs if perm For a machine with many CPUs it is possible to use several \fBndbd\fR processes which belong to different node groups; however, such a configuration is still considered experimental and is not supported for MySQL 5\&.1 in a production setting\&. See -Section\ \&17.12, \(lqKnown Limitations of MySQL Cluster\(rq\&. +Section\ \&17.1.5, \(lqKnown Limitations of MySQL Cluster\(rq\&. .SH "COPYRIGHT" .br .PP @@ -829,6 +774,11 @@ Bug#24631 \%http://bugs.mysql.com/24631 .RE .IP " 2." 4 +Bug#45588 +.RS 4 +\%http://bugs.mysql.com/45588 +.RE +.IP " 3." 4 ndbd Error Messages .RS 4 \%http://dev.mysql.com/doc/ndbapi/en/ndbd-error-messages.html === modified file 'man/ndbd_redo_log_reader.1' --- a/man/ndbd_redo_log_reader.1 2009-09-16 12:03:18 +0000 +++ b/man/ndbd_redo_log_reader.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBndbd_redo_log_reader\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBNDBD_REDO_LOG_REA" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBNDBD_REDO_LOG_REA" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -37,6 +37,10 @@ The C++ source files for can be found in the directory /storage/ndb/src/kernel/blocks/dblqh/redoLogReader\&. .PP +The following table includes options that are specific to the MySQL Cluster program +\fBndbd_redo_log_reader\fR\&. Additional descriptions follow the table\&. For options common to all MySQL Cluster programs, see +Section\ \&17.4.2, \(lqOptions Common to MySQL Cluster Programs\(rq\&. +.PP \fBUsage\fR: .sp .if n \{\ @@ -56,105 +60,83 @@ ndb_\fI#\fR_fs/D\fI#\fR/LCP/\fI#\fR/T\fI represents a number (not necessarily the same number)\&. For more information, see \m[blue]\fBCluster Data Node FileSystemDir Files\fR\m[]\&\s-2\u[1]\d\s+2\&. .PP -\fBAdditional Options\fR: +The name of the file to be read may be followed by one or more of the options listed here: +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} .TS allbox tab(:); -l l -l l. +l l s +l l s +^ l l +^ l l. T{ \fBCommand Line Format\fR T}:T{ \-noprint T} T{ -\fBPermitted Values \fR +\ \& T}:T{ -[\fInested\ table\fR]* +\fBPermitted Values \fR T} -.TE -.sp 1 -.sp -.it 1 an-trap -.nr an-no-space-flag 1 -.nr an-break-flag 1 -.br -.B *[nested\ table] -.TS -allbox tab(:); -l l -l l. -T{ +:T{ \fBType\fR T}:T{ boolean T} -T{ +:T{ \fBDefault\fR T}:T{ FALSE T} .TE .sp 1 +\fB\-noprint\fR: Do not print the contents of the log file\&. +.RE +.sp +.RS 4 +.ie n \{\ +\h'-04'\(bu\h'+03'\c +.\} +.el \{\ +.sp -1 +.IP \(bu 2.3 +.\} .TS allbox tab(:); -l l -l l. +l l s +l l s +^ l l +^ l l. T{ \fBCommand Line Format\fR T}:T{ \-nocheck T} T{ -\fBPermitted Values \fR +\ \& T}:T{ -[\fInested\ table\fR]* +\fBPermitted Values \fR T} -.TE -.sp 1 -.sp -.it 1 an-trap -.nr an-no-space-flag 1 -.nr an-break-flag 1 -.br -.B *[nested\ table] -.TS -allbox tab(:); -l l -l l. -T{ +:T{ \fBType\fR T}:T{ boolean T} -T{ +:T{ \fBDefault\fR T}:T{ FALSE T} .TE .sp 1 -.PP -The name of the file to be read may be followed by one or more of the options listed here: -.sp -.RS 4 -.ie n \{\ -\h'-04'\(bu\h'+03'\c -.\} -.el \{\ -.sp -1 -.IP \(bu 2.3 -.\} -\fB\-noprint\fR: Do not print the contents of the log file\&. -.RE -.sp -.RS 4 -.ie n \{\ -\h'-04'\(bu\h'+03'\c -.\} -.el \{\ -.sp -1 -.IP \(bu 2.3 -.\} \fB\-nocheck\fR: Do not check the log file for errors\&. .RE .sp @@ -184,7 +166,7 @@ You should have received a copy of the G .IP " 1." 4 Cluster Data Node FileSystemDir Files .RS 4 -\%http://dev.mysql.com/doc/ndbapi/en/ndb-internals-ndbd-filesystem.html#ndb-internals-ndbd-filesystemdir-files +\%http://dev.mysql.com/doc/ndbapi/en/ndb-internals-ndbd-filesystemdir-files.html .RE .SH "SEE ALSO" For more information, please refer to the MySQL Reference Manual, === modified file 'man/ndbmtd.8' --- a/man/ndbmtd.8 2009-09-16 12:03:18 +0000 +++ b/man/ndbmtd.8 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBndbmtd\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBNDBMTD\FR" "8" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBNDBMTD\FR" "8" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- @@ -49,8 +49,8 @@ Command\-line options and configuration \fBndbd\fR also apply to \fBndbmtd\fR\&. For more information about these options and parameters, see -Section\ \&17.6.3.2, \(lqProgram Options for ndbd and ndbmtd\(rq, and -Section\ \&17.3.4.6, \(lqDefining MySQL Cluster Data Nodes\(rq, respectively\&. +\fBndbd\fR(8), and +Section\ \&17.3.2.6, \(lqDefining MySQL Cluster Data Nodes\(rq, respectively\&. .PP \fBndbmtd\fR is also file system\-compatible with @@ -69,25 +69,13 @@ simply by stopping the node and then sta \fBndbd\fR in place of the multi\-threaded binary\&. It is not necessary when switching between the two to start the data node binary using \fB\-\-initial\fR\&. -.if n \{\ -.sp -.\} -.RS 4 -.it 1 an-trap -.nr an-no-space-flag 1 -.nr an-break-flag 1 -.br -.ps +1 -\fBImportant\fR -.ps -1 -.br .PP -We do not currently recommend using +Prior to MySQL Cluster NDB 7\&.0\&.6, there were known issues when using \fBndbmtd\fR -with MySQL Cluster Disk Data tables in production, due to known issues which we are working to fix in a future MySQL Cluster release\&. (\m[blue]\fBBug#41915\fR\m[]\&\s-2\u[1]\d\s+2, +with MySQL Cluster Disk Data tables\&. If you wish to use multi\-threaded data nodes with disk\-based +NDB +tables, you should insure that you are running MySQL Cluster NDB 7\&.0\&.6 or later\&. (\m[blue]\fBBug#41915\fR\m[]\&\s-2\u[1]\d\s+2, \m[blue]\fBBug#44915\fR\m[]\&\s-2\u[2]\d\s+2) -.sp .5v -.RE .PP Using \fBndbmtd\fR @@ -129,6 +117,8 @@ failures\&. .RE .PP These differences are discussed in more detail in the next few paragraphs\&. +.\" execution threads (MySQL Cluster) +.\" MySQL Cluster: execution threads .\" ndbmtd: MaxNoOfExecutionThreads .\" MaxNoOfExecutionThreads: ndbmtd .\" ndbmtd: trace files @@ -148,7 +138,7 @@ file, it is exclusive to and does not apply to \fBndbd\fR\&. .PP -This parameter takes an integer value from 2 to 8 inclusive\&. Generally, you should set this to the number of CPU cores on the data node host, as shown in the following table: +This parameter takes an integer value from 2 to 8 inclusive\&. Generally, you should set this parameter equal to the number of CPU cores on the data node host, as shown in the following table: .TS allbox tab(:); lB lB. === modified file 'man/perror.1' --- a/man/perror.1 2009-09-16 12:03:18 +0000 +++ b/man/perror.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBperror\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBPERROR\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBPERROR\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/replace.1' --- a/man/replace.1 2009-09-16 12:03:18 +0000 +++ b/man/replace.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBreplace\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBREPLACE\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBREPLACE\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/resolve_stack_dump.1' --- a/man/resolve_stack_dump.1 2009-09-16 12:03:18 +0000 +++ b/man/resolve_stack_dump.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBresolve_stack_dump\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBRESOLVE_STACK_DUM" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBRESOLVE_STACK_DUM" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'man/resolveip.1' --- a/man/resolveip.1 2009-09-16 12:03:18 +0000 +++ b/man/resolveip.1 2009-12-01 07:24:05 +0000 @@ -2,12 +2,12 @@ .\" Title: \fBresolveip\fR .\" Author: [FIXME: author] [see http://docbook.sf.net/el/author] .\" Generator: DocBook XSL Stylesheets v1.75.2 <http://docbook.sf.net/> -.\" Date: 08/12/2009 +.\" Date: 11/04/2009 .\" Manual: MySQL Database System .\" Source: MySQL 5.1 .\" Language: English .\" -.TH "\FBRESOLVEIP\FR" "1" "08/12/2009" "MySQL 5\&.1" "MySQL Database System" +.TH "\FBRESOLVEIP\FR" "1" "11/04/2009" "MySQL 5\&.1" "MySQL Database System" .\" ----------------------------------------------------------------- .\" * set default formatting .\" ----------------------------------------------------------------- === modified file 'scripts/fill_help_tables.sql' --- a/scripts/fill_help_tables.sql 2009-09-16 12:03:18 +0000 +++ b/scripts/fill_help_tables.sql 2009-12-01 07:24:05 +0000 @@ -87,7 +87,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (18,25,'SHOW CREATE PROCEDURE','Syntax:\nSHOW CREATE PROCEDURE proc_name\n\nThis statement is a MySQL extension. It returns the exact string that\ncan be used to re-create the named stored procedure. A similar\nstatement, SHOW CREATE FUNCTION, displays information about stored\nfunctions (see [HELP SHOW CREATE FUNCTION]).\n\nBoth statements require that you be the owner of the routine or have\nSELECT access to the mysql.proc table. If you do not have privileges\nfor the routine itself, the value displayed for the Create Procedure or\nCreate Function field will be NULL.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/show-create-procedure.html\n\n','mysql> SHOW CREATE PROCEDURE test.simpleproc\\G\n*************************** 1. row ***************************\n Procedure: simpleproc\n sql_mode:\n Create Procedure: CREATE PROCEDURE `simpleproc`(OUT param1 INT)\n BEGIN\n SELECT COUNT(*) INTO param1 FROM t;\n END\ncharacter_set_client: latin1\ncollation_connection: latin1_swedish_ci\n Database Collation: latin1_swedish_ci\n\nmysql> SHOW CREATE FUNCTION test.hello\\G\n*************************** 1. row ***************************\n Function: hello\n sql_mode:\n Create Function: CREATE FUNCTION `hello`(s CHAR(20))\n RETURNS CHAR(50)\n RETURN CONCAT(\'Hello, \',s,\'!\')\ncharacter_set_client: latin1\ncollation_connection: latin1_swedish_ci\n Database Collation: latin1_swedish_ci\n','http://dev.mysql.com/doc/refman/5.1/en/show-create-procedure.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (19,20,'INTEGER','INTEGER[(M)] [UNSIGNED] [ZEROFILL]\n\nThis type is a synonym for INT.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/numeric-type-overview.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/numeric-type-overview.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (20,35,'LOWER','Syntax:\nLOWER(str)\n\nReturns the string str with all characters changed to lowercase\naccording to the current character set mapping. The default is latin1\n(cp1252 West European).\n\nmysql> SELECT LOWER(\'QUADRATICALLY\');\n -> \'quadratically\'\n\nLOWER() (and UPPER()) are ineffective when applied to binary strings\n(BINARY, VARBINARY, BLOB). To perform lettercase conversion, convert\nthe string to a nonbinary string:\n\nmysql> SET @str = BINARY \'New York\';\nmysql> SELECT LOWER(@str), LOWER(CONVERT(@str USING latin1));\n+-------------+-----------------------------------+\n| LOWER(@str) | LOWER(CONVERT(@str USING latin1)) |\n+-------------+-----------------------------------+\n| New York | new york |\n+-------------+-----------------------------------+\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/string-functions.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/string-functions.html'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (21,25,'SHOW COLUMNS','Syntax:\nSHOW [FULL] COLUMNS {FROM | IN} tbl_name [{FROM | IN} db_name]\n [LIKE \'pattern\' | WHERE expr]\n\nSHOW COLUMNS displays information about the columns in a given table.\nIt also works for views. The LIKE clause, if present, indicates which\ncolumn names to match. The WHERE clause can be given to select rows\nusing more general conditions, as discussed in\nhttp://dev.mysql.com/doc/refman/5.1/en/extended-show.html.\n\nmysql> SHOW COLUMNS FROM City;\n+------------+----------+------+-----+---------+----------------+\n| Field | Type | Null | Key | Default | Extra |\n+------------+----------+------+-----+---------+----------------+\n| Id | int(11) | NO | PRI | NULL | auto_increment |\n| Name | char(35) | NO | | | |\n| Country | char(3) | NO | UNI | | |\n| District | char(20) | YES | MUL | | |\n| Population | int(11) | NO | | 0 | |\n+------------+----------+------+-----+---------+----------------+\n5 rows in set (0.00 sec)\n\nIf the data types differ from what you expect them to be based on a\nCREATE TABLE statement, note that MySQL sometimes changes data types\nwhen you create or alter a table. The conditions under which this\noccurs are described in\nhttp://dev.mysql.com/doc/refman/5.1/en/silent-column-changes.html.\n\nT… FULL keyword causes the output to include the column collation and\ncomments, as well as the privileges you have for each column.\n\nYou can use db_name.tbl_name as an alternative to the tbl_name FROM\ndb_name syntax. In other words, these two statements are equivalent:\n\nmysql> SHOW COLUMNS FROM mytable FROM mydb;\nmysql> SHOW COLUMNS FROM mydb.mytable;\n\nSHOW COLUMNS displays the following values for each table column:\n\nField indicates the column name.\n\nType indicates the column data type.\n\nCollation indicates the collation for nonbinary string columns, or NULL\nfor other columns. This value is displayed only if you use the FULL\nkeyword.\n\nThe Null field contains YES if NULL values can be stored in the column,\nNO if not.\n\nThe Key field indicates whether the column is indexed:\n\no If Key is empty, the column either is not indexed or is indexed only\n as a secondary column in a multiple-column, nonunique index.\n\no If Key is PRI, the column is a PRIMARY KEY or is one of the columns\n in a multiple-column PRIMARY KEY.\n\no If Key is UNI, the column is the first column of a unique-valued\n index that cannot contain NULL values.\n\no If Key is MUL, multiple occurrences of a given value are allowed\n within the column. The column is the first column of a nonunique\n index or a unique-valued index that can contain NULL values.\n\nIf more than one of the Key values applies to a given column of a\ntable, Key displays the one with the highest priority, in the order\nPRI, UNI, MUL.\n\nA UNIQUE index may be displayed as PRI if it cannot contain NULL values\nand there is no PRIMARY KEY in the table. A UNIQUE index may display as\nMUL if several columns form a composite UNIQUE index; although the\ncombination of the columns is unique, each column can still hold\nmultiple occurrences of a given value.\n\nThe Default field indicates the default value that is assigned to the\ncolumn.\n\nThe Extra field contains any additional information that is available\nabout a given column. The value is auto_increment if the column was\ncreated with the AUTO_INCREMENT keyword and empty otherwise.\n\nPrivileges indicates the privileges you have for the column. This value\nis displayed only if you use the FULL keyword.\n\nComment indicates any comment the column has. This value is displayed\nonly if you use the FULL keyword.\n\nSHOW FIELDS is a synonym for SHOW COLUMNS. You can also list a table\'s\ncolumns with the mysqlshow db_name tbl_name command.\n\nThe DESCRIBE statement provides information similar to SHOW COLUMNS.\nSee [HELP DESCRIBE].\n\nThe SHOW CREATE TABLE, SHOW TABLE STATUS, and SHOW INDEX statements\nalso provide information about tables. See [HELP SHOW].\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/show-columns.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/show-columns.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (21,25,'SHOW COLUMNS','Syntax:\nSHOW [FULL] COLUMNS {FROM | IN} tbl_name [{FROM | IN} db_name]\n [LIKE \'pattern\' | WHERE expr]\n\nSHOW COLUMNS displays information about the columns in a given table.\nIt also works for views. The LIKE clause, if present, indicates which\ncolumn names to match. The WHERE clause can be given to select rows\nusing more general conditions, as discussed in\nhttp://dev.mysql.com/doc/refman/5.1/en/extended-show.html.\n\nmysql> SHOW COLUMNS FROM City;\n+------------+----------+------+-----+---------+----------------+\n| Field | Type | Null | Key | Default | Extra |\n+------------+----------+------+-----+---------+----------------+\n| Id | int(11) | NO | PRI | NULL | auto_increment |\n| Name | char(35) | NO | | | |\n| Country | char(3) | NO | UNI | | |\n| District | char(20) | YES | MUL | | |\n| Population | int(11) | NO | | 0 | |\n+------------+----------+------+-----+---------+----------------+\n5 rows in set (0.00 sec)\n\nIf the data types differ from what you expect them to be based on a\nCREATE TABLE statement, note that MySQL sometimes changes data types\nwhen you create or alter a table. The conditions under which this\noccurs are described in\nhttp://dev.mysql.com/doc/refman/5.1/en/silent-column-changes.html.\n\nT… FULL keyword causes the output to include the column collation and\ncomments, as well as the privileges you have for each column.\n\nYou can use db_name.tbl_name as an alternative to the tbl_name FROM\ndb_name syntax. In other words, these two statements are equivalent:\n\nmysql> SHOW COLUMNS FROM mytable FROM mydb;\nmysql> SHOW COLUMNS FROM mydb.mytable;\n\nSHOW COLUMNS displays the following values for each table column:\n\nField indicates the column name.\n\nType indicates the column data type.\n\nCollation indicates the collation for nonbinary string columns, or NULL\nfor other columns. This value is displayed only if you use the FULL\nkeyword.\n\nThe Null field contains YES if NULL values can be stored in the column,\nNO if not.\n\nThe Key field indicates whether the column is indexed:\n\no If Key is empty, the column either is not indexed or is indexed only\n as a secondary column in a multiple-column, nonunique index.\n\no If Key is PRI, the column is a PRIMARY KEY or is one of the columns\n in a multiple-column PRIMARY KEY.\n\no If Key is UNI, the column is the first column of a unique-valued\n index that cannot contain NULL values.\n\no If Key is MUL, multiple occurrences of a given value are allowed\n within the column. The column is the first column of a nonunique\n index or a unique-valued index that can contain NULL values.\n\nIf more than one of the Key values applies to a given column of a\ntable, Key displays the one with the highest priority, in the order\nPRI, UNI, MUL.\n\nA UNIQUE index may be displayed as PRI if it cannot contain NULL values\nand there is no PRIMARY KEY in the table. A UNIQUE index may display as\nMUL if several columns form a composite UNIQUE index; although the\ncombination of the columns is unique, each column can still hold\nmultiple occurrences of a given value.\n\nThe Default field indicates the default value that is assigned to the\ncolumn.\n\nThe Extra field contains any additional information that is available\nabout a given column. The value is nonempty in these cases:\nauto_increment for columns that have the AUTO_INCREMENT attribute; as\nof MySQL 5.1.23, on update CURRENT_TIMESTAMP for TIMESTAMP columns that\nhave the ON UPDATE CURRENT_TIMESTAMP attribute.\n\nPrivileges indicates the privileges you have for the column. This value\nis displayed only if you use the FULL keyword.\n\nComment indicates any comment the column has. This value is displayed\nonly if you use the FULL keyword.\n\nSHOW FIELDS is a synonym for SHOW COLUMNS. You can also list a table\'s\ncolumns with the mysqlshow db_name tbl_name command.\n\nThe DESCRIBE statement provides information similar to SHOW COLUMNS.\nSee [HELP DESCRIBE].\n\nThe SHOW CREATE TABLE, SHOW TABLE STATUS, and SHOW INDEX statements\nalso provide information about tables. See [HELP SHOW].\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/show-columns.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/show-columns.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (22,37,'CREATE TRIGGER','Syntax:\nCREATE\n [DEFINER = { user | CURRENT_USER }]\n TRIGGER trigger_name trigger_time trigger_event\n ON tbl_name FOR EACH ROW trigger_stmt\n\nThis statement creates a new trigger. A trigger is a named database\nobject that is associated with a table, and that activates when a\nparticular event occurs for the table. The trigger becomes associated\nwith the table named tbl_name, which must refer to a permanent table.\nYou cannot associate a trigger with a TEMPORARY table or a view.\n\nCREATE TRIGGER requires the TRIGGER privilege for the table associated\nwith the trigger. (Before MySQL 5.1.6, this statement requires the\nSUPER privilege.)\n\nThe DEFINER clause determines the security context to be used when\nchecking access privileges at trigger activation time.\n\ntrigger_time is the trigger action time. It can be BEFORE or AFTER to\nindicate that the trigger activates before or after each row to be\nmodified.\n\ntrigger_event indicates the kind of statement that activates the\ntrigger. The trigger_event can be one of the following:\n\no INSERT: The trigger is activated whenever a new row is inserted into\n the table; for example, through INSERT, LOAD DATA, and REPLACE\n statements.\n\no UPDATE: The trigger is activated whenever a row is modified; for\n example, through UPDATE statements.\n\no DELETE: The trigger is activated whenever a row is deleted from the\n table; for example, through DELETE and REPLACE statements. However,\n DROP TABLE and TRUNCATE statements on the table do not activate this\n trigger, because they do not use DELETE. Dropping a partition does\n not activate DELETE triggers, either. See [HELP TRUNCATE TABLE].\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/create-trigger.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/create-trigger.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (23,30,'MONTH','Syntax:\nMONTH(date)\n\nReturns the month for date, in the range 1 to 12 for January to\nDecember, or 0 for dates such as \'0000-00-00\' or \'2008-00-00\' that have\na zero month part.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','mysql> SELECT MONTH(\'2008-02-03\');\n -> 2\n','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (24,20,'TINYINT','TINYINT[(M)] [UNSIGNED] [ZEROFILL]\n\nA very small integer. The signed range is -128 to 127. The unsigned\nrange is 0 to 255.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/numeric-type-overview.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/numeric-type-overview.html'); @@ -153,7 +153,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (84,25,'EXECUTE STATEMENT','Syntax:\nEXECUTE stmt_name\n [USING @var_name [, @var_name] ...]\n\nAfter preparing a statement with PREPARE, you execute it with an\nEXECUTE statement that refers to the prepared statement name. If the\nprepared statement contains any parameter markers, you must supply a\nUSING clause that lists user variables containing the values to be\nbound to the parameters. Parameter values can be supplied only by user\nvariables, and the USING clause must name exactly as many variables as\nthe number of parameter markers in the statement.\n\nYou can execute a given prepared statement multiple times, passing\ndifferent variables to it or setting the variables to different values\nbefore each execution.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/execute.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/execute.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (85,37,'DROP INDEX','Syntax:\nDROP [ONLINE|OFFLINE] INDEX index_name ON tbl_name\n\nDROP INDEX drops the index named index_name from the table tbl_name.\nThis statement is mapped to an ALTER TABLE statement to drop the index.\nSee [HELP ALTER TABLE].\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/drop-index.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/drop-index.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (86,35,'MATCH AGAINST','Syntax:\nMATCH (col1,col2,...) AGAINST (expr [search_modifier])\n\nMySQL has support for full-text indexing and searching:\n\no A full-text index in MySQL is an index of type FULLTEXT.\n\no Full-text indexes can be used only with MyISAM tables, and can be\n created only for CHAR, VARCHAR, or TEXT columns.\n\no A FULLTEXT index definition can be given in the CREATE TABLE\n statement when a table is created, or added later using ALTER TABLE\n or CREATE INDEX.\n\no For large data sets, it is much faster to load your data into a table\n that has no FULLTEXT index and then create the index after that, than\n to load data into a table that has an existing FULLTEXT index.\n\nFull-text searching is performed using MATCH() ... AGAINST syntax.\nMATCH() takes a comma-separated list that names the columns to be\nsearched. AGAINST takes a string to search for, and an optional\nmodifier that indicates what type of search to perform. The search\nstring must be a literal string, not a variable or a column name. There\nare three types of full-text searches:\n\no A boolean search interprets the search string using the rules of a\n special query language. The string contains the words to search for.\n It can also contain operators that specify requirements such that a\n word must be present or absent in matching rows, or that it should be\n weighted higher or lower than usual. Common words such as "some" or\n "then" are stopwords and do not match if present in the search\n string. The IN BOOLEAN MODE modifier specifies a boolean search. For\n more information, see\n http://dev.mysql.com/doc/refman/5.1/en/fulltext-boolean.html.\n\no A natural language search interprets the search string as a phrase in\n natural human language (a phrase in free text). There are no special\n operators. The stopword list applies. In addition, words that are\n present in 50% or more of the rows are considered common and do not\n match. Full-text searches are natural language searches if the IN\n NATURAL LANGUAGE MODE modifier is given or if no modifier is given.\n\no A query expansion search is a modification of a natural language\n search. The search string is used to perform a natural language\n search. Then words from the most relevant rows returned by the search\n are added to the search string and the search is done again. The\n query returns the rows from the second search. The IN NATURAL\n LANGUAGE MODE WITH QUERY EXPANSION or WITH QUERY EXPANSION modifier\n specifies a query expansion search. For more information, see\n http://dev.mysql.com/doc/refman/5.1/en/fulltext-query-expansion.html.\n\nThe IN NATURAL LANGUAGE MODE and IN NATURAL LANGUAGE MODE WITH QUERY\nEXPANSION modifiers were added in MySQL 5.1.7.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/fulltext-search.html\n\n','mysql> SELECT id, body, MATCH (title,body) AGAINST\n -> (\'Security implications of running MySQL as root\'\n -> IN NATURAL LANGUAGE MODE) AS score\n -> FROM articles WHERE MATCH (title,body) AGAINST\n -> (\'Security implications of running MySQL as root\'\n -> IN NATURAL LANGUAGE MODE);\n+----+-------------------------------------+-----------------+\n| id | body | score |\n+----+-------------------------------------+-----------------+\n| 4 | 1. Never run mysqld as root. 2. ... | 1.5219271183014 |\n| 6 | When configured properly, MySQL ... | 1.3114095926285 |\n+----+-------------------------------------+-----------------+\n2 rows in set (0.00 sec)\n','http://dev.mysql.com/doc/refman/5.1/en/fulltext-search.html'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (87,37,'CREATE EVENT','Syntax:\nCREATE\n [DEFINER = { user | CURRENT_USER }]\n EVENT\n [IF NOT EXISTS]\n event_name\n ON SCHEDULE schedule\n [ON COMPLETION [NOT] PRESERVE]\n [ENABLE | DISABLE | DISABLE ON SLAVE]\n [COMMENT \'comment\']\n DO sql_statement;\n\nschedule:\n AT timestamp [+ INTERVAL interval] ...\n | EVERY interval\n [STARTS timestamp [+ INTERVAL interval] ...]\n [ENDS timestamp [+ INTERVAL interval] ...]\n\ninterval:\n quantity {YEAR | QUARTER | MONTH | DAY | HOUR | MINUTE |\n WEEK | SECOND | YEAR_MONTH | DAY_HOUR | DAY_MINUTE |\n DAY_SECOND | HOUR_MINUTE | HOUR_SECOND | MINUTE_SECOND}\n\nThis statement creates and schedules a new event. It requires the EVENT\nprivilege for the schema in which the event is to be created.\n\nThe minimum requirements for a valid CREATE EVENT statement are as\nfollows:\n\no The keywords CREATE EVENT plus an event name, which uniquely\n identifies the event in the current schema. (Prior to MySQL 5.1.12,\n the event name needed to be unique only among events created by the\n same user on a given database.)\n\no An ON SCHEDULE clause, which determines when and how often the event\n executes.\n\no A DO clause, which contains the SQL statement to be executed by an\n event.\n\nThis is an example of a minimal CREATE EVENT statement:\n\nCREATE EVENT myevent\n ON SCHEDULE AT CURRENT_TIMESTAMP + INTERVAL 1 HOUR\n DO\n UPDATE myschema.mytable SET mycol = mycol + 1;\n\nThe previous statement creates an event named myevent. This event\nexecutes once --- one hour following its creation --- by running an SQL\nstatement that increments the value of the myschema.mytable table\'s\nmycol column by 1.\n\nThe event_name must be a valid MySQL identifier with a maximum length\nof 64 characters. It may be delimited using back ticks, and may be\nqualified with the name of a database schema. An event is associated\nwith both a MySQL user (the definer) and a schema, and its name must be\nunique among names of events within that schema. In general, the rules\ngoverning event names are the same as those for names of stored\nroutines. See http://dev.mysql.com/doc/refman/5.1/en/identifiers.html.\n\nIf no schema is indicated as part of event_name, the default (current)\nschema is assumed.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/create-event.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/create-event.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (87,37,'CREATE EVENT','Syntax:\nCREATE\n [DEFINER = { user | CURRENT_USER }]\n EVENT\n [IF NOT EXISTS]\n event_name\n ON SCHEDULE schedule\n [ON COMPLETION [NOT] PRESERVE]\n [ENABLE | DISABLE | DISABLE ON SLAVE]\n [COMMENT \'comment\']\n DO sql_statement;\n\nschedule:\n AT timestamp [+ INTERVAL interval] ...\n | EVERY interval\n [STARTS timestamp [+ INTERVAL interval] ...]\n [ENDS timestamp [+ INTERVAL interval] ...]\n\ninterval:\n quantity {YEAR | QUARTER | MONTH | DAY | HOUR | MINUTE |\n WEEK | SECOND | YEAR_MONTH | DAY_HOUR | DAY_MINUTE |\n DAY_SECOND | HOUR_MINUTE | HOUR_SECOND | MINUTE_SECOND}\n\nThis statement creates and schedules a new event. It requires the EVENT\nprivilege for the schema in which the event is to be created.\n\nThe minimum requirements for a valid CREATE EVENT statement are as\nfollows:\n\no The keywords CREATE EVENT plus an event name, which uniquely\n identifies the event within a database schema. (Prior to MySQL\n 5.1.12, the event name needed to be unique only among events created\n by the same user within a schema.)\n\no An ON SCHEDULE clause, which determines when and how often the event\n executes.\n\no A DO clause, which contains the SQL statement to be executed by an\n event.\n\nThis is an example of a minimal CREATE EVENT statement:\n\nCREATE EVENT myevent\n ON SCHEDULE AT CURRENT_TIMESTAMP + INTERVAL 1 HOUR\n DO\n UPDATE myschema.mytable SET mycol = mycol + 1;\n\nThe previous statement creates an event named myevent. This event\nexecutes once --- one hour following its creation --- by running an SQL\nstatement that increments the value of the myschema.mytable table\'s\nmycol column by 1.\n\nThe event_name must be a valid MySQL identifier with a maximum length\nof 64 characters. Event names are not case sensitive, so you cannot\nhave two events named myevent and MyEvent in the same schema. In\ngeneral, the rules governing event names are the same as those for\nnames of stored routines. See\nhttp://dev.mysql.com/doc/refman/5.1/en/identifiers.html.\n\nAn event is associated with a schema. If no schema is indicated as part\nof event_name, the default (current) schema is assumed. To create an\nevent in a specific schema, qualify the event name with a schema using\nschema_name.event_name syntax.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/create-event.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/create-event.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (88,4,'ABS','Syntax:\nABS(X)\n\nReturns the absolute value of X.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/mathematical-functions.html\n\n','mysql> SELECT ABS(2);\n -> 2\nmysql> SELECT ABS(-32);\n -> 32\n','http://dev.mysql.com/doc/refman/5.1/en/mathematical-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (89,31,'POLYFROMWKB','PolyFromWKB(wkb[,srid]), PolygonFromWKB(wkb[,srid])\n\nConstructs a POLYGON value using its WKB representation and SRID.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/creating-spatial-values.html#gis-wkb…','','http://dev.mysql.com/doc/refman/5.1/en/creating-spatial-values.html#gis-wkb…'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (90,35,'NOT LIKE','Syntax:\nexpr NOT LIKE pat [ESCAPE \'escape_char\']\n\nThis is the same as NOT (expr LIKE pat [ESCAPE \'escape_char\']).\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/string-comparison-functions.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/string-comparison-functions.html'); @@ -163,7 +163,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (94,16,'MAX','Syntax:\nMAX([DISTINCT] expr)\n\nReturns the maximum value of expr. MAX() may take a string argument; in\nsuch cases, it returns the maximum string value. See\nhttp://dev.mysql.com/doc/refman/5.1/en/mysql-indexes.html. The DISTINCT\nkeyword can be used to find the maximum of the distinct values of expr,\nhowever, this produces the same result as omitting DISTINCT.\n\nMAX() returns NULL if there were no matching rows.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/group-by-functions.html\n\n','mysql> SELECT student_name, MIN(test_score), MAX(test_score)\n -> FROM student\n -> GROUP BY student_name;\n','http://dev.mysql.com/doc/refman/5.1/en/group-by-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (95,21,'CREATE FUNCTION UDF','Syntax:\nCREATE [AGGREGATE] FUNCTION function_name RETURNS {STRING|INTEGER|REAL|DECIMAL}\n SONAME shared_library_name\n\nA user-defined function (UDF) is a way to extend MySQL with a new\nfunction that works like a native (built-in) MySQL function such as\nABS() or CONCAT().\n\nfunction_name is the name that should be used in SQL statements to\ninvoke the function. The RETURNS clause indicates the type of the\nfunction\'s return value. DECIMAL is a legal value after RETURNS, but\ncurrently DECIMAL functions return string values and should be written\nlike STRING functions.\n\nshared_library_name is the basename of the shared object file that\ncontains the code that implements the function. The file must be\nlocated in the plugin directory. This directory is given by the value\nof the plugin_dir system variable.\n\n*Note*: This is a change in MySQL 5.1. For earlier versions of MySQL,\nthe shared object can be located in any directory that is searched by\nyour system\'s dynamic linker.\n\nTo create a function, you must have the INSERT privilege for the mysql\ndatabase. This is necessary because CREATE FUNCTION adds a row to the\nmysql.func system table that records the function\'s name, type, and\nshared library name. If you do not have this table, you should run the\nmysql_upgrade command to create it. See\nhttp://dev.mysql.com/doc/refman/5.1/en/mysql-upgrade.html.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/create-function-udf.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/create-function-udf.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (96,4,'*','Syntax:\n*\n\nMultiplication:\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/arithmetic-functions.html\n\n','mysql> SELECT 3*5;\n -> 15\nmysql> SELECT 18014398509481984*18014398509481984.0;\n -> 324518553658426726783156020576256.0\nmysql> SELECT 18014398509481984*18014398509481984;\n -> 0\n','http://dev.mysql.com/doc/refman/5.1/en/arithmetic-functions.html'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (97,20,'TIMESTAMP','TIMESTAMP\n\nA timestamp. The range is \'1970-01-01 00:00:01\' UTC to \'2038-01-09\n03:14:07\' UTC. TIMESTAMP values are stored as the number of seconds\nsince the epoch (\'1970-01-01 00:00:00\' UTC). A TIMESTAMP cannot\nrepresent the value \'1970-01-01 00:00:00\' because that is equivalent to\n0 seconds from the epoch and the value 0 is reserved for representing\n\'0000-00-00 00:00:00\', the "zero" TIMESTAMP value.\n\nA TIMESTAMP column is useful for recording the date and time of an\nINSERT or UPDATE operation. By default, the first TIMESTAMP column in a\ntable is automatically set to the date and time of the most recent\noperation if you do not assign it a value yourself. You can also set\nany TIMESTAMP column to the current date and time by assigning it a\nNULL value. Variations on automatic initialization and update\nproperties are described in\nhttp://dev.mysql.com/doc/refman/5.1/en/timestamp.html.\n\nA TIMESTAMP value is returned as a string in the format \'YYYY-MM-DD\nHH:MM:SS\' with a display width fixed at 19 characters. To obtain the\nvalue as a number, you should add +0 to the timestamp column.\n\n*Note*: The TIMESTAMP format that was used prior to MySQL 4.1 is not\nsupported in MySQL 5.1; see MySQL 3.23, 4.0, 4.1 Reference Manual for\ninformation regarding the old format.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-type-overview.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-type-overview.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (97,20,'TIMESTAMP','TIMESTAMP\n\nA timestamp. The range is \'1970-01-01 00:00:01\' UTC to \'2038-01-19\n03:14:07\' UTC. TIMESTAMP values are stored as the number of seconds\nsince the epoch (\'1970-01-01 00:00:00\' UTC). A TIMESTAMP cannot\nrepresent the value \'1970-01-01 00:00:00\' because that is equivalent to\n0 seconds from the epoch and the value 0 is reserved for representing\n\'0000-00-00 00:00:00\', the "zero" TIMESTAMP value.\n\nA TIMESTAMP column is useful for recording the date and time of an\nINSERT or UPDATE operation. By default, the first TIMESTAMP column in a\ntable is automatically set to the date and time of the most recent\noperation if you do not assign it a value yourself. You can also set\nany TIMESTAMP column to the current date and time by assigning it a\nNULL value. Variations on automatic initialization and update\nproperties are described in\nhttp://dev.mysql.com/doc/refman/5.1/en/timestamp.html.\n\nA TIMESTAMP value is returned as a string in the format \'YYYY-MM-DD\nHH:MM:SS\' with a display width fixed at 19 characters. To obtain the\nvalue as a number, you should add +0 to the timestamp column.\n\n*Note*: The TIMESTAMP format that was used prior to MySQL 4.1 is not\nsupported in MySQL 5.1; see MySQL 3.23, 4.0, 4.1 Reference Manual for\ninformation regarding the old format.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-type-overview.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-type-overview.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (98,11,'DES_DECRYPT','Syntax:\nDES_DECRYPT(crypt_str[,key_str])\n\nDecrypts a string encrypted with DES_ENCRYPT(). If an error occurs,\nthis function returns NULL.\n\nThis function works only if MySQL has been configured with SSL support.\nSee http://dev.mysql.com/doc/refman/5.1/en/secure-connections.html.\n\nIf no key_str argument is given, DES_DECRYPT() examines the first byte\nof the encrypted string to determine the DES key number that was used\nto encrypt the original string, and then reads the key from the DES key\nfile to decrypt the message. For this to work, the user must have the\nSUPER privilege. The key file can be specified with the --des-key-file\nserver option.\n\nIf you pass this function a key_str argument, that string is used as\nthe key for decrypting the message.\n\nIf the crypt_str argument does not appear to be an encrypted string,\nMySQL returns the given crypt_str.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/encryption-functions.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/encryption-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (99,25,'CACHE INDEX','Syntax:\nCACHE INDEX\n tbl_index_list [, tbl_index_list] ...\n IN key_cache_name\n\ntbl_index_list:\n tbl_name [[INDEX|KEY] (index_name[, index_name] ...)]\n\nThe CACHE INDEX statement assigns table indexes to a specific key\ncache. It is used only for MyISAM tables.\n\nThe following statement assigns indexes from the tables t1, t2, and t3\nto the key cache named hot_cache:\n\nmysql> CACHE INDEX t1, t2, t3 IN hot_cache;\n+---------+--------------------+----------+----------+\n| Table | Op | Msg_type | Msg_text |\n+---------+--------------------+----------+----------+\n| test.t1 | assign_to_keycache | status | OK |\n| test.t2 | assign_to_keycache | status | OK |\n| test.t3 | assign_to_keycache | status | OK |\n+---------+--------------------+----------+----------+\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/cache-index.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/cache-index.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (100,12,'ENDPOINT','EndPoint(ls)\n\nReturns the Point that is the endpoint of the LineString value ls.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/geometry-property-functions.html#lin…','mysql> SET @ls = \'LineString(1 1,2 2,3 3)\';\nmysql> SELECT AsText(EndPoint(GeomFromText(@ls)));\n+-------------------------------------+\n| AsText(EndPoint(GeomFromText(@ls))) |\n+-------------------------------------+\n| POINT(3 3) |\n+-------------------------------------+\n','http://dev.mysql.com/doc/refman/5.1/en/geometry-property-functions.html#lin…'); @@ -202,7 +202,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (133,25,'SHOW STATUS','Syntax:\nSHOW [GLOBAL | SESSION] STATUS\n [LIKE \'pattern\' | WHERE expr]\n\nSHOW STATUS provides server status information. This information also\ncan be obtained using the mysqladmin extended-status command. The LIKE\nclause, if present, indicates which variable names to match. The WHERE\nclause can be given to select rows using more general conditions, as\ndiscussed in http://dev.mysql.com/doc/refman/5.1/en/extended-show.html.\nThis statement does not require any privilege. It requires only the\nability to connect to the server.\nWith a LIKE clause, the statement displays only rows for those\nvariables with names that match the pattern:\n\nmysql> SHOW STATUS LIKE \'Key%\';\n+--------------------+----------+\n| Variable_name | Value |\n+--------------------+----------+\n| Key_blocks_used | 14955 |\n| Key_read_requests | 96854827 |\n| Key_reads | 162040 |\n| Key_write_requests | 7589728 |\n| Key_writes | 3813196 |\n+--------------------+----------+\n\nWith the GLOBAL modifier, SHOW STATUS displays the status values for\nall connections to MySQL. With SESSION, it displays the status values\nfor the current connection. If no modifier is present, the default is\nSESSION. LOCAL is a synonym for SESSION.\n\nSome status variables have only a global value. For these, you get the\nsame value for both GLOBAL and SESSION. The scope for each status\nvariable is listed at\nhttp://dev.mysql.com/doc/refman/5.1/en/server-status-variables.html.\n\…: http://dev.mysql.com/doc/refman/5.1/en/show-status.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/show-status.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (134,35,'EXTRACTVALUE','Syntax:\nExtractValue(xml_frag, xpath_expr)\n\nExtractValue() takes two string arguments, a fragment of XML markup\nxml_frag and an XPath expression xpath_expr (also known as a locator);\nit returns the text (CDATA) of the first text node which is a child of\nthe element(s) matched by the XPath expression. It is the equivalent of\nperforming a match using the xpath_expr after appending /text(). In\nother words, ExtractValue(\'<a>Sakila</a>\', \'/a/b\') and\nExtractValue(\'<a>Sakila</a>\', \'/a/b/text()\') produce the same\nresult.\n\nIf multiple matches are found, then the content of the first child text\nnode of each matching element is returned (in the order matched) as a\nsingle, space-delimited string.\n\nIf no matching text node is found for the expression (including the\nimplicit /text()) --- for whatever reason, as long as xpath_expr is\nvalid, and xml_frag consists of elements which are properly nested and\nclosed --- an empty string is returned. No distinction is made between\na match on an empty element and no match at all. This is by design.\n\nIf you need to determine whether no matching element was found in\nxml_frag or such an element was found but contained no child text\nnodes, you should test the result of an expression that uses the XPath\ncount() function. For example, both of these statements return an empty\nstring, as shown here:\n\nmysql> SELECT ExtractValue(\'<a></a>\', \'/a/b\');\n+-------------------------------------+\n| ExtractValue(\'<a></a>\', \'/a/b\') |\n+-------------------------------------+\n| |\n+-------------------------------------+\n1 row in set (0.00 sec)\n\nmysql> SELECT ExtractValue(\'<a><c/></a>\', \'/a/b\');\n+-------------------------------------+\n| ExtractValue(\'<a><c/></a>\', \'/a/b\') |\n+-------------------------------------+\n| |\n+-------------------------------------+\n1 row in set (0.00 sec)\n\nHowever, you can determine whether there was actually a matching\nelement using the following:\n\nmysql> SELECT ExtractValue(\'<a></a>\', \'count(/a/b)\');\n+-------------------------------------+\n| ExtractValue(\'<a></a>\', \'count(/a/b)\') |\n+-------------------------------------+\n| 1 |\n+-------------------------------------+\n1 row in set (0.00 sec)\n\nmysql> SELECT ExtractValue(\'<a><c/></a>\', \'count(/a/b)\');\n+-------------------------------------+\n| ExtractValue(\'<a><c/></a>\', \'count(/a/b)\') |\n+-------------------------------------+\n| 0 |\n+-------------------------------------+\n1 row in set (0.01 sec)\n\n*Important*: ExtractValue() returns only CDATA, and does not return any\ntags that might be contained within a matching tag, nor any of their\ncontent (see the result returned as val1 in the following example).\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/xml-functions.html\n\n','mysql> SELECT\n -> ExtractValue(\'<a>cccddd</a>\', \'/a\') AS val1,\n -> ExtractValue(\'<a>cccddd</a>\', \'/a/b\') AS val2,\n -> ExtractValue(\'<a>cccddd</a>\', \'//b\') AS val3,\n -> ExtractValue(\'<a>cccddd</a>\', \'/b\') AS val4,\n -> ExtractValue(\'<a>cccdddeee</a>\', \'//b\') AS val5;\n\n+------+------+------+------+---------+\n| val1 | val2 | val3 | val4 | val5 |\n+------+------+------+------+---------+\n| ccc | ddd | ddd | | ddd eee |\n+------+------+------+------+---------+\n','http://dev.mysql.com/doc/refman/5.1/en/xml-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (135,11,'OLD_PASSWORD','Syntax:\nOLD_PASSWORD(str)\n\nOLD_PASSWORD() was added to MySQL when the implementation of PASSWORD()\nwas changed to improve security. OLD_PASSWORD() returns the value of\nthe old (pre-4.1) implementation of PASSWORD() as a binary string, and\nis intended to permit you to reset passwords for any pre-4.1 clients\nthat need to connect to your version 5.1 MySQL server without locking\nthem out. See\nhttp://dev.mysql.com/doc/refman/5.1/en/password-hashing.html.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/encryption-functions.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/encryption-functions.html'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (136,22,'SET VARIABLE','Syntax:\nSET var_name = expr [, var_name = expr] ...\n\nThe SET statement in stored programs is an extended version of the\ngeneral SET statement (see [HELP SET]). Referenced variables may be\nones declared inside a stored program, global system variables, or\nuser-defined variables.\n\nThe SET statement in stored programs is implemented as part of the\npre-existing SET syntax. This allows an extended syntax of SET a=x,\nb=y, ... where different variable types (locally declared variables,\nglobal and session server variables, user-defined variables) can be\nmixed. This also allows combinations of local variables and some\noptions that make sense only for system variables; in that case, the\noptions are recognized but ignored.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/set-statement.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/set-statement.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (136,22,'SET VARIABLE','Syntax:\nSET var_name = expr [, var_name = expr] ...\n\nThe SET statement in stored programs is an extended version of the\ngeneral SET statement (see [HELP SET]). Each var_name may refer to a\nlocal variable declared inside a stored program, a system variable, or\na user-defined variable.\n\nThe SET statement in stored programs is implemented as part of the\npre-existing SET syntax. This allows an extended syntax of SET a=x,\nb=y, ... where different variable types (locally declared variables,\nglobal and session system variables, user-defined variables) can be\nmixed. This also allows combinations of local variables and some\noptions that make sense only for system variables; in that case, the\noptions are recognized but ignored.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/set-statement.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/set-statement.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (137,35,'FORMAT','Syntax:\nFORMAT(X,D)\n\nFormats the number X to a format like \'#,###,###.##\', rounded to D\ndecimal places, and returns the result as a string. If D is 0, the\nresult has no decimal point or fractional part.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/string-functions.html\n\n','mysql> SELECT FORMAT(12332.123456, 4);\n -> \'12,332.1235\'\nmysql> SELECT FORMAT(12332.1,4);\n -> \'12,332.1000\'\nmysql> SELECT FORMAT(12332.2,0);\n -> \'12,332\'\n','http://dev.mysql.com/doc/refman/5.1/en/string-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (138,13,'||','Syntax:\nOR, ||\n\nLogical OR. When both operands are non-NULL, the result is 1 if any\noperand is nonzero, and 0 otherwise. With a NULL operand, the result is\n1 if the other operand is nonzero, and NULL otherwise. If both operands\nare NULL, the result is NULL.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/logical-operators.html\n\n','mysql> SELECT 1 || 1;\n -> 1\nmysql> SELECT 1 || 0;\n -> 1\nmysql> SELECT 0 || 0;\n -> 0\nmysql> SELECT 0 || NULL;\n -> NULL\nmysql> SELECT 1 || NULL;\n -> 1\n','http://dev.mysql.com/doc/refman/5.1/en/logical-operators.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (139,35,'BIT_LENGTH','Syntax:\nBIT_LENGTH(str)\n\nReturns the length of the string str in bits.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/string-functions.html\n\n','mysql> SELECT BIT_LENGTH(\'text\');\n -> 32\n','http://dev.mysql.com/doc/refman/5.1/en/string-functions.html'); @@ -233,7 +233,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (164,30,'SEC_TO_TIME','Syntax:\nSEC_TO_TIME(seconds)\n\nReturns the seconds argument, converted to hours, minutes, and seconds,\nas a TIME value. The range of the result is constrained to that of the\nTIME data type. A warning occurs if the argument corresponds to a value\noutside that range.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','mysql> SELECT SEC_TO_TIME(2378);\n -> \'00:39:38\'\nmysql> SELECT SEC_TO_TIME(2378) + 0;\n -> 3938\n','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (165,20,'FLOAT','FLOAT[(M,D)] [UNSIGNED] [ZEROFILL]\n\nA small (single-precision) floating-point number. Allowable values are\n-3.402823466E+38 to -1.175494351E-38, 0, and 1.175494351E-38 to\n3.402823466E+38. These are the theoretical limits, based on the IEEE\nstandard. The actual range might be slightly smaller depending on your\nhardware or operating system.\n\nM is the total number of digits and D is the number of digits following\nthe decimal point. If M and D are omitted, values are stored to the\nlimits allowed by the hardware. A single-precision floating-point\nnumber is accurate to approximately 7 decimal places.\n\nUNSIGNED, if specified, disallows negative values.\n\nUsing FLOAT might give you some unexpected problems because all\ncalculations in MySQL are done with double precision. See\nhttp://dev.mysql.com/doc/refman/5.1/en/no-matching-rows.html.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/numeric-type-overview.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/numeric-type-overview.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (166,35,'LOCATE','Syntax:\nLOCATE(substr,str), LOCATE(substr,str,pos)\n\nThe first syntax returns the position of the first occurrence of\nsubstring substr in string str. The second syntax returns the position\nof the first occurrence of substring substr in string str, starting at\nposition pos. Returns 0 if substr is not in str.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/string-functions.html\n\n','mysql> SELECT LOCATE(\'bar\', \'foobarbar\');\n -> 4\nmysql> SELECT LOCATE(\'xbar\', \'foobar\');\n -> 0\nmysql> SELECT LOCATE(\'bar\', \'foobarbar\', 5);\n -> 7\n','http://dev.mysql.com/doc/refman/5.1/en/string-functions.html'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (167,25,'SHOW EVENTS','Syntax:\nSHOW EVENTS [{FROM | IN} schema_name]\n [LIKE \'pattern\' | WHERE expr]\n\nIn its simplest form, SHOW EVENTS lists all of the events in the\ncurrent schema:\n\nmysql> SELECT CURRENT_USER(), SCHEMA();\n+----------------+----------+\n| CURRENT_USER() | SCHEMA() |\n+----------------+----------+\n| jon@ghidora | myschema |\n+----------------+----------+\n1 row in set (0.00 sec)\n\nmysql> SHOW EVENTS\\G\n*************************** 1. row ***************************\n Db: myschema\n Name: e_daily\n Definer: jon@ghidora\n Time zone: SYSTEM\n Type: RECURRING\n Execute at: NULL\n Interval value: 10\n Interval field: SECOND\n Starts: 2006-02-09 10:41:23\n Ends: 0000-00-00 00:00:00\n Status: ENABLED\n Originator: 0\ncharacter_set_client: latin1\ncollation_connection: latin1_swedish_ci\n Database Collation: latin1_swedish_ci\n\nThe LIKE clause, if present, indicates which event names to match. The\nWHERE clause can be given to select rows using more general conditions,\nas discussed in\nhttp://dev.mysql.com/doc/refman/5.1/en/extended-show.html.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/show-events.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/show-events.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (167,25,'SHOW EVENTS','Syntax:\nSHOW EVENTS [{FROM | IN} schema_name]\n [LIKE \'pattern\' | WHERE expr]\n\nIn its simplest form, SHOW EVENTS lists all of the events in the\ncurrent schema:\n\nmysql> SELECT CURRENT_USER(), SCHEMA();\n+----------------+----------+\n| CURRENT_USER() | SCHEMA() |\n+----------------+----------+\n| jon@ghidora | myschema |\n+----------------+----------+\n1 row in set (0.00 sec)\n\nmysql> SHOW EVENTS\\G\n*************************** 1. row ***************************\n Db: myschema\n Name: e_daily\n Definer: jon@ghidora\n Time zone: SYSTEM\n Type: RECURRING\n Execute at: NULL\n Interval value: 10\n Interval field: SECOND\n Starts: 2006-02-09 10:41:23\n Ends: NULL\n Status: ENABLED\n Originator: 0\ncharacter_set_client: latin1\ncollation_connection: latin1_swedish_ci\n Database Collation: latin1_swedish_ci\n\nTo see events for a specific schema, use the FROM clause. For example,\nto see events for the test schema, use the following statement:\n\nSHOW EVENTS FROM test;\n\nThe LIKE clause, if present, indicates which event names to match. The\nWHERE clause can be given to select rows using more general conditions,\nas discussed in\nhttp://dev.mysql.com/doc/refman/5.1/en/extended-show.html.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/show-events.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/show-events.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (168,15,'CHARSET','Syntax:\nCHARSET(str)\n\nReturns the character set of the string argument.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/information-functions.html\n\n','mysql> SELECT CHARSET(\'abc\');\n -> \'latin1\'\nmysql> SELECT CHARSET(CONVERT(\'abc\' USING utf8));\n -> \'utf8\'\nmysql> SELECT CHARSET(USER());\n -> \'utf8\'\n','http://dev.mysql.com/doc/refman/5.1/en/information-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (169,30,'SUBDATE','Syntax:\nSUBDATE(date,INTERVAL expr unit), SUBDATE(expr,days)\n\nWhen invoked with the INTERVAL form of the second argument, SUBDATE()\nis a synonym for DATE_SUB(). For information on the INTERVAL unit\nargument, see the discussion for DATE_ADD().\n\nmysql> SELECT DATE_SUB(\'2008-01-02\', INTERVAL 31 DAY);\n -> \'2007-12-02\'\nmysql> SELECT SUBDATE(\'2008-01-02\', INTERVAL 31 DAY);\n -> \'2007-12-02\'\n\nThe second form allows the use of an integer value for days. In such\ncases, it is interpreted as the number of days to be subtracted from\nthe date or datetime expression expr.\n\nmysql> SELECT SUBDATE(\'2008-01-02 12:00:00\', 31);\n -> \'2007-12-02 12:00:00\'\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (170,30,'DAYOFYEAR','Syntax:\nDAYOFYEAR(date)\n\nReturns the day of the year for date, in the range 1 to 366.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','mysql> SELECT DAYOFYEAR(\'2007-02-03\');\n -> 34\n','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); @@ -250,7 +250,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (181,24,'NUMGEOMETRIES','NumGeometries(gc)\n\nReturns the number of geometries in the GeometryCollection value gc.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/geometry-property-functions.html#geo…','mysql> SET @gc = \'GeometryCollection(Point(1 1),LineString(2 2, 3 3))\';\nmysql> SELECT NumGeometries(GeomFromText(@gc));\n+----------------------------------+\n| NumGeometries(GeomFromText(@gc)) |\n+----------------------------------+\n| 2 |\n+----------------------------------+\n','http://dev.mysql.com/doc/refman/5.1/en/geometry-property-functions.html#geo…'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (182,30,'MONTHNAME','Syntax:\nMONTHNAME(date)\n\nReturns the full name of the month for date. As of MySQL 5.1.12, the\nlanguage used for the name is controlled by the value of the\nlc_time_names system variable\n(http://dev.mysql.com/doc/refman/5.1/en/locale-support.html).\n\n…: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','mysql> SELECT MONTHNAME(\'2008-02-03\');\n -> \'February\'\n','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (183,36,'PROCEDURE ANALYSE','Syntax:\nANALYSE([max_elements[,max_memory]])\n\nANALYSE() is defined in the sql/sql_analyse.cc source file, which\nserves as an example of how to create a procedure for use with the\nPROCEDURE clause of SELECT statements. ANALYSE() is built in and is\navailable by default; other procedures can be created using the format\ndemonstrated in the source file.\n\nANALYSE() examines the result from a query and returns an analysis of\nthe results that suggests optimal data types for each column that may\nhelp reduce table sizes. To obtain this analysis, append PROCEDURE\nANALYSE to the end of a SELECT statement:\n\nSELECT ... FROM ... WHERE ... PROCEDURE ANALYSE([max_elements,[max_memory]])\n\nFor example:\n\nSELECT col1, col2 FROM table1 PROCEDURE ANALYSE(10, 2000);\n\nThe results show some statistics for the values returned by the query,\nand propose an optimal data type for the columns. This can be helpful\nfor checking your existing tables, or after importing new data. You may\nneed to try different settings for the arguments so that PROCEDURE\nANALYSE() does not suggest the ENUM data type when it is not\nappropriate.\n\nThe arguments are optional and are used as follows:\n\no max_elements (default 256) is the maximum number of distinct values\n that ANALYSE() notices per column. This is used by ANALYSE() to check\n whether the optimal data type should be of type ENUM; if there are\n more than max_elements distinct values, then ENUM is not a suggested\n type.\n\no max_memory (default 8192) is the maximum amount of memory that\n ANALYSE() should allocate per column while trying to find all\n distinct values.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/procedure-analyse.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/procedure-analyse.html'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (184,25,'CHANGE MASTER TO','Syntax:\nCHANGE MASTER TO master_def [, master_def] ...\n\nmaster_def:\n MASTER_BIND = \'interface_name\'\n | MASTER_HOST = \'host_name\'\n | MASTER_USER = \'user_name\'\n | MASTER_PASSWORD = \'password\'\n | MASTER_PORT = port_num\n | MASTER_CONNECT_RETRY = interval\n | MASTER_HEARTBEAT_PERIOD = interval\n | MASTER_LOG_FILE = \'master_log_name\'\n | MASTER_LOG_POS = master_log_pos\n | RELAY_LOG_FILE = \'relay_log_name\'\n | RELAY_LOG_POS = relay_log_pos\n | MASTER_SSL = {0|1}\n | MASTER_SSL_CA = \'ca_file_name\'\n | MASTER_SSL_CAPATH = \'ca_directory_name\'\n | MASTER_SSL_CERT = \'cert_file_name\'\n | MASTER_SSL_KEY = \'key_file_name\'\n | MASTER_SSL_CIPHER = \'cipher_list\'\n | MASTER_SSL_VERIFY_SERVER_CERT = {0|1}\n\nCHANGE MASTER TO changes the parameters that the slave server uses for\nconnecting to and communicating with the master server. It also updates\nthe contents of the master.info and relay-log.info files.\n\nMASTER_USER, MASTER_PASSWORD, MASTER_SSL, MASTER_SSL_CA,\nMASTER_SSL_CAPATH, MASTER_SSL_CERT, MASTER_SSL_KEY, MASTER_SSL_CIPHER,\nand MASTER_SSL_VERIFY_SERVER_CERT provide information to the slave\nabout how to connect to its master. MASTER_SSL_VERIFY_SERVER_CERT was\nadded in MySQL 5.1.18. It is used as described for the\n--ssl-verify-server-cert option in\nhttp://dev.mysql.com/doc/refman/5.1/en/ssl-options.html.\n\nMASTER_CONN… specifies how many seconds to wait between connect\nretries. The default is 60. The number of reconnection attempts is\nlimited by the --master-retry-count server option; for more\ninformation, see\nhttp://dev.mysql.com/doc/refman/5.1/en/replication-options.html.\n\nThe SSL options (MASTER_SSL, MASTER_SSL_CA, MASTER_SSL_CAPATH,\nMASTER_SSL_CERT, MASTER_SSL_KEY, MASTER_SSL_CIPHER), and\nMASTER_SSL_VERIFY_SERVER_CERT can be changed even on slaves that are\ncompiled without SSL support. They are saved to the master.info file,\nbut are ignored unless you use a server that has SSL support enabled.\n\nIf you don\'t specify a given parameter, it keeps its old value, except\nas indicated in the following discussion. For example, if the password\nto connect to your MySQL master has changed, you just need to issue\nthese statements to tell the slave about the new password:\n\nSTOP SLAVE; -- if replication was running\nCHANGE MASTER TO MASTER_PASSWORD=\'new3cret\';\nSTART SLAVE; -- if you want to restart replication\n\nThere is no need to specify the parameters that do not change (host,\nport, user, and so forth).\n\nMASTER_HOST and MASTER_PORT are the host name (or IP address) of the\nmaster host and its TCP/IP port.\n\nThe next two options are available only in MySQL Cluster NDB 6.3 and\n6.4; they are not supported in mainline MySQL 5.1:\n\no MASTER_BIND is for use on replication slaves having multiple network\n interfaces, and determines which of the slave\'s network interfaces is\n chosen for connecting to the master. It is also possible to determine\n which network interface is to be used in such cases by starting the\n slave mysqld process with the --master-bind option.\n\n The ability to bind a replication slave to specific network interface\n was added in MySQL Cluster NDB 6.3.4.\n\no MASTER_HEARTBEAT_PERIOD is used to set the interval in seconds\n between replication heartbeats. Whenever the master\'s binlog is\n updated with an event, the waiting period for the next heartbeat is\n reset. interval is a decimal value having the range 0 to 4294967\n seconds and a resolution to hundredths of a second; the smallest\n nonzero value is 0.001. Heartbeats are sent by the master only if\n there are no unsent events in the binlog file for a period longer\n than interval.\n\n Setting interval to 0 disables heartbeats altogether. The default\n value for interval is equal to the value of slave_net_timeout divided\n by 2.\n\n *Note*: Setting @@global.slave_net_timeout to a value less than that\n of the current heartbeat interval results in a warning being issued.\n\n Issuing RESET SLAVE resets the heartbeat interval to the default.\n\n MASTER_HEARTBEAT_PERIOD was added in MySQL Cluster NDB 6.3.4.\n\n*Note*: Replication cannot use Unix socket files. You must be able to\nconnect to the master MySQL server using TCP/IP.\n\nIf you specify MASTER_HOST or MASTER_PORT, the slave assumes that the\nmaster server is different from before (even if you specify a host or\nport value that is the same as the current value.) In this case, the\nold values for the master binary log name and position are considered\nno longer applicable, so if you do not specify MASTER_LOG_FILE and\nMASTER_LOG_POS in the statement, MASTER_LOG_FILE=\'\' and\nMASTER_LOG_POS=4 are silently appended to it.\n\nSetting MASTER_HOST=\'\' --- that is, setting its value explicitly to an\nempty string --- is not the same as not setting it at all. Setting this\noption to an empty string causes START SLAVE subsequently to fail. This\nissue is addressed in MySQL 6.0. (Bug#28796\n(http://bugs.mysql.com/28796))\n\nMASTER_LOG_FILE and MASTER_LOG_POS are the coordinates at which the\nslave I/O thread should begin reading from the master the next time the\nthread starts. If you specify either of them, you cannot specify\nRELAY_LOG_FILE or RELAY_LOG_POS. If neither of MASTER_LOG_FILE or\nMASTER_LOG_POS are specified, the slave uses the last coordinates of\nthe slave SQL thread before CHANGE MASTER TO was issued. This ensures\nthat there is no discontinuity in replication, even if the slave SQL\nthread was late compared to the slave I/O thread, when you merely want\nto change, say, the password to use.\n\nCHANGE MASTER TO deletes all relay log files and starts a new one,\nunless you specify RELAY_LOG_FILE or RELAY_LOG_POS. In that case, relay\nlogs are kept; the relay_log_purge global variable is set silently to\n0.\n\nCHANGE MASTER TO is useful for setting up a slave when you have the\nsnapshot of the master and have recorded the log and the offset\ncorresponding to it. After loading the snapshot into the slave, you can\nrun CHANGE MASTER TO MASTER_LOG_FILE=\'log_name_on_master\',\nMASTER_LOG_POS=log_offset_on_master on the slave.\n\nThe following example changes the master and master\'s binary log\ncoordinates. This is used when you want to set up the slave to\nreplicate the master:\n\nCHANGE MASTER TO\n MASTER_HOST=\'master2.mycompany.com\',\n MASTER_USER=\'replication\',\n MASTER_PASSWORD=\'bigs3cret\',\n MASTER_PORT=3306,\n MASTER_LOG_FILE=\'master2-bin.001\',\n MASTER_LOG_POS=4,\n MASTER_CONNECT_RETRY=10;\n\nThe next example shows an operation that is less frequently employed.\nIt is used when the slave has relay logs that you want it to execute\nagain for some reason. To do this, the master need not be reachable.\nYou need only use CHANGE MASTER TO and start the SQL thread (START\nSLAVE SQL_THREAD):\n\nCHANGE MASTER TO\n RELAY_LOG_FILE=\'slave-relay-bin.006\',\n RELAY_LOG_POS=4025;\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/change-master-to.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/change-master-to.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (184,25,'CHANGE MASTER TO','Syntax:\nCHANGE MASTER TO master_def [, master_def] ...\n\nmaster_def:\n MASTER_BIND = \'interface_name\'\n | MASTER_HOST = \'host_name\'\n | MASTER_USER = \'user_name\'\n | MASTER_PASSWORD = \'password\'\n | MASTER_PORT = port_num\n | MASTER_CONNECT_RETRY = interval\n | MASTER_HEARTBEAT_PERIOD = interval\n | MASTER_LOG_FILE = \'master_log_name\'\n | MASTER_LOG_POS = master_log_pos\n | RELAY_LOG_FILE = \'relay_log_name\'\n | RELAY_LOG_POS = relay_log_pos\n | MASTER_SSL = {0|1}\n | MASTER_SSL_CA = \'ca_file_name\'\n | MASTER_SSL_CAPATH = \'ca_directory_name\'\n | MASTER_SSL_CERT = \'cert_file_name\'\n | MASTER_SSL_KEY = \'key_file_name\'\n | MASTER_SSL_CIPHER = \'cipher_list\'\n | MASTER_SSL_VERIFY_SERVER_CERT = {0|1}\n\nCHANGE MASTER TO changes the parameters that the slave server uses for\nconnecting to and communicating with the master server. It also updates\nthe contents of the master.info and relay-log.info files.\n\nMASTER_USER, MASTER_PASSWORD, MASTER_SSL, MASTER_SSL_CA,\nMASTER_SSL_CAPATH, MASTER_SSL_CERT, MASTER_SSL_KEY, MASTER_SSL_CIPHER,\nand MASTER_SSL_VERIFY_SERVER_CERT provide information to the slave\nabout how to connect to its master. MASTER_SSL_VERIFY_SERVER_CERT was\nadded in MySQL 5.1.18. It is used as described for the\n--ssl-verify-server-cert option in\nhttp://dev.mysql.com/doc/refman/5.1/en/ssl-options.html.\n\nMASTER_CONN… specifies how many seconds to wait between connect\nretries. The default is 60. The number of reconnection attempts is\nlimited by the --master-retry-count server option; for more\ninformation, see\nhttp://dev.mysql.com/doc/refman/5.1/en/replication-options.html.\n\nThe SSL options (MASTER_SSL, MASTER_SSL_CA, MASTER_SSL_CAPATH,\nMASTER_SSL_CERT, MASTER_SSL_KEY, MASTER_SSL_CIPHER), and\nMASTER_SSL_VERIFY_SERVER_CERT can be changed even on slaves that are\ncompiled without SSL support. They are saved to the master.info file,\nbut are ignored unless you use a server that has SSL support enabled.\n\nIf you do not specify a given parameter, it keeps its old value, except\nas indicated in the following discussion. For example, if the password\nto connect to your MySQL master has changed, you just need to issue\nthese statements to tell the slave about the new password:\n\nSTOP SLAVE; -- if replication was running\nCHANGE MASTER TO MASTER_PASSWORD=\'new3cret\';\nSTART SLAVE; -- if you want to restart replication\n\nThere is no need to specify the parameters that do not change (host,\nport, user, and so forth).\n\nMASTER_HOST and MASTER_PORT are the host name (or IP address) of the\nmaster host and its TCP/IP port.\n\nThe next two options (MASTER_BIND and MASTER_HEARTBEAT_PERIOD) are\navailable in MySQL Cluster NDB 6.3 and later, but are not supported in\nmainline MySQL 5.1:\n\no MASTER_BIND is for use on replication slaves having multiple network\n interfaces, and determines which of the slave\'s network interfaces is\n chosen for connecting to the master. It is also possible to determine\n which network interface is to be used in such cases by starting the\n slave mysqld process with the --master-bind option.\n\n The ability to bind a replication slave to specific network interface\n was added in MySQL Cluster NDB 6.3.4.\n\no MASTER_HEARTBEAT_PERIOD is used to set the interval in seconds\n between replication heartbeats. Whenever the master\'s binlog is\n updated with an event, the waiting period for the next heartbeat is\n reset. interval is a decimal value having the range 0 to 4294967\n seconds and a resolution to hundredths of a second; the smallest\n nonzero value is 0.001. Heartbeats are sent by the master only if\n there are no unsent events in the binlog file for a period longer\n than interval.\n\n Setting interval to 0 disables heartbeats altogether. The default\n value for interval is equal to the value of slave_net_timeout divided\n by 2.\n\n Setting @@global.slave_net_timeout to a value less than that of the\n current heartbeat interval results in a warning being issued. The\n effect of issuing RESET SLAVE on the heartbeat interval is to reset\n it to the default value.\n\n MASTER_HEARTBEAT_PERIOD was added in MySQL Cluster NDB 6.3.4.\n\n*Note*: Replication cannot use Unix socket files. You must be able to\nconnect to the master MySQL server using TCP/IP.\n\nIf you specify MASTER_HOST or MASTER_PORT, the slave assumes that the\nmaster server is different from before (even if you specify a host or\nport value that is the same as the current value.) In this case, the\nold values for the master binary log name and position are considered\nno longer applicable, so if you do not specify MASTER_LOG_FILE and\nMASTER_LOG_POS in the statement, MASTER_LOG_FILE=\'\' and\nMASTER_LOG_POS=4 are silently appended to it.\n\nSetting MASTER_HOST=\'\' --- that is, setting its value explicitly to an\nempty string --- is not the same as not setting it at all. Setting this\noption to an empty string causes START SLAVE subsequently to fail. This\nissue is addressed in MySQL 5.5. (Bug#28796\n(http://bugs.mysql.com/28796))\n\nMASTER_LOG_FILE and MASTER_LOG_POS are the coordinates at which the\nslave I/O thread should begin reading from the master the next time the\nthread starts. If you specify either of them, you cannot specify\nRELAY_LOG_FILE or RELAY_LOG_POS. If neither of MASTER_LOG_FILE or\nMASTER_LOG_POS are specified, the slave uses the last coordinates of\nthe slave SQL thread before CHANGE MASTER TO was issued. This ensures\nthat there is no discontinuity in replication, even if the slave SQL\nthread was late compared to the slave I/O thread, when you merely want\nto change, say, the password to use.\n\nCHANGE MASTER TO deletes all relay log files and starts a new one,\nunless you specify RELAY_LOG_FILE or RELAY_LOG_POS. In that case, relay\nlogs are kept; the relay_log_purge global variable is set silently to\n0.\n\nCHANGE MASTER TO is useful for setting up a slave when you have the\nsnapshot of the master and have recorded the log and the offset\ncorresponding to it. After loading the snapshot into the slave, you can\nrun CHANGE MASTER TO MASTER_LOG_FILE=\'log_name_on_master\',\nMASTER_LOG_POS=log_offset_on_master on the slave.\n\nThe following example changes the master and master\'s binary log\ncoordinates. This is used when you want to set up the slave to\nreplicate the master:\n\nCHANGE MASTER TO\n MASTER_HOST=\'master2.mycompany.com\',\n MASTER_USER=\'replication\',\n MASTER_PASSWORD=\'bigs3cret\',\n MASTER_PORT=3306,\n MASTER_LOG_FILE=\'master2-bin.001\',\n MASTER_LOG_POS=4,\n MASTER_CONNECT_RETRY=10;\n\nThe next example shows an operation that is less frequently employed.\nIt is used when the slave has relay logs that you want it to execute\nagain for some reason. To do this, the master need not be reachable.\nYou need only use CHANGE MASTER TO and start the SQL thread (START\nSLAVE SQL_THREAD):\n\nCHANGE MASTER TO\n RELAY_LOG_FILE=\'slave-relay-bin.006\',\n RELAY_LOG_POS=4025;\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/change-master-to.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/change-master-to.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (185,37,'DROP DATABASE','Syntax:\nDROP {DATABASE | SCHEMA} [IF EXISTS] db_name\n\nDROP DATABASE drops all tables in the database and deletes the\ndatabase. Be very careful with this statement! To use DROP DATABASE,\nyou need the DROP privilege on the database. DROP SCHEMA is a synonym\nfor DROP DATABASE.\n\n*Important*: When a database is dropped, user privileges on the\ndatabase are not automatically dropped. See [HELP GRANT].\n\nIF EXISTS is used to prevent an error from occurring if the database\ndoes not exist.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/drop-database.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/drop-database.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (186,6,'MBREQUAL','MBREqual(g1,g2)\n\nReturns 1 or 0 to indicate whether the Minimum Bounding Rectangles of\nthe two geometries g1 and g2 are the same.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/relations-on-geometry-mbr.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/relations-on-geometry-mbr.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (187,30,'TIMESTAMP FUNCTION','Syntax:\nTIMESTAMP(expr), TIMESTAMP(expr1,expr2)\n\nWith a single argument, this function returns the date or datetime\nexpression expr as a datetime value. With two arguments, it adds the\ntime expression expr2 to the date or datetime expression expr1 and\nreturns the result as a datetime value.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','mysql> SELECT TIMESTAMP(\'2003-12-31\');\n -> \'2003-12-31 00:00:00\'\nmysql> SELECT TIMESTAMP(\'2003-12-31 12:00:00\',\'12:00:00\');\n -> \'2004-01-01 00:00:00\'\n','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); @@ -322,7 +322,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (253,20,'VARCHAR','[NATIONAL] VARCHAR(M) [CHARACTER SET charset_name] [COLLATE\ncollation_name]\n\nA variable-length string. M represents the maximum column length in\ncharacters. The range of M is 0 to 65,535. The effective maximum length\nof a VARCHAR is subject to the maximum row size (65,535 bytes, which is\nshared among all columns) and the character set used. For example, utf8\ncharacters can require up to three bytes per character, so a VARCHAR\ncolumn that uses the utf8 character set can be declared to be a maximum\nof 21,844 characters.\n\nMySQL stores VARCHAR values as a one-byte or two-byte length prefix\nplus data. The length prefix indicates the number of bytes in the\nvalue. A VARCHAR column uses one length byte if values require no more\nthan 255 bytes, two length bytes if values may require more than 255\nbytes.\n\n*Note*: MySQL 5.1 follows the standard SQL specification, and does not\nremove trailing spaces from VARCHAR values.\n\nVARCHAR is shorthand for CHARACTER VARYING. NATIONAL VARCHAR is the\nstandard SQL way to define that a VARCHAR column should use some\npredefined character set. MySQL 4.1 and up uses utf8 as this predefined\ncharacter set.\nhttp://dev.mysql.com/doc/refman/5.1/en/charset-national.html. NVARCHAR\nis shorthand for NATIONAL VARCHAR.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/string-type-overview.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/string-type-overview.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (254,35,'UNHEX','Syntax:\n\nUNHEX(str)\n\nPerforms the inverse operation of HEX(str). That is, it interprets each\npair of hexadecimal digits in the argument as a number and converts it\nto the character represented by the number. The resulting characters\nare returned as a binary string.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/string-functions.html\n\n','mysql> SELECT UNHEX(\'4D7953514C\');\n -> \'MySQL\'\nmysql> SELECT 0x4D7953514C;\n -> \'MySQL\'\nmysql> SELECT UNHEX(HEX(\'string\'));\n -> \'string\'\nmysql> SELECT HEX(UNHEX(\'1267\'));\n -> \'1267\'\n','http://dev.mysql.com/doc/refman/5.1/en/string-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (255,4,'- UNARY','Syntax:\n-\n\nUnary minus. This operator changes the sign of the argument.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/arithmetic-functions.html\n\n','mysql> SELECT - 2;\n -> -2\n','http://dev.mysql.com/doc/refman/5.1/en/arithmetic-functions.html'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (256,22,'SELECT INTO','Syntax:\nSELECT col_name [, col_name] ...\n INTO var_name [, var_name] ...\n table_expr\n\nSELECT ... INTO syntax enables selected columns to be stored directly\ninto variables. The query should return a single row. If the query\nreturns no rows, a warning with error code 1329 occurs (No data), and\nthe variable values remain unchanged. If the query returns multiple\nrows, error 1172 occurs (Result consisted of more than one row). If it\nis possible that the statement may retrieve multiple rows, you can use\nLIMIT 1 to limit the result set to a single row.\n\nIn the context of such statements that occur as part of events executed\nby the Event Scheduler, diagnostics messages (not only errors, but also\nwarnings) are written to the error log, and, on Windows, to the\napplication event log. For additional information, see\nhttp://dev.mysql.com/doc/refman/5.1/en/events-status-info.html.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/select-into-statement.html\n\n','SELECT id,data INTO x,y FROM test.t1 LIMIT 1;\n','http://dev.mysql.com/doc/refman/5.1/en/select-into-statement.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (256,22,'SELECT INTO','Syntax:\nSELECT col_name [, col_name] ...\n INTO var_name [, var_name] ...\n table_expr\n\nSELECT ... INTO syntax enables selected columns to be stored directly\ninto variables. The query should return a single row. If the query\nreturns no rows, a warning with error code 1329 occurs (No data), and\nthe variable values remain unchanged. If the query returns multiple\nrows, error 1172 occurs (Result consisted of more than one row). If it\nis possible that the statement may retrieve multiple rows, you can use\nLIMIT 1 to limit the result set to a single row.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/select-into-statement.html\n\n','SELECT id,data INTO x,y FROM test.t1 LIMIT 1;\n','http://dev.mysql.com/doc/refman/5.1/en/select-into-statement.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (257,16,'STD','Syntax:\nSTD(expr)\n\nReturns the population standard deviation of expr. This is an extension\nto standard SQL. The standard SQL function STDDEV_POP() can be used\ninstead.\n\nThis function returns NULL if there were no matching rows.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/group-by-functions.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/group-by-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (258,4,'COS','Syntax:\nCOS(X)\n\nReturns the cosine of X, where X is given in radians.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/mathematical-functions.html\n\n','mysql> SELECT COS(PI());\n -> -1\n','http://dev.mysql.com/doc/refman/5.1/en/mathematical-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (259,30,'DATE FUNCTION','Syntax:\nDATE(expr)\n\nExtracts the date part of the date or datetime expression expr.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','mysql> SELECT DATE(\'2003-12-31 01:02:03\');\n -> \'2003-12-31\'\n','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); @@ -384,7 +384,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (315,15,'FOUND_ROWS','Syntax:\nFOUND_ROWS()\n\nA SELECT statement may include a LIMIT clause to restrict the number of\nrows the server returns to the client. In some cases, it is desirable\nto know how many rows the statement would have returned without the\nLIMIT, but without running the statement again. To obtain this row\ncount, include a SQL_CALC_FOUND_ROWS option in the SELECT statement,\nand then invoke FOUND_ROWS() afterward:\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/information-functions.html\n\n','mysql> SELECT SQL_CALC_FOUND_ROWS * FROM tbl_name\n -> WHERE id > 100 LIMIT 10;\nmysql> SELECT FOUND_ROWS();\n','http://dev.mysql.com/doc/refman/5.1/en/information-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (316,15,'SYSTEM_USER','Syntax:\nSYSTEM_USER()\n\nSYSTEM_USER() is a synonym for USER().\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/information-functions.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/information-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (317,29,'CROSSES','Crosses(g1,g2)\n\nReturns 1 if g1 spatially crosses g2. Returns NULL if g1 is a Polygon\nor a MultiPolygon, or if g2 is a Point or a MultiPoint. Otherwise,\nreturns 0.\n\nThe term spatially crosses denotes a spatial relation between two given\ngeometries that has the following properties:\n\no The two geometries intersect\n\no Their intersection results in a geometry that has a dimension that is\n one less than the maximum dimension of the two given geometries\n\no Their intersection is not equal to either of the two given geometries\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/functions-that-test-spatial-relation…','','http://dev.mysql.com/doc/refman/5.1/en/functions-that-test-spatial-relation…'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (318,26,'TRUNCATE TABLE','Syntax:\nTRUNCATE [TABLE] tbl_name\n\nTRUNCATE TABLE empties a table completely. Logically, this is\nequivalent to a DELETE statement that deletes all rows, but there are\npractical differences under some circumstances.\n\nFor an InnoDB table, InnoDB processes TRUNCATE TABLE by deleting rows\none by one if there are any FOREIGN KEY constraints that reference the\ntable. If there are no FOREIGN KEY constraints, InnoDB performs fast\ntruncation by dropping the original table and creating an empty one\nwith the same definition, which is much faster than deleting rows one\nby one. The AUTO_INCREMENT counter is reset by TRUNCATE TABLE,\nregardless of whether there is a FOREIGN KEY constraint.\n\nIn the case that FOREIGN KEY constraints reference the table, InnoDB\ndeletes rows one by one and processes the constraints on each one. If\nthe FOREIGN KEY constraint specifies DELETE CASCADE, rows from the\nchild (referenced) table are deleted, and the truncated table becomes\nempty. If the FOREIGN KEY constraint does not specify CASCADE, the\nTRUNCATE statement deletes rows one by one and stops if it encounters a\nparent row that is referenced by the child, returning this error:\n\nERROR 1451 (23000): Cannot delete or update a parent row: a foreign\nkey constraint fails (`test`.`child`, CONSTRAINT `child_ibfk_1`\nFOREIGN KEY (`parent_id`) REFERENCES `parent` (`id`))\n\nThis is the same as a DELETE statement with no WHERE clause.\n\nBeginning with MySQL 5.1.32, TRUNCATE is treated for purposes of binary\nlogging and replication as DROP TABLE followed by CREATE TABLE --- that\nis, as DDL rather than DML. This is due to the fact that, when using\nInnoDB and other transactional storage engines where the transaction\nisolation level does not allow for statement-based logging (READ\nCOMMITTED or READ UNCOMMITTED), the statement was not logged and\nreplicated when using STATEMENT or MIXED logging mode. (Bug#36763\n(http://bugs.mysql.com/36763)) However, it is still applied on\nreplication slaves using InnoDB in the manner described previously.\n\nThe count of rows affected by TRUNCATE TABLE is accurate only when it\nis mapped to a DELETE statement.\n\nFor other storage engines, TRUNCATE TABLE differs from DELETE in the\nfollowing ways in MySQL 5.1:\n\no Truncate operations drop and re-create the table, which is much\n faster than deleting rows one by one, particularly for large tables.\n\no Truncate operations cause an implicit commit.\n\no Truncation operations cannot be performed if the session holds an\n active table lock.\n\no Truncation operations do not return a meaningful value for the number\n of deleted rows. The usual result is "0 rows affected," which should\n be interpreted as "no information."\n\no As long as the table format file tbl_name.frm is valid, the table can\n be re-created as an empty table with TRUNCATE TABLE, even if the data\n or index files have become corrupted.\n\no The table handler does not remember the last used AUTO_INCREMENT\n value, but starts counting from the beginning. This is true even for\n MyISAM and InnoDB, which normally do not reuse sequence values.\n\no When used with partitioned tables, TRUNCATE TABLE preserves the\n partitioning; that is, the data and index files are dropped and\n re-created, while the partition definitions (.par) file is\n unaffected.\n\no Since truncation of a table does not make any use of DELETE, the\n TRUNCATE statement does not invoke ON DELETE triggers.\n\nTRUNCATE TABLE requires the DROP privilege as of MySQL 5.1.16. (Before\n5.1.16, it requires the DELETE privilege.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/truncate.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/truncate.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (318,26,'TRUNCATE TABLE','Syntax:\nTRUNCATE [TABLE] tbl_name\n\nTRUNCATE TABLE empties a table completely. It requires the DROP\nprivilege as of MySQL 5.1.16. (Before 5.1.16, it requires the DELETE\nprivilege.\n\nLogically, TRUNCATE TABLE is equivalent to a DELETE statement that\ndeletes all rows, but there are practical differences under some\ncircumstances.\n\nFor an InnoDB table, InnoDB processes TRUNCATE TABLE by deleting rows\none by one if there are any FOREIGN KEY constraints that reference the\ntable. If there are no FOREIGN KEY constraints, InnoDB performs fast\ntruncation by dropping the original table and creating an empty one\nwith the same definition, which is much faster than deleting rows one\nby one. The AUTO_INCREMENT counter is reset by TRUNCATE TABLE,\nregardless of whether there is a FOREIGN KEY constraint.\n\nIn the case that FOREIGN KEY constraints reference the table, InnoDB\ndeletes rows one by one and processes the constraints on each one. If\nthe FOREIGN KEY constraint specifies DELETE CASCADE, rows from the\nchild (referenced) table are deleted, and the truncated table becomes\nempty. If the FOREIGN KEY constraint does not specify CASCADE, the\nTRUNCATE statement deletes rows one by one and stops if it encounters a\nparent row that is referenced by the child, returning this error:\n\nERROR 1451 (23000): Cannot delete or update a parent row: a foreign\nkey constraint fails (`test`.`child`, CONSTRAINT `child_ibfk_1`\nFOREIGN KEY (`parent_id`) REFERENCES `parent` (`id`))\n\nThis is the same as a DELETE statement with no WHERE clause.\n\nThe count of rows affected by TRUNCATE TABLE is accurate only when it\nis mapped to a DELETE statement.\n\nFor other storage engines, TRUNCATE TABLE differs from DELETE in the\nfollowing ways in MySQL 5.1:\n\no Truncate operations drop and re-create the table, which is much\n faster than deleting rows one by one, particularly for large tables.\n\no Truncate operations cause an implicit commit.\n\no Truncation operations cannot be performed if the session holds an\n active table lock.\n\no Truncation operations do not return a meaningful value for the number\n of deleted rows. The usual result is "0 rows affected," which should\n be interpreted as "no information."\n\no As long as the table format file tbl_name.frm is valid, the table can\n be re-created as an empty table with TRUNCATE TABLE, even if the data\n or index files have become corrupted.\n\no The table handler does not remember the last used AUTO_INCREMENT\n value, but starts counting from the beginning. This is true even for\n MyISAM and InnoDB, which normally do not reuse sequence values.\n\no When used with partitioned tables, TRUNCATE TABLE preserves the\n partitioning; that is, the data and index files are dropped and\n re-created, while the partition definitions (.par) file is\n unaffected.\n\no Since truncation of a table does not make any use of DELETE, the\n TRUNCATE statement does not invoke ON DELETE triggers.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/truncate.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/truncate.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (319,16,'BIT_XOR','Syntax:\nBIT_XOR(expr)\n\nReturns the bitwise XOR of all bits in expr. The calculation is\nperformed with 64-bit (BIGINT) precision.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/group-by-functions.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/group-by-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (320,30,'CURRENT_DATE','Syntax:\nCURRENT_DATE, CURRENT_DATE()\n\nCURRENT_DATE and CURRENT_DATE() are synonyms for CURDATE().\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (321,25,'START SLAVE','Syntax:\nSTART SLAVE [thread_type [, thread_type] ... ]\nSTART SLAVE [SQL_THREAD] UNTIL\n MASTER_LOG_FILE = \'log_name\', MASTER_LOG_POS = log_pos\nSTART SLAVE [SQL_THREAD] UNTIL\n RELAY_LOG_FILE = \'log_name\', RELAY_LOG_POS = log_pos\n\nthread_type: IO_THREAD | SQL_THREAD\n\nSTART SLAVE with no thread_type options starts both of the slave\nthreads. The I/O thread reads queries from the master server and stores\nthem in the relay log. The SQL thread reads the relay log and executes\nthe queries. START SLAVE requires the SUPER privilege.\n\nIf START SLAVE succeeds in starting the slave threads, it returns\nwithout any error. However, even in that case, it might be that the\nslave threads start and then later stop (for example, because they do\nnot manage to connect to the master or read its binary logs, or some\nother problem). START SLAVE does not warn you about this. You must\ncheck the slave\'s error log for error messages generated by the slave\nthreads, or check that they are running satisfactorily with SHOW SLAVE\nSTATUS.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/start-slave.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/start-slave.html'); @@ -410,7 +410,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (341,22,'LOOP','Syntax:\n[begin_label:] LOOP\n statement_list\nEND LOOP [end_label]\n\nLOOP implements a simple loop construct, enabling repeated execution of\nthe statement list, which consists of one or more statements, each\nterminated by a semicolon (;) statement delimiter. The statements\nwithin the loop are repeated until the loop is exited; usually this is\naccomplished with a LEAVE statement.\n\nA LOOP statement can be labeled. end_label cannot be given unless\nbegin_label also is present. If both are present, they must be the\nsame.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/loop-statement.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/loop-statement.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (342,4,'TRUNCATE','Syntax:\nTRUNCATE(X,D)\n\nReturns the number X, truncated to D decimal places. If D is 0, the\nresult has no decimal point or fractional part. D can be negative to\ncause D digits left of the decimal point of the value X to become zero.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/mathematical-functions.html\n\n','mysql> SELECT TRUNCATE(1.223,1);\n -> 1.2\nmysql> SELECT TRUNCATE(1.999,1);\n -> 1.9\nmysql> SELECT TRUNCATE(1.999,0);\n -> 1\nmysql> SELECT TRUNCATE(-1.999,1);\n -> -1.9\nmysql> SELECT TRUNCATE(122,-2);\n -> 100\nmysql> SELECT TRUNCATE(10.28*100,0);\n -> 1028\n','http://dev.mysql.com/doc/refman/5.1/en/mathematical-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (343,30,'TIMESTAMPADD','Syntax:\nTIMESTAMPADD(unit,interval,datetime_expr)\n\nAdds the integer expression interval to the date or datetime expression\ndatetime_expr. The unit for interval is given by the unit argument,\nwhich should be one of the following values: FRAC_SECOND\n(microseconds), SECOND, MINUTE, HOUR, DAY, WEEK, MONTH, QUARTER, or\nYEAR.\n\nBeginning with MySQL 5.1.24, it is possible to use MICROSECOND in place\nof FRAC_SECOND with this function, and FRAC_SECOND is deprecated.\n\nThe unit value may be specified using one of keywords as shown, or with\na prefix of SQL_TSI_. For example, DAY and SQL_TSI_DAY both are legal.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','mysql> SELECT TIMESTAMPADD(MINUTE,1,\'2003-01-02\');\n -> \'2003-01-02 00:01:00\'\nmysql> SELECT TIMESTAMPADD(WEEK,1,\'2003-01-02\');\n -> \'2003-01-09\'\n','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (344,25,'SHOW','SHOW has many forms that provide information about databases, tables,\ncolumns, or status information about the server. This section describes\nthose following:\n\nSHOW AUTHORS\nSHOW CHARACTER SET [like_or_where]\nSHOW COLLATION [like_or_where]\nSHOW [FULL] COLUMNS FROM tbl_name [FROM db_name] [like_or_where]\nSHOW CONTRIBUTORS\nSHOW CREATE DATABASE db_name\nSHOW CREATE EVENT event_name\nSHOW CREATE FUNCTION func_name\nSHOW CREATE PROCEDURE proc_name\nSHOW CREATE TABLE tbl_name\nSHOW CREATE TRIGGER trigger_name\nSHOW CREATE VIEW view_name\nSHOW DATABASES [like_or_where]\nSHOW ENGINE engine_name {STATUS | MUTEX}\nSHOW [STORAGE] ENGINES\nSHOW ERRORS [LIMIT [offset,] row_count]\nSHOW [FULL] EVENTS\nSHOW FUNCTION CODE func_name\nSHOW FUNCTION STATUS [like_or_where]\nSHOW GRANTS FOR user\nSHOW INDEX FROM tbl_name [FROM db_name]\nSHOW INNODB STATUS\nSHOW OPEN TABLES [FROM db_name] [like_or_where]\nSHOW PLUGINS\nSHOW PROCEDURE CODE proc_name\nSHOW PROCEDURE STATUS [like_or_where]\nSHOW PRIVILEGES\nSHOW [FULL] PROCESSLIST\nSHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n]\nSHOW PROFILES\nSHOW SCHEDULER STATUS\nSHOW [GLOBAL | SESSION] STATUS [like_or_where]\nSHOW TABLE STATUS [FROM db_name] [like_or_where]\nSHOW TABLES [FROM db_name] [like_or_where]\nSHOW TRIGGERS [FROM db_name] [like_or_where]\nSHOW [GLOBAL | SESSION] VARIABLES [like_or_where]\nSHOW WARNINGS [LIMIT [offset,] row_count]\n\nlike_or_where:\n LIKE \'pattern\'\n | WHERE expr\n\nIf the syntax for a given SHOW statement includes a LIKE \'pattern\'\npart, \'pattern\' is a string that can contain the SQL "%" and "_"\nwildcard characters. The pattern is useful for restricting statement\noutput to matching values.\n\nSeveral SHOW statements also accept a WHERE clause that provides more\nflexibility in specifying which rows to display. See\nhttp://dev.mysql.com/doc/refman/5.1/en/extended-show.html.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/show.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/show.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (344,25,'SHOW','SHOW has many forms that provide information about databases, tables,\ncolumns, or status information about the server. This section describes\nthose following:\n\nSHOW AUTHORS\nSHOW CHARACTER SET [like_or_where]\nSHOW COLLATION [like_or_where]\nSHOW [FULL] COLUMNS FROM tbl_name [FROM db_name] [like_or_where]\nSHOW CONTRIBUTORS\nSHOW CREATE DATABASE db_name\nSHOW CREATE EVENT event_name\nSHOW CREATE FUNCTION func_name\nSHOW CREATE PROCEDURE proc_name\nSHOW CREATE TABLE tbl_name\nSHOW CREATE TRIGGER trigger_name\nSHOW CREATE VIEW view_name\nSHOW DATABASES [like_or_where]\nSHOW ENGINE engine_name {STATUS | MUTEX}\nSHOW [STORAGE] ENGINES\nSHOW ERRORS [LIMIT [offset,] row_count]\nSHOW EVENTS\nSHOW FUNCTION CODE func_name\nSHOW FUNCTION STATUS [like_or_where]\nSHOW GRANTS FOR user\nSHOW INDEX FROM tbl_name [FROM db_name]\nSHOW INNODB STATUS\nSHOW OPEN TABLES [FROM db_name] [like_or_where]\nSHOW PLUGINS\nSHOW PROCEDURE CODE proc_name\nSHOW PROCEDURE STATUS [like_or_where]\nSHOW PRIVILEGES\nSHOW [FULL] PROCESSLIST\nSHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n]\nSHOW PROFILES\nSHOW SCHEDULER STATUS\nSHOW [GLOBAL | SESSION] STATUS [like_or_where]\nSHOW TABLE STATUS [FROM db_name] [like_or_where]\nSHOW TABLES [FROM db_name] [like_or_where]\nSHOW TRIGGERS [FROM db_name] [like_or_where]\nSHOW [GLOBAL | SESSION] VARIABLES [like_or_where]\nSHOW WARNINGS [LIMIT [offset,] row_count]\n\nlike_or_where:\n LIKE \'pattern\'\n | WHERE expr\n\nIf the syntax for a given SHOW statement includes a LIKE \'pattern\'\npart, \'pattern\' is a string that can contain the SQL "%" and "_"\nwildcard characters. The pattern is useful for restricting statement\noutput to matching values.\n\nSeveral SHOW statements also accept a WHERE clause that provides more\nflexibility in specifying which rows to display. See\nhttp://dev.mysql.com/doc/refman/5.1/en/extended-show.html.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/show.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/show.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (345,17,'GREATEST','Syntax:\nGREATEST(value1,value2,...)\n\nWith two or more arguments, returns the largest (maximum-valued)\nargument. The arguments are compared using the same rules as for\nLEAST().\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/comparison-operators.html\n\n','mysql> SELECT GREATEST(2,0);\n -> 2\nmysql> SELECT GREATEST(34.0,3.0,5.0,767.0);\n -> 767.0\nmysql> SELECT GREATEST(\'B\',\'A\',\'C\');\n -> \'C\'\n','http://dev.mysql.com/doc/refman/5.1/en/comparison-operators.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (346,25,'SHOW VARIABLES','Syntax:\nSHOW [GLOBAL | SESSION] VARIABLES\n [LIKE \'pattern\' | WHERE expr]\n\nSHOW VARIABLES shows the values of MySQL system variables. This\ninformation also can be obtained using the mysqladmin variables\ncommand. The LIKE clause, if present, indicates which variable names to\nmatch. The WHERE clause can be given to select rows using more general\nconditions, as discussed in\nhttp://dev.mysql.com/doc/refman/5.1/en/extended-show.html. This\nstatement does not require any privilege. It requires only the ability\nto connect to the server.\n\nWith the GLOBAL modifier, SHOW VARIABLES displays the values that are\nused for new connections to MySQL. With SESSION, it displays the values\nthat are in effect for the current connection. If no modifier is\npresent, the default is SESSION. LOCAL is a synonym for SESSION.\nWith a LIKE clause, the statement displays only rows for those\nvariables with names that match the pattern. To obtain the row for a\nspecific variable, use a LIKE clause as shown:\n\nSHOW VARIABLES LIKE \'max_join_size\';\nSHOW SESSION VARIABLES LIKE \'max_join_size\';\n\nTo get a list of variables whose name match a pattern, use the "%"\nwildcard character in a LIKE clause:\n\nSHOW VARIABLES LIKE \'%size%\';\nSHOW GLOBAL VARIABLES LIKE \'%size%\';\n\nWildcard characters can be used in any position within the pattern to\nbe matched. Strictly speaking, because "_" is a wildcard that matches\nany single character, you should escape it as "\\_" to match it\nliterally. In practice, this is rarely necessary.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/show-variables.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/show-variables.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (347,25,'BINLOG','Syntax:\nBINLOG \'str\'\n\nBINLOG is an internal-use statement. It is generated by the mysqlbinlog\nprogram as the printable representation of certain events in binary log\nfiles. (See http://dev.mysql.com/doc/refman/5.1/en/mysqlbinlog.html.)\nThe \'str\' value is a base 64-encoded string the that server decodes to\ndetermine the data change indicated by the corresponding event. This\nstatement requires the SUPER privilege. It was added in MySQL 5.1.5.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/binlog.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/binlog.html'); @@ -419,7 +419,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (350,4,'ATAN2','Syntax:\nATAN(Y,X), ATAN2(Y,X)\n\nReturns the arc tangent of the two variables X and Y. It is similar to\ncalculating the arc tangent of Y / X, except that the signs of both\narguments are used to determine the quadrant of the result.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/mathematical-functions.html\n\n','mysql> SELECT ATAN(-2,2);\n -> -0.78539816339745\nmysql> SELECT ATAN2(PI(),0);\n -> 1.5707963267949\n','http://dev.mysql.com/doc/refman/5.1/en/mathematical-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (351,6,'MBRCONTAINS','MBRContains(g1,g2)\n\nReturns 1 or 0 to indicate whether the Minimum Bounding Rectangle of g1\ncontains the Minimum Bounding Rectangle of g2. This tests the opposite\nrelationship as MBRWithin().\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/relations-on-geometry-mbr.html\n\n','mysql> SET @g1 = GeomFromText(\'Polygon((0 0,0 3,3 3,3 0,0 0))\');\nmysql> SET @g2 = GeomFromText(\'Point(1 1)\');\nmysql> SELECT MBRContains(@g1,@g2), MBRContains(@g2,@g1);\n----------------------+----------------------+\n| MBRContains(@g1,@g2) | MBRContains(@g2,@g1) |\n+----------------------+----------------------+\n| 1 | 0 |\n+----------------------+----------------------+\n','http://dev.mysql.com/doc/refman/5.1/en/relations-on-geometry-mbr.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (352,30,'HOUR','Syntax:\nHOUR(time)\n\nReturns the hour for time. The range of the return value is 0 to 23 for\ntime-of-day values. However, the range of TIME values actually is much\nlarger, so HOUR can return values greater than 23.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','mysql> SELECT HOUR(\'10:05:03\');\n -> 10\nmysql> SELECT HOUR(\'272:59:59\');\n -> 272\n','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (353,26,'SELECT','Syntax:\nSELECT\n [ALL | DISTINCT | DISTINCTROW ]\n [HIGH_PRIORITY]\n [STRAIGHT_JOIN]\n [SQL_SMALL_RESULT] [SQL_BIG_RESULT] [SQL_BUFFER_RESULT]\n [SQL_CACHE | SQL_NO_CACHE] [SQL_CALC_FOUND_ROWS]\n select_expr [, select_expr ...]\n [FROM table_references\n [WHERE where_condition]\n [GROUP BY {col_name | expr | position}\n [ASC | DESC], ... [WITH ROLLUP]]\n [HAVING where_condition]\n [ORDER BY {col_name | expr | position}\n [ASC | DESC], ...]\n [LIMIT {[offset,] row_count | row_count OFFSET offset}]\n [PROCEDURE procedure_name(argument_list)]\n [INTO OUTFILE \'file_name\' export_options\n | INTO DUMPFILE \'file_name\'\n | INTO var_name [, var_name]]\n [FOR UPDATE | LOCK IN SHARE MODE]]\n\nSELECT is used to retrieve rows selected from one or more tables, and\ncan include UNION statements and subqueries. See [HELP UNION], and\nhttp://dev.mysql.com/doc/refman/5.1/en/subqueries.html.\n\nThe most commonly used clauses of SELECT statements are these:\n\no Each select_expr indicates a column that you want to retrieve. There\n must be at least one select_expr.\n\no table_references indicates the table or tables from which to retrieve\n rows. Its syntax is described in [HELP JOIN].\n\no The WHERE clause, if given, indicates the condition or conditions\n that rows must satisfy to be selected. where_condition is an\n expression that evaluates to true for each row to be selected. The\n statement selects all rows if there is no WHERE clause.\n\n In the WHERE clause, you can use any of the functions and operators\n that MySQL supports, except for aggregate (summary) functions. See\n http://dev.mysql.com/doc/refman/5.1/en/functions.html.\n\nSELECT can also be used to retrieve rows computed without reference to\nany table.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/select.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/select.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (353,26,'SELECT','Syntax:\nSELECT\n [ALL | DISTINCT | DISTINCTROW ]\n [HIGH_PRIORITY]\n [STRAIGHT_JOIN]\n [SQL_SMALL_RESULT] [SQL_BIG_RESULT] [SQL_BUFFER_RESULT]\n [SQL_CACHE | SQL_NO_CACHE] [SQL_CALC_FOUND_ROWS]\n select_expr [, select_expr ...]\n [FROM table_references\n [WHERE where_condition]\n [GROUP BY {col_name | expr | position}\n [ASC | DESC], ... [WITH ROLLUP]]\n [HAVING where_condition]\n [ORDER BY {col_name | expr | position}\n [ASC | DESC], ...]\n [LIMIT {[offset,] row_count | row_count OFFSET offset}]\n [PROCEDURE procedure_name(argument_list)]\n [INTO OUTFILE \'file_name\'\n [CHARACTER SET charset_name]\n export_options\n | INTO DUMPFILE \'file_name\'\n | INTO var_name [, var_name]]\n [FOR UPDATE | LOCK IN SHARE MODE]]\n\nSELECT is used to retrieve rows selected from one or more tables, and\ncan include UNION statements and subqueries. See [HELP UNION], and\nhttp://dev.mysql.com/doc/refman/5.1/en/subqueries.html.\n\nThe most commonly used clauses of SELECT statements are these:\n\no Each select_expr indicates a column that you want to retrieve. There\n must be at least one select_expr.\n\no table_references indicates the table or tables from which to retrieve\n rows. Its syntax is described in [HELP JOIN].\n\no The WHERE clause, if given, indicates the condition or conditions\n that rows must satisfy to be selected. where_condition is an\n expression that evaluates to true for each row to be selected. The\n statement selects all rows if there is no WHERE clause.\n\n In the WHERE clause, you can use any of the functions and operators\n that MySQL supports, except for aggregate (summary) functions. See\n http://dev.mysql.com/doc/refman/5.1/en/functions.html.\n\nSELECT can also be used to retrieve rows computed without reference to\nany table.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/select.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/select.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (354,4,'COT','Syntax:\nCOT(X)\n\nReturns the cotangent of X.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/mathematical-functions.html\n\n','mysql> SELECT COT(12);\n -> -1.5726734063977\nmysql> SELECT COT(0);\n -> NULL\n','http://dev.mysql.com/doc/refman/5.1/en/mathematical-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (355,25,'SHOW CREATE EVENT','Syntax:\nSHOW CREATE EVENT event_name\n\nThis statement displays the CREATE EVENT statement needed to re-create\na given event. For example (using the same event e_daily defined and\nthen altered in [HELP SHOW EVENTS]):\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/show-create-event.html\n\n','mysql> SHOW CREATE EVENT test.e_daily\\G\n*************************** 1. row ***************************\n Event: e_daily\n sql_mode:\n time_zone: SYSTEM\n Create Event: CREATE EVENT `e_daily`\n ON SCHEDULE EVERY 1 DAY\n STARTS CURRENT_TIMESTAMP + INTERVAL 6 HOUR\n ON COMPLETION NOT PRESERVE\n ENABLE\n COMMENT \'Saves total number of sessions then\n clears the table each day\'\n DO BEGIN\n INSERT INTO site_activity.totals (time, total)\n SELECT CURRENT_TIMESTAMP, COUNT(*)\n FROM site_activity.sessions;\n DELETE FROM site_activity.sessions;\n END\ncharacter_set_client: latin1\ncollation_connection: latin1_swedish_ci\n Database Collation: latin1_swedish_ci\n','http://dev.mysql.com/doc/refman/5.1/en/show-create-event.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (356,19,'BACKUP TABLE','Syntax:\nBACKUP TABLE tbl_name [, tbl_name] ... TO \'/path/to/backup/directory\'\n\n*Note*: This statement is deprecated. We are working on a better\nreplacement for it that will provide online backup capabilities. In the\nmeantime, the mysqlhotcopy script can be used instead.\n\nBACKUP TABLE copies to the backup directory the minimum number of table\nfiles needed to restore the table, after flushing any buffered changes\nto disk. The statement works only for MyISAM tables. It copies the .frm\ndefinition and .MYD data files. The .MYI index file can be rebuilt from\nthose two files. The directory should be specified as a full path name.\nTo restore the table, use RESTORE TABLE.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/backup-table.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/backup-table.html'); @@ -486,7 +486,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (417,35,'BIN','Syntax:\nBIN(N)\n\nReturns a string representation of the binary value of N, where N is a\nlonglong (BIGINT) number. This is equivalent to CONV(N,10,2). Returns\nNULL if N is NULL.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/string-functions.html\n\n','mysql> SELECT BIN(12);\n -> \'1100\'\n','http://dev.mysql.com/doc/refman/5.1/en/string-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (418,5,'INSTALL PLUGIN','Syntax:\nINSTALL PLUGIN plugin_name SONAME \'plugin_library\'\n\nThis statement installs a plugin.\n\nplugin_name is the name of the plugin as defined in the plugin\ndeclaration structure contained in the library file. Plugin names are\nnot case sensitive. For maximal compatibility, plugin names should be\nlimited to ASCII letters, digits, and underscore, because they are used\nin C source files, shell command lines, M4 and Bourne shell scripts,\nand SQL environments.\n\nplugin_library is the name of the shared library that contains the\nplugin code. The name includes the file name extension (for example,\nlibmyplugin.so or libmyplugin.dylib).\n\nThe shared library must be located in the plugin directory (that is,\nthe directory named by the plugin_dir system variable). The library\nmust be in the plugin directory itself, not in a subdirectory. By\ndefault, plugin_dir is plugin directory under the directory named by\nthe pkglibdir configuration variable, but it can be changed by setting\nthe value of plugin_dir at server startup. For example, set its value\nin a my.cnf file:\n\n[mysqld]\nplugin_dir=/path/to/plugin/directory\n\nIf the value of plugin_dir is a relative path name, it is taken to be\nrelative to the MySQL base directory (the value of the basedir system\nvariable).\n\nINSTALL PLUGIN adds a line to the mysql.plugin table that describes the\nplugin. This table contains the plugin name and library file name.\n\nAs of MySQL 5.1.33, INSTALL PLUGIN causes the server to read option\n(my.cnf) files just as during server startup. This enables the plugin\nto pick up any relevant options from those files. It is possible to add\nplugin options to an option file even before loading a plugin (if the\nloose prefix is used). It is also possible to uninstall a plugin, edit\nmy.cnf, and install the plugin again. Restarting the plugin this way\nenables it to the new option values without a server restart.\n\nBefore MySQL 5.1.33, a plugin is started with each option set to its\ndefault value.\n\nINSTALL PLUGIN also loads and initializes the plugin code to make the\nplugin available for use. A plugin is initialized by executing its\ninitialization function, which handles any setup that the plugin must\nperform before it can be used.\n\nTo use INSTALL PLUGIN, you must have the INSERT privilege for the\nmysql.plugin table.\n\nAt server startup, the server loads and initializes any plugin that is\nlisted in the mysql.plugin table. This means that a plugin is installed\nwith INSTALL PLUGIN only once, not every time the server starts. Plugin\nloading at startup does not occur if the server is started with the\n--skip-grant-tables option.\n\nWhen the server shuts down, it executes the deinitialization function\nfor each plugin that is loaded so that the plugin has a change to\nperform any final cleanup.\n\nFor options that control individual plugin loading at server startup,\nsee http://dev.mysql.com/doc/refman/5.1/en/server-plugin-options.html.\nIf you need to load plugins for a single server startup when the\n--skip-grant-tables option is given (which tells the server not to read\nsystem tables), use the --plugin-load option. See\nhttp://dev.mysql.com/doc/refman/5.1/en/server-options.html.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/install-plugin.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/install-plugin.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (419,22,'DECLARE CURSOR','Syntax:\nDECLARE cursor_name CURSOR FOR select_statement\n\nThis statement declares a cursor. Multiple cursors may be declared in a\nstored program, but each cursor in a given block must have a unique\nname.\n\nThe SELECT statement cannot have an INTO clause.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/declare-cursor.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/declare-cursor.html'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (420,26,'LOAD DATA','Syntax:\nLOAD DATA [LOW_PRIORITY | CONCURRENT] [LOCAL] INFILE \'file_name\'\n [REPLACE | IGNORE]\n INTO TABLE tbl_name\n [CHARACTER SET charset_name]\n [{FIELDS | COLUMNS}\n [TERMINATED BY \'string\']\n [[OPTIONALLY] ENCLOSED BY \'char\']\n [ESCAPED BY \'char\']\n ]\n [LINES\n [STARTING BY \'string\']\n [TERMINATED BY \'string\']\n ]\n [IGNORE number LINES]\n [(col_name_or_user_var,...)]\n [SET col_name = expr,...]\n\nThe LOAD DATA INFILE statement reads rows from a text file into a table\nat a very high speed. The file name must be given as a literal string.\n\nLOAD DATA INFILE is the complement of SELECT ... INTO OUTFILE. (See\nhttp://dev.mysql.com/doc/refman/5.1/en/select.html.) To write data from\na table to a file, use SELECT ... INTO OUTFILE. To read the file back\ninto a table, use LOAD DATA INFILE. The syntax of the FIELDS and LINES\nclauses is the same for both statements. Both clauses are optional, but\nFIELDS must precede LINES if both are specified.\n\nFor more information about the efficiency of INSERT versus LOAD DATA\nINFILE and speeding up LOAD DATA INFILE, see\nhttp://dev.mysql.com/doc/refman/5.1/en/insert-speed.html.\n\nThe character set indicated by the character_set_database system\nvariable is used to interpret the information in the file. SET NAMES\nand the setting of character_set_client do not affect interpretation of\ninput. If the contents of the input file use a character set that\ndiffers from the default, it is usually preferable to specify the\ncharacter set of the file by using the CHARACTER SET clause, which is\navailable as of MySQL 5.1.17. A character set of binary specifies "no\nconversion."\n\nLOAD DATA INFILE interprets all fields in the file as having the same\ncharacter set, regardless of the data types of the columns into which\nfield values are loaded. For proper interpretation of file contents,\nyou must ensure that it was written with the correct character set. For\nexample, if you write a data file with mysqldump -T or by issuing a\nSELECT ... INTO OUTFILE statement in mysql, be sure to use a\n--default-character-set option with mysqldump or mysql so that output\nis written in the character set to be used when the file is loaded with\nLOAD DATA INFILE.\n\nNote that it is currently not possible to load data files that use the\nucs2, utf16, or utf32 character set.\n\nAs of MySQL 5.1.6, the character_set_filesystem system variable\ncontrols the interpretation of the file name.\n\nYou can also load data files by using the mysqlimport utility; it\noperates by sending a LOAD DATA INFILE statement to the server. The\n--local option causes mysqlimport to read data files from the client\nhost. You can specify the --compress option to get better performance\nover slow networks if the client and server support the compressed\nprotocol. See http://dev.mysql.com/doc/refman/5.1/en/mysqlimport.html.\n\nIf you use LOW_PRIORITY, execution of the LOAD DATA statement is\ndelayed until no other clients are reading from the table. This affects\nonly storage engines that use only table-level locking (MyISAM, MEMORY,\nMERGE).\n\nIf you specify CONCURRENT with a MyISAM table that satisfies the\ncondition for concurrent inserts (that is, it contains no free blocks\nin the middle), other threads can retrieve data from the table while\nLOAD DATA is executing. Using this option affects the performance of\nLOAD DATA a bit, even if no other thread is using the table at the same\ntime.\n\nCONCURRENT is not replicated when using statement-based replication;\nhowever, it is replicated when using row-based replication. See\nhttp://dev.mysql.com/doc/refman/5.1/en/replication-features-load-data.…, for more information.\n\n*Note*: Prior to MySQL 5.1.23, LOAD DATA performed very poorly when\nimporting into partitioned tables. The statement now uses buffering to\nimprove performance; however, the buffer uses 130 KB memory per\npartition to achieve this. (Bug#26527 (http://bugs.mysql.com/26527))\n\nThe LOCAL keyword, if specified, is interpreted with respect to the\nclient end of the connection:\n\no If LOCAL is specified, the file is read by the client program on the\n client host and sent to the server. The file can be given as a full\n path name to specify its exact location. If given as a relative path\n name, the name is interpreted relative to the directory in which the\n client program was started.\n\no If LOCAL is not specified, the file must be located on the server\n host and is read directly by the server. The server uses the\n following rules to locate the file:\n\n o If the file name is an absolute path name, the server uses it as\n given.\n\n o If the file name is a relative path name with one or more leading\n components, the server searches for the file relative to the\n server\'s data directory.\n\n o If a file name with no leading components is given, the server\n looks for the file in the database directory of the default\n database.\n\nNote that, in the non-LOCAL case, these rules mean that a file named as\n./myfile.txt is read from the server\'s data directory, whereas the file\nnamed as myfile.txt is read from the database directory of the default\ndatabase. For example, if db1 is the default database, the following\nLOAD DATA statement reads the file data.txt from the database directory\nfor db1, even though the statement explicitly loads the file into a\ntable in the db2 database:\n\nLOAD DATA INFILE \'data.txt\' INTO TABLE db2.my_table;\n\nWindows path names are specified using forward slashes rather than\nbackslashes. If you do use backslashes, you must double them.\n\nFor security reasons, when reading text files located on the server,\nthe files must either reside in the database directory or be readable\nby all. Also, to use LOAD DATA INFILE on server files, you must have\nthe FILE privilege. See\nhttp://dev.mysql.com/doc/refman/5.1/en/privileges-provided.html. For\nnon-LOCAL load operations, if the secure_file_priv system variable is\nset to a nonempty directory name, the file to be loaded must be located\nin that directory.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/load-data.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/load-data.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (420,26,'LOAD DATA','Syntax:\nLOAD DATA [LOW_PRIORITY | CONCURRENT] [LOCAL] INFILE \'file_name\'\n [REPLACE | IGNORE]\n INTO TABLE tbl_name\n [CHARACTER SET charset_name]\n [{FIELDS | COLUMNS}\n [TERMINATED BY \'string\']\n [[OPTIONALLY] ENCLOSED BY \'char\']\n [ESCAPED BY \'char\']\n ]\n [LINES\n [STARTING BY \'string\']\n [TERMINATED BY \'string\']\n ]\n [IGNORE number LINES]\n [(col_name_or_user_var,...)]\n [SET col_name = expr,...]\n\nThe LOAD DATA INFILE statement reads rows from a text file into a table\nat a very high speed. The file name must be given as a literal string.\n\nLOAD DATA INFILE is the complement of SELECT ... INTO OUTFILE. (See\nhttp://dev.mysql.com/doc/refman/5.1/en/select.html.) To write data from\na table to a file, use SELECT ... INTO OUTFILE. To read the file back\ninto a table, use LOAD DATA INFILE. The syntax of the FIELDS and LINES\nclauses is the same for both statements. Both clauses are optional, but\nFIELDS must precede LINES if both are specified.\n\nFor more information about the efficiency of INSERT versus LOAD DATA\nINFILE and speeding up LOAD DATA INFILE, see\nhttp://dev.mysql.com/doc/refman/5.1/en/insert-speed.html.\n\nThe character set indicated by the character_set_database system\nvariable is used to interpret the information in the file. SET NAMES\nand the setting of character_set_client do not affect interpretation of\ninput. If the contents of the input file use a character set that\ndiffers from the default, it is usually preferable to specify the\ncharacter set of the file by using the CHARACTER SET clause, which is\navailable as of MySQL 5.1.17. A character set of binary specifies "no\nconversion."\n\nLOAD DATA INFILE interprets all fields in the file as having the same\ncharacter set, regardless of the data types of the columns into which\nfield values are loaded. For proper interpretation of file contents,\nyou must ensure that it was written with the correct character set. For\nexample, if you write a data file with mysqldump -T or by issuing a\nSELECT ... INTO OUTFILE statement in mysql, be sure to use a\n--default-character-set option with mysqldump or mysql so that output\nis written in the character set to be used when the file is loaded with\nLOAD DATA INFILE.\n\nNote that it is currently not possible to load data files that use the\nucs2, utf16, or utf32 character set.\n\nAs of MySQL 5.1.6, the character_set_filesystem system variable\ncontrols the interpretation of the file name.\n\nYou can also load data files by using the mysqlimport utility; it\noperates by sending a LOAD DATA INFILE statement to the server. The\n--local option causes mysqlimport to read data files from the client\nhost. You can specify the --compress option to get better performance\nover slow networks if the client and server support the compressed\nprotocol. See http://dev.mysql.com/doc/refman/5.1/en/mysqlimport.html.\n\nIf you use LOW_PRIORITY, execution of the LOAD DATA statement is\ndelayed until no other clients are reading from the table. This affects\nonly storage engines that use only table-level locking (MyISAM, MEMORY,\nMERGE).\n\nIf you specify CONCURRENT with a MyISAM table that satisfies the\ncondition for concurrent inserts (that is, it contains no free blocks\nin the middle), other threads can retrieve data from the table while\nLOAD DATA is executing. Using this option affects the performance of\nLOAD DATA a bit, even if no other thread is using the table at the same\ntime.\n\nCONCURRENT is not replicated when using statement-based replication;\nhowever, it is replicated when using row-based replication. See\nhttp://dev.mysql.com/doc/refman/5.1/en/replication-features-load-data.…, for more information.\n\n*Note*: Prior to MySQL 5.1.23, LOAD DATA performed very poorly when\nimporting into partitioned tables. The statement now uses buffering to\nimprove performance; however, the buffer uses 130 KB memory per\npartition to achieve this. (Bug#26527 (http://bugs.mysql.com/26527))\n\nThe LOCAL keyword, if specified, is interpreted with respect to the\nclient end of the connection:\n\no If LOCAL is specified, the file is read by the client program on the\n client host and sent to the server. The file can be given as a full\n path name to specify its exact location. If given as a relative path\n name, the name is interpreted relative to the directory in which the\n client program was started.\n\no If LOCAL is not specified, the file must be located on the server\n host and is read directly by the server. The server uses the\n following rules to locate the file:\n\n o If the file name is an absolute path name, the server uses it as\n given.\n\n o If the file name is a relative path name with one or more leading\n components, the server searches for the file relative to the\n server\'s data directory.\n\n o If a file name with no leading components is given, the server\n looks for the file in the database directory of the default\n database.\n\nNote that, in the non-LOCAL case, these rules mean that a file named as\n./myfile.txt is read from the server\'s data directory, whereas the file\nnamed as myfile.txt is read from the database directory of the default\ndatabase. For example, if db1 is the default database, the following\nLOAD DATA statement reads the file data.txt from the database directory\nfor db1, even though the statement explicitly loads the file into a\ntable in the db2 database:\n\nLOAD DATA INFILE \'data.txt\' INTO TABLE db2.my_table;\n\nWindows path names are specified using forward slashes rather than\nbackslashes. If you do use backslashes, you must double them.\n\n*Note*: A regression in MySQL 5.1.40 caused the database referenced in\na fully qualified table name to be ignored by LOAD DATA when using\nreplication with either STATEMENT or MIXED as the binary logging\nformat; this could lead to problems if the table was not in the current\ndatabase. As a workaround, you can specify the correct database with\nthe USE statement prior to executing LOAD DATA. If necessary, you can\nreset the current database with a second USE statement following the\nLOAD DATA statement. This issue was fixed in MySQL 5.1.41. (Bug#48297\n(http://bugs.mysql.com/48297))\n\nFor security reasons, when reading text files located on the server,\nthe files must either reside in the database directory or be readable\nby all. Also, to use LOAD DATA INFILE on server files, you must have\nthe FILE privilege. See\nhttp://dev.mysql.com/doc/refman/5.1/en/privileges-provided.html. For\nnon-LOCAL load operations, if the secure_file_priv system variable is\nset to a nonempty directory name, the file to be loaded must be located\nin that directory.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/load-data.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/load-data.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (421,23,'MULTILINESTRING','MultiLineString(ls1,ls2,...)\n\nConstructs a MultiLineString value using LineString or WKB LineString\narguments.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/creating-spatial-values.html#gis-mys…','','http://dev.mysql.com/doc/refman/5.1/en/creating-spatial-values.html#gis-mys…'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (422,30,'LOCALTIME','Syntax:\nLOCALTIME, LOCALTIME()\n\nLOCALTIME and LOCALTIME() are synonyms for NOW().\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (423,3,'MPOINTFROMTEXT','MPointFromText(wkt[,srid]), MultiPointFromText(wkt[,srid])\n\nConstructs a MULTIPOINT value using its WKT representation and SRID.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/creating-spatial-values.html#gis-wkt…','','http://dev.mysql.com/doc/refman/5.1/en/creating-spatial-values.html#gis-wkt…'); @@ -497,7 +497,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (428,20,'CHAR','[NATIONAL] CHAR[(M)] [CHARACTER SET charset_name] [COLLATE\ncollation_name]\n\nA fixed-length string that is always right-padded with spaces to the\nspecified length when stored. M represents the column length in\ncharacters. The range of M is 0 to 255. If M is omitted, the length is\n1.\n\n*Note*: Trailing spaces are removed when CHAR values are retrieved\nunless the PAD_CHAR_TO_FULL_LENGTH SQL mode is enabled.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/string-type-overview.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/string-type-overview.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (429,30,'UTC_DATE','Syntax:\nUTC_DATE, UTC_DATE()\n\nReturns the current UTC date as a value in \'YYYY-MM-DD\' or YYYYMMDD\nformat, depending on whether the function is used in a string or\nnumeric context.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','mysql> SELECT UTC_DATE(), UTC_DATE() + 0;\n -> \'2003-08-14\', 20030814\n','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (430,34,'DIMENSION','Dimension(g)\n\nReturns the inherent dimension of the geometry value g. The result can\nbe -1, 0, 1, or 2. The meaning of these values is given in\nhttp://dev.mysql.com/doc/refman/5.1/en/gis-class-geometry.html.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/geometry-property-functions.html#gen…','mysql> SELECT Dimension(GeomFromText(\'LineString(1 1,2 2)\'));\n+------------------------------------------------+\n| Dimension(GeomFromText(\'LineString(1 1,2 2)\')) |\n+------------------------------------------------+\n| 1 |\n+------------------------------------------------+\n','http://dev.mysql.com/doc/refman/5.1/en/geometry-property-functions.html#gen…'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (431,16,'COUNT DISTINCT','Syntax:\nCOUNT(DISTINCT expr,[expr...])\n\nReturns a count of the number of different non-NULL values.\n\nCOUNT(DISTINCT) returns 0 if there were no matching rows.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/group-by-functions.html\n\n','mysql> SELECT COUNT(DISTINCT results) FROM student;\n','http://dev.mysql.com/doc/refman/5.1/en/group-by-functions.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (431,16,'COUNT DISTINCT','Syntax:\nCOUNT(DISTINCT expr,[expr...])\n\nReturns a count of the number of rows with different non-NULL expr\nvalues.\n\nCOUNT(DISTINCT) returns 0 if there were no matching rows.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/group-by-functions.html\n\n','mysql> SELECT COUNT(DISTINCT results) FROM student;\n','http://dev.mysql.com/doc/refman/5.1/en/group-by-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (432,20,'BIT','BIT[(M)]\n\nA bit-field type. M indicates the number of bits per value, from 1 to\n64. The default is 1 if M is omitted.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/numeric-type-overview.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/numeric-type-overview.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (433,29,'EQUALS','Equals(g1,g2)\n\nReturns 1 or 0 to indicate whether g1 is spatially equal to g2.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/functions-that-test-spatial-relation…','','http://dev.mysql.com/doc/refman/5.1/en/functions-that-test-spatial-relation…'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (434,25,'SHOW CREATE VIEW','Syntax:\nSHOW CREATE VIEW view_name\n\nThis statement shows a CREATE VIEW statement that creates the given\nview.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/show-create-view.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/show-create-view.html'); @@ -532,7 +532,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (463,37,'MERGE','The MERGE storage engine, also known as the MRG_MyISAM engine, is a\ncollection of identical MyISAM tables that can be used as one.\n"Identical" means that all tables have identical column and index\ninformation. You cannot merge MyISAM tables in which the columns are\nlisted in a different order, do not have exactly the same columns, or\nhave the indexes in different order. However, any or all of the MyISAM\ntables can be compressed with myisampack. See\nhttp://dev.mysql.com/doc/refman/5.1/en/myisampack.html. Differences in\ntable options such as AVG_ROW_LENGTH, MAX_ROWS, or PACK_KEYS do not\nmatter.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/merge-storage-engine.html\n\n','mysql> CREATE TABLE t1 (\n -> a INT NOT NULL AUTO_INCREMENT PRIMARY KEY,\n -> message CHAR(20)) ENGINE=MyISAM;\nmysql> CREATE TABLE t2 (\n -> a INT NOT NULL AUTO_INCREMENT PRIMARY KEY,\n -> message CHAR(20)) ENGINE=MyISAM;\nmysql> INSERT INTO t1 (message) VALUES (\'Testing\'),(\'table\'),(\'t1\');\nmysql> INSERT INTO t2 (message) VALUES (\'Testing\'),(\'table\'),(\'t2\');\nmysql> CREATE TABLE total (\n -> a INT NOT NULL AUTO_INCREMENT,\n -> message CHAR(20), INDEX(a))\n -> ENGINE=MERGE UNION=(t1,t2) INSERT_METHOD=LAST;\n','http://dev.mysql.com/doc/refman/5.1/en/merge-storage-engine.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (464,37,'CREATE TABLE','Syntax:\nCREATE [TEMPORARY] TABLE [IF NOT EXISTS] tbl_name\n (create_definition,...)\n [table_options]\n [partition_options]\n\nOr:\n\nCREATE [TEMPORARY] TABLE [IF NOT EXISTS] tbl_name\n [(create_definition,...)]\n [table_options]\n [partition_options]\n select_statement\n\nOr:\n\nCREATE [TEMPORARY] TABLE [IF NOT EXISTS] tbl_name\n { LIKE old_tbl_name | (LIKE old_tbl_name) }\n\ncreate_definition:\n col_name column_definition\n | [CONSTRAINT [symbol]] PRIMARY KEY [index_type] (index_col_name,...)\n [index_option] ...\n | {INDEX|KEY} [index_name] [index_type] (index_col_name,...)\n [index_option] ...\n | [CONSTRAINT [symbol]] UNIQUE [INDEX|KEY]\n [index_name] [index_type] (index_col_name,...)\n [index_option] ...\n | {FULLTEXT|SPATIAL} [INDEX|KEY] [index_name] (index_col_name,...)\n [index_option] ...\n | [CONSTRAINT [symbol]] FOREIGN KEY\n [index_name] (index_col_name,...) reference_definition\n | CHECK (expr)\n\ncolumn_definition:\n data_type [NOT NULL | NULL] [DEFAULT default_value]\n [AUTO_INCREMENT] [UNIQUE [KEY] | [PRIMARY] KEY]\n [COMMENT \'string\']\n [COLUMN_FORMAT {FIXED|DYNAMIC|DEFAULT}]\n [STORAGE {DISK|MEMORY|DEFAULT}]\n [reference_definition]\n\ndata_type:\n BIT[(length)]\n | TINYINT[(length)] [UNSIGNED] [ZEROFILL]\n | SMALLINT[(length)] [UNSIGNED] [ZEROFILL]\n | MEDIUMINT[(length)] [UNSIGNED] [ZEROFILL]\n | INT[(length)] [UNSIGNED] [ZEROFILL]\n | INTEGER[(length)] [UNSIGNED] [ZEROFILL]\n | BIGINT[(length)] [UNSIGNED] [ZEROFILL]\n | REAL[(length,decimals)] [UNSIGNED] [ZEROFILL]\n | DOUBLE[(length,decimals)] [UNSIGNED] [ZEROFILL]\n | FLOAT[(length,decimals)] [UNSIGNED] [ZEROFILL]\n | DECIMAL[(length[,decimals])] [UNSIGNED] [ZEROFILL]\n | NUMERIC[(length[,decimals])] [UNSIGNED] [ZEROFILL]\n | DATE\n | TIME\n | TIMESTAMP\n | DATETIME\n | YEAR\n | CHAR[(length)]\n [CHARACTER SET charset_name] [COLLATE collation_name]\n | VARCHAR(length)\n [CHARACTER SET charset_name] [COLLATE collation_name]\n | BINARY[(length)]\n | VARBINARY(length)\n | TINYBLOB\n | BLOB\n | MEDIUMBLOB\n | LONGBLOB\n | TINYTEXT [BINARY]\n [CHARACTER SET charset_name] [COLLATE collation_name]\n | TEXT [BINARY]\n [CHARACTER SET charset_name] [COLLATE collation_name]\n | MEDIUMTEXT [BINARY]\n [CHARACTER SET charset_name] [COLLATE collation_name]\n | LONGTEXT [BINARY]\n [CHARACTER SET charset_name] [COLLATE collation_name]\n | ENUM(value1,value2,value3,...)\n [CHARACTER SET charset_name] [COLLATE collation_name]\n | SET(value1,value2,value3,...)\n [CHARACTER SET charset_name] [COLLATE collation_name]\n | spatial_type\n\nindex_col_name:\n col_name [(length)] [ASC | DESC]\n\nindex_type:\n USING {BTREE | HASH | RTREE}\n\nindex_option:\n KEY_BLOCK_SIZE [=] value\n | index_type\n | WITH PARSER parser_name\n\nreference_definition:\n REFERENCES tbl_name (index_col_name,...)\n [MATCH FULL | MATCH PARTIAL | MATCH SIMPLE]\n [ON DELETE reference_option]\n [ON UPDATE reference_option]\n\nreference_option:\n RESTRICT | CASCADE | SET NULL | NO ACTION\n\ntable_options:\n table_option [[,] table_option] ...\n\ntable_option:\n ENGINE [=] engine_name\n | AUTO_INCREMENT [=] value\n | AVG_ROW_LENGTH [=] value\n | [DEFAULT] CHARACTER SET [=] charset_name\n | CHECKSUM [=] {0 | 1}\n | [DEFAULT] COLLATE [=] collation_name\n | COMMENT [=] \'string\'\n | CONNECTION [=] \'connect_string\'\n | DATA DIRECTORY [=] \'absolute path to directory\'\n | DELAY_KEY_WRITE [=] {0 | 1}\n | INDEX DIRECTORY [=] \'absolute path to directory\'\n | INSERT_METHOD [=] { NO | FIRST | LAST }\n | KEY_BLOCK_SIZE [=] value\n | MAX_ROWS [=] value\n | MIN_ROWS [=] value\n | PACK_KEYS [=] {0 | 1 | DEFAULT}\n | PASSWORD [=] \'string\'\n | ROW_FORMAT [=] {DEFAULT|DYNAMIC|FIXED|COMPRESSED|REDUNDANT|COMPACT}\n | TABLESPACE tablespace_name [STORAGE {DISK|MEMORY|DEFAULT}]\n | UNION [=] (tbl_name[,tbl_name]...)\n\npartition_options:\n PARTITION BY\n { [LINEAR] HASH(expr)\n | [LINEAR] KEY(column_list)\n | RANGE(expr)\n | LIST(expr) }\n [PARTITIONS num]\n [SUBPARTITION BY\n { [LINEAR] HASH(expr)\n | [LINEAR] KEY(column_list) }\n [SUBPARTITIONS num]\n ]\n [(partition_definition [, partition_definition] ...)]\n\npartition_definition:\n PARTITION partition_name\n [VALUES {LESS THAN {(expr) | MAXVALUE} | IN (value_list)}]\n [[STORAGE] ENGINE [=] engine_name]\n [COMMENT [=] \'comment_text\' ]\n [DATA DIRECTORY [=] \'data_dir\']\n [INDEX DIRECTORY [=] \'index_dir\']\n [MAX_ROWS [=] max_number_of_rows]\n [MIN_ROWS [=] min_number_of_rows]\n [TABLESPACE [=] tablespace_name]\n [NODEGROUP [=] node_group_id]\n [(subpartition_definition [, subpartition_definition] ...)]\n\nsubpartition_definition:\n SUBPARTITION logical_name\n [[STORAGE] ENGINE [=] engine_name]\n [COMMENT [=] \'comment_text\' ]\n [DATA DIRECTORY [=] \'data_dir\']\n [INDEX DIRECTORY [=] \'index_dir\']\n [MAX_ROWS [=] max_number_of_rows]\n [MIN_ROWS [=] min_number_of_rows]\n [TABLESPACE [=] tablespace_name]\n [NODEGROUP [=] node_group_id]\n\nselect_statement:\n [IGNORE | REPLACE] [AS] SELECT ... (Some legal select statement)\n\nCREATE TABLE creates a table with the given name. You must have the\nCREATE privilege for the table.\n\nRules for allowable table names are given in\nhttp://dev.mysql.com/doc/refman/5.1/en/identifiers.html. By default,\nthe table is created in the default database. An error occurs if the\ntable exists, if there is no default database, or if the database does\nnot exist.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/create-table.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/create-table.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (465,17,'>','Syntax:\n>\n\nGreater than:\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/comparison-operators.html\n\n','mysql> SELECT 2 > 2;\n -> 0\n','http://dev.mysql.com/doc/refman/5.1/en/comparison-operators.html'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (466,19,'ANALYZE TABLE','Syntax:\nANALYZE [NO_WRITE_TO_BINLOG | LOCAL] TABLE\n tbl_name [, tbl_name] ...\n\nANALYZE TABLE analyzes and stores the key distribution for a table.\nDuring the analysis, the table is locked with a read lock for MyISAM.\nFor InnoDB the table is locked with a write lock. This statement works\nwith MyISAM, and InnoDB tables. For MyISAM tables, this statement is\nequivalent to using myisamchk --analyze.\n\nFor more information on how the analysis works within InnoDB, see\nhttp://dev.mysql.com/doc/refman/5.1/en/innodb-restrictions.html.\n\nMy… uses the stored key distribution to decide the order in which\ntables should be joined when you perform a join on something other than\na constant. In addition, key distributions can be used when deciding\nwhich indexes to use for a specific table within a query.\n\nThis statement requires SELECT and INSERT privileges for the table.\n\nBeginning with MySQL 5.1.27, ANALYZE TABLE is also supported for\npartitioned tables. Also beginning with MySQL 5.1.27, you can use ALTER\nTABLE ... ANALYZE PARTITION to analyze one or more partitions; for more\ninformation, see [HELP ALTER TABLE], and\nhttp://dev.mysql.com/doc/refman/5.1/en/partitioning-maintenance.html.\…: http://dev.mysql.com/doc/refman/5.1/en/analyze-table.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/analyze-table.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (466,19,'ANALYZE TABLE','Syntax:\nANALYZE [NO_WRITE_TO_BINLOG | LOCAL] TABLE\n tbl_name [, tbl_name] ...\n\nANALYZE TABLE analyzes and stores the key distribution for a table.\nDuring the analysis, the table is locked with a read lock for MyISAM.\nFor InnoDB the table is locked with a write lock. This statement works\nwith MyISAM and InnoDB tables. For MyISAM tables, this statement is\nequivalent to using myisamchk --analyze.\n\nFor more information on how the analysis works within InnoDB, see\nhttp://dev.mysql.com/doc/refman/5.1/en/innodb-restrictions.html.\n\nMy… uses the stored key distribution to decide the order in which\ntables should be joined when you perform a join on something other than\na constant. In addition, key distributions can be used when deciding\nwhich indexes to use for a specific table within a query.\n\nThis statement requires SELECT and INSERT privileges for the table.\n\nBeginning with MySQL 5.1.27, ANALYZE TABLE is also supported for\npartitioned tables. Also beginning with MySQL 5.1.27, you can use ALTER\nTABLE ... ANALYZE PARTITION to analyze one or more partitions; for more\ninformation, see [HELP ALTER TABLE], and\nhttp://dev.mysql.com/doc/refman/5.1/en/partitioning-maintenance.html.\…: http://dev.mysql.com/doc/refman/5.1/en/analyze-table.html\n\n','','http://dev.mysql.com/doc/refman/5.1/en/analyze-table.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (467,30,'MICROSECOND','Syntax:\nMICROSECOND(expr)\n\nReturns the microseconds from the time or datetime expression expr as a\nnumber in the range from 0 to 999999.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html\n\n','mysql> SELECT MICROSECOND(\'12:00:00.123456\');\n -> 123456\nmysql> SELECT MICROSECOND(\'2009-12-31 23:59:59.000010\');\n -> 10\n','http://dev.mysql.com/doc/refman/5.1/en/date-and-time-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (468,37,'CONSTRAINT','InnoDB supports foreign key constraints. The syntax for a foreign key\nconstraint definition in InnoDB looks like this:\n\n[CONSTRAINT [symbol]] FOREIGN KEY\n [index_name] (index_col_name, ...)\n REFERENCES tbl_name (index_col_name,...)\n [ON DELETE reference_option]\n [ON UPDATE reference_option]\n\nreference_option:\n RESTRICT | CASCADE | SET NULL | NO ACTION\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/innodb-foreign-key-constraints.html\…','CREATE TABLE product (category INT NOT NULL, id INT NOT NULL,\n price DECIMAL,\n PRIMARY KEY(category, id)) ENGINE=INNODB;\nCREATE TABLE customer (id INT NOT NULL,\n PRIMARY KEY (id)) ENGINE=INNODB;\nCREATE TABLE product_order (no INT NOT NULL AUTO_INCREMENT,\n product_category INT NOT NULL,\n product_id INT NOT NULL,\n customer_id INT NOT NULL,\n PRIMARY KEY(no),\n INDEX (product_category, product_id),\n FOREIGN KEY (product_category, product_id)\n REFERENCES product(category, id)\n ON UPDATE CASCADE ON DELETE RESTRICT,\n INDEX (customer_id),\n FOREIGN KEY (customer_id)\n REFERENCES customer(id)) ENGINE=INNODB;\n','http://dev.mysql.com/doc/refman/5.1/en/innodb-foreign-key-constraints.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (469,37,'CREATE SERVER','Syntax:\nCREATE SERVER server_name\n FOREIGN DATA WRAPPER wrapper_name\n OPTIONS (option [, option] ...)\n\noption:\n { HOST character-literal\n | DATABASE character-literal\n | USER character-literal\n | PASSWORD character-literal\n | SOCKET character-literal\n | OWNER character-literal\n | PORT numeric-literal }\n\nThis statement creates the definition of a server for use with the\nFEDERATED storage engine. The CREATE SERVER statement creates a new row\nwithin the servers table within the mysql database. This statement\nrequires the SUPER privilege.\n\nThe server_name should be a unique reference to the server. Server\ndefinitions are global within the scope of the server, it is not\npossible to qualify the server definition to a specific database.\nserver_name has a maximum length of 64 characters (names longer than 64\ncharacters are silently truncated), and is case insensitive. You may\nspecify the name as a quoted string.\n\nThe wrapper_name should be mysql, and may be quoted with single quotes.\nOther values for wrapper_name are not currently supported.\n\nFor each option you must specify either a character literal or numeric\nliteral. Character literals are UTF-8, support a maximum length of 64\ncharacters and default to a blank (empty) string. String literals are\nsilently truncated to 64 characters. Numeric literals must be a number\nbetween 0 and 9999, default value is 0.\n\n*Note*: Note that the OWNER option is currently not applied, and has no\neffect on the ownership or operation of the server connection that is\ncreated.\n\nThe CREATE SERVER statement creates an entry in the mysql.server table\nthat can later be used with the CREATE TABLE statement when creating a\nFEDERATED table. The options that you specify will be used to populate\nthe columns in the mysql.server table. The table columns are\nServer_name, Host, Db, Username, Password, Port and Socket.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/create-server.html\n\n','CREATE SERVER s\nFOREIGN DATA WRAPPER mysql\nOPTIONS (USER \'Remote\', HOST \'192.168.1.106\', DATABASE \'test\');\n','http://dev.mysql.com/doc/refman/5.1/en/create-server.html'); @@ -567,7 +567,7 @@ insert into help_topic (help_topic_id,he insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (498,4,'RADIANS','Syntax:\nRADIANS(X)\n\nReturns the argument X, converted from degrees to radians. (Note that\nπ radians equals 180 degrees.)\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/mathematical-functions.html\n\n','mysql> SELECT RADIANS(90);\n -> 1.5707963267949\n','http://dev.mysql.com/doc/refman/5.1/en/mathematical-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (499,15,'COLLATION','Syntax:\nCOLLATION(str)\n\nReturns the collation of the string argument.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/information-functions.html\n\n','mysql> SELECT COLLATION(\'abc\');\n -> \'latin1_swedish_ci\'\nmysql> SELECT COLLATION(_utf8\'abc\');\n -> \'utf8_general_ci\'\n','http://dev.mysql.com/doc/refman/5.1/en/information-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (500,17,'COALESCE','Syntax:\nCOALESCE(value,...)\n\nReturns the first non-NULL value in the list, or NULL if there are no\nnon-NULL values.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/comparison-operators.html\n\n','mysql> SELECT COALESCE(NULL,1);\n -> 1\nmysql> SELECT COALESCE(NULL,NULL,NULL);\n -> NULL\n','http://dev.mysql.com/doc/refman/5.1/en/comparison-operators.html'); -insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (501,15,'VERSION','Syntax:\nVERSION()\n\nReturns a string that indicates the MySQL server version. The string\nuses the utf8 character set.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/information-functions.html\n\n','mysql> SELECT VERSION();\n -> \'5.1.39-standard\'\n','http://dev.mysql.com/doc/refman/5.1/en/information-functions.html'); +insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (501,15,'VERSION','Syntax:\nVERSION()\n\nReturns a string that indicates the MySQL server version. The string\nuses the utf8 character set.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/information-functions.html\n\n','mysql> SELECT VERSION();\n -> \'5.1.41-standard\'\n','http://dev.mysql.com/doc/refman/5.1/en/information-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (502,35,'MAKE_SET','Syntax:\nMAKE_SET(bits,str1,str2,...)\n\nReturns a set value (a string containing substrings separated by ","\ncharacters) consisting of the strings that have the corresponding bit\nin bits set. str1 corresponds to bit 0, str2 to bit 1, and so on. NULL\nvalues in str1, str2, ... are not appended to the result.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/string-functions.html\n\n','mysql> SELECT MAKE_SET(1,\'a\',\'b\',\'c\');\n -> \'a\'\nmysql> SELECT MAKE_SET(1 | 4,\'hello\',\'nice\',\'world\');\n -> \'hello,world\'\nmysql> SELECT MAKE_SET(1 | 4,\'hello\',\'nice\',NULL,\'world\');\n -> \'hello\'\nmysql> SELECT MAKE_SET(0,\'a\',\'b\',\'c\');\n -> \'\'\n','http://dev.mysql.com/doc/refman/5.1/en/string-functions.html'); insert into help_topic (help_topic_id,help_category_id,name,description,example,url) values (503,35,'FIND_IN_SET','Syntax:\nFIND_IN_SET(str,strlist)\n\nReturns a value in the range of 1 to N if the string str is in the\nstring list strlist consisting of N substrings. A string list is a\nstring composed of substrings separated by "," characters. If the first\nargument is a constant string and the second is a column of type SET,\nthe FIND_IN_SET() function is optimized to use bit arithmetic. Returns\n0 if str is not in strlist or if strlist is the empty string. Returns\nNULL if either argument is NULL. This function does not work properly\nif the first argument contains a comma (",") character.\n\nURL: http://dev.mysql.com/doc/refman/5.1/en/string-functions.html\n\n','mysql> SELECT FIND_IN_SET(\'b\',\'a,b,c,d\');\n -> 2\n','http://dev.mysql.com/doc/refman/5.1/en/string-functions.html'); @@ -1293,6 +1293,7 @@ insert into help_relation (help_topic_id insert into help_relation (help_topic_id,help_keyword_id) values (143,115); insert into help_relation (help_topic_id,help_keyword_id) values (468,115); insert into help_relation (help_topic_id,help_keyword_id) values (180,115); +insert into help_relation (help_topic_id,help_keyword_id) values (353,115); insert into help_relation (help_topic_id,help_keyword_id) values (420,115); insert into help_relation (help_topic_id,help_keyword_id) values (473,115); insert into help_relation (help_topic_id,help_keyword_id) values (197,116); @@ -1550,11 +1551,12 @@ insert into help_relation (help_topic_id insert into help_relation (help_topic_id,help_keyword_id) values (383,238); insert into help_relation (help_topic_id,help_keyword_id) values (330,239); insert into help_relation (help_topic_id,help_keyword_id) values (344,239); -insert into help_relation (help_topic_id,help_keyword_id) values (253,239); insert into help_relation (help_topic_id,help_keyword_id) values (152,239); insert into help_relation (help_topic_id,help_keyword_id) values (428,239); insert into help_relation (help_topic_id,help_keyword_id) values (464,239); +insert into help_relation (help_topic_id,help_keyword_id) values (253,239); insert into help_relation (help_topic_id,help_keyword_id) values (209,239); +insert into help_relation (help_topic_id,help_keyword_id) values (353,239); insert into help_relation (help_topic_id,help_keyword_id) values (420,239); insert into help_relation (help_topic_id,help_keyword_id) values (184,240); insert into help_relation (help_topic_id,help_keyword_id) values (273,241); === modified file 'support-files/MacOSX/ReadMe.txt' --- a/support-files/MacOSX/ReadMe.txt 2009-05-25 09:59:47 +0000 +++ b/support-files/MacOSX/ReadMe.txt 2009-12-01 07:24:05 +0000 @@ -1,5 +1,47 @@ -2.5. Installing MySQL on Mac OS X +2.7. Installing MySQL on Mac OS X + + MySQL for Mac OS X is available in a number of different forms: + + * Native Package Installer format, which uses the native Mac OS + X installer to walk you through the installation of MySQL. For + more information, see Section 2.7.1, "Installing MySQL Using + the Installation Package." You can use the package installer + with Mac OS X 10.3 and later, and available for both PowerPC + and Intel architectures, and both 32-bit and 64-bit + architectures. There is no Universal Binary available using + the package installation method. The user you use to perform + the installation must have administrator privileges. + + * Tar package format, which uses a file packaged using the Unix + tar and gzip commands. To use this method, you will need to + open a Terminal window. You do not need administrator + privileges using this method, as you can install the MySQL + server anywhere using this method. For more information on + using this method, you can use the generic instructions for + using a tarball, Section 2.2, "Installing MySQL from Generic + Binaries on Unix/Linux."You can use the package installer with + Mac OS X 10.3 and later, and available for both PowerPC and + Intel architectures, and both 32-bit and 64-bit architectures. + A Universal Binary, incorporating both Power PC and Intel + architectures and 32-bit and 64-bit binaries is available. + In addition to the core installation, the Package Installer + also includes Section 2.7.2, "Installing the MySQL Startup + Item" and Section 2.7.3, "Installing and Using the MySQL + Preference Pane," both of which simplify the management of + your installation. + + * Mac OS X server includes a version of MySQL as standard. If + you want to use a more recent version than that supplied with + the Mac OS X server release, you can make use of the package + or tar formats. For more information on using the MySQL + bundled with Mac OS X, see Section 2.7.4, "Using MySQL on Mac + OS X Server." + + For additional information on using MySQL on Mac OS X, see Section + 2.7.5, "MySQL Installation on Mac OS X Notes." + +2.7.1. Installing MySQL Using the Installation Package You can install MySQL on Mac OS X 10.3.x ("Panther") or newer using a Mac OS X binary package in PKG format instead of the @@ -11,8 +53,6 @@ first need to mount by double-clicking its icon in the Finder. It should then mount the image and display its contents. - To obtain MySQL, see Section 2.1.3, "How to Get MySQL." - Note Before proceeding with the installation, be sure to shut down all @@ -20,144 +60,319 @@ Note Application (on Mac OS X Server) or via mysqladmin shutdown on the command line. - To actually install the MySQL PKG file, double-click on the - package icon. This launches the Mac OS X Package Installer, which - guides you through the installation of MySQL. + When installing from the package version, you should also install + the MySQL Preference Pane, which will allow you to control the + startup and execution of your MySQL server from System + Preferences. For more information, see Section 2.7.3, "Installing + and Using the MySQL Preference Pane." + + When installing using the package installer, the files are + installed into a directory within /usr/local matching the name of + the installation version and platform. For example, the installer + file mysql-5.1.39-osx10.5-x86_64.pkg installs MySQL into + /usr/local/mysql-5.1.39-osx10.5-x86_64 . The installation layout + of the directory is as shown in the following table: + Directory Contents of Directory + bin Client programs and the mysqld server + data Log files, databases + docs Manual in Info format + include Include (header) files + lib Libraries + man Unix manual pages + mysql-test MySQL test suite + scripts Contains the mysql_install_db script + share/mysql Error message files + sql-bench Benchmarks + support-files Scripts and sample configuration files + /tmp/mysql.sock The location of the MySQL Unix socket + + During the package installer process, a symbolic link from + /usr/local/mysql to the version/platform specific directory + created during installation will be created automatically. + + 1. Download and open the MySQL package installer, which is + provided on a disk image (.dmg). Double-click to open the disk + image, which includes the main MySQL installation package, the + MySQLStartupItem.pkg installation package, and the + MySQL.prefPane. + + 2. Double-click on the MySQL installer package. It will be named + according to the version of MySQL you have downloaded. For + example, if you have downloaded MySQL 5.1.39, double-click + mysql-5.1.39-osx10.5-x86.pkg. + + 3. You will be presented with the openin installer dialog. Click + Continue to begihn installation. + MySQL Package Installer: Step 1 + + 4. A copy of the installation instructions and other important + information relevant to this installation are display. Click + Continue . + + 5. If you have downloaded the community version of MySQL, you + will be shown a copy of the relevent GNU General Public + License. Click Continue . + + 6. Select the drive you want to use to install the MySQL Startup + Item. The drive must have a valid, bootable, Mac OS X + operating system installed. Click Continue. + MySQL Package Installer: Step 4 + + 7. You will be asked to confirm the details of the installation, + including the space required for the installation. To change + the drive on which the startup item is installed you can click + either Go Back or Change Install Location.... To install the + startup item, click Install. - Due to a bug in the Mac OS X package installer, you may see this - error message in the destination disk selection dialog: -You cannot install this software on this disk. (null) - - If this error occurs, simply click the Go Back button once to - return to the previous screen. Then click Continue to advance to - the destination disk selection again, and you should be able to - choose the destination disk correctly. We have reported this bug - to Apple and it is investigating this problem. - - The Mac OS X PKG of MySQL installs itself into - /usr/local/mysql-VERSION and also installs a symbolic link, - /usr/local/mysql, that points to the new location. If a directory - named /usr/local/mysql exists, it is renamed to - /usr/local/mysql.bak first. Additionally, the installer creates - the grant tables in the mysql database by executing - mysql_install_db. - - The installation layout is similar to that of a tar file binary - distribution; all MySQL binaries are located in the directory - /usr/local/mysql/bin. The MySQL socket file is created as - /tmp/mysql.sock by default. See Section 2.1.5, "Installation - Layouts." - - MySQL installation requires a Mac OS X user account named mysql. A - user account with this name should exist by default on Mac OS X - 10.2 and up. + 8. Once the installation has been completed successfully, you + will be given an Install Succeeded message. - If you are running Mac OS X Server, a version of MySQL should - already be installed. The following table shows the versions of - MySQL that ship with Mac OS X Server versions. - Mac OS X Server Version MySQL Version - 10.2-10.2.2 3.23.51 - 10.2.3-10.2.6 3.23.53 - 10.3 4.0.14 - 10.3.2 4.0.16 - 10.4.0 4.1.10a + Once you have completed the basic installation, you must complete + the post-installation steps as specifed in Section 2.13, + "Post-Installation Setup and Testing." - This manual section covers the installation of the official MySQL - Mac OS X PKG only. Make sure to read Apple's help information - about installing MySQL: Run the "Help View" application, select - "Mac OS X Server" help, do a search for "MySQL," and read the item - entitled "Installing MySQL." - - If you previously used Marc Liyanage's MySQL packages for Mac OS X - from http://www.entropy.ch, you can simply follow the update - instructions for packages using the binary installation layout as - given on his pages. - - If you are upgrading from Marc's 3.23.x versions or from the Mac - OS X Server version of MySQL to the official MySQL PKG, you also - need to convert the existing MySQL privilege tables to the current - format, because some new security privileges have been added. See - Section 4.4.8, "mysql_upgrade --- Check Tables for MySQL Upgrade." - - If you want MySQL to start automatically during system startup, - you also need to install the MySQL Startup Item. It is part of the - Mac OS X installation disk images as a separate installation - package. Simply double-click the MySQLStartupItem.pkg icon and - follow the instructions to install it. The Startup Item need be - installed only once. There is no need to install it each time you - upgrade the MySQL package later. + For convenience, you may also want to install the Section 2.7.2, + "Installing the MySQL Startup Item" and Section 2.7.3, "Installing + and Using the MySQL Preference Pane." + +2.7.2. Installing the MySQL Startup Item + + The MySQL Installation Package includes a startup item that can be + used to automatically startup and shutdown MySQL during boot. + + To install the MySQL Startup Item: + + 1. Download and open the MySQL package installer, which is + provided on a disk image (.dmg). Double-click to open the disk + image, which includes the main MySQL installation package, the + MySQLStartupItem.pkg installation package, and the + MySQL.prefPane. + + 2. Double-click on the MySQLStartItem.pkg file to start the + installation process. + + 3. You will be presented with the Install MySQL Startup Item + dialog. + MySQL Startup Item Installer: Step 1 + Click Continue to continue the installation process. + + 4. A copy of the installation instructions and other important + information relevant to this installation are display. Click + Continue . + + 5. Select the drive you want to use to install the MySQL Startup + Item. The drive must have a valid, bootable, Mac OS X + operating system installed. Click Continue. + MySQL Startup Item Installer: Step 3 + + 6. You will be asked to confirm the details of the installation. + To change the drive on which the startup item is installed you + can click either Go Back or Change Install Location.... To + install the startup item, click Install. + + 7. Once the installation has been completed successfully, you + will be given an Install Succeeded message. + MySQL Startup Item Installer: Step 5 The Startup Item for MySQL is installed into - /Library/StartupItems/MySQLCOM. (Before MySQL 4.1.2, the location - was /Library/StartupItems/MySQL, but that collided with the MySQL - Startup Item installed by Mac OS X Server.) Startup Item - installation adds a variable MYSQLCOM=-YES- to the system - configuration file /etc/hostconfig. If you want to disable the - automatic startup of MySQL, simply change this variable to - MYSQLCOM=-NO-. - - On Mac OS X Server, the default MySQL installation uses the - variable MYSQL in the /etc/hostconfig file. The MySQL Startup Item - installer disables this variable by setting it to MYSQL=-NO-. This - avoids boot time conflicts with the MYSQLCOM variable used by the - MySQL Startup Item. However, it does not shut down a running MySQL - server. You should do that yourself. + /Library/StartupItems/MySQLCOM. The Startup Item installation adds + a variable MYSQLCOM=-YES- to the system configuration file + /etc/hostconfig. If you want to disable the automatic startup of + MySQL, simply change this variable to MYSQLCOM=-NO-. After the installation, you can start up MySQL by running the following commands in a terminal window. You must have administrator privileges to perform this task. - If you have installed the Startup Item, use this command: + If you have installed the Startup Item, use this command to start + the server: shell> sudo /Library/StartupItems/MySQLCOM/MySQLCOM start -(Enter your password, if necessary) -(Press Control-D or enter "exit" to exit the shell) - If you don't use the Startup Item, enter the following command - sequence: -shell> cd /usr/local/mysql -shell> sudo ./bin/mysqld_safe -(Enter your password, if necessary) -(Press Control-Z) -shell> bg -(Press Control-D or enter "exit" to exit the shell) + You may be prompted for your password to complete the startup. - You should be able to connect to the MySQL server, for example, by - running /usr/local/mysql/bin/mysql. + If you have installed the Startup Item, use this command to stop + the server: +shell> sudo /Library/StartupItems/MySQLCOM/MySQLCOM stop + + You may be prompted for your password to complete the shutdown. + +2.7.3. Installing and Using the MySQL Preference Pane + + The MySQL Package installer disk image also includes a custom + MySQL Preference Pane that enables you to start, stop and control + automated startup during boot of your MySQL installation. + + To install the MySQL Preference Pane: + + 1. Download and open the MySQL package installer package, which + is provided on a disk image (.dmg). Double-click to open the + disk image, which includes the main MySQL installation + package, the MySQLStartupItem.pkg installation package, and + the MySQL.prefPane. + + 2. Double click on MySQL.prefPane. The MySQL System Preferences + will open. + + 3. If this is the first time you have installed the preference + pane, you will be asked to confirm installation and whether + you want to install the preference pane for all users, or only + the current user. To install the preference pane for all users + you will need administrator privileges. If necessary, you will + be prompted for the username and password for a user with + administrator privileges. + + 4. If you already have the MySQL Preference Pane installed, you + will be asked to confirm whether you want to overwrite the + existing MySQL Preference Pane. Note - The accounts that are listed in the MySQL grant tables initially - have no passwords. After starting the server, you should set up - passwords for them using the instructions in Section 2.11, - "Post-Installation Setup and Testing." + The MySQL Preference Pane only starts and stops MySQL installation + installed from the MySQL package installation that have been + installed in the default location. + + Once the MySQL Preference Pane has been installed, you can control + your MySQL server instance using the preference pane. To use the + preference pane, open the System Preferences... from the Apple + menu. Select the MySQL preference pane by clicking on the MySQL + logo within the Other section of the preference panes list. + MySQL Preference Pane + + The MySQL Preference Pane shows the current status of the MySQL + server, showing stopped (in red) if the server is not running and + running (in green) if the server has already been started. The + preference pane will also show the current setting for whether the + MySQL server has been set to start up automatically. + + * To start MySQL using the preference pane: + Click Start MySQL Server. You may be prompted for the username + and password of a user with administrator privileges to start + the MySQL server. + + * To stop MySQL using the preference pane: + Click Stop MySQL Server. You may be prompted for the username + and password of a user with administrator privileges to + shutdown the MySQL server. + + * To automatically start the MySQL server when the system boots: + Check the checkbox next to Automatically Start MySQL Server on + Startup. + + * To disable the automatic starting of the MySQL server when the + system boots: + Uncheck the checkbox next to Automatically Start MySQL Server + on Startup. + + You can close the System Preferences... once you have completed + your settings. - You might want to add aliases to your shell's resource file to - make it easier to access commonly used programs such as mysql and - mysqladmin from the command line. The syntax for bash is: +2.7.4. Using MySQL on Mac OS X Server + + If you are running Mac OS X Server, a version of MySQL should + already be installed. The following table shows the versions of + MySQL that ship with Mac OS X Server versions. + Mac OS X Server Version MySQL Version + 10.2-10.2.2 3.23.51 + 10.2.3-10.2.6 3.23.53 + 10.3 4.0.14 + 10.3.2 4.0.16 + 10.4.0 4.1.10a + 10.5.0 5.0.45 + 10.6.0 5.0.82 + + The installation layout of MySQL on Mac OS X Server is as shown in + the table below: + Directory Contents of Directory + /usr/bin Client programs + /var/mysql Log files, databases + /usr/libexec The mysqld server + /usr/share/man Unix manual pages + /usr/share/mysql/mysql-test MySQL test suite + /usr/share/mysql Contains the mysql_install_db script + /var/mysql/mysql.sock The location of the MySQL Unix socket + +Note + + The MySQL server bundled with Mac OS X Server does not include the + MySQL client libraries and header files required if you want to + access and use MySQL from a third-party driver, such as Perl DBI + or PHP. For more information on obtaining and installing MySQL + libraries, see Mac OS X Server version 10.5: MySQL libraries + available for download (http://support.apple.com/kb/TA25017) + Alternatively, you can ignore the bundled MySQL server and install + MySQL from the package or tarball installation. + + For more information on managing the bundled MySQL instance in Mac + OS X Server 10.5, see Mac OS X Server: Web Technologies + Administration For Version 10.5 Leopard + (http://images.apple.com/server/macosx/docs/Web_Technologies_Admin + _v10.5.pdf). For more information on managing the bundled MySQL + instance in Mac OS X Server 10.6, see Mac OS X Server: Web + Technologies Administration Version 10.6 Snow Leopard + (http://manuals.info.apple.com/en_US/WebTech_v10.6.pdf) + +2.7.5. MySQL Installation on Mac OS X Notes + + You should keep the following issues and notes in mind: + + * The default location for the MySQL Unix socket is different on + Mac OS X and Mac OS X Server depending on the installation + type you chose. The default locations by installation are as + follows: + + Package Installer from MySQL /tmp/mysql.sock + Tarball from MySQL /tmp/mysql.sock + MySQL Bundled with Mac OS X Server /var/mysql/mysql.sock + To prevent issues, you should either change the configuration + of the socket used within your application (for example, + changing php.ini), or you should configure the socket location + using a MySQL configuration file and the socket option. For + more information, see Section 5.1.2, "Server Command Options." + + * You may need (or want) to create a specific mysql user to own + the MySQL directory and data. On Mac OS X 10.4 and lower you + can do this by using the Netinfo Manager application, located + within the Utilities folder within the Applications folder. On + Mac OS X 10.5 and later you can do this through the Directory + Utility. From Mac OS X 10.5 and later (including Mac OS X + Server 10.5) the mysql should already exist. For use in single + user mode, an entry for _mysql (note the underscore prefix) + should already exist within the system /etc/passwd file. + + * Due to a bug in the Mac OS X package installer, you may see + this error message in the destination disk selection dialog: +You cannot install this software on this disk. (null) + If this error occurs, simply click the Go Back button once to + return to the previous screen. Then click Continue to advance + to the destination disk selection again, and you should be + able to choose the destination disk correctly. We have + reported this bug to Apple and it is investigating this + problem. + + * Because the MySQL package installer installs the MySQL + contents into a version and platform specific directory, you + can use this to upgrade and migrate your database between + versions. You will need to either copy the data directory from + the old version to the new version, or alternatively specify + an alternative datadir value to set location of the data + directory. + + * You might want to add aliases to your shell's resource file to + make it easier to access commonly used programs such as mysql + and mysqladmin from the command line. The syntax for bash is: alias mysql=/usr/local/mysql/bin/mysql alias mysqladmin=/usr/local/mysql/bin/mysqladmin - - For tcsh, use: + For tcsh, use: alias mysql /usr/local/mysql/bin/mysql alias mysqladmin /usr/local/mysql/bin/mysqladmin - - Even better, add /usr/local/mysql/bin to your PATH environment - variable. You can do this by modifying the appropriate startup - file for your shell. For more information, see Section 4.2.1, - "Invoking MySQL Programs." - - If you are upgrading an existing installation, note that - installing a new MySQL PKG does not remove the directory of an - older installation. Unfortunately, the Mac OS X Installer does not - yet offer the functionality required to properly upgrade - previously installed packages. - - To use your existing databases with the new installation, you'll - need to copy the contents of the old data directory to the new - data directory. Make sure that neither the old server nor the new - one is running when you do this. After you have copied over the - MySQL database files from the previous installation and have - successfully started the new server, you should consider removing - the old installation files to save disk space. Additionally, you - should also remove older versions of the Package Receipt - directories located in /Library/Receipts/mysql-VERSION.pkg. + Even better, add /usr/local/mysql/bin to your PATH environment + variable. You can do this by modifying the appropriate startup + file for your shell. For more information, see Section 4.2.1, + "Invoking MySQL Programs." + + * After you have copied over the MySQL database files from the + previous installation and have successfully started the new + server, you should consider removing the old installation + files to save disk space. Additionally, you should also remove + older versions of the Package Receipt directories located in + /Library/Receipts/mysql-VERSION.pkg.

1 0

[Maria-developers] Running sql scripts with '@'
by Viljo Hakala 30 Nov '09

30 Nov '09

Wouldn't it be nice to run .sql -scripts just like: mysql> @script.sql I have found many Oracle/MySQL dbas have a tendency to use '@' to run a script.

1 0

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (monty:2770)
by Michael Widenius 30 Nov '09

30 Nov '09

#At lp:maria based on revid:monty@askmonty.org-20091126201933-dgoynszp2z90gknl 2770 Michael Widenius 2009-11-30 Remove compiler warnings (Including some warnings from -Wstrict-aliasing) Don't use static link by default (in compile-pentium) as some new systems doesn't have all static libraries available Change type for functions in plugin.h:str_mysql_ftparser_param() to const unsigned char and string lengths to size_t. One effect of the above change is that one needs to include mysql_global.h or define size_t before including plugin.h This fixes a case where mysql_client_test failed with newer gcc that enables strict-aliasing by default modified: BUILD/compile-pentium client/mysql_upgrade.c cmd-line-utils/readline/config_readline.h cmd-line-utils/readline/display.c cmd-line-utils/readline/history.c cmd-line-utils/readline/rlmbutil.h cmd-line-utils/readline/text.c cmd-line-utils/readline/xmalloc.c configure.in include/mysql/plugin.h include/mysql/plugin.h.pp libmysql/libmysql.c mysql-test/t/information_schema.test mysql-test/t/not_partition.test mysys/lf_hash.c mysys/my_redel.c regex/engine.c regex/engine.ih sql/sp_head.cc sql/sql_base.cc sql/sql_builtin.cc.in sql/sql_parse.cc sql/sql_select.cc sql/sql_show.cc sql/table.cc storage/maria/ha_maria.cc storage/maria/lockman.c storage/maria/ma_check.c storage/maria/ma_check_standalone.h storage/maria/ma_ft_boolean_search.c storage/maria/ma_ft_nlq_search.c storage/maria/ma_ft_parser.c storage/maria/ma_ftdefs.h storage/maria/ma_sort.c storage/maria/ma_state.c storage/maria/maria_def.h storage/maria/maria_ftdump.c storage/maria/trnman.c storage/myisam/ft_boolean_search.c storage/myisam/ft_nlq_search.c storage/myisam/ft_parser.c storage/myisam/ft_stopwords.c storage/myisam/ftdefs.h storage/myisam/ha_myisam.cc storage/myisam/mi_check.c storage/myisam/myisam_ftdump.c storage/myisam/myisamchk.c storage/myisam/myisamdef.h storage/myisam/myisamlog.c storage/myisam/sort.c storage/xtradb/fil/fil0fil.c storage/xtradb/trx/trx0i_s.c per-file messages: BUILD/compile-pentium Don't use static link by default as some new systems doesn't have all static libraries available client/mysql_upgrade.c Remove not used variable cmd-line-utils/readline/config_readline.h Define some constants to get rid of compiler warnings on Linux cmd-line-utils/readline/display.c Get rid of compiler warnings cmd-line-utils/readline/history.c Got rid of compiler warnings: - Defining some strings as const - Added cast cmd-line-utils/readline/rlmbutil.h Added cast to get rid of compiler warnings cmd-line-utils/readline/text.c Remove not needed initialization to get rid of compiler warnings cmd-line-utils/readline/xmalloc.c Changed types to 'const char* to get rid of compiler warnings configure.in Ensure that we use MariaDB as suffix include/mysql/plugin.h Changed types to 'const unsigned char* to get rid of compiler warnings (in other parts of the code) Change length for not \0 terminated string to size_t include/mysql/plugin.h.pp Update related to plugin.h libmysql/libmysql.c Fixed bug that caused core dump with newer gcc when strict aliasing is not turned off mysql-test/t/information_schema.test Test is depending on innodb mysql-test/t/not_partition.test Fixed wrong directory name (Not noticed before as we don't ususally run this test) mysys/lf_hash.c Got rid of compiler warnings from -Wstrict-aliasing mysys/my_redel.c Removed not used variable regex/engine.c Changed types to 'const char* to get rid of compiler warnings regex/engine.ih Changed types to 'const char* to get rid of compiler warnings sql/sp_head.cc Got rid of compiler warning from -Wstrict-aliasing sql/sql_base.cc Got rid of compiler warnings from -Wstrict-aliasing (The original code was probably wrong as nj_col->table_field was sql/sql_builtin.cc.in plugin.h needs to have size_t defined sql/sql_parse.cc Remove used variable sql/sql_select.cc Got rid of compiler warnings from -Wstrict-aliasing sql/sql_show.cc Added #ifdef to get rid of compiler warning when not using partition engine sql/table.cc Got rid of compiler warning from -Wstrict-aliasing storage/maria/ha_maria.cc Got rid of compiler warnings from -Wstrict-aliasing: - Use the thd_killed() API function storage/maria/lockman.c Got rid of compiler warnings from -Wstrict-aliasing storage/maria/ma_check.c Got rid of compiler warnings from -Wstrict-aliasing Change to use new version of _ma_killed_ptr; Don't call it as often as before storage/maria/ma_check_standalone.h Update to compatible _ma_killed_ptr() from ha_maria.cc storage/maria/ma_ft_boolean_search.c Changed pointers from char -> const char* and length to size_t (to get rid of compiler warnings and casts) storage/maria/ma_ft_nlq_search.c Got rid of compiler warnings from -Wstrict-aliasing Ensure that 'subkeys' is 32 bit storage/maria/ma_ft_parser.c Changed pointers from char -> const char* and length to size_t (to get rid of compiler warnings and casts) storage/maria/ma_ftdefs.h Changed pointers from char -> const char* and length to size_t (to get rid of compiler warnings and casts) storage/maria/ma_sort.c Change to use new version of _ma_killed_ptr; Don't call it as often as before storage/maria/ma_state.c Got rid of compiler warnings from -Wstrict-aliasing storage/maria/maria_def.h Redefine ma_killed_ptr() storage/maria/maria_ftdump.c Got rid of compiler warnings from -Wstrict-aliasing storage/maria/trnman.c Got rid of compiler warnings from -Wstrict-aliasing storage/myisam/ft_boolean_search.c Changed pointers from char -> const char* and length to size_t (to get rid of compiler warnings and casts) storage/myisam/ft_nlq_search.c Got rid of compiler warnings from -Wstrict-aliasing storage/myisam/ft_parser.c Changed pointers from char -> const char* and length to size_t (to get rid of compiler warnings and casts) storage/myisam/ft_stopwords.c Changed pointers from char -> const char* and length to size_t (to get rid of compiler warnings and casts) storage/myisam/ftdefs.h Changed pointers from char -> const char* and length to size_t (to get rid of compiler warnings and casts) storage/myisam/ha_myisam.cc Got rid of compiler warnings from -Wstrict-aliasing: - Use the thd_killed() API function storage/myisam/mi_check.c Use new killed_ptr() function storage/myisam/myisam_ftdump.c Got rid of compiler warnings from -Wstrict-aliasing storage/myisam/myisamchk.c Update to compatible killed_ptr() from ha_myisam.cc storage/myisam/myisamdef.h Redefine killed_ptr() storage/myisam/myisamlog.c Got rid of compiler warnings from -Wstrict-aliasing storage/myisam/sort.c Change to use new version of killed_ptr; Don't call it as often as before storage/xtradb/fil/fil0fil.c Fixedc ompiler warning storage/xtradb/trx/trx0i_s.c Include mysql_plugin.h later to ensure that size_t is defined === modified file 'BUILD/compile-pentium' --- a/BUILD/compile-pentium 2007-04-11 12:12:00 +0000 +++ b/BUILD/compile-pentium 2009-11-29 23:08:56 +0000 @@ -4,7 +4,7 @@ path=`dirname $0` . "$path/SETUP.sh" extra_flags="$pentium_cflags $fast_cflags" -extra_configs="$pentium_configs $static_link" +extra_configs="$pentium_configs" strip=yes . "$path/FINISH.sh" === modified file 'client/mysql_upgrade.c' --- a/client/mysql_upgrade.c 2009-11-06 17:22:32 +0000 +++ b/client/mysql_upgrade.c 2009-11-29 23:08:56 +0000 @@ -552,7 +552,6 @@ static int upgrade_already_done(void) FILE *in; char upgrade_info_file[FN_REFLEN]= {0}; char buf[sizeof(MYSQL_SERVER_VERSION)+1]; - char *res; if (get_upgrade_info_file_name(upgrade_info_file)) return 0; /* Could not get filename => not sure */ === modified file 'cmd-line-utils/readline/config_readline.h' --- a/cmd-line-utils/readline/config_readline.h 2005-04-20 10:02:07 +0000 +++ b/cmd-line-utils/readline/config_readline.h 2009-11-29 23:08:56 +0000 @@ -7,6 +7,11 @@ # include <config.h> #endif +/* to get wcwidth() defined */ +#define _XOPEN_SOURCE 600 +#define _XOPEN_SOURCE_EXTENDED +#define _XOPEN_ + /* Ultrix botches type-ahead when switching from canonical to non-canonical mode, at least through version 4.3 === modified file 'cmd-line-utils/readline/display.c' --- a/cmd-line-utils/readline/display.c 2009-09-07 20:50:10 +0000 +++ b/cmd-line-utils/readline/display.c 2009-11-29 23:08:56 +0000 @@ -461,12 +461,12 @@ rl_redisplay () register char *line; int inv_botlin, lb_linenum, o_cpos; int newlines, lpos, temp, modmark; - char *prompt_this_line; + const char *prompt_this_line; #if defined (HANDLE_MULTIBYTE) - int num, n0; + int num, n0= 0; wchar_t wc; size_t wc_bytes; - int wc_width; + int wc_width= 0; mbstate_t ps; int _rl_wrapped_multicolumn = 0; #endif @@ -824,7 +824,7 @@ rl_redisplay () cpos_buffer_position = out; lb_linenum = newlines; } - for (i = in; i < in+wc_bytes; i++) + for (i = in; (size_t) i < in+wc_bytes; i++) line[out++] = rl_line_buffer[i]; for (i = 0; i < wc_width; i++) CHECK_LPOS(); === modified file 'cmd-line-utils/readline/history.c' --- a/cmd-line-utils/readline/history.c 2008-04-28 16:24:05 +0000 +++ b/cmd-line-utils/readline/history.c 2009-11-29 23:08:56 +0000 @@ -211,14 +211,14 @@ history_get (offset) HIST_ENTRY * alloc_history_entry (string, ts) - char *string; + const char *string; char *ts; { HIST_ENTRY *temp; temp = (HIST_ENTRY *)xmalloc (sizeof (HIST_ENTRY)); - temp->line = string ? savestring (string) : string; + temp->line = string ? savestring ((char*) string) : (char*) string; temp->data = (char *)NULL; temp->timestamp = ts; === modified file 'cmd-line-utils/readline/rlmbutil.h' --- a/cmd-line-utils/readline/rlmbutil.h 2007-11-19 13:38:08 +0000 +++ b/cmd-line-utils/readline/rlmbutil.h 2009-11-29 23:08:56 +0000 @@ -109,8 +109,8 @@ extern int _rl_is_mbchar_matched PARAMS( extern wchar_t _rl_char_value PARAMS((char *, int)); extern int _rl_walphabetic PARAMS((wchar_t)); -#define _rl_to_wupper(wc) (iswlower (wc) ? towupper (wc) : (wc)) -#define _rl_to_wlower(wc) (iswupper (wc) ? towlower (wc) : (wc)) +#define _rl_to_wupper(wc) (iswlower (wc) ? (wchar_t) towupper (wc) : (wc)) +#define _rl_to_wlower(wc) (iswupper (wc) ? (wchar_t) towlower (wc) : (wc)) #define MB_NEXTCHAR(b,s,c,f) \ ((MB_CUR_MAX > 1 && rl_byte_oriented == 0) \ === modified file 'cmd-line-utils/readline/text.c' --- a/cmd-line-utils/readline/text.c 2009-09-07 20:50:10 +0000 +++ b/cmd-line-utils/readline/text.c 2009-11-29 23:08:56 +0000 @@ -614,7 +614,7 @@ rl_arrow_keys (count, c) #ifdef HANDLE_MULTIBYTE static char pending_bytes[MB_LEN_MAX]; static int pending_bytes_length = 0; -static mbstate_t ps = {0}; +static mbstate_t ps; #endif /* Insert the character C at the current location, moving point forward. === modified file 'cmd-line-utils/readline/xmalloc.c' --- a/cmd-line-utils/readline/xmalloc.c 2008-01-23 16:43:46 +0000 +++ b/cmd-line-utils/readline/xmalloc.c 2009-11-29 23:08:56 +0000 @@ -42,7 +42,7 @@ static void memory_error_and_abort (fname) - char *fname; + const char *fname; { fprintf (stderr, "%s: out of virtual memory\n", fname); exit (2); === modified file 'configure.in' --- a/configure.in 2009-11-07 15:56:51 +0000 +++ b/configure.in 2009-11-29 23:08:56 +0000 @@ -15,7 +15,7 @@ AC_CANONICAL_SYSTEM # MySQL version number. # # Note: the following line must be parseable by win/configure.js:GetVersion() -AM_INIT_AUTOMAKE(mysql, 5.1.39-maria-beta) +AM_INIT_AUTOMAKE(mysql, 5.1.39-MariaDB-beta) AM_CONFIG_HEADER([include/config.h:config.h.in]) PROTOCOL_VERSION=10 === modified file 'include/mysql/plugin.h' --- a/include/mysql/plugin.h 2009-09-07 20:50:10 +0000 +++ b/include/mysql/plugin.h 2009-11-29 23:08:56 +0000 @@ -567,9 +567,9 @@ typedef struct st_mysql_ftparser_boolean typedef struct st_mysql_ftparser_param { int (*mysql_parse)(struct st_mysql_ftparser_param *, - char *doc, int doc_len); + const unsigned char *doc, size_t doc_len); int (*mysql_add_word)(struct st_mysql_ftparser_param *, - char *word, int word_len, + const unsigned char *word, size_t word_len, MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info); void *ftparser_state; void *mysql_ftparam; === modified file 'include/mysql/plugin.h.pp' --- a/include/mysql/plugin.h.pp 2008-10-10 15:28:41 +0000 +++ b/include/mysql/plugin.h.pp 2009-11-29 23:08:56 +0000 @@ -73,9 +73,9 @@ typedef struct st_mysql_ftparser_boolean typedef struct st_mysql_ftparser_param { int (*mysql_parse)(struct st_mysql_ftparser_param *, - char *doc, int doc_len); + const unsigned char *doc, size_t doc_len); int (*mysql_add_word)(struct st_mysql_ftparser_param *, - char *word, int word_len, + const unsigned char *word, size_t word_len, MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info); void *ftparser_state; void *mysql_ftparam; === modified file 'libmysql/libmysql.c' --- a/libmysql/libmysql.c 2009-11-06 17:22:32 +0000 +++ b/libmysql/libmysql.c 2009-11-29 23:08:56 +0000 @@ -2284,7 +2284,7 @@ mysql_stmt_param_metadata(MYSQL_STMT *st /* Store type of parameter in network buffer. */ -static void store_param_type(char **pos, MYSQL_BIND *param) +static void store_param_type(uchar **pos, MYSQL_BIND *param) { uint typecode= param->buffer_type | (param->is_unsigned ? 32768 : 0); int2store(*pos, typecode); @@ -2564,7 +2564,7 @@ int cli_stmt_execute(MYSQL_STMT *stmt) that is sent to the server. */ for (param= stmt->params; param < param_end ; param++) - store_param_type((char**) &net->write_pos, param); + store_param_type(&net->write_pos, param); } for (param= stmt->params; param < param_end; param++) === modified file 'mysql-test/t/information_schema.test' --- a/mysql-test/t/information_schema.test 2009-09-29 20:19:43 +0000 +++ b/mysql-test/t/information_schema.test 2009-11-29 23:08:56 +0000 @@ -5,6 +5,9 @@ # on the presence of the log tables (which are CSV-based). --source include/have_csv.inc +# Check that innodb/xtradb is incompiled in as result depends on it +-- source include/have_innodb.inc + # Save the initial number of concurrent sessions --source include/count_sessions.inc === modified file 'mysql-test/t/not_partition.test' --- a/mysql-test/t/not_partition.test 2009-01-08 14:16:44 +0000 +++ b/mysql-test/t/not_partition.test 2009-11-29 23:08:56 +0000 @@ -15,7 +15,7 @@ let $MYSQLD_DATADIR= `SELECT @@datadir`; # Bug#39893: Crash if select on a partitioned table, # when partitioning is disabled FLUSH TABLES; ---copy_file $MYSQLTEST_VARDIR/std_data_ln/parts/t1.frm $MYSQLD_DATADIR/test/t1.frm +--copy_file $MYSQLTEST_VARDIR/std_data/parts/t1.frm $MYSQLD_DATADIR/test/t1.frm SELECT * FROM t1; TRUNCATE TABLE t1; ANALYZE TABLE t1; === modified file 'mysys/lf_hash.c' --- a/mysys/lf_hash.c 2009-01-15 21:27:36 +0000 +++ b/mysys/lf_hash.c 2009-11-29 23:08:56 +0000 @@ -124,8 +124,8 @@ retry: we found a deleted node - be nice, help the other thread and remove this deleted node */ - if (my_atomic_casptr((void **)cursor->prev, - (void **)&cursor->curr, cursor->next)) + if (my_atomic_casptr((void **) cursor->prev, + (void **)(char*) &cursor->curr, cursor->next)) _lf_alloc_free(pins, cursor->curr); else { @@ -171,7 +171,8 @@ static LF_SLIST *linsert(LF_SLIST * vola node->link= (intptr)cursor.curr; DBUG_ASSERT(node->link != (intptr)node); /* no circular references */ DBUG_ASSERT(cursor.prev != &node->link); /* no circular references */ - if (my_atomic_casptr((void **)cursor.prev, (void **)&cursor.curr, node)) + if (my_atomic_casptr((void **) cursor.prev, + (void **)(char*) &cursor.curr, node)) { res= 1; /* inserted ok */ break; @@ -218,13 +219,13 @@ static int ldelete(LF_SLIST * volatile * else { /* mark the node deleted */ - if (my_atomic_casptr((void **)&(cursor.curr->link), - (void **)&cursor.next, + if (my_atomic_casptr((void **) (char*) &(cursor.curr->link), + (void **) (char*) &cursor.next, (void *)(((intptr)cursor.next) | 1))) { /* and remove it from the list */ if (my_atomic_casptr((void **)cursor.prev, - (void **)&cursor.curr, cursor.next)) + (void **)(char*)&cursor.curr, cursor.next)) _lf_alloc_free(pins, cursor.curr); else { @@ -493,7 +494,7 @@ static int initialize_bucket(LF_HASH *ha my_free((void *)dummy, MYF(0)); dummy= cur; } - my_atomic_casptr((void **)node, (void **)&tmp, dummy); + my_atomic_casptr((void **)node, (void **)(char*) &tmp, dummy); /* note that if the CAS above failed (after linsert() succeeded), it would mean that some other thread has executed linsert() for === modified file 'mysys/my_redel.c' --- a/mysys/my_redel.c 2009-11-06 17:22:32 +0000 +++ b/mysys/my_redel.c 2009-11-29 23:08:56 +0000 @@ -77,9 +77,6 @@ end: int my_copystat(const char *from, const char *to, int MyFlags) { struct stat statbuf; -#if !defined(__WIN__) && !defined(__NETWARE__) - int res; -#endif if (stat((char*) from, &statbuf)) { === modified file 'regex/engine.c' --- a/regex/engine.c 2005-09-29 01:20:31 +0000 +++ b/regex/engine.c 2009-11-29 23:08:56 +0000 @@ -33,11 +33,11 @@ struct match { struct re_guts *g; int eflags; my_regmatch_t *pmatch; /* [nsub+1] (0 element unused) */ - char *offp; /* offsets work from here */ - char *beginp; /* start of string -- virtual NUL precedes */ - char *endp; /* end of string -- virtual NUL here */ - char *coldp; /* can be no match starting before here */ - char **lastpos; /* [nplus+1] */ + const char *offp; /* offsets work from here */ + const char *beginp; /* start of string -- virtual NUL precedes */ + const char *endp; /* end of string -- virtual NUL here */ + const char *coldp; /* can be no match starting before here */ + const char **lastpos; /* [nplus+1] */ STATEVARS; states st; /* current states */ states fresh; /* states for a fresh start */ @@ -66,20 +66,20 @@ static int /* 0 success, REG_NOMATCH f matcher(charset,g, str, nmatch, pmatch, eflags) CHARSET_INFO *charset; register struct re_guts *g; -char *str; +const char *str; size_t nmatch; my_regmatch_t pmatch[]; int eflags; { - register char *endp; + register const char *endp; register uint i; struct match mv; register struct match *m = &mv; - register char *dp; + register const char *dp; register const sopno gf = g->firststate+1; /* +1 for OEND */ register const sopno gl = g->laststate; - char *start; - char *stop; + const char *start; + const char *stop; /* simplify the situation where possible */ if (g->cflags&REG_NOSUB) @@ -163,7 +163,7 @@ int eflags; dp = dissect(charset, m, m->coldp, endp, gf, gl); } else { if (g->nplus > 0 && m->lastpos == NULL) - m->lastpos = (char **)malloc((g->nplus+1) * + m->lastpos = (const char **)malloc((g->nplus+1) * sizeof(char *)); if (g->nplus > 0 && m->lastpos == NULL) { free(m->pmatch); @@ -235,28 +235,28 @@ int eflags; == static char *dissect(register struct match *m, char *start, \ == char *stop, sopno startst, sopno stopst); */ -static char * /* == stop (success) always */ +static const char * /* == stop (success) always */ dissect(charset, m, start, stop, startst, stopst) CHARSET_INFO *charset; register struct match *m; -char *start; -char *stop; +const char *start; +const char *stop; sopno startst; sopno stopst; { register uint i; register sopno ss; /* start sop of current subRE */ register sopno es; /* end sop of current subRE */ - register char *sp; /* start of string matched by it */ - register char *stp; /* string matched by it cannot pass here */ - register char *rest; /* start of rest of string */ - register char *tail; /* string unmatched by rest of RE */ + register const char *sp; /* start of string matched by it */ + register const char *stp; /* string matched by it cannot pass here */ + register const char *rest; /* start of rest of string */ + register const char *tail; /* string unmatched by rest of RE */ register sopno ssub; /* start sop of subsubRE */ register sopno esub; /* end sop of subsubRE */ - register char *ssp; /* start of string matched by subsubRE */ - register char *sep; /* end of string matched by subsubRE */ - register char *oldssp; /* previous ssp */ - register char *dp; /* used in debug mode to check asserts */ + register const char *ssp; /* start of string matched by subsubRE */ + register const char *sep; /* end of string matched by subsubRE */ + register const char *oldssp; /* previous ssp */ + register const char *dp; /* used in debug mode to check asserts */ AT("diss", start, stop, startst, stopst); sp = start; @@ -424,23 +424,23 @@ sopno stopst; == static char *backref(register struct match *m, char *start, \ == char *stop, sopno startst, sopno stopst, sopno lev); */ -static char * /* == stop (success) or NULL (failure) */ +static const char * /* == stop (success) or NULL (failure) */ backref(charset,m, start, stop, startst, stopst, lev) CHARSET_INFO *charset; register struct match *m; -char *start; -char *stop; +const char *start; +const char *stop; sopno startst; sopno stopst; sopno lev; /* PLUS nesting level */ { register uint i; register sopno ss; /* start sop of current subRE */ - register char *sp; /* start of string matched by it */ + register const char *sp; /* start of string matched by it */ register sopno ssub; /* start sop of subsubRE */ register sopno esub; /* end sop of subsubRE */ - register char *ssp; /* start of string matched by subsubRE */ - register char *dp; + register const char *ssp; /* start of string matched by subsubRE */ + register const char *dp; register size_t len; register int hard; register sop s; @@ -630,24 +630,24 @@ sopno lev; /* PLUS nesting level */ == static char *fast(register struct match *m, char *start, \ == char *stop, sopno startst, sopno stopst); */ -static char * /* where tentative match ended, or NULL */ +static const char * /* where tentative match ended, or NULL */ fast(charset, m, start, stop, startst, stopst) CHARSET_INFO *charset; register struct match *m; -char *start; -char *stop; +const char *start; +const char *stop; sopno startst; sopno stopst; { register states st = m->st; register states fresh = m->fresh; register states tmp = m->tmp; - register char *p = start; + register const char *p = start; register int c = (start == m->beginp) ? OUT : *(start-1); register int lastc; /* previous c */ register int flagch; register int i; - register char *coldp; /* last p after which no match was underway */ + register const char *coldp; /* last p after which no match was underway */ CLEAR(st); SET1(st, startst); @@ -722,24 +722,24 @@ sopno stopst; == static char *slow(register struct match *m, char *start, \ == char *stop, sopno startst, sopno stopst); */ -static char * /* where it ended */ +static const char * /* where it ended */ slow(charset, m, start, stop, startst, stopst) CHARSET_INFO *charset; register struct match *m; -char *start; -char *stop; +const char *start; +const char *stop; sopno startst; sopno stopst; { register states st = m->st; register states empty = m->empty; register states tmp = m->tmp; - register char *p = start; + register const char *p = start; register int c = (start == m->beginp) ? OUT : *(start-1); register int lastc; /* previous c */ register int flagch; register int i; - register char *matchp; /* last p at which a match ended */ + register const char *matchp; /* last p at which a match ended */ AT("slow", start, stop, startst, stopst); CLEAR(st); === modified file 'regex/engine.ih' --- a/regex/engine.ih 2005-09-29 00:08:24 +0000 +++ b/regex/engine.ih 2009-11-29 23:08:56 +0000 @@ -4,11 +4,11 @@ extern "C" { #endif /* === engine.c === */ -static int matcher(CHARSET_INFO *charset,register struct re_guts *g, char *string, size_t nmatch, my_regmatch_t pmatch[], int eflags); -static char *dissect(CHARSET_INFO *charset,register struct match *m, char *start, char *stop, sopno startst, sopno stopst); -static char *backref(CHARSET_INFO *charset, register struct match *m, char *start, char *stop, sopno startst, sopno stopst, sopno lev); -static char *fast(CHARSET_INFO *charset, register struct match *m, char *start, char *stop, sopno startst, sopno stopst); -static char *slow(CHARSET_INFO *charset, register struct match *m, char *start, char *stop, sopno startst, sopno stopst); +static int matcher(CHARSET_INFO *charset,register struct re_guts *g, const char *string, size_t nmatch, my_regmatch_t pmatch[], int eflags); +static const char *dissect(CHARSET_INFO *charset,register struct match *m, const char *start, const char *stop, sopno startst, sopno stopst); +static const char *backref(CHARSET_INFO *charset, register struct match *m, const char *start, const char *stop, sopno startst, sopno stopst, sopno lev); +static const char *fast(CHARSET_INFO *charset, register struct match *m, const char *start, const char *stop, sopno startst, sopno stopst); +static const char *slow(CHARSET_INFO *charset, register struct match *m, const char *start, const char *stop, sopno startst, sopno stopst); static states step(register struct re_guts *g, sopno start, sopno stop, register states bef, int ch, register states aft); #define BOL (OUT+1) #define EOL (BOL+1) === modified file 'sql/sp_head.cc' --- a/sql/sp_head.cc 2009-09-15 10:46:35 +0000 +++ b/sql/sp_head.cc 2009-11-29 23:08:56 +0000 @@ -1924,9 +1924,10 @@ sp_head::execute_procedure(THD *thd, Lis if (spvar->mode == sp_param_out) { Item_null *null_item= new Item_null(); + Item *tmp_item= (Item*) null_item; if (!null_item || - nctx->set_variable(thd, i, (Item **)&null_item)) + nctx->set_variable(thd, i, &tmp_item)) { err_status= TRUE; break; === modified file 'sql/sql_base.cc' --- a/sql/sql_base.cc 2009-10-15 21:38:29 +0000 +++ b/sql/sql_base.cc 2009-11-29 23:08:56 +0000 @@ -5823,6 +5823,7 @@ find_field_in_natural_join(THD *thd, TAB { /* This is a base table. */ DBUG_ASSERT(nj_col->view_field == NULL); + Item *ref= 0; /* This fix_fields is not necessary (initially this item is fixed by the Item_field constructor; after reopen_tables the Item_func_eq @@ -5830,12 +5831,13 @@ find_field_in_natural_join(THD *thd, TAB reopening for columns that was dropped by the concurrent connection. */ if (!nj_col->table_field->fixed && - nj_col->table_field->fix_fields(thd, (Item **)&nj_col->table_field)) + nj_col->table_field->fix_fields(thd, &ref)) { DBUG_PRINT("info", ("column '%s' was dropped by the concurrent connection", nj_col->table_field->name)); DBUG_RETURN(NULL); } + DBUG_ASSERT(ref == 0); // Should not have changed DBUG_ASSERT(nj_col->table_ref->table == nj_col->table_field->field->table); found_field= nj_col->table_field->field; update_field_dependencies(thd, found_field, nj_col->table_ref->table); === modified file 'sql/sql_builtin.cc.in' --- a/sql/sql_builtin.cc.in 2006-12-31 01:29:11 +0000 +++ b/sql/sql_builtin.cc.in 2009-11-29 23:08:56 +0000 @@ -13,6 +13,7 @@ along with this program; if not, write to the Free Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA */ +#include <my_global.h> #include <mysql/plugin.h> typedef struct st_mysql_plugin builtin_plugin[]; === modified file 'sql/sql_parse.cc' --- a/sql/sql_parse.cc 2009-11-06 17:22:32 +0000 +++ b/sql/sql_parse.cc 2009-11-29 23:08:56 +0000 @@ -464,7 +464,7 @@ pthread_handler_t handle_bootstrap(void thd->init_for_queries(); while (fgets(buff, thd->net.max_packet, file)) { - char *query, *res; + char *query; /* strlen() can't be deleted because fgets() doesn't return length */ ulong length= (ulong) strlen(buff); while (buff[length-1] != '\n' && !feof(file)) === modified file 'sql/sql_select.cc' --- a/sql/sql_select.cc 2009-11-06 17:22:32 +0000 +++ b/sql/sql_select.cc 2009-11-29 23:08:56 +0000 @@ -3586,8 +3586,9 @@ add_key_fields(JOIN *join, KEY_FIELD **k { if (!field->eq(item->field)) { + Item *tmp_item= (Item*) item; add_key_field(key_fields, *and_level, cond_func, field, - TRUE, (Item **) &item, 1, usable_tables, + TRUE, &tmp_item, 1, usable_tables, sargables); } } @@ -15748,7 +15749,11 @@ static bool add_ref_to_table_cond(THD *t DBUG_RETURN(TRUE); if (!cond->fixed) - cond->fix_fields(thd, (Item**)&cond); + { + Item *tmp_item= (Item*) cond; + cond->fix_fields(thd, &tmp_item); + DBUG_ASSERT(cond == tmp_item); + } if (join_tab->select) { error=(int) cond->add(join_tab->select->cond); === modified file 'sql/sql_show.cc' --- a/sql/sql_show.cc 2009-11-26 20:19:33 +0000 +++ b/sql/sql_show.cc 2009-11-29 23:08:56 +0000 @@ -3551,7 +3551,9 @@ static int get_schema_tables_record(THD TABLE_SHARE *share= show_table->s; handler *file= show_table->file; handlerton *tmp_db_type= share->db_type(); +#ifdef WITH_PARTITION_STORAGE_ENGINE bool is_partitioned= FALSE; +#endif if (share->tmp_table == SYSTEM_TMP_TABLE) table->field[3]->store(STRING_WITH_LEN("SYSTEM VIEW"), cs); else if (share->tmp_table) === modified file 'sql/table.cc' --- a/sql/table.cc 2009-10-15 21:38:29 +0000 +++ b/sql/table.cc 2009-11-29 23:08:56 +0000 @@ -2077,8 +2077,9 @@ ulong get_form_pos(File file, uchar *hea else { char *str; + const char **tmp = (const char**) (char*) buf; str=(char *) (buf+a_length); - fix_type_pointers((const char ***) &buf,save_names,1,&str); + fix_type_pointers(&tmp, save_names, 1, &str); } DBUG_RETURN(ret_value); } === modified file 'storage/maria/ha_maria.cc' --- a/storage/maria/ha_maria.cc 2009-11-07 10:34:19 +0000 +++ b/storage/maria/ha_maria.cc 2009-11-29 23:08:56 +0000 @@ -665,10 +665,9 @@ int maria_check_definition(MARIA_KEYDEF extern "C" { -volatile int *_ma_killed_ptr(HA_CHECK *param) +int _ma_killed_ptr(HA_CHECK *param) { - /* In theory Unsafe conversion, but should be ok for now */ - return (int*) &(((THD *) (param->thd))->killed); + return thd_killed((THD*)param->thd); } === modified file 'storage/maria/lockman.c' --- a/storage/maria/lockman.c 2008-02-21 00:51:51 +0000 +++ b/storage/maria/lockman.c 2009-11-29 23:08:56 +0000 @@ -360,7 +360,7 @@ retry: else { if (my_atomic_casptr((void **)cursor->prev, - (void **)&cursor->curr, cursor->next)) + (void **)(char*) &cursor->curr, cursor->next)) _lf_alloc_free(pins, cursor->curr); else { @@ -421,7 +421,8 @@ static int lockinsert(LOCK * volatile *h node->link= (intptr)cursor.curr; DBUG_ASSERT(node->link != (intptr)node); DBUG_ASSERT(cursor.prev != &node->link); - if (!my_atomic_casptr((void **)cursor.prev, (void **)&cursor.curr, node)) + if (!my_atomic_casptr((void **)cursor.prev, + (void **)(char*) &cursor.curr, node)) { res= REPEAT_ONCE_MORE; node->flags&= ~ACTIVE; @@ -498,11 +499,11 @@ static int lockdelete(LOCK * volatile *h then we can delete. Good news is - this is only required when rolling back a savepoint. */ - if (my_atomic_casptr((void **)&(cursor.curr->link), - (void **)&cursor.next, 1+(char *)cursor.next)) + if (my_atomic_casptr((void **)(char*)&(cursor.curr->link), + (void **)(char*)&cursor.next, 1+(char *)cursor.next)) { if (my_atomic_casptr((void **)cursor.prev, - (void **)&cursor.curr, cursor.next)) + (void **)(char*)&cursor.curr, cursor.next)) _lf_alloc_free(pins, cursor.curr); else lockfind(head, node, &cursor, pins); @@ -573,7 +574,7 @@ static void initialize_bucket(LOCKMAN *l my_free((void *)dummy, MYF(0)); dummy= cur; } - my_atomic_casptr((void **)node, (void **)&tmp, dummy); + my_atomic_casptr((void **)node, (void **)(char*) &tmp, dummy); } static inline uint calc_hash(uint64 resource) === modified file 'storage/maria/ma_check.c' --- a/storage/maria/ma_check.c 2009-05-06 12:03:24 +0000 +++ b/storage/maria/ma_check.c 2009-11-29 23:08:56 +0000 @@ -215,7 +215,7 @@ int maria_chk_del(HA_CHECK *param, regis empty=0; for (i= share->state.state.del ; i > 0L && next_link != HA_OFFSET_ERROR ; i--) { - if (*_ma_killed_ptr(param)) + if (_ma_killed_ptr(param)) DBUG_RETURN(1); if (test_flag & T_VERBOSE) printf(" %9s",llstr(next_link,buff)); @@ -310,7 +310,7 @@ static int check_k_link(HA_CHECK *param, records= (ha_rows) (share->state.state.key_file_length / block_size); while (next_link != HA_OFFSET_ERROR && records > 0) { - if (*_ma_killed_ptr(param)) + if (_ma_killed_ptr(param)) DBUG_RETURN(1); if (param->testflag & T_VERBOSE) printf("%16s",llstr(next_link,llbuff)); @@ -876,10 +876,10 @@ static int chk_index(HA_CHECK *param, MA tmp_key.data= tmp_key_buff; for ( ;; ) { - if (*_ma_killed_ptr(param)) - goto err; if (nod_flag) { + if (_ma_killed_ptr(param)) + goto err; next_page= _ma_kpos(nod_flag,keypos); if (chk_index_down(param,info,keyinfo,next_page, temp_buff,keys,key_checksum,level+1)) @@ -1180,7 +1180,7 @@ static int check_static_record(HA_CHECK pos= 0; while (pos < share->state.state.data_file_length) { - if (*_ma_killed_ptr(param)) + if (_ma_killed_ptr(param)) return -1; if (my_b_read(&param->read_cache, record, share->base.pack_reclength)) @@ -1230,7 +1230,7 @@ static int check_dynamic_record(HA_CHECK { my_bool got_error= 0; int flag; - if (*_ma_killed_ptr(param)) + if (_ma_killed_ptr(param)) DBUG_RETURN(-1); flag= block_info.second_read=0; @@ -1451,7 +1451,7 @@ static int check_compressed_record(HA_CH pos= share->pack.header_length; /* Skip header */ while (pos < share->state.state.data_file_length) { - if (*_ma_killed_ptr(param)) + if (_ma_killed_ptr(param)) DBUG_RETURN(-1); if (_ma_read_cache(&param->read_cache, block_info.header, pos, @@ -1815,7 +1815,7 @@ static int check_block_record(HA_CHECK * LINT_INIT(row_count); LINT_INIT(empty_space); - if (*_ma_killed_ptr(param)) + if (_ma_killed_ptr(param)) { _ma_scan_end_block_record(info); return -1; @@ -4631,7 +4631,7 @@ static int sort_get_next_record(MARIA_SO char llbuff[22],llbuff2[22]; DBUG_ENTER("sort_get_next_record"); - if (*_ma_killed_ptr(param)) + if (_ma_killed_ptr(param)) DBUG_RETURN(1); switch (sort_info->org_data_file_type) { === modified file 'storage/maria/ma_check_standalone.h' --- a/storage/maria/ma_check_standalone.h 2007-10-03 16:10:32 +0000 +++ b/storage/maria/ma_check_standalone.h 2009-11-29 23:08:56 +0000 @@ -30,11 +30,9 @@ Check if check/repair operation was killed by a signal */ -static int not_killed= 0; - -volatile int *_ma_killed_ptr(HA_CHECK *param __attribute__((unused))) +int _ma_killed_ptr(HA_CHECK *param __attribute__((unused))) { - return &not_killed; /* always NULL */ + return 0; } /* print warnings and errors */ === modified file 'storage/maria/ma_ft_boolean_search.c' --- a/storage/maria/ma_ft_boolean_search.c 2009-02-19 09:01:25 +0000 +++ b/storage/maria/ma_ft_boolean_search.c 2009-11-29 23:08:56 +0000 @@ -180,7 +180,7 @@ typedef struct st_my_ftb_param static int ftb_query_add_word(MYSQL_FTPARSER_PARAM *param, - char *word, int word_len, + const uchar *word, size_t word_len, MYSQL_FTPARSER_BOOLEAN_INFO *info) { MY_FTB_PARAM *ftb_param= param->mysql_ftparam; @@ -282,19 +282,19 @@ static int ftb_query_add_word(MYSQL_FTPA static int ftb_parse_query_internal(MYSQL_FTPARSER_PARAM *param, - char *query, int len) + const uchar *query, size_t len) { MY_FTB_PARAM *ftb_param= param->mysql_ftparam; MYSQL_FTPARSER_BOOLEAN_INFO info; CHARSET_INFO *cs= ftb_param->ftb->charset; - uchar **start= (uchar**) &query; - uchar *end= (uchar*) query + len; + const uchar **start= &query; + const uchar *end= query + len; FT_WORD w; info.prev= ' '; info.quot= 0; while (maria_ft_get_word(cs, start, end, &w, &info)) - param->mysql_add_word(param, (char *) w.pos, w.len, &info); + param->mysql_add_word(param, w.pos, w.len, &info); return(0); } @@ -615,7 +615,7 @@ typedef struct st_my_ftb_phrase_param static int ftb_phrase_add_word(MYSQL_FTPARSER_PARAM *param, - char *word, int word_len, + const uchar *word, size_t word_len, MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info __attribute__((unused))) { MY_FTB_PHRASE_PARAM *phrase_param= param->mysql_ftparam; @@ -647,15 +647,15 @@ static int ftb_phrase_add_word(MYSQL_FTP static int ftb_check_phrase_internal(MYSQL_FTPARSER_PARAM *param, - char *document, int len) + const uchar *document, size_t len) { FT_WORD word; MY_FTB_PHRASE_PARAM *phrase_param= param->mysql_ftparam; - const uchar *docend= (uchar*) document + len; - while (maria_ft_simple_get_word(phrase_param->cs, (uchar**) &document, + const uchar *docend= document + len; + while (maria_ft_simple_get_word(phrase_param->cs, &document, docend, &word, FALSE)) { - param->mysql_add_word(param, (char*) word.pos, word.len, 0); + param->mysql_add_word(param, word.pos, word.len, 0); if (phrase_param->match) break; } @@ -872,7 +872,7 @@ typedef struct st_my_ftb_find_param static int ftb_find_relevance_add_word(MYSQL_FTPARSER_PARAM *param, - char *word, int len, + const uchar *word, size_t len, MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info __attribute__((unused))) { MY_FTB_FIND_PARAM *ftb_param= param->mysql_ftparam; @@ -933,15 +933,14 @@ static int ftb_find_relevance_add_word(M static int ftb_find_relevance_parse(MYSQL_FTPARSER_PARAM *param, - char *doc, int len) + const uchar *doc, size_t len) { MY_FTB_FIND_PARAM *ftb_param= param->mysql_ftparam; FT_INFO *ftb= ftb_param->ftb; - uchar *end= (uchar*) doc + len; + const uchar *end= doc + len; FT_WORD w; - while (maria_ft_simple_get_word(ftb->charset, (uchar**) &doc, - end, &w, TRUE)) - param->mysql_add_word(param, (char *) w.pos, w.len, 0); + while (maria_ft_simple_get_word(ftb->charset, &doc, end, &w, TRUE)) + param->mysql_add_word(param, w.pos, w.len, 0); return(0); } === modified file 'storage/maria/ma_ft_nlq_search.c' --- a/storage/maria/ma_ft_nlq_search.c 2009-01-09 04:23:25 +0000 +++ b/storage/maria/ma_ft_nlq_search.c 2009-11-29 23:08:56 +0000 @@ -63,7 +63,8 @@ static int FT_SUPERDOC_cmp(void* cmp_arg static int walk_and_match(FT_WORD *word, uint32 count, ALL_IN_ONE *aio) { - int subkeys, r; + int32 subkeys; + int r; uint doc_cnt; FT_SUPERDOC sdoc, *sptr; TREE_ELEMENT *selem; @@ -127,7 +128,7 @@ static int walk_and_match(FT_WORD *word, goto do_skip; } #if HA_FT_WTYPE == HA_KEYTYPE_FLOAT - tmp_weight=*(float*)&subkeys; + tmp_weight=*(float*) (char*) &subkeys; #else #error #endif === modified file 'storage/maria/ma_ft_parser.c' --- a/storage/maria/ma_ft_parser.c 2009-02-19 09:01:25 +0000 +++ b/storage/maria/ma_ft_parser.c 2009-11-29 23:08:56 +0000 @@ -109,10 +109,11 @@ my_bool maria_ft_boolean_check_syntax_st 3 - right bracket 4 - stopword found */ -uchar maria_ft_get_word(CHARSET_INFO *cs, uchar **start, uchar *end, +uchar maria_ft_get_word(CHARSET_INFO *cs, const uchar **start, + const uchar *end, FT_WORD *word, MYSQL_FTPARSER_BOOLEAN_INFO *param) { - uchar *doc=*start; + const uchar *doc= *start; int ctype; uint mwc, length; int mbl; @@ -203,11 +204,11 @@ ret: return param->type; } -uchar maria_ft_simple_get_word(CHARSET_INFO *cs, uchar **start, +uchar maria_ft_simple_get_word(CHARSET_INFO *cs, const uchar **start, const uchar *end, FT_WORD *word, my_bool skip_stopwords) { - uchar *doc= *start; + const uchar *doc= *start; uint mwc, length; int ctype, mbl; DBUG_ENTER("maria_ft_simple_get_word"); @@ -259,8 +260,9 @@ void maria_ft_parse_init(TREE *wtree, CH static int maria_ft_add_word(MYSQL_FTPARSER_PARAM *param, - char *word, int word_len, - MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info __attribute__((unused))) + const uchar *word, size_t word_len, + MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info + __attribute__((unused))) { TREE *wtree; FT_WORD w; @@ -276,7 +278,7 @@ static int maria_ft_add_word(MYSQL_FTPAR w.pos= ptr; } else - w.pos= (uchar *) word; + w.pos= word; w.len= word_len; if (!tree_insert(wtree, &w, 0, wtree->custom_arg)) { @@ -288,17 +290,17 @@ static int maria_ft_add_word(MYSQL_FTPAR static int maria_ft_parse_internal(MYSQL_FTPARSER_PARAM *param, - char *doc_arg, int doc_len) + const uchar *doc_arg, size_t doc_len) { - uchar *doc= (uchar*) doc_arg; - uchar *end= doc + doc_len; + const uchar *doc= doc_arg; + const uchar *end= doc + doc_len; MY_FT_PARSER_PARAM *ft_param=param->mysql_ftparam; TREE *wtree= ft_param->wtree; FT_WORD w; DBUG_ENTER("maria_ft_parse_internal"); while (maria_ft_simple_get_word(wtree->custom_arg, &doc, end, &w, TRUE)) - if (param->mysql_add_word(param, (char *) w.pos, w.len, 0)) + if (param->mysql_add_word(param, w.pos, w.len, 0)) DBUG_RETURN(1); DBUG_RETURN(0); } @@ -378,7 +380,7 @@ MYSQL_FTPARSER_PARAM *maria_ftparser_cal mysql_add_word != 0 - parser is initialized, or no initialization needed. */ info->ftparser_param[ftparser_nr].mysql_add_word= - (int (*)(struct st_mysql_ftparser_param *, char *, int, + (int (*)(struct st_mysql_ftparser_param *, const uchar *, size_t, MYSQL_FTPARSER_BOOLEAN_INFO *)) 1; if (parser->init && parser->init(&info->ftparser_param[ftparser_nr])) return 0; === modified file 'storage/maria/ma_ftdefs.h' --- a/storage/maria/ma_ftdefs.h 2009-02-12 17:51:00 +0000 +++ b/storage/maria/ma_ftdefs.h 2009-11-29 23:08:56 +0000 @@ -96,7 +96,7 @@ #define FTB_RQUOT (ft_boolean_syntax[11]) typedef struct st_maria_ft_word { - uchar * pos; + const uchar * pos; uint len; double weight; } FT_WORD; @@ -106,9 +106,9 @@ int is_stopword(char *word, uint len); MARIA_KEY *_ma_ft_make_key(MARIA_HA *, MARIA_KEY *, uint , uchar *, FT_WORD *, my_off_t); -uchar maria_ft_get_word(CHARSET_INFO *, uchar **, uchar *, FT_WORD *, - MYSQL_FTPARSER_BOOLEAN_INFO *); -uchar maria_ft_simple_get_word(CHARSET_INFO *, uchar **, const uchar *, +uchar maria_ft_get_word(CHARSET_INFO *, const uchar **, const uchar *, + FT_WORD *, MYSQL_FTPARSER_BOOLEAN_INFO *); +uchar maria_ft_simple_get_word(CHARSET_INFO *, const uchar **, const uchar *, FT_WORD *, my_bool); typedef struct _st_maria_ft_seg_iterator { === modified file 'storage/maria/ma_sort.c' --- a/storage/maria/ma_sort.c 2009-01-09 04:23:25 +0000 +++ b/storage/maria/ma_sort.c 2009-11-29 23:08:56 +0000 @@ -920,7 +920,6 @@ merge_buffers(MARIA_SORT_PARAM *info, ui uchar *strpos; BUFFPEK *buffpek,**refpek; QUEUE queue; - volatile int *killed= _ma_killed_ptr(info->sort_info->param); DBUG_ENTER("merge_buffers"); count=error=0; @@ -953,10 +952,6 @@ merge_buffers(MARIA_SORT_PARAM *info, ui { for (;;) { - if (*killed) - { - error=1; goto err; - } buffpek=(BUFFPEK*) queue_top(&queue); if (to_file) { @@ -976,6 +971,12 @@ merge_buffers(MARIA_SORT_PARAM *info, ui buffpek->key+=sort_length; if (! --buffpek->mem_count) { + /* It's enough to check for killedptr before a slow operation */ + if (_ma_killed_ptr(info->sort_info->param)) + { + error=1; + goto err; + } if (!(error=(int) info->read_to_buffer(from_file,buffpek,sort_length))) { uchar *base= buffpek->base; === modified file 'storage/maria/ma_state.c' --- a/storage/maria/ma_state.c 2009-10-06 06:57:22 +0000 +++ b/storage/maria/ma_state.c 2009-11-29 23:08:56 +0000 @@ -528,7 +528,7 @@ void _ma_remove_table_from_trnman(MARIA_ safe_mutex_assert_owner(&share->intern_lock); - for (prev= (MARIA_USED_TABLES**) &trn->used_tables, tables= *prev; + for (prev= (MARIA_USED_TABLES**) (char*) &trn->used_tables, tables= *prev; tables; tables= *prev) { === modified file 'storage/maria/maria_def.h' --- a/storage/maria/maria_def.h 2009-10-06 06:57:22 +0000 +++ b/storage/maria/maria_def.h 2009-11-29 23:08:56 +0000 @@ -1160,7 +1160,7 @@ int _ma_flush_table_files(MARIA_HA *info Functions needed by _ma_check (are overridden in MySQL/ha_maria.cc). See ma_check_standalone.h . */ -volatile int *_ma_killed_ptr(HA_CHECK *param); +int _ma_killed_ptr(HA_CHECK *param); void _ma_check_print_error _VARARGS((HA_CHECK *param, const char *fmt, ...)) ATTRIBUTE_FORMAT(printf, 2, 3); void _ma_check_print_warning _VARARGS((HA_CHECK *param, const char *fmt, ...)) === modified file 'storage/maria/maria_ftdump.c' --- a/storage/maria/maria_ftdump.c 2008-08-25 11:49:47 +0000 +++ b/storage/maria/maria_ftdump.c 2009-11-29 23:08:56 +0000 @@ -116,7 +116,7 @@ int main(int argc,char *argv[]) subkeys=ft_sintXkorr(info->lastkey_buff + keylen + 1); if (subkeys >= 0) - weight=*(float*)&subkeys; + weight=*(float*) (char*) &subkeys; #ifdef HAVE_SNPRINTF snprintf(buf,MAX_LEN,"%.*s",(int) keylen,info->lastkey_buff+1); === modified file 'storage/maria/trnman.c' --- a/storage/maria/trnman.c 2009-03-05 13:46:45 +0000 +++ b/storage/maria/trnman.c 2009-11-29 23:08:56 +0000 @@ -300,8 +300,8 @@ TRN *trnman_new_trn(WT_THD *wt) (ABA isn't possible, we're behind a mutex */ my_atomic_rwlock_wrlock(&LOCK_pool); - while (tmp.trn && !my_atomic_casptr((void **)&pool, &tmp.v, - (void *)tmp.trn->next)) + while (tmp.trn && !my_atomic_casptr((void **)(char*) &pool, &tmp.v, + (void *)tmp.trn->next)) /* no-op */; my_atomic_rwlock_wrunlock(&LOCK_pool); @@ -545,7 +545,7 @@ static void trnman_free_trn(TRN *trn) down after the loop at -O2 */ *(TRN * volatile *)&(trn->next)= tmp.trn; - } while (!my_atomic_casptr((void **)&pool, &tmp.v, trn)); + } while (!my_atomic_casptr((void **)(char*)&pool, &tmp.v, trn)); my_atomic_rwlock_wrunlock(&LOCK_pool); } === modified file 'storage/myisam/ft_boolean_search.c' --- a/storage/myisam/ft_boolean_search.c 2009-10-15 21:38:29 +0000 +++ b/storage/myisam/ft_boolean_search.c 2009-11-29 23:08:56 +0000 @@ -180,7 +180,7 @@ typedef struct st_my_ftb_param static int ftb_query_add_word(MYSQL_FTPARSER_PARAM *param, - char *word, int word_len, + const uchar *word, size_t word_len, MYSQL_FTPARSER_BOOLEAN_INFO *info) { MY_FTB_PARAM *ftb_param= param->mysql_ftparam; @@ -282,19 +282,19 @@ static int ftb_query_add_word(MYSQL_FTPA static int ftb_parse_query_internal(MYSQL_FTPARSER_PARAM *param, - char *query, int len) + const uchar *query, size_t len) { MY_FTB_PARAM *ftb_param= param->mysql_ftparam; MYSQL_FTPARSER_BOOLEAN_INFO info; CHARSET_INFO *cs= ftb_param->ftb->charset; - uchar **start= (uchar**) &query; - uchar *end= (uchar*) query + len; + const uchar **start= &query; + const uchar *end= query + len; FT_WORD w; info.prev= ' '; info.quot= 0; while (ft_get_word(cs, start, end, &w, &info)) - param->mysql_add_word(param, (char*) w.pos, w.len, &info); + param->mysql_add_word(param, w.pos, w.len, &info); return(0); } @@ -616,7 +616,7 @@ typedef struct st_my_ftb_phrase_param static int ftb_phrase_add_word(MYSQL_FTPARSER_PARAM *param, - char *word, int word_len, + const uchar *word, size_t word_len, MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info __attribute__((unused))) { MY_FTB_PHRASE_PARAM *phrase_param= param->mysql_ftparam; @@ -648,15 +648,15 @@ static int ftb_phrase_add_word(MYSQL_FTP static int ftb_check_phrase_internal(MYSQL_FTPARSER_PARAM *param, - char *document, int len) + const uchar *document, size_t len) { FT_WORD word; MY_FTB_PHRASE_PARAM *phrase_param= param->mysql_ftparam; const uchar *docend= (uchar*) document + len; - while (ft_simple_get_word(phrase_param->cs, (uchar**) &document, docend, + while (ft_simple_get_word(phrase_param->cs, &document, docend, &word, FALSE)) { - param->mysql_add_word(param, (char*) word.pos, word.len, 0); + param->mysql_add_word(param, word.pos, word.len, 0); if (phrase_param->match) break; } @@ -874,7 +874,7 @@ typedef struct st_my_ftb_find_param static int ftb_find_relevance_add_word(MYSQL_FTPARSER_PARAM *param, - char *word, int len, + const uchar *word, size_t len, MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info __attribute__((unused))) { MY_FTB_FIND_PARAM *ftb_param= param->mysql_ftparam; @@ -888,8 +888,8 @@ static int ftb_find_relevance_add_word(M for (a= 0, b= ftb->queue.elements, c= (a+b)/2; b-a>1; c= (a+b)/2) { ftbw= ftb->list[c]; - if (ha_compare_text(ftb->charset, (uchar*)word, len, - (uchar*)ftbw->word+1, ftbw->len-1, + if (ha_compare_text(ftb->charset, word, len, + ftbw->word+1, ftbw->len-1, (my_bool) (ftbw->flags & FTB_FLAG_TRUNC), 0) < 0) b= c; else @@ -915,8 +915,8 @@ static int ftb_find_relevance_add_word(M for (; c >= 0; c--) { ftbw= ftb->list[c]; - if (ha_compare_text(ftb->charset, (uchar*)word, len, - (uchar*)ftbw->word + 1,ftbw->len - 1, + if (ha_compare_text(ftb->charset, word, len, + ftbw->word + 1,ftbw->len - 1, (my_bool)(ftbw->flags & FTB_FLAG_TRUNC), 0)) { if (ftb->with_scan & FTB_FLAG_TRUNC) @@ -935,14 +935,14 @@ static int ftb_find_relevance_add_word(M static int ftb_find_relevance_parse(MYSQL_FTPARSER_PARAM *param, - char *doc, int len) + const uchar *doc, size_t len) { MY_FTB_FIND_PARAM *ftb_param= param->mysql_ftparam; FT_INFO *ftb= ftb_param->ftb; - uchar *end= (uchar*) doc + len; + const uchar *end= doc + len; FT_WORD w; - while (ft_simple_get_word(ftb->charset, (uchar**) &doc, end, &w, TRUE)) - param->mysql_add_word(param, (char*) w.pos, w.len, 0); + while (ft_simple_get_word(ftb->charset, &doc, end, &w, TRUE)) + param->mysql_add_word(param, w.pos, w.len, 0); return(0); } === modified file 'storage/myisam/ft_nlq_search.c' --- a/storage/myisam/ft_nlq_search.c 2008-04-28 16:24:05 +0000 +++ b/storage/myisam/ft_nlq_search.c 2009-11-29 23:08:56 +0000 @@ -63,7 +63,8 @@ static int FT_SUPERDOC_cmp(void* cmp_arg static int walk_and_match(FT_WORD *word, uint32 count, ALL_IN_ONE *aio) { - int subkeys, r; + int32 subkeys; + int r; uint keylen, doc_cnt; FT_SUPERDOC sdoc, *sptr; TREE_ELEMENT *selem; @@ -123,7 +124,8 @@ static int walk_and_match(FT_WORD *word, goto do_skip; } #if HA_FT_WTYPE == HA_KEYTYPE_FLOAT - tmp_weight=*(float*)&subkeys; + /* The weight we read was actually a float */ + tmp_weight=*(float*) (char*) &subkeys; #else #error #endif === modified file 'storage/myisam/ft_parser.c' --- a/storage/myisam/ft_parser.c 2009-02-12 14:08:56 +0000 +++ b/storage/myisam/ft_parser.c 2009-11-29 23:08:56 +0000 @@ -106,10 +106,10 @@ my_bool ft_boolean_check_syntax_string(c 3 - right bracket 4 - stopword found */ -uchar ft_get_word(CHARSET_INFO *cs, uchar **start, uchar *end, +uchar ft_get_word(CHARSET_INFO *cs, const uchar **start, const uchar *end, FT_WORD *word, MYSQL_FTPARSER_BOOLEAN_INFO *param) { - uchar *doc=*start; + const uchar *doc= *start; int ctype; uint mwc, length; int mbl; @@ -201,10 +201,11 @@ ret: return param->type; } -uchar ft_simple_get_word(CHARSET_INFO *cs, uchar **start, const uchar *end, - FT_WORD *word, my_bool skip_stopwords) +uchar ft_simple_get_word(CHARSET_INFO *cs, const uchar **start, + const uchar *end, FT_WORD *word, + my_bool skip_stopwords) { - uchar *doc= *start; + const uchar *doc= *start; uint mwc, length; int mbl; int ctype; @@ -216,7 +217,7 @@ uchar ft_simple_get_word(CHARSET_INFO *c { if (doc >= end) DBUG_RETURN(0); - mbl= cs->cset->ctype(cs, &ctype, (uchar*)doc, (uchar*)end); + mbl= cs->cset->ctype(cs, &ctype, doc, end); if (true_word_char(ctype, *doc)) break; } @@ -225,7 +226,7 @@ uchar ft_simple_get_word(CHARSET_INFO *c for (word->pos= doc; doc < end; length++, doc+= (mbl > 0 ? mbl : (mbl < 0 ? -mbl : 1))) { - mbl= cs->cset->ctype(cs, &ctype, (uchar*)doc, (uchar*)end); + mbl= cs->cset->ctype(cs, &ctype, doc, end); if (true_word_char(ctype, *doc)) mwc= 0; else if (!misc_word_char(*doc) || mwc) @@ -238,7 +239,7 @@ uchar ft_simple_get_word(CHARSET_INFO *c if (skip_stopwords == FALSE || (length >= ft_min_word_len && length < ft_max_word_len && - !is_stopword((char*) word->pos, word->len))) + !is_stopword(word->pos, word->len))) { *start= doc; DBUG_RETURN(1); @@ -257,8 +258,9 @@ void ft_parse_init(TREE *wtree, CHARSET_ static int ft_add_word(MYSQL_FTPARSER_PARAM *param, - char *word, int word_len, - MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info __attribute__((unused))) + const uchar *word, size_t word_len, + MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info + __attribute__((unused))) { TREE *wtree; FT_WORD w; @@ -286,10 +288,10 @@ static int ft_add_word(MYSQL_FTPARSER_PA static int ft_parse_internal(MYSQL_FTPARSER_PARAM *param, - char *doc_arg, int doc_len) + const uchar *doc_arg, size_t doc_len) { - uchar *doc= (uchar*) doc_arg; - uchar *end= doc + doc_len; + const uchar *doc= doc_arg; + const uchar *end= doc + doc_len; MY_FT_PARSER_PARAM *ft_param=param->mysql_ftparam; TREE *wtree= ft_param->wtree; FT_WORD w; @@ -302,7 +304,7 @@ static int ft_parse_internal(MYSQL_FTPAR } -int ft_parse(TREE *wtree, uchar *doc, int doclen, +int ft_parse(TREE *wtree, const uchar *doc, int doclen, struct st_mysql_ftparser *parser, MYSQL_FTPARSER_PARAM *param, MEM_ROOT *mem_root) { @@ -377,7 +379,7 @@ MYSQL_FTPARSER_PARAM *ftparser_call_init mysql_add_word != 0 - parser is initialized, or no initialization needed. */ info->ftparser_param[ftparser_nr].mysql_add_word= - (int (*)(struct st_mysql_ftparser_param *, char *, int, + (int (*)(struct st_mysql_ftparser_param *, const uchar *, size_t, MYSQL_FTPARSER_BOOLEAN_INFO *)) 1; if (parser->init && parser->init(&info->ftparser_param[ftparser_nr])) return 0; === modified file 'storage/myisam/ft_stopwords.c' --- a/storage/myisam/ft_stopwords.c 2008-04-28 16:24:05 +0000 +++ b/storage/myisam/ft_stopwords.c 2009-11-29 23:08:56 +0000 @@ -66,7 +66,8 @@ int ft_init_stopwords() { File fd; uint len; - uchar *buffer, *start, *end; + uchar *buffer; + const uchar *start, *end; FT_WORD w; int error=-1; @@ -109,7 +110,7 @@ err0: } -int is_stopword(char *word, uint len) +int is_stopword(const uchar *word, size_t len) { FT_STOPWORD sw; sw.pos=word; === modified file 'storage/myisam/ftdefs.h' --- a/storage/myisam/ftdefs.h 2009-01-26 06:35:15 +0000 +++ b/storage/myisam/ftdefs.h 2009-11-29 23:08:56 +0000 @@ -96,18 +96,18 @@ #define FTB_RQUOT (ft_boolean_syntax[11]) typedef struct st_ft_word { - uchar * pos; + const uchar *pos; uint len; double weight; } FT_WORD; -int is_stopword(char *word, uint len); +int is_stopword(const uchar *word, size_t len); uint _ft_make_key(MI_INFO *, uint , uchar *, FT_WORD *, my_off_t); -uchar ft_get_word(CHARSET_INFO *, uchar **, uchar *, FT_WORD *, +uchar ft_get_word(CHARSET_INFO *, const uchar **, const uchar *, FT_WORD *, MYSQL_FTPARSER_BOOLEAN_INFO *); -uchar ft_simple_get_word(CHARSET_INFO *, uchar **, const uchar *, +uchar ft_simple_get_word(CHARSET_INFO *, const uchar **, const uchar *, FT_WORD *, my_bool); typedef struct _st_ft_seg_iterator { @@ -121,7 +121,7 @@ void _mi_ft_segiterator_dummy_init(const uint _mi_ft_segiterator(FT_SEG_ITERATOR *); void ft_parse_init(TREE *, CHARSET_INFO *); -int ft_parse(TREE *, uchar *, int, struct st_mysql_ftparser *parser, +int ft_parse(TREE *, const uchar *, int, struct st_mysql_ftparser *parser, MYSQL_FTPARSER_PARAM *, MEM_ROOT *); FT_WORD * ft_linearize(TREE *, MEM_ROOT *); FT_WORD * _mi_ft_parserecord(MI_INFO *, uint, const uchar *, MEM_ROOT *); === modified file 'storage/myisam/ha_myisam.cc' --- a/storage/myisam/ha_myisam.cc 2009-10-06 14:53:46 +0000 +++ b/storage/myisam/ha_myisam.cc 2009-11-29 23:08:56 +0000 @@ -497,10 +497,9 @@ int check_definition(MI_KEYDEF *t1_keyin extern "C" { -volatile int *killed_ptr(HA_CHECK *param) +int killed_ptr(HA_CHECK *param) { - /* In theory Unsafe conversion, but should be ok for now */ - return (int*) &(((THD *)(param->thd))->killed); + return thd_killed((THD*)param->thd); } void mi_check_print_error(HA_CHECK *param, const char *fmt,...) === modified file 'storage/myisam/mi_check.c' --- a/storage/myisam/mi_check.c 2009-10-15 21:38:29 +0000 +++ b/storage/myisam/mi_check.c 2009-11-29 23:08:56 +0000 @@ -166,7 +166,7 @@ int chk_del(HA_CHECK *param, register MI empty=0; for (i= info->state->del ; i > 0L && next_link != HA_OFFSET_ERROR ; i--) { - if (*killed_ptr(param)) + if (killed_ptr(param)) DBUG_RETURN(1); if (test_flag & T_VERBOSE) printf(" %9s",llstr(next_link,buff)); @@ -261,7 +261,7 @@ static int check_k_link(HA_CHECK *param, records= (ha_rows) (info->state->key_file_length / block_size); while (next_link != HA_OFFSET_ERROR && records > 0) { - if (*killed_ptr(param)) + if (killed_ptr(param)) DBUG_RETURN(1); if (param->testflag & T_VERBOSE) printf("%16s",llstr(next_link,llbuff)); @@ -778,7 +778,7 @@ static int chk_index(HA_CHECK *param, MI } for ( ;; ) { - if (*killed_ptr(param)) + if (killed_ptr(param)) goto err; memcpy((char*) info->lastkey,(char*) key,key_length); info->lastkey_length=key_length; @@ -990,7 +990,7 @@ int chk_data_link(HA_CHECK *param, MI_IN bzero((char*) key_checksum, info->s->base.keys * sizeof(key_checksum[0])); while (pos < info->state->data_file_length) { - if (*killed_ptr(param)) + if (killed_ptr(param)) goto err2; switch (info->s->data_file_type) { case BLOCK_RECORD: @@ -3247,7 +3247,7 @@ static int sort_get_next_record(MI_SORT_ char llbuff[22],llbuff2[22]; DBUG_ENTER("sort_get_next_record"); - if (*killed_ptr(param)) + if (killed_ptr(param)) DBUG_RETURN(1); switch (share->data_file_type) { === modified file 'storage/myisam/myisam_ftdump.c' --- a/storage/myisam/myisam_ftdump.c 2007-05-10 09:59:39 +0000 +++ b/storage/myisam/myisam_ftdump.c 2009-11-29 23:08:56 +0000 @@ -113,7 +113,7 @@ int main(int argc,char *argv[]) subkeys=ft_sintXkorr(info->lastkey+keylen+1); if (subkeys >= 0) - weight=*(float*)&subkeys; + weight= *(float*) (char*) &subkeys; #ifdef HAVE_SNPRINTF snprintf(buf,MAX_LEN,"%.*s",(int) keylen,info->lastkey+1); === modified file 'storage/myisam/myisamchk.c' --- a/storage/myisam/myisamchk.c 2009-09-19 21:21:29 +0000 +++ b/storage/myisam/myisamchk.c 2009-11-29 23:08:56 +0000 @@ -1745,11 +1745,9 @@ err: sorting */ -static int not_killed= 0; - -volatile int *killed_ptr(HA_CHECK *param __attribute__((unused))) +int killed_ptr(HA_CHECK *param __attribute__((unused))) { - return &not_killed; /* always NULL */ + return 0; } /* print warnings and errors */ === modified file 'storage/myisam/myisamdef.h' --- a/storage/myisam/myisamdef.h 2009-10-06 06:57:22 +0000 +++ b/storage/myisam/myisamdef.h 2009-11-29 23:08:56 +0000 @@ -725,7 +725,7 @@ my_bool mi_dynmap_file(MI_INFO *info, my void mi_remap_file(MI_INFO *info, my_off_t size); /* Functions needed by mi_check */ -volatile int *killed_ptr(HA_CHECK *param); +int killed_ptr(HA_CHECK *param); void mi_check_print_error _VARARGS((HA_CHECK *param, const char *fmt, ...)); void mi_check_print_warning _VARARGS((HA_CHECK *param, const char *fmt, ...)); void mi_check_print_info _VARARGS((HA_CHECK *param, const char *fmt, ...)); === modified file 'storage/myisam/myisamlog.c' --- a/storage/myisam/myisamlog.c 2008-02-18 22:35:17 +0000 +++ b/storage/myisam/myisamlog.c 2009-11-29 23:08:56 +0000 @@ -385,7 +385,7 @@ static int examine_log(char * file_name, file_info.name=0; file_info.show_name=0; file_info.record=0; - if (read_string(&cache,(uchar**) &file_info.name, + if (read_string(&cache,(uchar**) (char*) &file_info.name, (uint) mi_uint2korr(head))) goto err; { === modified file 'storage/myisam/sort.c' --- a/storage/myisam/sort.c 2009-10-15 21:38:29 +0000 +++ b/storage/myisam/sort.c 2009-11-29 23:08:56 +0000 @@ -900,7 +900,6 @@ merge_buffers(MI_SORT_PARAM *info, uint uchar *strpos; BUFFPEK *buffpek,**refpek; QUEUE queue; - volatile int *killed= killed_ptr(info->sort_info->param); DBUG_ENTER("merge_buffers"); count=error=0; @@ -933,10 +932,6 @@ merge_buffers(MI_SORT_PARAM *info, uint { for (;;) { - if (*killed) - { - error=1; goto err; - } buffpek=(BUFFPEK*) queue_top(&queue); if (to_file) { @@ -956,6 +951,12 @@ merge_buffers(MI_SORT_PARAM *info, uint buffpek->key+=sort_length; if (! --buffpek->mem_count) { + /* It's enough to check for killedptr before a slow operation */ + if (killed_ptr(info->sort_info->param)) + { + error=1; + goto err; + } if (!(error=(int) info->read_to_buffer(from_file,buffpek,sort_length))) { uchar *base= buffpek->base; === modified file 'storage/xtradb/fil/fil0fil.c' --- a/storage/xtradb/fil/fil0fil.c 2009-11-13 21:26:08 +0000 +++ b/storage/xtradb/fil/fil0fil.c 2009-11-29 23:08:56 +0000 @@ -3139,7 +3139,7 @@ skip_info: if (mach_read_from_4(page + FIL_PAGE_OFFSET) || !offset) { mach_write_to_4(page + FIL_PAGE_ARCH_LOG_NO_OR_SPACE_ID, id); - for (i = 0; i < n_index; i++) { + for (i = 0; (ulint) i < n_index; i++) { if (offset / UNIV_PAGE_SIZE == root_page[i]) { /* this is index root page */ mach_write_to_4(page + FIL_PAGE_DATA + PAGE_BTR_SEG_LEAF === modified file 'storage/xtradb/trx/trx0i_s.c' --- a/storage/xtradb/trx/trx0i_s.c 2009-09-07 10:22:53 +0000 +++ b/storage/xtradb/trx/trx0i_s.c 2009-11-29 23:08:56 +0000 @@ -28,11 +28,11 @@ table cache" for later retrieval. Created July 17, 2007 Vasil Dimov *******************************************************/ -#include <mysql/plugin.h> #include "mysql_addons.h" #include "univ.i" +#include <mysql/plugin.h> #include "buf0buf.h" #include "dict0dict.h" #include "ha0storage.h"

3 5

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (monty:2770)
by Michael Widenius 30 Nov '09

30 Nov '09

#At lp:maria based on revid:monty@askmonty.org-20091130111302-6zcyj6ucz6rnphdq 2770 Michael Widenius 2009-11-30 Fixes after comments from last push: - Removed some not needed casts - Change plugin.h to be 'binary compatible' with old versions - Added mysql_ft_size_t typedef to plugin.h to make it trivial to change string lengths to size_t on next ABI change - Made some fixes suggested by Kristian to make things more portable and future safe (when it comes to strict aliasing) modified: include/ft_global.h include/mysql/plugin.h include/mysql/plugin.h.pp mysql-test/t/information_schema.test sql/sp_head.cc sql/sql_select.cc sql/table.cc storage/maria/ma_ft_boolean_search.c storage/maria/ma_ft_nlq_search.c storage/maria/ma_ft_parser.c storage/maria/ma_ftdefs.h storage/maria/maria_ftdump.c storage/myisam/ft_boolean_search.c storage/myisam/ft_nlq_search.c storage/myisam/ft_parser.c storage/myisam/myisam_ftdump.c per-file messages: include/ft_global.h Introduced FT_WEIGTH, to handle fulltext weights in a slightly more portable manner include/mysql/plugin.h Change plugin.h to be 'binary compatible' with old versions Added mysql_ft_size_t typedef to plugin.h to make it trivial to change string lengths to size_t on next ABI change Changed flags to unsigned (as flags should always be unsigned) mysql-test/t/information_schema.test Fixed typo sql/sp_head.cc Removed cast sql/sql_select.cc Removed cast sql/table.cc Removed cast storage/maria/ma_ft_boolean_search.c Use mysql_ft_size_t instead of size_t for plugin.h code Changed some other string lengths to size_t storage/maria/ma_ft_nlq_search.c Use FT_WEIGTH to make code more portable storage/maria/ma_ft_parser.c Use mysql_ft_size_t instead of size_t for plugin.h code Changed some other string lengths to size_t storage/maria/ma_ftdefs.h Changed some string lengths to size_t storage/maria/maria_ftdump.c Use FT_WEIGTH to make code more portable storage/myisam/ft_boolean_search.c Use mysql_ft_size_t instead of size_t for plugin.h code storage/myisam/ft_nlq_search.c Use FT_WEIGTH to make code more portable storage/myisam/ft_parser.c Use mysql_ft_size_t instead of size_t for plugin.h code storage/myisam/myisam_ftdump.c Use FT_WEIGTH to make code more portable === modified file 'include/ft_global.h' --- a/include/ft_global.h 2007-06-27 14:49:12 +0000 +++ b/include/ft_global.h 2009-11-30 13:36:06 +0000 @@ -76,6 +76,8 @@ my_bool ft_boolean_check_syntax_string(c extern const HA_KEYSEG ft_keysegs[FT_SEGS]; +typedef union {int32 i; float f;} FT_WEIGTH; + #ifdef __cplusplus } #endif === modified file 'include/mysql/plugin.h' --- a/include/mysql/plugin.h 2009-11-29 23:08:56 +0000 +++ b/include/mysql/plugin.h 2009-11-30 13:36:06 +0000 @@ -564,19 +564,22 @@ typedef struct st_mysql_ftparser_boolean nothing. See enum_ftparser_mode above. */ +/* TODO: Change the following int to size_t at next ABI update */ +typedef int mysql_ft_size_t; + typedef struct st_mysql_ftparser_param { int (*mysql_parse)(struct st_mysql_ftparser_param *, - const unsigned char *doc, size_t doc_len); + const unsigned char *doc, mysql_ft_size_t doc_len); int (*mysql_add_word)(struct st_mysql_ftparser_param *, - const unsigned char *word, size_t word_len, + const unsigned char *word, mysql_ft_size_t word_len, MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info); void *ftparser_state; void *mysql_ftparam; struct charset_info_st *cs; - char *doc; - int length; - int flags; + const unsigned char *doc; + mysql_ft_size_t length; + unsigned int flags; enum enum_ftparser_mode mode; } MYSQL_FTPARSER_PARAM; === modified file 'include/mysql/plugin.h.pp' --- a/include/mysql/plugin.h.pp 2009-11-29 23:08:56 +0000 +++ b/include/mysql/plugin.h.pp 2009-11-30 13:36:06 +0000 @@ -70,19 +70,20 @@ typedef struct st_mysql_ftparser_boolean char prev; char *quot; } MYSQL_FTPARSER_BOOLEAN_INFO; +typedef int mysql_ft_size_t; typedef struct st_mysql_ftparser_param { int (*mysql_parse)(struct st_mysql_ftparser_param *, - const unsigned char *doc, size_t doc_len); + const unsigned char *doc, mysql_ft_size_t doc_len); int (*mysql_add_word)(struct st_mysql_ftparser_param *, - const unsigned char *word, size_t word_len, + const unsigned char *word, mysql_ft_size_t word_len, MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info); void *ftparser_state; void *mysql_ftparam; struct charset_info_st *cs; - char *doc; - int length; - int flags; + const unsigned char *doc; + mysql_ft_size_t length; + unsigned int flags; enum enum_ftparser_mode mode; } MYSQL_FTPARSER_PARAM; struct st_mysql_ftparser === modified file 'mysql-test/t/information_schema.test' --- a/mysql-test/t/information_schema.test 2009-11-29 23:08:56 +0000 +++ b/mysql-test/t/information_schema.test 2009-11-30 13:36:06 +0000 @@ -5,7 +5,7 @@ # on the presence of the log tables (which are CSV-based). --source include/have_csv.inc -# Check that innodb/xtradb is incompiled in as result depends on it +# Check that InnoDB/XtraDB was compiled in as result depends on it -- source include/have_innodb.inc # Save the initial number of concurrent sessions === modified file 'sql/sp_head.cc' --- a/sql/sp_head.cc 2009-11-29 23:08:56 +0000 +++ b/sql/sp_head.cc 2009-11-30 13:36:06 +0000 @@ -1924,7 +1924,7 @@ sp_head::execute_procedure(THD *thd, Lis if (spvar->mode == sp_param_out) { Item_null *null_item= new Item_null(); - Item *tmp_item= (Item*) null_item; + Item *tmp_item= null_item; if (!null_item || nctx->set_variable(thd, i, &tmp_item)) === modified file 'sql/sql_select.cc' --- a/sql/sql_select.cc 2009-11-29 23:08:56 +0000 +++ b/sql/sql_select.cc 2009-11-30 13:36:06 +0000 @@ -3586,7 +3586,7 @@ add_key_fields(JOIN *join, KEY_FIELD **k { if (!field->eq(item->field)) { - Item *tmp_item= (Item*) item; + Item *tmp_item= item; add_key_field(key_fields, *and_level, cond_func, field, TRUE, &tmp_item, 1, usable_tables, sargables); === modified file 'sql/table.cc' --- a/sql/table.cc 2009-11-29 23:08:56 +0000 +++ b/sql/table.cc 2009-11-30 13:36:06 +0000 @@ -2077,7 +2077,7 @@ ulong get_form_pos(File file, uchar *hea else { char *str; - const char **tmp = (const char**) (char*) buf; + const char **tmp = (const char**) buf; str=(char *) (buf+a_length); fix_type_pointers(&tmp, save_names, 1, &str); } === modified file 'storage/maria/ma_ft_boolean_search.c' --- a/storage/maria/ma_ft_boolean_search.c 2009-11-29 23:08:56 +0000 +++ b/storage/maria/ma_ft_boolean_search.c 2009-11-30 13:36:06 +0000 @@ -180,7 +180,7 @@ typedef struct st_my_ftb_param static int ftb_query_add_word(MYSQL_FTPARSER_PARAM *param, - const uchar *word, size_t word_len, + const uchar *word, mysql_ft_size_t word_len, MYSQL_FTPARSER_BOOLEAN_INFO *info) { MY_FTB_PARAM *ftb_param= param->mysql_ftparam; @@ -282,7 +282,7 @@ static int ftb_query_add_word(MYSQL_FTPA static int ftb_parse_query_internal(MYSQL_FTPARSER_PARAM *param, - const uchar *query, size_t len) + const uchar *query, mysql_ft_size_t len) { MY_FTB_PARAM *ftb_param= param->mysql_ftparam; MYSQL_FTPARSER_BOOLEAN_INFO info; @@ -299,7 +299,7 @@ static int ftb_parse_query_internal(MYSQ } -static int _ftb_parse_query(FTB *ftb, uchar *query, uint len, +static int _ftb_parse_query(FTB *ftb, uchar *query, size_t len, struct st_mysql_ftparser *parser) { MYSQL_FTPARSER_PARAM *param; @@ -321,7 +321,7 @@ static int _ftb_parse_query(FTB *ftb, uc param->mysql_add_word= ftb_query_add_word; param->mysql_ftparam= (void *)&ftb_param; param->cs= ftb->charset; - param->doc= (char*) query; + param->doc= query; param->length= len; param->flags= 0; param->mode= MYSQL_FTPARSER_FULL_BOOLEAN_INFO; @@ -539,8 +539,8 @@ static void _ftb_init_index_search(FT_IN FT_INFO * maria_ft_init_boolean_search(MARIA_HA *info, uint keynr, - uchar *query, - uint query_len, CHARSET_INFO *cs) + uchar *query, size_t query_len, + CHARSET_INFO *cs) { FTB *ftb; FTB_EXPR *ftbe; @@ -615,8 +615,9 @@ typedef struct st_my_ftb_phrase_param static int ftb_phrase_add_word(MYSQL_FTPARSER_PARAM *param, - const uchar *word, size_t word_len, - MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info __attribute__((unused))) + const uchar *word, mysql_ft_size_t word_len, + MYSQL_FTPARSER_BOOLEAN_INFO + *boolean_info __attribute__((unused))) { MY_FTB_PHRASE_PARAM *phrase_param= param->mysql_ftparam; FT_WORD *w= (FT_WORD *)phrase_param->document->data; @@ -647,7 +648,8 @@ static int ftb_phrase_add_word(MYSQL_FTP static int ftb_check_phrase_internal(MYSQL_FTPARSER_PARAM *param, - const uchar *document, size_t len) + const uchar *document, + mysql_ft_size_t len) { FT_WORD word; MY_FTB_PHRASE_PARAM *phrase_param= param->mysql_ftparam; @@ -678,8 +680,8 @@ static int ftb_check_phrase_internal(MYS -1 is returned if error occurs. */ -static int _ftb_check_phrase(FTB *ftb, const uchar *document, uint len, - FTB_EXPR *ftbe, struct st_mysql_ftparser *parser) +static int _ftb_check_phrase(FTB *ftb, const uchar *document, size_t len, + FTB_EXPR *ftbe, struct st_mysql_ftparser *parser) { MY_FTB_PHRASE_PARAM ftb_param; MYSQL_FTPARSER_PARAM *param; @@ -699,7 +701,7 @@ static int _ftb_check_phrase(FTB *ftb, c param->mysql_add_word= ftb_phrase_add_word; param->mysql_ftparam= (void *)&ftb_param; param->cs= ftb->charset; - param->doc= (char *) document; + param->doc= document; param->length= len; param->flags= 0; param->mode= MYSQL_FTPARSER_WITH_STOPWORDS; @@ -872,8 +874,9 @@ typedef struct st_my_ftb_find_param static int ftb_find_relevance_add_word(MYSQL_FTPARSER_PARAM *param, - const uchar *word, size_t len, - MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info __attribute__((unused))) + const uchar *word, mysql_ft_size_t len, + MYSQL_FTPARSER_BOOLEAN_INFO + *boolean_info __attribute__((unused))) { MY_FTB_FIND_PARAM *ftb_param= param->mysql_ftparam; FT_INFO *ftb= ftb_param->ftb; @@ -933,7 +936,7 @@ static int ftb_find_relevance_add_word(M static int ftb_find_relevance_parse(MYSQL_FTPARSER_PARAM *param, - const uchar *doc, size_t len) + const uchar *doc, mysql_ft_size_t len) { MY_FTB_FIND_PARAM *ftb_param= param->mysql_ftparam; FT_INFO *ftb= ftb_param->ftb; @@ -997,7 +1000,7 @@ float maria_ft_boolean_find_relevance(FT { if (!ftsi.pos) continue; - param->doc= (char *)ftsi.pos; + param->doc= ftsi.pos; param->length= ftsi.len; if (unlikely(parser->parse(param))) return 0; === modified file 'storage/maria/ma_ft_nlq_search.c' --- a/storage/maria/ma_ft_nlq_search.c 2009-11-29 23:08:56 +0000 +++ b/storage/maria/ma_ft_nlq_search.c 2009-11-30 13:36:06 +0000 @@ -51,6 +51,7 @@ typedef struct st_ft_superdoc double tmp_weight; } FT_SUPERDOC; + static int FT_SUPERDOC_cmp(void* cmp_arg __attribute__((unused)), FT_SUPERDOC *p1, FT_SUPERDOC *p2) { @@ -63,7 +64,7 @@ static int FT_SUPERDOC_cmp(void* cmp_arg static int walk_and_match(FT_WORD *word, uint32 count, ALL_IN_ONE *aio) { - int32 subkeys; + FT_WEIGTH subkeys; int r; uint doc_cnt; FT_SUPERDOC sdoc, *sptr; @@ -91,9 +92,9 @@ static int walk_and_match(FT_WORD *word, /* Skip rows inserted by current inserted */ for (r= _ma_search(info, &key, SEARCH_FIND, key_root) ; !r && - (subkeys=ft_sintXkorr(info->last_key.data + - info->last_key.data_length + - info->last_key.ref_length - extra)) > 0 && + (subkeys.i= ft_sintXkorr(info->last_key.data + + info->last_key.data_length + + info->last_key.ref_length - extra)) > 0 && info->cur_row.lastpos >= info->state->data_file_length ; r= _ma_search_next(info, &info->last_key, SEARCH_BIGGER, key_root)) ; @@ -112,7 +113,7 @@ static int walk_and_match(FT_WORD *word, key.data+1, key.data_length-1, 0, 0)) break; - if (subkeys<0) + if (subkeys.i < 0) { if (doc_cnt) DBUG_RETURN(1); /* index is corrupted */ @@ -128,7 +129,8 @@ static int walk_and_match(FT_WORD *word, goto do_skip; } #if HA_FT_WTYPE == HA_KEYTYPE_FLOAT - tmp_weight=*(float*) (char*) &subkeys; + /* The weight we read was actually a float */ + tmp_weight= subkeys.f; #else #error #endif @@ -163,9 +165,9 @@ static int walk_and_match(FT_WORD *word, else r= _ma_search(info, &info->last_key, SEARCH_BIGGER, key_root); do_skip: - while ((subkeys=ft_sintXkorr(info->last_key.data + - info->last_key.data_length + - info->last_key.ref_length - extra)) > 0 && + while ((subkeys.i= ft_sintXkorr(info->last_key.data + + info->last_key.data_length + + info->last_key.ref_length - extra)) > 0 && !r && info->cur_row.lastpos >= info->state->data_file_length) r= _ma_search_next(info, &info->last_key, SEARCH_BIGGER, key_root); @@ -206,7 +208,7 @@ static int FT_DOC_cmp(void *unused __att FT_INFO *maria_ft_init_nlq_search(MARIA_HA *info, uint keynr, uchar *query, - uint query_len, uint flags, uchar *record) + size_t query_len, uint flags, uchar *record) { TREE wtree; ALL_IN_ONE aio; === modified file 'storage/maria/ma_ft_parser.c' --- a/storage/maria/ma_ft_parser.c 2009-11-29 23:08:56 +0000 +++ b/storage/maria/ma_ft_parser.c 2009-11-30 13:36:06 +0000 @@ -260,7 +260,7 @@ void maria_ft_parse_init(TREE *wtree, CH static int maria_ft_add_word(MYSQL_FTPARSER_PARAM *param, - const uchar *word, size_t word_len, + const uchar *word, mysql_ft_size_t word_len, MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info __attribute__((unused))) { @@ -290,7 +290,8 @@ static int maria_ft_add_word(MYSQL_FTPAR static int maria_ft_parse_internal(MYSQL_FTPARSER_PARAM *param, - const uchar *doc_arg, size_t doc_len) + const uchar *doc_arg, + mysql_ft_size_t doc_len) { const uchar *doc= doc_arg; const uchar *end= doc + doc_len; @@ -306,8 +307,8 @@ static int maria_ft_parse_internal(MYSQL } -int maria_ft_parse(TREE *wtree, uchar *doc, int doclen, - struct st_mysql_ftparser *parser, +int maria_ft_parse(TREE *wtree, uchar *doc, size_t doclen, + struct st_mysql_ftparser *parser, MYSQL_FTPARSER_PARAM *param, MEM_ROOT *mem_root) { MY_FT_PARSER_PARAM my_param; @@ -320,7 +321,7 @@ int maria_ft_parse(TREE *wtree, uchar *d param->mysql_add_word= maria_ft_add_word; param->mysql_ftparam= &my_param; param->cs= wtree->custom_arg; - param->doc= (char *) doc; + param->doc= doc; param->length= doclen; param->mode= MYSQL_FTPARSER_SIMPLE_MODE; DBUG_RETURN(parser->parse(param)); @@ -380,8 +381,8 @@ MYSQL_FTPARSER_PARAM *maria_ftparser_cal mysql_add_word != 0 - parser is initialized, or no initialization needed. */ info->ftparser_param[ftparser_nr].mysql_add_word= - (int (*)(struct st_mysql_ftparser_param *, const uchar *, size_t, - MYSQL_FTPARSER_BOOLEAN_INFO *)) 1; + (int (*)(struct st_mysql_ftparser_param *, const uchar *, + mysql_ft_size_t, MYSQL_FTPARSER_BOOLEAN_INFO *)) 1; if (parser->init && parser->init(&info->ftparser_param[ftparser_nr])) return 0; } === modified file 'storage/maria/ma_ftdefs.h' --- a/storage/maria/ma_ftdefs.h 2009-11-29 23:08:56 +0000 +++ b/storage/maria/ma_ftdefs.h 2009-11-30 13:36:06 +0000 @@ -122,15 +122,17 @@ void _ma_ft_segiterator_dummy_init(const uint _ma_ft_segiterator(FT_SEG_ITERATOR *); void maria_ft_parse_init(TREE *, CHARSET_INFO *); -int maria_ft_parse(TREE *, uchar *, int, struct st_mysql_ftparser *parser, +int maria_ft_parse(TREE *, uchar *, size_t, struct st_mysql_ftparser *parser, MYSQL_FTPARSER_PARAM *, MEM_ROOT *); FT_WORD * maria_ft_linearize(TREE *, MEM_ROOT *); FT_WORD * _ma_ft_parserecord(MARIA_HA *, uint, const uchar *, MEM_ROOT *); uint _ma_ft_parse(TREE *, MARIA_HA *, uint, const uchar *, MYSQL_FTPARSER_PARAM *, MEM_ROOT *); -FT_INFO *maria_ft_init_nlq_search(MARIA_HA *, uint, uchar *, uint, uint, uchar *); -FT_INFO *maria_ft_init_boolean_search(MARIA_HA *, uint, uchar *, uint, CHARSET_INFO *); +FT_INFO *maria_ft_init_nlq_search(MARIA_HA *, uint, uchar *, size_t, uint, + uchar *); +FT_INFO *maria_ft_init_boolean_search(MARIA_HA *, uint, uchar *, size_t, + CHARSET_INFO *); extern const struct _ft_vft _ma_ft_vft_nlq; int maria_ft_nlq_read_next(FT_INFO *, char *); === modified file 'storage/maria/maria_ftdump.c' --- a/storage/maria/maria_ftdump.c 2009-11-29 23:08:56 +0000 +++ b/storage/maria/maria_ftdump.c 2009-11-30 13:36:06 +0000 @@ -53,7 +53,7 @@ static struct my_option my_long_options[ int main(int argc,char *argv[]) { - int error=0, subkeys; + int error=0; uint keylen, keylen2=0, inx, doc_cnt=0; float weight= 1.0; double gws, min_gws=0, avg_gws=0; @@ -112,11 +112,12 @@ int main(int argc,char *argv[]) while (!(error=maria_rnext(info,NULL,inx))) { + FT_WEIGTH subkeys; keylen=*(info->lastkey_buff); - subkeys=ft_sintXkorr(info->lastkey_buff + keylen + 1); - if (subkeys >= 0) - weight=*(float*) (char*) &subkeys; + subkeys.i= ft_sintXkorr(info->lastkey_buff + keylen + 1); + if (subkeys.i >= 0) + weight= subkeys.f; #ifdef HAVE_SNPRINTF snprintf(buf,MAX_LEN,"%.*s",(int) keylen,info->lastkey_buff+1); @@ -153,14 +154,15 @@ int main(int argc,char *argv[]) keylen2=keylen; doc_cnt=0; } - doc_cnt+= (subkeys >= 0 ? 1 : -subkeys); + doc_cnt+= (subkeys.i >= 0 ? 1 : -subkeys.i); } if (dump) { - if (subkeys>=0) + if (subkeys.i >= 0) printf("%9lx %20.7f %s\n", (long) info->cur_row.lastpos,weight,buf); else - printf("%9lx => %17d %s\n",(long) info->cur_row.lastpos,-subkeys,buf); + printf("%9lx => %17d %s\n",(long) info->cur_row.lastpos,-subkeys.i, + buf); } if (verbose && (total%HOW_OFTEN_TO_WRITE)==0) printf("%10ld\r",total); === modified file 'storage/myisam/ft_boolean_search.c' --- a/storage/myisam/ft_boolean_search.c 2009-11-29 23:08:56 +0000 +++ b/storage/myisam/ft_boolean_search.c 2009-11-30 13:36:06 +0000 @@ -180,7 +180,7 @@ typedef struct st_my_ftb_param static int ftb_query_add_word(MYSQL_FTPARSER_PARAM *param, - const uchar *word, size_t word_len, + const uchar *word, mysql_ft_size_t word_len, MYSQL_FTPARSER_BOOLEAN_INFO *info) { MY_FTB_PARAM *ftb_param= param->mysql_ftparam; @@ -282,7 +282,7 @@ static int ftb_query_add_word(MYSQL_FTPA static int ftb_parse_query_internal(MYSQL_FTPARSER_PARAM *param, - const uchar *query, size_t len) + const uchar *query, mysql_ft_size_t len) { MY_FTB_PARAM *ftb_param= param->mysql_ftparam; MYSQL_FTPARSER_BOOLEAN_INFO info; @@ -616,7 +616,7 @@ typedef struct st_my_ftb_phrase_param static int ftb_phrase_add_word(MYSQL_FTPARSER_PARAM *param, - const uchar *word, size_t word_len, + const uchar *word, mysql_ft_size_t word_len, MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info __attribute__((unused))) { MY_FTB_PHRASE_PARAM *phrase_param= param->mysql_ftparam; @@ -648,7 +648,8 @@ static int ftb_phrase_add_word(MYSQL_FTP static int ftb_check_phrase_internal(MYSQL_FTPARSER_PARAM *param, - const uchar *document, size_t len) + const uchar *document, + mysql_ft_size_t len) { FT_WORD word; MY_FTB_PHRASE_PARAM *phrase_param= param->mysql_ftparam; @@ -874,8 +875,9 @@ typedef struct st_my_ftb_find_param static int ftb_find_relevance_add_word(MYSQL_FTPARSER_PARAM *param, - const uchar *word, size_t len, - MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info __attribute__((unused))) + const uchar *word, mysql_ft_size_t len, + MYSQL_FTPARSER_BOOLEAN_INFO + *boolean_info __attribute__((unused))) { MY_FTB_FIND_PARAM *ftb_param= param->mysql_ftparam; FT_INFO *ftb= ftb_param->ftb; @@ -935,7 +937,7 @@ static int ftb_find_relevance_add_word(M static int ftb_find_relevance_parse(MYSQL_FTPARSER_PARAM *param, - const uchar *doc, size_t len) + const uchar *doc, mysql_ft_size_t len) { MY_FTB_FIND_PARAM *ftb_param= param->mysql_ftparam; FT_INFO *ftb= ftb_param->ftb; === modified file 'storage/myisam/ft_nlq_search.c' --- a/storage/myisam/ft_nlq_search.c 2009-11-29 23:08:56 +0000 +++ b/storage/myisam/ft_nlq_search.c 2009-11-30 13:36:06 +0000 @@ -63,7 +63,7 @@ static int FT_SUPERDOC_cmp(void* cmp_arg static int walk_and_match(FT_WORD *word, uint32 count, ALL_IN_ONE *aio) { - int32 subkeys; + FT_WEIGTH subkeys; int r; uint keylen, doc_cnt; FT_SUPERDOC sdoc, *sptr; @@ -91,7 +91,7 @@ static int walk_and_match(FT_WORD *word, /* Skip rows inserted by current inserted */ for (r=_mi_search(info, keyinfo, keybuff, keylen, SEARCH_FIND, key_root) ; !r && - (subkeys=ft_sintXkorr(info->lastkey+info->lastkey_length-extra)) > 0 && + (subkeys.i= ft_sintXkorr(info->lastkey+info->lastkey_length-extra)) > 0 && info->lastpos >= info->state->data_file_length ; r= _mi_search_next(info, keyinfo, info->lastkey, info->lastkey_length, SEARCH_BIGGER, key_root)) @@ -108,7 +108,7 @@ static int walk_and_match(FT_WORD *word, info->lastkey_length-extra-1, keybuff+1,keylen-1,0,0)) break; - if (subkeys<0) + if (subkeys.i < 0) { if (doc_cnt) DBUG_RETURN(1); /* index is corrupted */ @@ -125,7 +125,7 @@ static int walk_and_match(FT_WORD *word, } #if HA_FT_WTYPE == HA_KEYTYPE_FLOAT /* The weight we read was actually a float */ - tmp_weight=*(float*) (char*) &subkeys; + tmp_weight= subkeys.f; #else #error #endif @@ -162,7 +162,7 @@ static int walk_and_match(FT_WORD *word, r=_mi_search(info, keyinfo, info->lastkey, info->lastkey_length, SEARCH_BIGGER, key_root); do_skip: - while ((subkeys=ft_sintXkorr(info->lastkey+info->lastkey_length-extra)) > 0 && + while ((subkeys.i= ft_sintXkorr(info->lastkey+info->lastkey_length-extra)) > 0 && !r && info->lastpos >= info->state->data_file_length) r= _mi_search_next(info, keyinfo, info->lastkey, info->lastkey_length, SEARCH_BIGGER, key_root); === modified file 'storage/myisam/ft_parser.c' --- a/storage/myisam/ft_parser.c 2009-11-29 23:08:56 +0000 +++ b/storage/myisam/ft_parser.c 2009-11-30 13:36:06 +0000 @@ -258,7 +258,7 @@ void ft_parse_init(TREE *wtree, CHARSET_ static int ft_add_word(MYSQL_FTPARSER_PARAM *param, - const uchar *word, size_t word_len, + const uchar *word, mysql_ft_size_t word_len, MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info __attribute__((unused))) { @@ -288,7 +288,7 @@ static int ft_add_word(MYSQL_FTPARSER_PA static int ft_parse_internal(MYSQL_FTPARSER_PARAM *param, - const uchar *doc_arg, size_t doc_len) + const uchar *doc_arg, mysql_ft_size_t doc_len) { const uchar *doc= doc_arg; const uchar *end= doc + doc_len; @@ -379,8 +379,8 @@ MYSQL_FTPARSER_PARAM *ftparser_call_init mysql_add_word != 0 - parser is initialized, or no initialization needed. */ info->ftparser_param[ftparser_nr].mysql_add_word= - (int (*)(struct st_mysql_ftparser_param *, const uchar *, size_t, - MYSQL_FTPARSER_BOOLEAN_INFO *)) 1; + (int (*)(struct st_mysql_ftparser_param *, const uchar *, + mysql_ft_size_t, MYSQL_FTPARSER_BOOLEAN_INFO *)) 1; if (parser->init && parser->init(&info->ftparser_param[ftparser_nr])) return 0; } === modified file 'storage/myisam/myisam_ftdump.c' --- a/storage/myisam/myisam_ftdump.c 2009-11-29 23:08:56 +0000 +++ b/storage/myisam/myisam_ftdump.c 2009-11-30 13:36:06 +0000 @@ -53,7 +53,7 @@ static struct my_option my_long_options[ int main(int argc,char *argv[]) { - int error=0, subkeys; + int error=0; uint keylen, keylen2=0, inx, doc_cnt=0; float weight= 1.0; double gws, min_gws=0, avg_gws=0; @@ -109,11 +109,12 @@ int main(int argc,char *argv[]) while (!(error=mi_rnext(info,NULL,inx))) { + FT_WEIGTH subkeys; keylen=*(info->lastkey); - subkeys=ft_sintXkorr(info->lastkey+keylen+1); - if (subkeys >= 0) - weight= *(float*) (char*) &subkeys; + subkeys.i =ft_sintXkorr(info->lastkey+keylen+1); + if (subkeys.i >= 0) + weight= subkeys.f; #ifdef HAVE_SNPRINTF snprintf(buf,MAX_LEN,"%.*s",(int) keylen,info->lastkey+1); @@ -150,14 +151,14 @@ int main(int argc,char *argv[]) keylen2=keylen; doc_cnt=0; } - doc_cnt+= (subkeys >= 0 ? 1 : -subkeys); + doc_cnt+= (subkeys.i >= 0 ? 1 : -subkeys.i); } if (dump) { - if (subkeys>=0) + if (subkeys.i >= 0) printf("%9lx %20.7f %s\n", (long) info->lastpos,weight,buf); else - printf("%9lx => %17d %s\n",(long) info->lastpos,-subkeys,buf); + printf("%9lx => %17d %s\n",(long) info->lastpos,-subkeys.i,buf); } if (verbose && (total%HOW_OFTEN_TO_WRITE)==0) printf("%10ld\r",total);

1 0

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (knielsen:2778)
by knielsen＠knielsen-hq.org 30 Nov '09

30 Nov '09

#At lp:maria 2778 knielsen(a)knielsen-hq.org 2009-11-30 After-merge fixes for MySQL 5.1.41 merge into MariaDB: Another warning also needed during shutdown. modified: mysql-test/mysql-test-run.pl === modified file 'mysql-test/mysql-test-run.pl' --- a/mysql-test/mysql-test-run.pl 2009-11-27 13:20:59 +0000 +++ b/mysql-test/mysql-test-run.pl 2009-11-30 13:24:30 +0000 @@ -3999,6 +3999,7 @@ sub extract_warning_lines ($) { qr/Slave I\/O: Get master COLLATION_SERVER failed with error:.*/, qr/Slave I\/O: Get master TIME_ZONE failed with error:.*/, qr/Slave I\/O: error reconnecting to master '.*' - retry-time: [1-3] retries/, + qr/Error reading packet/, ); my $match_count= 0;

1 0

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (monty:2777)
by Michael Widenius 30 Nov '09

30 Nov '09

#At lp:maria based on revid:knielsen@knielsen-hq.org-20091127132059-3su1w7xhsbbtpg6f 2777 Michael Widenius 2009-11-30 Added more general support for sorting 2 characters as one (contractions) Added support for Croatian sorting orders utf8_croatian_ci and ucs2_croatian_ci. Patch done by Alexander Barkov. See http://www.collation-charts.org/articles/croatian.htm modified: include/m_ctype.h mysql-test/r/ctype_uca.result mysql-test/t/ctype_uca.test mysys/charset-def.c strings/ctype-mb.c strings/ctype-uca.c strings/ctype-ucs2.c per-file messages: mysql-test/r/ctype_uca.result Added testing of Croatian sort order mysql-test/t/ctype_uca.test Added testing of Croatian sort order === modified file 'include/m_ctype.h' --- a/include/m_ctype.h 2009-09-07 20:50:10 +0000 +++ b/include/m_ctype.h 2009-11-30 12:42:24 +0000 @@ -49,6 +49,24 @@ typedef struct unicase_info_st extern MY_UNICASE_INFO *my_unicase_default[256]; extern MY_UNICASE_INFO *my_unicase_turkish[256]; +#define MY_UCA_MAX_CONTRACTION 4 +#define MY_UCA_MAX_WEIGHT_SIZE 8 + +typedef struct my_contraction_t +{ + my_wc_t ch[MY_UCA_MAX_CONTRACTION]; /* Character sequence */ + uint16 weight[MY_UCA_MAX_WEIGHT_SIZE];/* Its weight string, 0-terminated */ +} MY_CONTRACTION; + + +typedef struct my_contraction_list_t +{ + size_t nitems; /* Number of items in the list */ + MY_CONTRACTION *item; /* List of contractions */ + char *flags; /* Character flags, e.g. "is contraction head") */ +} MY_CONTRACTIONS; + + typedef struct uni_ctype_st { uchar pctype; @@ -262,7 +280,7 @@ typedef struct charset_info_st uchar *to_lower; uchar *to_upper; uchar *sort_order; - uint16 *contractions; + MY_CONTRACTIONS *contractions; uint16 **sort_order_big; uint16 *tab_to_uni; MY_UNI_IDX *tab_from_uni; @@ -475,6 +493,13 @@ my_bool my_charset_is_ascii_based(CHARSE my_bool my_charset_is_8bit_pure_ascii(CHARSET_INFO *cs); uint my_charset_repertoire(CHARSET_INFO *cs); +my_bool my_uca_have_contractions(CHARSET_INFO *cs); +my_bool my_uca_can_be_contraction_head(CHARSET_INFO *cs, my_wc_t wc); +my_bool my_uca_can_be_contraction_tail(CHARSET_INFO *cs, my_wc_t wc); +uint16 *my_uca_contraction2_weight(CHARSET_INFO *cs, my_wc_t wc1, my_wc_t wc2); + + + #define _MY_U 01 /* Upper case */ #define _MY_L 02 /* Lower case */ === modified file 'mysql-test/r/ctype_uca.result' --- a/mysql-test/r/ctype_uca.result 2008-03-26 09:51:16 +0000 +++ b/mysql-test/r/ctype_uca.result 2009-11-30 12:42:24 +0000 @@ -159,6 +159,7 @@ insert into t1 values (_ucs2 0x01fc),(_u insert into t1 values ('AA'),('Aa'),('aa'),('aA'); insert into t1 values ('CH'),('Ch'),('ch'),('cH'); insert into t1 values ('DZ'),('Dz'),('dz'),('dZ'); +insert into t1 values ('D��'),('D��'),('d��'),('d��'); insert into t1 values ('IJ'),('Ij'),('ij'),('iJ'); insert into t1 values ('LJ'),('Lj'),('lj'),('lJ'); insert into t1 values ('LL'),('Ll'),('ll'),('lL'); @@ -181,7 +182,7 @@ C,c,��,��,��,��,��,��,��,��,��,�� CH,Ch,cH,ch ��,�� D,d,��,�� -DZ,Dz,dZ,dz,��,��,��,��,��,�� +DZ,Dz,D��,D��,dZ,dz,d��,d��,��,��,��,��,��,�� ,�� @@ -286,7 +287,7 @@ C,c,��,��,��,��,��,��,��,��,��,�� CH,Ch,cH,ch ��,�� D,d,��,�� -DZ,Dz,dZ,dz,��,��,��,��,��,�� +DZ,Dz,D��,D��,dZ,dz,d��,d��,��,��,��,��,��,�� ,�� ,�� @@ -400,6 +401,7 @@ CH,Ch,cH,ch ��,�� D,d,��,�� DZ,Dz,dZ,dz,��,��,��,��,��,�� +D��,D��,d��,d�� ,�� @@ -513,7 +515,7 @@ C,c,��,��,��,��,��,��,��,��,��,�� CH,Ch,cH,ch ��,�� D,d,��,�� -DZ,Dz,dZ,dz,��,��,��,��,��,�� +DZ,Dz,D��,D��,dZ,dz,d��,d��,��,��,��,��,��,�� ,�� @@ -622,6 +624,7 @@ CH,Ch,cH,ch ��,�� D,d,��,�� DZ,Dz,dZ,dz,��,��,��,��,��,�� +D��,D��,d��,d�� ,�� @@ -729,7 +732,7 @@ CH,Ch,cH,ch ��,�� ,�� D,d,��,�� -DZ,Dz,dZ,dz,��,��,��,��,��,�� +DZ,Dz,D��,D��,dZ,dz,d��,d��,��,��,��,��,��,�� ,�� @@ -840,6 +843,7 @@ CH,Ch,cH,ch ��,�� D,d,��,�� DZ,Dz,dZ,dz +D��,D��,d��,d�� ,��,��,��,��,�� ,�� @@ -951,7 +955,7 @@ C,c,��,��,��,��,��,��,��,��,��,�� CH,Ch,cH,ch ��,�� D,d,��,�� -DZ,Dz,dZ,dz,��,��,��,��,��,�� +DZ,Dz,D��,D��,dZ,dz,d��,d��,��,��,��,��,��,�� ,�� @@ -1056,7 +1060,7 @@ C,c,��,��,��,��,��,��,��,��,��,�� CH,Ch,cH,ch ��,�� D,d,��,�� -DZ,Dz,dZ,dz,��,��,��,��,��,�� +DZ,Dz,D��,D��,dZ,dz,d��,d��,��,��,��,��,��,�� ,�� @@ -1164,7 +1168,7 @@ CH,Ch,cH,ch ��,�� ,�� D,d,��,�� -DZ,Dz,dZ,dz,��,��,��,��,��,�� +DZ,Dz,D��,D��,dZ,dz,d��,d��,��,��,��,��,��,�� ,�� @@ -1275,6 +1279,7 @@ cH ��,�� D,d,��,�� DZ,Dz,dZ,dz,��,��,��,��,��,�� +D��,D��,d��,d�� ,�� @@ -1382,7 +1387,7 @@ C,c,��,��,��,��,��,��,��,��,��,�� CH,Ch,cH,ch ��,�� D,d,��,�� -DZ,Dz,dZ,dz,��,��,��,��,��,�� +DZ,Dz,D��,D��,dZ,dz,d��,d��,��,��,��,��,��,�� ,�� @@ -1491,6 +1496,7 @@ cH ��,�� D,d,��,�� DZ,Dz,dZ,dz,��,��,��,��,��,�� +D��,D��,d��,d�� ,�� @@ -1599,6 +1605,7 @@ cH ��,�� D,d,��,�� DZ,Dz,dZ,dz,��,��,��,��,��,�� +D��,D��,d��,d�� ,�� @@ -1707,7 +1714,7 @@ cH CH,Ch,ch ��,�� D,d,��,�� -DZ,Dz,dZ,dz,��,��,��,��,��,�� +DZ,Dz,D��,D��,dZ,dz,d��,d��,��,��,��,��,��,�� ,�� @@ -1813,7 +1820,7 @@ C,c,��,��,��,��,��,��,��,��,��,�� CH,Ch,cH,ch ��,�� D,d,��,�� -DZ,Dz,dZ,dz,��,��,��,��,��,�� +DZ,Dz,D��,D��,dZ,dz,d��,d��,��,��,��,��,��,�� ,�� @@ -1921,7 +1928,7 @@ CH,Ch,cH,ch ��,�� ,�� D,d,��,�� -DZ,Dz,dZ,dz,��,��,��,��,��,�� +DZ,Dz,D��,D��,dZ,dz,d��,d��,��,��,��,��,��,�� ,�� @@ -2030,7 +2037,7 @@ C,c,��,��,��,��,��,��,��,��,��,�� CH,Ch,cH,ch ��,�� D,d,��,�� -DZ,Dz,dZ,dz,��,��,��,��,��,�� +DZ,Dz,D��,D��,dZ,dz,d��,d��,��,��,��,��,��,�� ,�� @@ -2121,6 +2128,118 @@ Z,z,��,��,��,��,��,�� +select group_concat(c1 order by c1) from t1 group by c1 collate utf8_croatian_ci; +group_concat(c1 order by c1) +�� +�� +A,a,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,�� +AA,Aa,aA,aa +��,��,��,��,��,�� +B,b +�� +�� +��,�� +C,c,��,��,��,��,��,�� +CH,Ch,cH,ch +��,�� +��,�� +��,�� +D,d,��,�� +DZ,Dz,dZ,dz,��,��,�� +d�� +D��,D��,d��,��,��,�� +��,�� +�� +�� +��,�� +��,�� +E,e,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,�� +��,�� +�� +�� +F,f +��,�� +G,g,��,��,��,��,��,��,��,��,��,��,��,�� +��,�� +�� +�� +��,�� +H,h,��,�� +��,�� +��,�� +I,i,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,�� +IJ,Ij,iJ,ij,��,�� +�� +�� +�� +J,j,��,��,�� +K,k,��,��,��,�� +��,�� +L,l,��,��,��,��,��,�� +��,�� +lJ +LL,Ll,lL,ll +LJ,Lj,lj,��,��,�� +��,�� +�� +�� +M,m +N,n,��,��,��,��,��,��,��,��,��,�� +nJ +NJ,Nj,nj,��,��,�� +�� +�� +��,�� +O,o,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,�� +OE,Oe,oE,oe,��,�� +��,��,��,�� +�� +�� +P,p +��,�� +Q,q +�� +R,r,��,��,��,��,��,�� +RR,Rr,rR,rr +�� +S,s,��,��,��,��,��,��,�� +SS,Ss,sS,ss,�� +��,�� +�� +�� +T,t,��,��,��,�� +�� +��,�� +�� +��,�� +�� +U,u,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,��,�� +�� +�� +V,v +�� +W,w,��,�� +X,x +Y,y,��,��,��,��,��,�� +��,�� +Z,z,��,��,��,�� +�� +��,�� +��,�� +��,��,�� +��,�� +�� +��,�� +��,�� +�� +��,�� +��,�� +��,�� +�� +�� +�� +�� +�� drop table t1; SET NAMES utf8; CREATE TABLE t1 (c varchar(255) NOT NULL COLLATE utf8_general_ci, INDEX (c)); === modified file 'mysql-test/t/ctype_uca.test' --- a/mysql-test/t/ctype_uca.test 2008-02-20 18:49:26 +0000 +++ b/mysql-test/t/ctype_uca.test 2009-11-30 12:42:24 +0000 @@ -186,6 +186,7 @@ insert into t1 values (_ucs2 0x01fc),(_u insert into t1 values ('AA'),('Aa'),('aa'),('aA'); insert into t1 values ('CH'),('Ch'),('ch'),('cH'); insert into t1 values ('DZ'),('Dz'),('dz'),('dZ'); +insert into t1 values ('D��'),('D��'),('d��'),('d��'); insert into t1 values ('IJ'),('Ij'),('ij'),('iJ'); insert into t1 values ('LJ'),('Lj'),('lj'),('lJ'); insert into t1 values ('LL'),('Ll'),('ll'),('lL'); @@ -213,6 +214,7 @@ select group_concat(c1 order by c1) from select group_concat(c1 order by c1) from t1 group by c1 collate utf8_roman_ci; select group_concat(c1 order by c1) from t1 group by c1 collate utf8_esperanto_ci; select group_concat(c1 order by c1) from t1 group by c1 collate utf8_hungarian_ci; +select group_concat(c1 order by c1) from t1 group by c1 collate utf8_croatian_ci; drop table t1; === modified file 'mysys/charset-def.c' --- a/mysys/charset-def.c 2007-06-21 20:10:40 +0000 +++ b/mysys/charset-def.c 2009-11-30 12:42:24 +0000 @@ -42,6 +42,7 @@ extern CHARSET_INFO my_charset_ucs2_roma extern CHARSET_INFO my_charset_ucs2_persian_uca_ci; extern CHARSET_INFO my_charset_ucs2_esperanto_uca_ci; extern CHARSET_INFO my_charset_ucs2_hungarian_uca_ci; +extern CHARSET_INFO my_charset_ucs2_croatian_uca_ci; #endif #ifdef HAVE_CHARSET_utf8 @@ -63,6 +64,7 @@ extern CHARSET_INFO my_charset_utf8_roma extern CHARSET_INFO my_charset_utf8_persian_uca_ci; extern CHARSET_INFO my_charset_utf8_esperanto_uca_ci; extern CHARSET_INFO my_charset_utf8_hungarian_uca_ci; +extern CHARSET_INFO my_charset_utf8_croatian_uca_ci; #ifdef HAVE_UTF8_GENERAL_CS extern CHARSET_INFO my_charset_utf8_general_cs; #endif @@ -152,6 +154,7 @@ my_bool init_compiled_charsets(myf flags add_compiled_collation(&my_charset_ucs2_persian_uca_ci); add_compiled_collation(&my_charset_ucs2_esperanto_uca_ci); add_compiled_collation(&my_charset_ucs2_hungarian_uca_ci); + add_compiled_collation(&my_charset_ucs2_croatian_uca_ci); #endif #endif @@ -186,6 +189,7 @@ my_bool init_compiled_charsets(myf flags add_compiled_collation(&my_charset_utf8_persian_uca_ci); add_compiled_collation(&my_charset_utf8_esperanto_uca_ci); add_compiled_collation(&my_charset_utf8_hungarian_uca_ci); + add_compiled_collation(&my_charset_utf8_croatian_uca_ci); #endif #endif === modified file 'strings/ctype-mb.c' --- a/strings/ctype-mb.c 2009-02-13 16:41:47 +0000 +++ b/strings/ctype-mb.c 2009-11-30 12:42:24 +0000 @@ -567,8 +567,7 @@ my_bool my_like_range_mb(CHARSET_INFO *c char *min_end= min_str + res_length; char *max_end= max_str + res_length; size_t maxcharlen= res_length / cs->mbmaxlen; - const char *contraction_flags= cs->contractions ? - ((const char*) cs->contractions) + 0x40*0x40 : NULL; + my_bool have_contractions= my_uca_have_contractions(cs); for (; ptr != end && min_str != min_end && maxcharlen ; maxcharlen--) { @@ -636,8 +635,8 @@ fill_max_and_min: 'ab\min\min\min\min' and 'ab\max\max\max\max'. */ - if (contraction_flags && ptr + 1 < end && - contraction_flags[(uchar) *ptr]) + if (have_contractions && ptr + 1 < end && + my_uca_can_be_contraction_head(cs, (uchar) *ptr)) { /* Ptr[0] is a contraction head. */ @@ -659,8 +658,8 @@ fill_max_and_min: is not a contraction, then we put only ptr[0], and continue with ptr[1] on the next loop. */ - if (contraction_flags[(uchar) ptr[1]] && - cs->contractions[(*ptr-0x40)*0x40 + ptr[1] - 0x40]) + if (my_uca_can_be_contraction_tail(cs, (uchar) ptr[1]) && + my_uca_contraction2_weight(cs, (uchar) ptr[0], (uchar) ptr[1])) { /* Contraction found */ if (maxcharlen == 1 || min_str + 1 >= min_end) === modified file 'strings/ctype-uca.c' --- a/strings/ctype-uca.c 2009-11-16 20:49:51 +0000 +++ b/strings/ctype-uca.c 2009-11-30 12:42:24 +0000 @@ -6713,6 +6713,16 @@ static const char hungarian[]= "&U < \\u00FC <<< \\u00DC << \\u0171 <<< \\u0170"; +static const char croatian[]= + +"&C < \\u010D <<< \\u010C < \\u0107 <<< \\u0106 " +"&D < d\\u017E <<< \\u01C6 <<< D\\u017E <<< \\u01C5 <<< D\\u017D <<< \\u01C4 " +" < \\u0111 <<< \\u0110 " +"&L < lj <<< \\u01C9 <<< Lj <<< \\u01C8 <<< LJ <<< \\u01C7 " +"&N < nj <<< \\u01CC <<< Nj <<< \\u01CB <<< NJ <<< \\u01CA " +"&S < \\u0161 <<< \\u0160 " +"&Z < \\u017E <<< \\u017D"; + /* Unicode Collation Algorithm: Collation element (weight) scanner, @@ -6726,7 +6736,7 @@ typedef struct my_uca_scanner_st const uchar *send; /* End of the input string */ uchar *uca_length; uint16 **uca_weight; - uint16 *contractions; + MY_CONTRACTIONS *contractions; uint16 implicit[2]; int page; int code; @@ -6747,6 +6757,164 @@ typedef struct my_uca_scanner_handler_st static uint16 nochar[]= {0,0}; +#define MY_UCA_CNT_FLAG_SIZE 4096 +#define MY_UCA_CNT_FLAG_MASK 4095 + +#define MY_UCA_CNT_HEAD 1 +#define MY_UCA_CNT_TAIL 2 + + + + +/********** Helper functions to handle contraction ************/ + + +/** + Mark a character as a contraction part + + @cs Pointer to CHARSET_INFO data + @wc Unicode code point + @flag flag: "is contraction head", "is contraction tail" +*/ + +static void +my_uca_add_contraction_flag(CHARSET_INFO *cs, my_wc_t wc, int flag) +{ + cs->contractions->flags[wc & MY_UCA_CNT_FLAG_MASK]|= flag; +} + + +/** + Add a new contraction into contraction list + + @cs Pointer to CHARSET_INFO data + @wc Unicode code points of the characters + @len Number of characters + + @return New contraction + @retval Pointer to a newly added contraction +*/ + +static MY_CONTRACTION * +my_uca_add_contraction(CHARSET_INFO *cs, + my_wc_t *wc, int len __attribute__((unused))) +{ + MY_CONTRACTIONS *list= cs->contractions; + MY_CONTRACTION *next= &list->item[list->nitems]; + DBUG_ASSERT(len == 2); /* We currently support only contraction2 */ + next->ch[0]= wc[0]; + next->ch[1]= wc[1]; + list->nitems++; + return next; +} + + +/** + Allocate and initialize memory for contraction list and flags + + @cs Pointer to CHARSET_INFO data + @alloc Memory allocation function (typically points to my_alloc_once) + @n Number of contractions + + @return Error code + @retval 0 - memory allocated successfully + @retval 1 - not enough memory +*/ + +static my_bool +my_uca_alloc_contractions(CHARSET_INFO *cs, void *(*alloc)(size_t), size_t n) +{ + uint size= n * sizeof(MY_CONTRACTION); + if (!(cs->contractions= (*alloc)(sizeof(MY_CONTRACTIONS)))) + return 1; + bzero(cs->contractions, sizeof(MY_CONTRACTIONS)); + if (!(cs->contractions->item= (*alloc)(size)) || + !(cs->contractions->flags= (char*) (*alloc)(MY_UCA_CNT_FLAG_SIZE))) + return 1; + bzero((void*) cs->contractions->item, size); + bzero((void*) cs->contractions->flags, MY_UCA_CNT_FLAG_SIZE); + return 0; +} + + +/** + Check if UCA data has contractions (public version) + + @cs Pointer to CHARSET_INFO data + @retval 0 - no contraction, 1 - have contractions. +*/ + +my_bool +my_uca_have_contractions(CHARSET_INFO *cs) +{ + return cs->contractions != NULL; +} + + +/** + Check if a character can be contraction head + + @cs Pointer to CHARSET_INFO data + @wc Code point + + @retval 0 - cannot be contraction head + @retval 1 - can be contraction head +*/ + +my_bool +my_uca_can_be_contraction_head(CHARSET_INFO *cs, my_wc_t wc) +{ + return cs->contractions->flags[wc & MY_UCA_CNT_FLAG_MASK] & MY_UCA_CNT_HEAD; +} + + +/** + Check if a character can be contraction tail + + @cs Pointer to CHARSET_INFO data + @wc Code point + + @retval 0 - cannot be contraction tail + @retval 1 - can be contraction tail +*/ + +my_bool +my_uca_can_be_contraction_tail(CHARSET_INFO *cs, my_wc_t wc) +{ + return cs->contractions->flags[wc & MY_UCA_CNT_FLAG_MASK] & MY_UCA_CNT_TAIL; +} + + +/** + Find a contraction and return its weight array + + @cs Pointer to CHARSET data + @wc1 First character + @wc2 Second character + + @return Weight array + @retval NULL - no contraction found + @retval ptr - contraction weight array +*/ + +uint16 * +my_uca_contraction2_weight(CHARSET_INFO *cs, my_wc_t wc1, my_wc_t wc2) +{ + MY_CONTRACTIONS *list= cs->contractions; + MY_CONTRACTION *c, *last; + for (c= list->item, last= &list->item[list->nitems]; c < last; c++) + { + if (c->ch[0] == wc1 && c->ch[1] == wc2) + { + return c->weight; + } + } + return NULL; +} + + + + #ifdef HAVE_CHARSET_ucs2 /* Initialize collation weight scanner @@ -6766,7 +6934,7 @@ static uint16 nochar[]= {0,0}; */ static void my_uca_scanner_init_ucs2(my_uca_scanner *scanner, - CHARSET_INFO *cs __attribute__((unused)), + CHARSET_INFO *cs, const uchar *str, size_t length) { scanner->wbeg= nochar; @@ -6777,6 +6945,7 @@ static void my_uca_scanner_init_ucs2(my_ scanner->uca_length= cs->sort_order; scanner->uca_weight= cs->sort_order_big; scanner->contractions= cs->contractions; + scanner->cs= cs; return; } @@ -6865,18 +7034,23 @@ static int my_uca_scanner_next_ucs2(my_u if (scanner->contractions && (scanner->sbeg <= scanner->send)) { - int cweight; + my_wc_t wc1= ((scanner->page << 8) | scanner->code); - if (!scanner->page && !scanner->sbeg[0] && - (scanner->sbeg[1] > 0x40) && (scanner->sbeg[1] < 0x80) && - (scanner->code > 0x40) && (scanner->code < 0x80) && - (cweight= scanner->contractions[(scanner->code-0x40)*0x40+scanner->sbeg[1]-0x40])) + if (my_uca_can_be_contraction_head(scanner->cs, wc1)) + { + uint16 *cweight; + my_wc_t wc2= (((my_wc_t) scanner->sbeg[0]) << 8) | scanner->sbeg[1]; + if (my_uca_can_be_contraction_tail(scanner->cs, wc2) && + (cweight= my_uca_contraction2_weight(scanner->cs, + scanner->code, + scanner->sbeg[1]))) { scanner->implicit[0]= 0; scanner->wbeg= scanner->implicit; scanner->sbeg+=2; - return cweight; + return *cweight; } + } } if (!ucaw[scanner->page]) @@ -6959,23 +7133,22 @@ static int my_uca_scanner_next_any(my_uc scanner->code= wc & 0xFF; scanner->sbeg+= mb_len; - if (scanner->contractions && !scanner->page && - (scanner->code > 0x40) && (scanner->code < 0x80)) + if (my_uca_have_contractions(scanner->cs) && + my_uca_can_be_contraction_head(scanner->cs, wc)) { - uint page1, code1, cweight; + my_wc_t wc2; + uint16 *cweight; - if (((mb_len= scanner->cs->cset->mb_wc(scanner->cs, &wc, + if (((mb_len= scanner->cs->cset->mb_wc(scanner->cs, &wc2, scanner->sbeg, scanner->send)) >=0) && - (!(page1= (wc >> 8))) && - ((code1= (wc & 0xFF)) > 0x40) && - (code1 < 0x80) && - (cweight= scanner->contractions[(scanner->code-0x40)*0x40 + code1-0x40])) + my_uca_can_be_contraction_tail(scanner->cs, wc2) && + (cweight= my_uca_contraction2_weight(scanner->cs, wc, wc2))) { scanner->implicit[0]= 0; scanner->wbeg= scanner->implicit; scanner->sbeg+= mb_len; - return cweight; + return *cweight; } } @@ -7012,6 +7185,33 @@ static my_uca_scanner_handler my_any_uca my_uca_scanner_next_any }; + + +/** + Helper function: + Find address of weights of the given character. + + @weights UCA weight array + @lengths UCA length array + @ch character Unicode code point + + @return Weight array + @retval pointer to weight array for the given character, + or NULL if this page does not have implicit weights. +*/ + +static inline uint16 * +my_char_weight_addr(CHARSET_INFO *cs, uint wc) +{ + uint page= (wc >> 8); + uint ofst= wc & 0xFF; + return cs->sort_order_big[page] ? + cs->sort_order_big[page] + ofst * cs->sort_order[page] : + NULL; +} + + + /* Compares two strings according to the collation @@ -7683,8 +7883,8 @@ ex: typedef struct my_coll_rule_item_st { - uint base; /* Base character */ - uint curr[2]; /* Current character */ + my_wc_t base; /* Base character */ + my_wc_t curr[2]; /* Current character */ int diff[3]; /* Primary, Secondary and Tertiary difference */ } MY_COLL_RULE; @@ -7834,6 +8034,7 @@ static int my_coll_rule_parse(MY_COLL_RU static my_bool create_tailoring(CHARSET_INFO *cs, void *(*alloc)(size_t)) { MY_COLL_RULE rule[MY_MAX_COLL_RULE]; + MY_COLL_RULE *r, *rfirst, *rlast; char errstr[128]; uchar *newlengths; uint16 **newweights; @@ -7858,6 +8059,9 @@ static my_bool create_tailoring(CHARSET_ return 1; } + rfirst= rule; + rlast= rule + rc; + if (!cs->caseinfo) cs->caseinfo= my_unicase_default; @@ -7941,44 +8145,21 @@ static my_bool create_tailoring(CHARSET_ /* Now process contractions */ if (ncontractions) { - /* - 8K for weights for basic latin letter pairs, - plus 256 bytes for "is contraction part" flags. - */ - uint size= 0x40*0x40*sizeof(uint16) + 256; - char *contraction_flags; - if (!(cs->contractions= (uint16*) (*alloc)(size))) - return 1; - bzero((void*)cs->contractions, size); - contraction_flags= ((char*) cs->contractions) + 0x40*0x40; - for (i=0; i < rc; i++) + if (my_uca_alloc_contractions(cs, alloc, ncontractions)) + return 1; + for (r= rfirst; r < rlast; r++) { - if (rule[i].curr[1]) + uint16 *to; + if (r->curr[1]) /* Contraction */ { - uint pageb= (rule[i].base >> 8) & 0xFF; - uint chb= rule[i].base & 0xFF; - uint16 *offsb= defweights[pageb] + chb*deflengths[pageb]; - uint offsc; - - if (offsb[1] || - rule[i].curr[0] < 0x40 || rule[i].curr[0] > 0x7f || - rule[i].curr[1] < 0x40 || rule[i].curr[1] > 0x7f) - { - /* - TODO: add error reporting; - We support only basic latin letters contractions at this point. - Also, We don't support contractions with weight longer than one. - Otherwise, we'd need much more memory. - */ - return 1; - } - offsc= (rule[i].curr[0]-0x40)*0x40+(rule[i].curr[1]-0x40); - - /* Copy base weight applying primary difference */ - cs->contractions[offsc]= offsb[0] + rule[i].diff[0]; - /* Mark both letters as "is contraction part */ - contraction_flags[rule[i].curr[0]]= 1; - contraction_flags[rule[i].curr[1]]= 1; + /* Mark both letters as "is contraction part" */ + my_uca_add_contraction_flag(cs, r->curr[0], MY_UCA_CNT_HEAD); + my_uca_add_contraction_flag(cs, r->curr[1], MY_UCA_CNT_TAIL); + to= my_uca_add_contraction(cs, r->curr, 2)->weight; + /* Copy weight from the reset character */ + to[0]= my_char_weight_addr(cs, r->base)[0]; + /* Apply primary difference */ + to[0]+= r->diff[0]; } } } @@ -8701,6 +8882,39 @@ CHARSET_INFO my_charset_ucs2_hungarian_u }; +CHARSET_INFO my_charset_ucs2_croatian_uca_ci= +{ + 149,0,0, /* number */ + MY_CS_COMPILED|MY_CS_STRNXFRM|MY_CS_UNICODE, + "ucs2", /* cs name */ + "ucs2_croatian_ci", /* name */ + "", /* comment */ + croatian, /* tailoring */ + NULL, /* ctype */ + NULL, /* to_lower */ + NULL, /* to_upper */ + NULL, /* sort_order */ + NULL, /* contractions */ + NULL, /* sort_order_big*/ + NULL, /* tab_to_uni */ + NULL, /* tab_from_uni */ + my_unicase_default, /* caseinfo */ + NULL, /* state_map */ + NULL, /* ident_map */ + 8, /* strxfrm_multiply */ + 1, /* caseup_multiply */ + 1, /* casedn_multiply */ + 2, /* mbminlen */ + 2, /* mbmaxlen */ + 9, /* min_sort_char */ + 0xFFFF, /* max_sort_char */ + ' ', /* pad char */ + 0, /* escape_with_backslash_is_dangerous */ + &my_charset_ucs2_handler, + &my_collation_ucs2_uca_handler +}; + + #endif @@ -9358,6 +9572,38 @@ CHARSET_INFO my_charset_utf8_hungarian_u &my_collation_any_uca_handler }; +CHARSET_INFO my_charset_utf8_croatian_uca_ci= +{ + 213,0,0, /* number */ + MY_CS_COMPILED|MY_CS_STRNXFRM|MY_CS_UNICODE, + "utf8", /* cs name */ + "utf8_croatian_ci", /* name */ + "", /* comment */ + croatian, /* tailoring */ + ctype_utf8, /* ctype */ + NULL, /* to_lower */ + NULL, /* to_upper */ + NULL, /* sort_order */ + NULL, /* contractions */ + NULL, /* sort_order_big*/ + NULL, /* tab_to_uni */ + NULL, /* tab_from_uni */ + my_unicase_default, /* caseinfo */ + NULL, /* state_map */ + NULL, /* ident_map */ + 8, /* strxfrm_multiply */ + 1, /* caseup_multiply */ + 1, /* casedn_multiply */ + 1, /* mbminlen */ + 3, /* mbmaxlen */ + 9, /* min_sort_char */ + 0xFFFF, /* max_sort_char */ + ' ', /* pad char */ + 0, /* escape_with_backslash_is_dangerous */ + &my_charset_utf8_handler, + &my_collation_any_uca_handler +}; + #endif /* HAVE_CHARSET_utf8 */ #endif /* HAVE_UCA_COLLATIONS */ === modified file 'strings/ctype-ucs2.c' --- a/strings/ctype-ucs2.c 2009-10-15 21:38:29 +0000 +++ b/strings/ctype-ucs2.c 2009-11-30 12:42:24 +0000 @@ -1526,8 +1526,7 @@ my_bool my_like_range_ucs2(CHARSET_INFO char *min_org=min_str; char *min_end=min_str+res_length; size_t charlen= res_length / cs->mbmaxlen; - const char *contraction_flags= cs->contractions ? - ((const char*) cs->contractions) + 0x40*0x40 : NULL; + my_bool have_contractions= my_uca_have_contractions(cs); for ( ; ptr + 1 < end && min_str + 1 < min_end && charlen > 0 ; ptr+=2, charlen--) @@ -1567,8 +1566,9 @@ fill_max_and_min: return 0; } - if (contraction_flags && ptr + 3 < end && - ptr[0] == '\0' && contraction_flags[(uchar) ptr[1]]) + if (have_contractions && ptr + 3 < end && + ptr[0] == '\0' && + my_uca_can_be_contraction_head(cs, (uchar) ptr[1])) { /* Contraction head found */ if (ptr[2] == '\0' && (ptr[3] == w_one || ptr[3] == w_many)) @@ -1581,8 +1581,9 @@ fill_max_and_min: Check if the second letter can be contraction part, and if two letters really produce a contraction. */ - if (ptr[2] == '\0' && contraction_flags[(uchar) ptr[3]] && - cs->contractions[(ptr[1]-0x40)*0x40 + ptr[3] - 0x40]) + if (ptr[2] == '\0' && + my_uca_can_be_contraction_tail(cs, (uchar) ptr[3]) && + my_uca_contraction2_weight(cs,(uchar) ptr[1], (uchar) ptr[3])) { /* Contraction found */ if (charlen == 1 || min_str + 2 >= min_end)

1 0

[Maria-developers] MWL#47: On the question of annotations persisting from one binary log to another
by Sergey Petrunya 30 Nov '09

30 Nov '09

Hi Alex, My opinion is as follows: When log-slave-updates is not set, the slave is not binlogging the updates it is applying, so the comments do not go to the binary log either. When log-slave-updates is set, the action should depend on the binlog-annotate-row-events setting, like it's done on the master. That is, if binlog-annotate-row-events is on, annotate events should be passed to slave's binary log, otherwise they should be discarded. There is one hole left: if one does mysqlbinlog /path/to/binlog | mysql --host=slave-which-has-log-slave-updates then the annotations will not make it to slave's binary log. I think that should be ok, but we better get a confirmation of this to be sure. One should also take care that the slave do not write annotations for SQL statements in form of BINLOG 'xaqw324435324hnjsfp' as they are not useful (or at least don't copy the entire 'xaqe23....' part). BR Sergey -- Sergey Petrunia, Software Developer Monty Program AB, http://askmonty.org Blog: http://s.petrunia.net/blog

1 0

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (monty:2769)
by Michael Widenius 30 Nov '09

30 Nov '09

#At lp:maria based on revid:bo.thorsen@canonical.com-20091126153249-0xfszhcynbym2lvr 2769 Michael Widenius 2009-11-30 [merge] Automatic merge modified: .bzrignore BUILD/compile-pentium client/mysql_upgrade.c cmd-line-utils/readline/config_readline.h cmd-line-utils/readline/display.c cmd-line-utils/readline/history.c cmd-line-utils/readline/rlmbutil.h cmd-line-utils/readline/text.c cmd-line-utils/readline/xmalloc.c configure.in include/mysql/plugin.h include/mysql/plugin.h.pp libmysql/libmysql.c mysql-test/t/information_schema.test mysql-test/t/not_partition.test mysys/lf_hash.c mysys/my_redel.c regex/engine.c regex/engine.ih sql/log_event.cc sql/mysqld.cc sql/scheduler.cc sql/sp_head.cc sql/sql_base.cc sql/sql_builtin.cc.in sql/sql_class.cc sql/sql_class.h sql/sql_insert.cc sql/sql_parse.cc sql/sql_select.cc sql/sql_show.cc sql/table.cc storage/maria/ha_maria.cc storage/maria/lockman.c storage/maria/ma_check.c storage/maria/ma_check_standalone.h storage/maria/ma_ft_boolean_search.c storage/maria/ma_ft_nlq_search.c storage/maria/ma_ft_parser.c storage/maria/ma_ftdefs.h storage/maria/ma_sort.c storage/maria/ma_state.c storage/maria/maria_def.h storage/maria/maria_ftdump.c storage/maria/trnman.c storage/myisam/ft_boolean_search.c storage/myisam/ft_nlq_search.c storage/myisam/ft_parser.c storage/myisam/ft_stopwords.c storage/myisam/ftdefs.h storage/myisam/ha_myisam.cc storage/myisam/mi_check.c storage/myisam/myisam_ftdump.c storage/myisam/myisamchk.c storage/myisam/myisamdef.h storage/myisam/myisamlog.c storage/myisam/sort.c storage/xtradb/fil/fil0fil.c storage/xtradb/trx/trx0i_s.c === modified file '.bzrignore' --- a/.bzrignore 2009-11-06 17:22:32 +0000 +++ b/.bzrignore 2009-11-29 23:16:14 +0000 @@ -1922,3 +1922,4 @@ libmysqld/examples/mysqltest.cc extra/libevent/event-config.h libmysqld/opt_table_elimination.cc libmysqld/ha_federatedx.cc +tmp === modified file 'BUILD/compile-pentium' --- a/BUILD/compile-pentium 2007-04-11 12:12:00 +0000 +++ b/BUILD/compile-pentium 2009-11-29 23:08:56 +0000 @@ -4,7 +4,7 @@ path=`dirname $0` . "$path/SETUP.sh" extra_flags="$pentium_cflags $fast_cflags" -extra_configs="$pentium_configs $static_link" +extra_configs="$pentium_configs" strip=yes . "$path/FINISH.sh" === modified file 'client/mysql_upgrade.c' --- a/client/mysql_upgrade.c 2009-11-06 17:22:32 +0000 +++ b/client/mysql_upgrade.c 2009-11-29 23:08:56 +0000 @@ -552,7 +552,6 @@ static int upgrade_already_done(void) FILE *in; char upgrade_info_file[FN_REFLEN]= {0}; char buf[sizeof(MYSQL_SERVER_VERSION)+1]; - char *res; if (get_upgrade_info_file_name(upgrade_info_file)) return 0; /* Could not get filename => not sure */ === modified file 'cmd-line-utils/readline/config_readline.h' --- a/cmd-line-utils/readline/config_readline.h 2005-04-20 10:02:07 +0000 +++ b/cmd-line-utils/readline/config_readline.h 2009-11-29 23:08:56 +0000 @@ -7,6 +7,11 @@ # include <config.h> #endif +/* to get wcwidth() defined */ +#define _XOPEN_SOURCE 600 +#define _XOPEN_SOURCE_EXTENDED +#define _XOPEN_ + /* Ultrix botches type-ahead when switching from canonical to non-canonical mode, at least through version 4.3 === modified file 'cmd-line-utils/readline/display.c' --- a/cmd-line-utils/readline/display.c 2009-09-07 20:50:10 +0000 +++ b/cmd-line-utils/readline/display.c 2009-11-29 23:08:56 +0000 @@ -461,12 +461,12 @@ rl_redisplay () register char *line; int inv_botlin, lb_linenum, o_cpos; int newlines, lpos, temp, modmark; - char *prompt_this_line; + const char *prompt_this_line; #if defined (HANDLE_MULTIBYTE) - int num, n0; + int num, n0= 0; wchar_t wc; size_t wc_bytes; - int wc_width; + int wc_width= 0; mbstate_t ps; int _rl_wrapped_multicolumn = 0; #endif @@ -824,7 +824,7 @@ rl_redisplay () cpos_buffer_position = out; lb_linenum = newlines; } - for (i = in; i < in+wc_bytes; i++) + for (i = in; (size_t) i < in+wc_bytes; i++) line[out++] = rl_line_buffer[i]; for (i = 0; i < wc_width; i++) CHECK_LPOS(); === modified file 'cmd-line-utils/readline/history.c' --- a/cmd-line-utils/readline/history.c 2008-04-28 16:24:05 +0000 +++ b/cmd-line-utils/readline/history.c 2009-11-29 23:08:56 +0000 @@ -211,14 +211,14 @@ history_get (offset) HIST_ENTRY * alloc_history_entry (string, ts) - char *string; + const char *string; char *ts; { HIST_ENTRY *temp; temp = (HIST_ENTRY *)xmalloc (sizeof (HIST_ENTRY)); - temp->line = string ? savestring (string) : string; + temp->line = string ? savestring ((char*) string) : (char*) string; temp->data = (char *)NULL; temp->timestamp = ts; === modified file 'cmd-line-utils/readline/rlmbutil.h' --- a/cmd-line-utils/readline/rlmbutil.h 2007-11-19 13:38:08 +0000 +++ b/cmd-line-utils/readline/rlmbutil.h 2009-11-29 23:08:56 +0000 @@ -109,8 +109,8 @@ extern int _rl_is_mbchar_matched PARAMS( extern wchar_t _rl_char_value PARAMS((char *, int)); extern int _rl_walphabetic PARAMS((wchar_t)); -#define _rl_to_wupper(wc) (iswlower (wc) ? towupper (wc) : (wc)) -#define _rl_to_wlower(wc) (iswupper (wc) ? towlower (wc) : (wc)) +#define _rl_to_wupper(wc) (iswlower (wc) ? (wchar_t) towupper (wc) : (wc)) +#define _rl_to_wlower(wc) (iswupper (wc) ? (wchar_t) towlower (wc) : (wc)) #define MB_NEXTCHAR(b,s,c,f) \ ((MB_CUR_MAX > 1 && rl_byte_oriented == 0) \ === modified file 'cmd-line-utils/readline/text.c' --- a/cmd-line-utils/readline/text.c 2009-09-07 20:50:10 +0000 +++ b/cmd-line-utils/readline/text.c 2009-11-29 23:08:56 +0000 @@ -614,7 +614,7 @@ rl_arrow_keys (count, c) #ifdef HANDLE_MULTIBYTE static char pending_bytes[MB_LEN_MAX]; static int pending_bytes_length = 0; -static mbstate_t ps = {0}; +static mbstate_t ps; #endif /* Insert the character C at the current location, moving point forward. === modified file 'cmd-line-utils/readline/xmalloc.c' --- a/cmd-line-utils/readline/xmalloc.c 2008-01-23 16:43:46 +0000 +++ b/cmd-line-utils/readline/xmalloc.c 2009-11-29 23:08:56 +0000 @@ -42,7 +42,7 @@ static void memory_error_and_abort (fname) - char *fname; + const char *fname; { fprintf (stderr, "%s: out of virtual memory\n", fname); exit (2); === modified file 'configure.in' --- a/configure.in 2009-11-07 15:56:51 +0000 +++ b/configure.in 2009-11-29 23:08:56 +0000 @@ -15,7 +15,7 @@ AC_CANONICAL_SYSTEM # MySQL version number. # # Note: the following line must be parseable by win/configure.js:GetVersion() -AM_INIT_AUTOMAKE(mysql, 5.1.39-maria-beta) +AM_INIT_AUTOMAKE(mysql, 5.1.39-MariaDB-beta) AM_CONFIG_HEADER([include/config.h:config.h.in]) PROTOCOL_VERSION=10 === modified file 'include/mysql/plugin.h' --- a/include/mysql/plugin.h 2009-09-07 20:50:10 +0000 +++ b/include/mysql/plugin.h 2009-11-29 23:08:56 +0000 @@ -567,9 +567,9 @@ typedef struct st_mysql_ftparser_boolean typedef struct st_mysql_ftparser_param { int (*mysql_parse)(struct st_mysql_ftparser_param *, - char *doc, int doc_len); + const unsigned char *doc, size_t doc_len); int (*mysql_add_word)(struct st_mysql_ftparser_param *, - char *word, int word_len, + const unsigned char *word, size_t word_len, MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info); void *ftparser_state; void *mysql_ftparam; === modified file 'include/mysql/plugin.h.pp' --- a/include/mysql/plugin.h.pp 2008-10-10 15:28:41 +0000 +++ b/include/mysql/plugin.h.pp 2009-11-29 23:08:56 +0000 @@ -73,9 +73,9 @@ typedef struct st_mysql_ftparser_boolean typedef struct st_mysql_ftparser_param { int (*mysql_parse)(struct st_mysql_ftparser_param *, - char *doc, int doc_len); + const unsigned char *doc, size_t doc_len); int (*mysql_add_word)(struct st_mysql_ftparser_param *, - char *word, int word_len, + const unsigned char *word, size_t word_len, MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info); void *ftparser_state; void *mysql_ftparam; === modified file 'libmysql/libmysql.c' --- a/libmysql/libmysql.c 2009-11-06 17:22:32 +0000 +++ b/libmysql/libmysql.c 2009-11-29 23:08:56 +0000 @@ -2284,7 +2284,7 @@ mysql_stmt_param_metadata(MYSQL_STMT *st /* Store type of parameter in network buffer. */ -static void store_param_type(char **pos, MYSQL_BIND *param) +static void store_param_type(uchar **pos, MYSQL_BIND *param) { uint typecode= param->buffer_type | (param->is_unsigned ? 32768 : 0); int2store(*pos, typecode); @@ -2564,7 +2564,7 @@ int cli_stmt_execute(MYSQL_STMT *stmt) that is sent to the server. */ for (param= stmt->params; param < param_end ; param++) - store_param_type((char**) &net->write_pos, param); + store_param_type(&net->write_pos, param); } for (param= stmt->params; param < param_end; param++) === modified file 'mysql-test/t/information_schema.test' --- a/mysql-test/t/information_schema.test 2009-09-29 20:19:43 +0000 +++ b/mysql-test/t/information_schema.test 2009-11-29 23:08:56 +0000 @@ -5,6 +5,9 @@ # on the presence of the log tables (which are CSV-based). --source include/have_csv.inc +# Check that innodb/xtradb is incompiled in as result depends on it +-- source include/have_innodb.inc + # Save the initial number of concurrent sessions --source include/count_sessions.inc === modified file 'mysql-test/t/not_partition.test' --- a/mysql-test/t/not_partition.test 2009-01-08 14:16:44 +0000 +++ b/mysql-test/t/not_partition.test 2009-11-29 23:08:56 +0000 @@ -15,7 +15,7 @@ let $MYSQLD_DATADIR= `SELECT @@datadir`; # Bug#39893: Crash if select on a partitioned table, # when partitioning is disabled FLUSH TABLES; ---copy_file $MYSQLTEST_VARDIR/std_data_ln/parts/t1.frm $MYSQLD_DATADIR/test/t1.frm +--copy_file $MYSQLTEST_VARDIR/std_data/parts/t1.frm $MYSQLD_DATADIR/test/t1.frm SELECT * FROM t1; TRUNCATE TABLE t1; ANALYZE TABLE t1; === modified file 'mysys/lf_hash.c' --- a/mysys/lf_hash.c 2009-01-15 21:27:36 +0000 +++ b/mysys/lf_hash.c 2009-11-29 23:08:56 +0000 @@ -124,8 +124,8 @@ retry: we found a deleted node - be nice, help the other thread and remove this deleted node */ - if (my_atomic_casptr((void **)cursor->prev, - (void **)&cursor->curr, cursor->next)) + if (my_atomic_casptr((void **) cursor->prev, + (void **)(char*) &cursor->curr, cursor->next)) _lf_alloc_free(pins, cursor->curr); else { @@ -171,7 +171,8 @@ static LF_SLIST *linsert(LF_SLIST * vola node->link= (intptr)cursor.curr; DBUG_ASSERT(node->link != (intptr)node); /* no circular references */ DBUG_ASSERT(cursor.prev != &node->link); /* no circular references */ - if (my_atomic_casptr((void **)cursor.prev, (void **)&cursor.curr, node)) + if (my_atomic_casptr((void **) cursor.prev, + (void **)(char*) &cursor.curr, node)) { res= 1; /* inserted ok */ break; @@ -218,13 +219,13 @@ static int ldelete(LF_SLIST * volatile * else { /* mark the node deleted */ - if (my_atomic_casptr((void **)&(cursor.curr->link), - (void **)&cursor.next, + if (my_atomic_casptr((void **) (char*) &(cursor.curr->link), + (void **) (char*) &cursor.next, (void *)(((intptr)cursor.next) | 1))) { /* and remove it from the list */ if (my_atomic_casptr((void **)cursor.prev, - (void **)&cursor.curr, cursor.next)) + (void **)(char*)&cursor.curr, cursor.next)) _lf_alloc_free(pins, cursor.curr); else { @@ -493,7 +494,7 @@ static int initialize_bucket(LF_HASH *ha my_free((void *)dummy, MYF(0)); dummy= cur; } - my_atomic_casptr((void **)node, (void **)&tmp, dummy); + my_atomic_casptr((void **)node, (void **)(char*) &tmp, dummy); /* note that if the CAS above failed (after linsert() succeeded), it would mean that some other thread has executed linsert() for === modified file 'mysys/my_redel.c' --- a/mysys/my_redel.c 2009-11-06 17:22:32 +0000 +++ b/mysys/my_redel.c 2009-11-29 23:08:56 +0000 @@ -77,9 +77,6 @@ end: int my_copystat(const char *from, const char *to, int MyFlags) { struct stat statbuf; -#if !defined(__WIN__) && !defined(__NETWARE__) - int res; -#endif if (stat((char*) from, &statbuf)) { === modified file 'regex/engine.c' --- a/regex/engine.c 2005-09-29 01:20:31 +0000 +++ b/regex/engine.c 2009-11-29 23:08:56 +0000 @@ -33,11 +33,11 @@ struct match { struct re_guts *g; int eflags; my_regmatch_t *pmatch; /* [nsub+1] (0 element unused) */ - char *offp; /* offsets work from here */ - char *beginp; /* start of string -- virtual NUL precedes */ - char *endp; /* end of string -- virtual NUL here */ - char *coldp; /* can be no match starting before here */ - char **lastpos; /* [nplus+1] */ + const char *offp; /* offsets work from here */ + const char *beginp; /* start of string -- virtual NUL precedes */ + const char *endp; /* end of string -- virtual NUL here */ + const char *coldp; /* can be no match starting before here */ + const char **lastpos; /* [nplus+1] */ STATEVARS; states st; /* current states */ states fresh; /* states for a fresh start */ @@ -66,20 +66,20 @@ static int /* 0 success, REG_NOMATCH f matcher(charset,g, str, nmatch, pmatch, eflags) CHARSET_INFO *charset; register struct re_guts *g; -char *str; +const char *str; size_t nmatch; my_regmatch_t pmatch[]; int eflags; { - register char *endp; + register const char *endp; register uint i; struct match mv; register struct match *m = &mv; - register char *dp; + register const char *dp; register const sopno gf = g->firststate+1; /* +1 for OEND */ register const sopno gl = g->laststate; - char *start; - char *stop; + const char *start; + const char *stop; /* simplify the situation where possible */ if (g->cflags&REG_NOSUB) @@ -163,7 +163,7 @@ int eflags; dp = dissect(charset, m, m->coldp, endp, gf, gl); } else { if (g->nplus > 0 && m->lastpos == NULL) - m->lastpos = (char **)malloc((g->nplus+1) * + m->lastpos = (const char **)malloc((g->nplus+1) * sizeof(char *)); if (g->nplus > 0 && m->lastpos == NULL) { free(m->pmatch); @@ -235,28 +235,28 @@ int eflags; == static char *dissect(register struct match *m, char *start, \ == char *stop, sopno startst, sopno stopst); */ -static char * /* == stop (success) always */ +static const char * /* == stop (success) always */ dissect(charset, m, start, stop, startst, stopst) CHARSET_INFO *charset; register struct match *m; -char *start; -char *stop; +const char *start; +const char *stop; sopno startst; sopno stopst; { register uint i; register sopno ss; /* start sop of current subRE */ register sopno es; /* end sop of current subRE */ - register char *sp; /* start of string matched by it */ - register char *stp; /* string matched by it cannot pass here */ - register char *rest; /* start of rest of string */ - register char *tail; /* string unmatched by rest of RE */ + register const char *sp; /* start of string matched by it */ + register const char *stp; /* string matched by it cannot pass here */ + register const char *rest; /* start of rest of string */ + register const char *tail; /* string unmatched by rest of RE */ register sopno ssub; /* start sop of subsubRE */ register sopno esub; /* end sop of subsubRE */ - register char *ssp; /* start of string matched by subsubRE */ - register char *sep; /* end of string matched by subsubRE */ - register char *oldssp; /* previous ssp */ - register char *dp; /* used in debug mode to check asserts */ + register const char *ssp; /* start of string matched by subsubRE */ + register const char *sep; /* end of string matched by subsubRE */ + register const char *oldssp; /* previous ssp */ + register const char *dp; /* used in debug mode to check asserts */ AT("diss", start, stop, startst, stopst); sp = start; @@ -424,23 +424,23 @@ sopno stopst; == static char *backref(register struct match *m, char *start, \ == char *stop, sopno startst, sopno stopst, sopno lev); */ -static char * /* == stop (success) or NULL (failure) */ +static const char * /* == stop (success) or NULL (failure) */ backref(charset,m, start, stop, startst, stopst, lev) CHARSET_INFO *charset; register struct match *m; -char *start; -char *stop; +const char *start; +const char *stop; sopno startst; sopno stopst; sopno lev; /* PLUS nesting level */ { register uint i; register sopno ss; /* start sop of current subRE */ - register char *sp; /* start of string matched by it */ + register const char *sp; /* start of string matched by it */ register sopno ssub; /* start sop of subsubRE */ register sopno esub; /* end sop of subsubRE */ - register char *ssp; /* start of string matched by subsubRE */ - register char *dp; + register const char *ssp; /* start of string matched by subsubRE */ + register const char *dp; register size_t len; register int hard; register sop s; @@ -630,24 +630,24 @@ sopno lev; /* PLUS nesting level */ == static char *fast(register struct match *m, char *start, \ == char *stop, sopno startst, sopno stopst); */ -static char * /* where tentative match ended, or NULL */ +static const char * /* where tentative match ended, or NULL */ fast(charset, m, start, stop, startst, stopst) CHARSET_INFO *charset; register struct match *m; -char *start; -char *stop; +const char *start; +const char *stop; sopno startst; sopno stopst; { register states st = m->st; register states fresh = m->fresh; register states tmp = m->tmp; - register char *p = start; + register const char *p = start; register int c = (start == m->beginp) ? OUT : *(start-1); register int lastc; /* previous c */ register int flagch; register int i; - register char *coldp; /* last p after which no match was underway */ + register const char *coldp; /* last p after which no match was underway */ CLEAR(st); SET1(st, startst); @@ -722,24 +722,24 @@ sopno stopst; == static char *slow(register struct match *m, char *start, \ == char *stop, sopno startst, sopno stopst); */ -static char * /* where it ended */ +static const char * /* where it ended */ slow(charset, m, start, stop, startst, stopst) CHARSET_INFO *charset; register struct match *m; -char *start; -char *stop; +const char *start; +const char *stop; sopno startst; sopno stopst; { register states st = m->st; register states empty = m->empty; register states tmp = m->tmp; - register char *p = start; + register const char *p = start; register int c = (start == m->beginp) ? OUT : *(start-1); register int lastc; /* previous c */ register int flagch; register int i; - register char *matchp; /* last p at which a match ended */ + register const char *matchp; /* last p at which a match ended */ AT("slow", start, stop, startst, stopst); CLEAR(st); === modified file 'regex/engine.ih' --- a/regex/engine.ih 2005-09-29 00:08:24 +0000 +++ b/regex/engine.ih 2009-11-29 23:08:56 +0000 @@ -4,11 +4,11 @@ extern "C" { #endif /* === engine.c === */ -static int matcher(CHARSET_INFO *charset,register struct re_guts *g, char *string, size_t nmatch, my_regmatch_t pmatch[], int eflags); -static char *dissect(CHARSET_INFO *charset,register struct match *m, char *start, char *stop, sopno startst, sopno stopst); -static char *backref(CHARSET_INFO *charset, register struct match *m, char *start, char *stop, sopno startst, sopno stopst, sopno lev); -static char *fast(CHARSET_INFO *charset, register struct match *m, char *start, char *stop, sopno startst, sopno stopst); -static char *slow(CHARSET_INFO *charset, register struct match *m, char *start, char *stop, sopno startst, sopno stopst); +static int matcher(CHARSET_INFO *charset,register struct re_guts *g, const char *string, size_t nmatch, my_regmatch_t pmatch[], int eflags); +static const char *dissect(CHARSET_INFO *charset,register struct match *m, const char *start, const char *stop, sopno startst, sopno stopst); +static const char *backref(CHARSET_INFO *charset, register struct match *m, const char *start, const char *stop, sopno startst, sopno stopst, sopno lev); +static const char *fast(CHARSET_INFO *charset, register struct match *m, const char *start, const char *stop, sopno startst, sopno stopst); +static const char *slow(CHARSET_INFO *charset, register struct match *m, const char *start, const char *stop, sopno startst, sopno stopst); static states step(register struct re_guts *g, sopno start, sopno stop, register states bef, int ch, register states aft); #define BOL (OUT+1) #define EOL (BOL+1) === modified file 'sql/log_event.cc' --- a/sql/log_event.cc 2009-11-26 15:32:49 +0000 +++ b/sql/log_event.cc 2009-11-30 11:13:02 +0000 @@ -2142,7 +2142,7 @@ static void write_str_with_code_and_len( */ DBUG_ASSERT(len <= 255); DBUG_ASSERT(src); - *((*dst)++)= code; + *((*dst)++)= (uchar) code; *((*dst)++)= (uchar) len; bmove(*dst, src, len); (*dst)+= len; === modified file 'sql/mysqld.cc' --- a/sql/mysqld.cc 2009-10-31 19:22:50 +0000 +++ b/sql/mysqld.cc 2009-11-26 20:19:33 +0000 @@ -957,6 +957,7 @@ static void close_connections(void) tmp->killed= THD::KILL_CONNECTION; thread_scheduler.post_kill_notification(tmp); + pthread_mutex_lock(&tmp->LOCK_thd_data); if (tmp->mysys_var) { tmp->mysys_var->abort=1; @@ -979,6 +980,7 @@ static void close_connections(void) } pthread_mutex_unlock(&tmp->mysys_var->mutex); } + pthread_mutex_unlock(&tmp->LOCK_thd_data); } (void) pthread_mutex_unlock(&LOCK_thread_count); // For unlink from list === modified file 'sql/scheduler.cc' --- a/sql/scheduler.cc 2009-09-07 20:50:10 +0000 +++ b/sql/scheduler.cc 2009-11-26 20:19:33 +0000 @@ -235,9 +235,7 @@ void thd_scheduler::thread_detach() if (thread_attached) { THD* thd = (THD*)list.data; - pthread_mutex_lock(&thd->LOCK_thd_data); - thd->mysys_var= NULL; - pthread_mutex_unlock(&thd->LOCK_thd_data); + thd->reset_globals(); thread_attached= FALSE; #ifndef DBUG_OFF /* === modified file 'sql/sp_head.cc' --- a/sql/sp_head.cc 2009-09-15 10:46:35 +0000 +++ b/sql/sp_head.cc 2009-11-29 23:08:56 +0000 @@ -1924,9 +1924,10 @@ sp_head::execute_procedure(THD *thd, Lis if (spvar->mode == sp_param_out) { Item_null *null_item= new Item_null(); + Item *tmp_item= (Item*) null_item; if (!null_item || - nctx->set_variable(thd, i, (Item **)&null_item)) + nctx->set_variable(thd, i, &tmp_item)) { err_status= TRUE; break; === modified file 'sql/sql_base.cc' --- a/sql/sql_base.cc 2009-10-15 21:38:29 +0000 +++ b/sql/sql_base.cc 2009-11-29 23:08:56 +0000 @@ -5823,6 +5823,7 @@ find_field_in_natural_join(THD *thd, TAB { /* This is a base table. */ DBUG_ASSERT(nj_col->view_field == NULL); + Item *ref= 0; /* This fix_fields is not necessary (initially this item is fixed by the Item_field constructor; after reopen_tables the Item_func_eq @@ -5830,12 +5831,13 @@ find_field_in_natural_join(THD *thd, TAB reopening for columns that was dropped by the concurrent connection. */ if (!nj_col->table_field->fixed && - nj_col->table_field->fix_fields(thd, (Item **)&nj_col->table_field)) + nj_col->table_field->fix_fields(thd, &ref)) { DBUG_PRINT("info", ("column '%s' was dropped by the concurrent connection", nj_col->table_field->name)); DBUG_RETURN(NULL); } + DBUG_ASSERT(ref == 0); // Should not have changed DBUG_ASSERT(nj_col->table_ref->table == nj_col->table_field->field->table); found_field= nj_col->table_field->field; update_field_dependencies(thd, found_field, nj_col->table_ref->table); === modified file 'sql/sql_builtin.cc.in' --- a/sql/sql_builtin.cc.in 2006-12-31 01:29:11 +0000 +++ b/sql/sql_builtin.cc.in 2009-11-29 23:08:56 +0000 @@ -13,6 +13,7 @@ along with this program; if not, write to the Free Software Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA */ +#include <my_global.h> #include <mysql/plugin.h> typedef struct st_mysql_plugin builtin_plugin[]; === modified file 'sql/sql_class.cc' --- a/sql/sql_class.cc 2009-09-15 10:46:35 +0000 +++ b/sql/sql_class.cc 2009-11-26 20:19:33 +0000 @@ -1166,6 +1166,19 @@ bool THD::store_globals() } +/** + Untie THD from current thread + + Used when using --thread-handling=pool-of-threads +*/ + +void THD::reset_globals() +{ + pthread_mutex_lock(&LOCK_thd_data); + mysys_var= 0; + pthread_mutex_unlock(&LOCK_thd_data); +} + /* Cleanup after query. === modified file 'sql/sql_class.h' --- a/sql/sql_class.h 2009-09-15 10:46:35 +0000 +++ b/sql/sql_class.h 2009-11-26 20:19:33 +0000 @@ -1906,6 +1906,7 @@ public: void cleanup(void); void cleanup_after_query(); bool store_globals(); + void reset_globals(); #ifdef SIGNAL_WITH_VIO_CLOSE inline void set_active_vio(Vio* vio) { === modified file 'sql/sql_insert.cc' --- a/sql/sql_insert.cc 2009-11-07 10:34:19 +0000 +++ b/sql/sql_insert.cc 2009-11-26 20:19:33 +0000 @@ -2279,6 +2279,7 @@ void kill_delayed_threads(void) while ((di= it++)) { di->thd.killed= THD::KILL_CONNECTION; + pthread_mutex_lock(&di->thd.LOCK_thd_data); if (di->thd.mysys_var) { pthread_mutex_lock(&di->thd.mysys_var->mutex); @@ -2297,6 +2298,7 @@ void kill_delayed_threads(void) } pthread_mutex_unlock(&di->thd.mysys_var->mutex); } + pthread_mutex_unlock(&di->thd.LOCK_thd_data); } VOID(pthread_mutex_unlock(&LOCK_delayed_insert)); // For unlink from list } === modified file 'sql/sql_parse.cc' --- a/sql/sql_parse.cc 2009-11-06 17:22:32 +0000 +++ b/sql/sql_parse.cc 2009-11-29 23:08:56 +0000 @@ -464,7 +464,7 @@ pthread_handler_t handle_bootstrap(void thd->init_for_queries(); while (fgets(buff, thd->net.max_packet, file)) { - char *query, *res; + char *query; /* strlen() can't be deleted because fgets() doesn't return length */ ulong length= (ulong) strlen(buff); while (buff[length-1] != '\n' && !feof(file)) === modified file 'sql/sql_select.cc' --- a/sql/sql_select.cc 2009-11-06 17:22:32 +0000 +++ b/sql/sql_select.cc 2009-11-29 23:08:56 +0000 @@ -3586,8 +3586,9 @@ add_key_fields(JOIN *join, KEY_FIELD **k { if (!field->eq(item->field)) { + Item *tmp_item= (Item*) item; add_key_field(key_fields, *and_level, cond_func, field, - TRUE, (Item **) &item, 1, usable_tables, + TRUE, &tmp_item, 1, usable_tables, sargables); } } @@ -15748,7 +15749,11 @@ static bool add_ref_to_table_cond(THD *t DBUG_RETURN(TRUE); if (!cond->fixed) - cond->fix_fields(thd, (Item**)&cond); + { + Item *tmp_item= (Item*) cond; + cond->fix_fields(thd, &tmp_item); + DBUG_ASSERT(cond == tmp_item); + } if (join_tab->select) { error=(int) cond->add(join_tab->select->cond); === modified file 'sql/sql_show.cc' --- a/sql/sql_show.cc 2009-11-02 09:30:21 +0000 +++ b/sql/sql_show.cc 2009-11-29 23:08:56 +0000 @@ -1747,6 +1747,7 @@ void mysqld_list_processes(THD *thd,cons if ((thd_info->db=tmp->db)) // Safe test thd_info->db=thd->strdup(thd_info->db); thd_info->command=(int) tmp->command; + pthread_mutex_lock(&tmp->LOCK_thd_data); if ((mysys_var= tmp->mysys_var)) pthread_mutex_lock(&mysys_var->mutex); thd_info->proc_info= (char*) (tmp->killed == THD::KILL_CONNECTION? "Killed" : 0); @@ -1766,6 +1767,7 @@ void mysqld_list_processes(THD *thd,cons #endif if (mysys_var) pthread_mutex_unlock(&mysys_var->mutex); + pthread_mutex_unlock(&tmp->LOCK_thd_data); thd_info->start_time= tmp->start_time; thd_info->query=0; @@ -3549,7 +3551,9 @@ static int get_schema_tables_record(THD TABLE_SHARE *share= show_table->s; handler *file= show_table->file; handlerton *tmp_db_type= share->db_type(); +#ifdef WITH_PARTITION_STORAGE_ENGINE bool is_partitioned= FALSE; +#endif if (share->tmp_table == SYSTEM_TMP_TABLE) table->field[3]->store(STRING_WITH_LEN("SYSTEM VIEW"), cs); else if (share->tmp_table) === modified file 'sql/table.cc' --- a/sql/table.cc 2009-10-15 21:38:29 +0000 +++ b/sql/table.cc 2009-11-29 23:08:56 +0000 @@ -2077,8 +2077,9 @@ ulong get_form_pos(File file, uchar *hea else { char *str; + const char **tmp = (const char**) (char*) buf; str=(char *) (buf+a_length); - fix_type_pointers((const char ***) &buf,save_names,1,&str); + fix_type_pointers(&tmp, save_names, 1, &str); } DBUG_RETURN(ret_value); } === modified file 'storage/maria/ha_maria.cc' --- a/storage/maria/ha_maria.cc 2009-11-07 10:34:19 +0000 +++ b/storage/maria/ha_maria.cc 2009-11-29 23:08:56 +0000 @@ -665,10 +665,9 @@ int maria_check_definition(MARIA_KEYDEF extern "C" { -volatile int *_ma_killed_ptr(HA_CHECK *param) +int _ma_killed_ptr(HA_CHECK *param) { - /* In theory Unsafe conversion, but should be ok for now */ - return (int*) &(((THD *) (param->thd))->killed); + return thd_killed((THD*)param->thd); } === modified file 'storage/maria/lockman.c' --- a/storage/maria/lockman.c 2008-02-21 00:51:51 +0000 +++ b/storage/maria/lockman.c 2009-11-29 23:08:56 +0000 @@ -360,7 +360,7 @@ retry: else { if (my_atomic_casptr((void **)cursor->prev, - (void **)&cursor->curr, cursor->next)) + (void **)(char*) &cursor->curr, cursor->next)) _lf_alloc_free(pins, cursor->curr); else { @@ -421,7 +421,8 @@ static int lockinsert(LOCK * volatile *h node->link= (intptr)cursor.curr; DBUG_ASSERT(node->link != (intptr)node); DBUG_ASSERT(cursor.prev != &node->link); - if (!my_atomic_casptr((void **)cursor.prev, (void **)&cursor.curr, node)) + if (!my_atomic_casptr((void **)cursor.prev, + (void **)(char*) &cursor.curr, node)) { res= REPEAT_ONCE_MORE; node->flags&= ~ACTIVE; @@ -498,11 +499,11 @@ static int lockdelete(LOCK * volatile *h then we can delete. Good news is - this is only required when rolling back a savepoint. */ - if (my_atomic_casptr((void **)&(cursor.curr->link), - (void **)&cursor.next, 1+(char *)cursor.next)) + if (my_atomic_casptr((void **)(char*)&(cursor.curr->link), + (void **)(char*)&cursor.next, 1+(char *)cursor.next)) { if (my_atomic_casptr((void **)cursor.prev, - (void **)&cursor.curr, cursor.next)) + (void **)(char*)&cursor.curr, cursor.next)) _lf_alloc_free(pins, cursor.curr); else lockfind(head, node, &cursor, pins); @@ -573,7 +574,7 @@ static void initialize_bucket(LOCKMAN *l my_free((void *)dummy, MYF(0)); dummy= cur; } - my_atomic_casptr((void **)node, (void **)&tmp, dummy); + my_atomic_casptr((void **)node, (void **)(char*) &tmp, dummy); } static inline uint calc_hash(uint64 resource) === modified file 'storage/maria/ma_check.c' --- a/storage/maria/ma_check.c 2009-05-06 12:03:24 +0000 +++ b/storage/maria/ma_check.c 2009-11-29 23:08:56 +0000 @@ -215,7 +215,7 @@ int maria_chk_del(HA_CHECK *param, regis empty=0; for (i= share->state.state.del ; i > 0L && next_link != HA_OFFSET_ERROR ; i--) { - if (*_ma_killed_ptr(param)) + if (_ma_killed_ptr(param)) DBUG_RETURN(1); if (test_flag & T_VERBOSE) printf(" %9s",llstr(next_link,buff)); @@ -310,7 +310,7 @@ static int check_k_link(HA_CHECK *param, records= (ha_rows) (share->state.state.key_file_length / block_size); while (next_link != HA_OFFSET_ERROR && records > 0) { - if (*_ma_killed_ptr(param)) + if (_ma_killed_ptr(param)) DBUG_RETURN(1); if (param->testflag & T_VERBOSE) printf("%16s",llstr(next_link,llbuff)); @@ -876,10 +876,10 @@ static int chk_index(HA_CHECK *param, MA tmp_key.data= tmp_key_buff; for ( ;; ) { - if (*_ma_killed_ptr(param)) - goto err; if (nod_flag) { + if (_ma_killed_ptr(param)) + goto err; next_page= _ma_kpos(nod_flag,keypos); if (chk_index_down(param,info,keyinfo,next_page, temp_buff,keys,key_checksum,level+1)) @@ -1180,7 +1180,7 @@ static int check_static_record(HA_CHECK pos= 0; while (pos < share->state.state.data_file_length) { - if (*_ma_killed_ptr(param)) + if (_ma_killed_ptr(param)) return -1; if (my_b_read(&param->read_cache, record, share->base.pack_reclength)) @@ -1230,7 +1230,7 @@ static int check_dynamic_record(HA_CHECK { my_bool got_error= 0; int flag; - if (*_ma_killed_ptr(param)) + if (_ma_killed_ptr(param)) DBUG_RETURN(-1); flag= block_info.second_read=0; @@ -1451,7 +1451,7 @@ static int check_compressed_record(HA_CH pos= share->pack.header_length; /* Skip header */ while (pos < share->state.state.data_file_length) { - if (*_ma_killed_ptr(param)) + if (_ma_killed_ptr(param)) DBUG_RETURN(-1); if (_ma_read_cache(&param->read_cache, block_info.header, pos, @@ -1815,7 +1815,7 @@ static int check_block_record(HA_CHECK * LINT_INIT(row_count); LINT_INIT(empty_space); - if (*_ma_killed_ptr(param)) + if (_ma_killed_ptr(param)) { _ma_scan_end_block_record(info); return -1; @@ -4631,7 +4631,7 @@ static int sort_get_next_record(MARIA_SO char llbuff[22],llbuff2[22]; DBUG_ENTER("sort_get_next_record"); - if (*_ma_killed_ptr(param)) + if (_ma_killed_ptr(param)) DBUG_RETURN(1); switch (sort_info->org_data_file_type) { === modified file 'storage/maria/ma_check_standalone.h' --- a/storage/maria/ma_check_standalone.h 2007-10-03 16:10:32 +0000 +++ b/storage/maria/ma_check_standalone.h 2009-11-29 23:08:56 +0000 @@ -30,11 +30,9 @@ Check if check/repair operation was killed by a signal */ -static int not_killed= 0; - -volatile int *_ma_killed_ptr(HA_CHECK *param __attribute__((unused))) +int _ma_killed_ptr(HA_CHECK *param __attribute__((unused))) { - return &not_killed; /* always NULL */ + return 0; } /* print warnings and errors */ === modified file 'storage/maria/ma_ft_boolean_search.c' --- a/storage/maria/ma_ft_boolean_search.c 2009-02-19 09:01:25 +0000 +++ b/storage/maria/ma_ft_boolean_search.c 2009-11-29 23:08:56 +0000 @@ -180,7 +180,7 @@ typedef struct st_my_ftb_param static int ftb_query_add_word(MYSQL_FTPARSER_PARAM *param, - char *word, int word_len, + const uchar *word, size_t word_len, MYSQL_FTPARSER_BOOLEAN_INFO *info) { MY_FTB_PARAM *ftb_param= param->mysql_ftparam; @@ -282,19 +282,19 @@ static int ftb_query_add_word(MYSQL_FTPA static int ftb_parse_query_internal(MYSQL_FTPARSER_PARAM *param, - char *query, int len) + const uchar *query, size_t len) { MY_FTB_PARAM *ftb_param= param->mysql_ftparam; MYSQL_FTPARSER_BOOLEAN_INFO info; CHARSET_INFO *cs= ftb_param->ftb->charset; - uchar **start= (uchar**) &query; - uchar *end= (uchar*) query + len; + const uchar **start= &query; + const uchar *end= query + len; FT_WORD w; info.prev= ' '; info.quot= 0; while (maria_ft_get_word(cs, start, end, &w, &info)) - param->mysql_add_word(param, (char *) w.pos, w.len, &info); + param->mysql_add_word(param, w.pos, w.len, &info); return(0); } @@ -615,7 +615,7 @@ typedef struct st_my_ftb_phrase_param static int ftb_phrase_add_word(MYSQL_FTPARSER_PARAM *param, - char *word, int word_len, + const uchar *word, size_t word_len, MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info __attribute__((unused))) { MY_FTB_PHRASE_PARAM *phrase_param= param->mysql_ftparam; @@ -647,15 +647,15 @@ static int ftb_phrase_add_word(MYSQL_FTP static int ftb_check_phrase_internal(MYSQL_FTPARSER_PARAM *param, - char *document, int len) + const uchar *document, size_t len) { FT_WORD word; MY_FTB_PHRASE_PARAM *phrase_param= param->mysql_ftparam; - const uchar *docend= (uchar*) document + len; - while (maria_ft_simple_get_word(phrase_param->cs, (uchar**) &document, + const uchar *docend= document + len; + while (maria_ft_simple_get_word(phrase_param->cs, &document, docend, &word, FALSE)) { - param->mysql_add_word(param, (char*) word.pos, word.len, 0); + param->mysql_add_word(param, word.pos, word.len, 0); if (phrase_param->match) break; } @@ -872,7 +872,7 @@ typedef struct st_my_ftb_find_param static int ftb_find_relevance_add_word(MYSQL_FTPARSER_PARAM *param, - char *word, int len, + const uchar *word, size_t len, MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info __attribute__((unused))) { MY_FTB_FIND_PARAM *ftb_param= param->mysql_ftparam; @@ -933,15 +933,14 @@ static int ftb_find_relevance_add_word(M static int ftb_find_relevance_parse(MYSQL_FTPARSER_PARAM *param, - char *doc, int len) + const uchar *doc, size_t len) { MY_FTB_FIND_PARAM *ftb_param= param->mysql_ftparam; FT_INFO *ftb= ftb_param->ftb; - uchar *end= (uchar*) doc + len; + const uchar *end= doc + len; FT_WORD w; - while (maria_ft_simple_get_word(ftb->charset, (uchar**) &doc, - end, &w, TRUE)) - param->mysql_add_word(param, (char *) w.pos, w.len, 0); + while (maria_ft_simple_get_word(ftb->charset, &doc, end, &w, TRUE)) + param->mysql_add_word(param, w.pos, w.len, 0); return(0); } === modified file 'storage/maria/ma_ft_nlq_search.c' --- a/storage/maria/ma_ft_nlq_search.c 2009-01-09 04:23:25 +0000 +++ b/storage/maria/ma_ft_nlq_search.c 2009-11-29 23:08:56 +0000 @@ -63,7 +63,8 @@ static int FT_SUPERDOC_cmp(void* cmp_arg static int walk_and_match(FT_WORD *word, uint32 count, ALL_IN_ONE *aio) { - int subkeys, r; + int32 subkeys; + int r; uint doc_cnt; FT_SUPERDOC sdoc, *sptr; TREE_ELEMENT *selem; @@ -127,7 +128,7 @@ static int walk_and_match(FT_WORD *word, goto do_skip; } #if HA_FT_WTYPE == HA_KEYTYPE_FLOAT - tmp_weight=*(float*)&subkeys; + tmp_weight=*(float*) (char*) &subkeys; #else #error #endif === modified file 'storage/maria/ma_ft_parser.c' --- a/storage/maria/ma_ft_parser.c 2009-02-19 09:01:25 +0000 +++ b/storage/maria/ma_ft_parser.c 2009-11-29 23:08:56 +0000 @@ -109,10 +109,11 @@ my_bool maria_ft_boolean_check_syntax_st 3 - right bracket 4 - stopword found */ -uchar maria_ft_get_word(CHARSET_INFO *cs, uchar **start, uchar *end, +uchar maria_ft_get_word(CHARSET_INFO *cs, const uchar **start, + const uchar *end, FT_WORD *word, MYSQL_FTPARSER_BOOLEAN_INFO *param) { - uchar *doc=*start; + const uchar *doc= *start; int ctype; uint mwc, length; int mbl; @@ -203,11 +204,11 @@ ret: return param->type; } -uchar maria_ft_simple_get_word(CHARSET_INFO *cs, uchar **start, +uchar maria_ft_simple_get_word(CHARSET_INFO *cs, const uchar **start, const uchar *end, FT_WORD *word, my_bool skip_stopwords) { - uchar *doc= *start; + const uchar *doc= *start; uint mwc, length; int ctype, mbl; DBUG_ENTER("maria_ft_simple_get_word"); @@ -259,8 +260,9 @@ void maria_ft_parse_init(TREE *wtree, CH static int maria_ft_add_word(MYSQL_FTPARSER_PARAM *param, - char *word, int word_len, - MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info __attribute__((unused))) + const uchar *word, size_t word_len, + MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info + __attribute__((unused))) { TREE *wtree; FT_WORD w; @@ -276,7 +278,7 @@ static int maria_ft_add_word(MYSQL_FTPAR w.pos= ptr; } else - w.pos= (uchar *) word; + w.pos= word; w.len= word_len; if (!tree_insert(wtree, &w, 0, wtree->custom_arg)) { @@ -288,17 +290,17 @@ static int maria_ft_add_word(MYSQL_FTPAR static int maria_ft_parse_internal(MYSQL_FTPARSER_PARAM *param, - char *doc_arg, int doc_len) + const uchar *doc_arg, size_t doc_len) { - uchar *doc= (uchar*) doc_arg; - uchar *end= doc + doc_len; + const uchar *doc= doc_arg; + const uchar *end= doc + doc_len; MY_FT_PARSER_PARAM *ft_param=param->mysql_ftparam; TREE *wtree= ft_param->wtree; FT_WORD w; DBUG_ENTER("maria_ft_parse_internal"); while (maria_ft_simple_get_word(wtree->custom_arg, &doc, end, &w, TRUE)) - if (param->mysql_add_word(param, (char *) w.pos, w.len, 0)) + if (param->mysql_add_word(param, w.pos, w.len, 0)) DBUG_RETURN(1); DBUG_RETURN(0); } @@ -378,7 +380,7 @@ MYSQL_FTPARSER_PARAM *maria_ftparser_cal mysql_add_word != 0 - parser is initialized, or no initialization needed. */ info->ftparser_param[ftparser_nr].mysql_add_word= - (int (*)(struct st_mysql_ftparser_param *, char *, int, + (int (*)(struct st_mysql_ftparser_param *, const uchar *, size_t, MYSQL_FTPARSER_BOOLEAN_INFO *)) 1; if (parser->init && parser->init(&info->ftparser_param[ftparser_nr])) return 0; === modified file 'storage/maria/ma_ftdefs.h' --- a/storage/maria/ma_ftdefs.h 2009-02-12 17:51:00 +0000 +++ b/storage/maria/ma_ftdefs.h 2009-11-29 23:08:56 +0000 @@ -96,7 +96,7 @@ #define FTB_RQUOT (ft_boolean_syntax[11]) typedef struct st_maria_ft_word { - uchar * pos; + const uchar * pos; uint len; double weight; } FT_WORD; @@ -106,9 +106,9 @@ int is_stopword(char *word, uint len); MARIA_KEY *_ma_ft_make_key(MARIA_HA *, MARIA_KEY *, uint , uchar *, FT_WORD *, my_off_t); -uchar maria_ft_get_word(CHARSET_INFO *, uchar **, uchar *, FT_WORD *, - MYSQL_FTPARSER_BOOLEAN_INFO *); -uchar maria_ft_simple_get_word(CHARSET_INFO *, uchar **, const uchar *, +uchar maria_ft_get_word(CHARSET_INFO *, const uchar **, const uchar *, + FT_WORD *, MYSQL_FTPARSER_BOOLEAN_INFO *); +uchar maria_ft_simple_get_word(CHARSET_INFO *, const uchar **, const uchar *, FT_WORD *, my_bool); typedef struct _st_maria_ft_seg_iterator { === modified file 'storage/maria/ma_sort.c' --- a/storage/maria/ma_sort.c 2009-01-09 04:23:25 +0000 +++ b/storage/maria/ma_sort.c 2009-11-29 23:08:56 +0000 @@ -920,7 +920,6 @@ merge_buffers(MARIA_SORT_PARAM *info, ui uchar *strpos; BUFFPEK *buffpek,**refpek; QUEUE queue; - volatile int *killed= _ma_killed_ptr(info->sort_info->param); DBUG_ENTER("merge_buffers"); count=error=0; @@ -953,10 +952,6 @@ merge_buffers(MARIA_SORT_PARAM *info, ui { for (;;) { - if (*killed) - { - error=1; goto err; - } buffpek=(BUFFPEK*) queue_top(&queue); if (to_file) { @@ -976,6 +971,12 @@ merge_buffers(MARIA_SORT_PARAM *info, ui buffpek->key+=sort_length; if (! --buffpek->mem_count) { + /* It's enough to check for killedptr before a slow operation */ + if (_ma_killed_ptr(info->sort_info->param)) + { + error=1; + goto err; + } if (!(error=(int) info->read_to_buffer(from_file,buffpek,sort_length))) { uchar *base= buffpek->base; === modified file 'storage/maria/ma_state.c' --- a/storage/maria/ma_state.c 2009-10-06 06:57:22 +0000 +++ b/storage/maria/ma_state.c 2009-11-29 23:08:56 +0000 @@ -528,7 +528,7 @@ void _ma_remove_table_from_trnman(MARIA_ safe_mutex_assert_owner(&share->intern_lock); - for (prev= (MARIA_USED_TABLES**) &trn->used_tables, tables= *prev; + for (prev= (MARIA_USED_TABLES**) (char*) &trn->used_tables, tables= *prev; tables; tables= *prev) { === modified file 'storage/maria/maria_def.h' --- a/storage/maria/maria_def.h 2009-10-06 06:57:22 +0000 +++ b/storage/maria/maria_def.h 2009-11-29 23:08:56 +0000 @@ -1160,7 +1160,7 @@ int _ma_flush_table_files(MARIA_HA *info Functions needed by _ma_check (are overridden in MySQL/ha_maria.cc). See ma_check_standalone.h . */ -volatile int *_ma_killed_ptr(HA_CHECK *param); +int _ma_killed_ptr(HA_CHECK *param); void _ma_check_print_error _VARARGS((HA_CHECK *param, const char *fmt, ...)) ATTRIBUTE_FORMAT(printf, 2, 3); void _ma_check_print_warning _VARARGS((HA_CHECK *param, const char *fmt, ...)) === modified file 'storage/maria/maria_ftdump.c' --- a/storage/maria/maria_ftdump.c 2008-08-25 11:49:47 +0000 +++ b/storage/maria/maria_ftdump.c 2009-11-29 23:08:56 +0000 @@ -116,7 +116,7 @@ int main(int argc,char *argv[]) subkeys=ft_sintXkorr(info->lastkey_buff + keylen + 1); if (subkeys >= 0) - weight=*(float*)&subkeys; + weight=*(float*) (char*) &subkeys; #ifdef HAVE_SNPRINTF snprintf(buf,MAX_LEN,"%.*s",(int) keylen,info->lastkey_buff+1); === modified file 'storage/maria/trnman.c' --- a/storage/maria/trnman.c 2009-03-05 13:46:45 +0000 +++ b/storage/maria/trnman.c 2009-11-29 23:08:56 +0000 @@ -300,8 +300,8 @@ TRN *trnman_new_trn(WT_THD *wt) (ABA isn't possible, we're behind a mutex */ my_atomic_rwlock_wrlock(&LOCK_pool); - while (tmp.trn && !my_atomic_casptr((void **)&pool, &tmp.v, - (void *)tmp.trn->next)) + while (tmp.trn && !my_atomic_casptr((void **)(char*) &pool, &tmp.v, + (void *)tmp.trn->next)) /* no-op */; my_atomic_rwlock_wrunlock(&LOCK_pool); @@ -545,7 +545,7 @@ static void trnman_free_trn(TRN *trn) down after the loop at -O2 */ *(TRN * volatile *)&(trn->next)= tmp.trn; - } while (!my_atomic_casptr((void **)&pool, &tmp.v, trn)); + } while (!my_atomic_casptr((void **)(char*)&pool, &tmp.v, trn)); my_atomic_rwlock_wrunlock(&LOCK_pool); } === modified file 'storage/myisam/ft_boolean_search.c' --- a/storage/myisam/ft_boolean_search.c 2009-10-15 21:38:29 +0000 +++ b/storage/myisam/ft_boolean_search.c 2009-11-29 23:08:56 +0000 @@ -180,7 +180,7 @@ typedef struct st_my_ftb_param static int ftb_query_add_word(MYSQL_FTPARSER_PARAM *param, - char *word, int word_len, + const uchar *word, size_t word_len, MYSQL_FTPARSER_BOOLEAN_INFO *info) { MY_FTB_PARAM *ftb_param= param->mysql_ftparam; @@ -282,19 +282,19 @@ static int ftb_query_add_word(MYSQL_FTPA static int ftb_parse_query_internal(MYSQL_FTPARSER_PARAM *param, - char *query, int len) + const uchar *query, size_t len) { MY_FTB_PARAM *ftb_param= param->mysql_ftparam; MYSQL_FTPARSER_BOOLEAN_INFO info; CHARSET_INFO *cs= ftb_param->ftb->charset; - uchar **start= (uchar**) &query; - uchar *end= (uchar*) query + len; + const uchar **start= &query; + const uchar *end= query + len; FT_WORD w; info.prev= ' '; info.quot= 0; while (ft_get_word(cs, start, end, &w, &info)) - param->mysql_add_word(param, (char*) w.pos, w.len, &info); + param->mysql_add_word(param, w.pos, w.len, &info); return(0); } @@ -616,7 +616,7 @@ typedef struct st_my_ftb_phrase_param static int ftb_phrase_add_word(MYSQL_FTPARSER_PARAM *param, - char *word, int word_len, + const uchar *word, size_t word_len, MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info __attribute__((unused))) { MY_FTB_PHRASE_PARAM *phrase_param= param->mysql_ftparam; @@ -648,15 +648,15 @@ static int ftb_phrase_add_word(MYSQL_FTP static int ftb_check_phrase_internal(MYSQL_FTPARSER_PARAM *param, - char *document, int len) + const uchar *document, size_t len) { FT_WORD word; MY_FTB_PHRASE_PARAM *phrase_param= param->mysql_ftparam; const uchar *docend= (uchar*) document + len; - while (ft_simple_get_word(phrase_param->cs, (uchar**) &document, docend, + while (ft_simple_get_word(phrase_param->cs, &document, docend, &word, FALSE)) { - param->mysql_add_word(param, (char*) word.pos, word.len, 0); + param->mysql_add_word(param, word.pos, word.len, 0); if (phrase_param->match) break; } @@ -874,7 +874,7 @@ typedef struct st_my_ftb_find_param static int ftb_find_relevance_add_word(MYSQL_FTPARSER_PARAM *param, - char *word, int len, + const uchar *word, size_t len, MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info __attribute__((unused))) { MY_FTB_FIND_PARAM *ftb_param= param->mysql_ftparam; @@ -888,8 +888,8 @@ static int ftb_find_relevance_add_word(M for (a= 0, b= ftb->queue.elements, c= (a+b)/2; b-a>1; c= (a+b)/2) { ftbw= ftb->list[c]; - if (ha_compare_text(ftb->charset, (uchar*)word, len, - (uchar*)ftbw->word+1, ftbw->len-1, + if (ha_compare_text(ftb->charset, word, len, + ftbw->word+1, ftbw->len-1, (my_bool) (ftbw->flags & FTB_FLAG_TRUNC), 0) < 0) b= c; else @@ -915,8 +915,8 @@ static int ftb_find_relevance_add_word(M for (; c >= 0; c--) { ftbw= ftb->list[c]; - if (ha_compare_text(ftb->charset, (uchar*)word, len, - (uchar*)ftbw->word + 1,ftbw->len - 1, + if (ha_compare_text(ftb->charset, word, len, + ftbw->word + 1,ftbw->len - 1, (my_bool)(ftbw->flags & FTB_FLAG_TRUNC), 0)) { if (ftb->with_scan & FTB_FLAG_TRUNC) @@ -935,14 +935,14 @@ static int ftb_find_relevance_add_word(M static int ftb_find_relevance_parse(MYSQL_FTPARSER_PARAM *param, - char *doc, int len) + const uchar *doc, size_t len) { MY_FTB_FIND_PARAM *ftb_param= param->mysql_ftparam; FT_INFO *ftb= ftb_param->ftb; - uchar *end= (uchar*) doc + len; + const uchar *end= doc + len; FT_WORD w; - while (ft_simple_get_word(ftb->charset, (uchar**) &doc, end, &w, TRUE)) - param->mysql_add_word(param, (char*) w.pos, w.len, 0); + while (ft_simple_get_word(ftb->charset, &doc, end, &w, TRUE)) + param->mysql_add_word(param, w.pos, w.len, 0); return(0); } === modified file 'storage/myisam/ft_nlq_search.c' --- a/storage/myisam/ft_nlq_search.c 2008-04-28 16:24:05 +0000 +++ b/storage/myisam/ft_nlq_search.c 2009-11-29 23:08:56 +0000 @@ -63,7 +63,8 @@ static int FT_SUPERDOC_cmp(void* cmp_arg static int walk_and_match(FT_WORD *word, uint32 count, ALL_IN_ONE *aio) { - int subkeys, r; + int32 subkeys; + int r; uint keylen, doc_cnt; FT_SUPERDOC sdoc, *sptr; TREE_ELEMENT *selem; @@ -123,7 +124,8 @@ static int walk_and_match(FT_WORD *word, goto do_skip; } #if HA_FT_WTYPE == HA_KEYTYPE_FLOAT - tmp_weight=*(float*)&subkeys; + /* The weight we read was actually a float */ + tmp_weight=*(float*) (char*) &subkeys; #else #error #endif === modified file 'storage/myisam/ft_parser.c' --- a/storage/myisam/ft_parser.c 2009-02-12 14:08:56 +0000 +++ b/storage/myisam/ft_parser.c 2009-11-29 23:08:56 +0000 @@ -106,10 +106,10 @@ my_bool ft_boolean_check_syntax_string(c 3 - right bracket 4 - stopword found */ -uchar ft_get_word(CHARSET_INFO *cs, uchar **start, uchar *end, +uchar ft_get_word(CHARSET_INFO *cs, const uchar **start, const uchar *end, FT_WORD *word, MYSQL_FTPARSER_BOOLEAN_INFO *param) { - uchar *doc=*start; + const uchar *doc= *start; int ctype; uint mwc, length; int mbl; @@ -201,10 +201,11 @@ ret: return param->type; } -uchar ft_simple_get_word(CHARSET_INFO *cs, uchar **start, const uchar *end, - FT_WORD *word, my_bool skip_stopwords) +uchar ft_simple_get_word(CHARSET_INFO *cs, const uchar **start, + const uchar *end, FT_WORD *word, + my_bool skip_stopwords) { - uchar *doc= *start; + const uchar *doc= *start; uint mwc, length; int mbl; int ctype; @@ -216,7 +217,7 @@ uchar ft_simple_get_word(CHARSET_INFO *c { if (doc >= end) DBUG_RETURN(0); - mbl= cs->cset->ctype(cs, &ctype, (uchar*)doc, (uchar*)end); + mbl= cs->cset->ctype(cs, &ctype, doc, end); if (true_word_char(ctype, *doc)) break; } @@ -225,7 +226,7 @@ uchar ft_simple_get_word(CHARSET_INFO *c for (word->pos= doc; doc < end; length++, doc+= (mbl > 0 ? mbl : (mbl < 0 ? -mbl : 1))) { - mbl= cs->cset->ctype(cs, &ctype, (uchar*)doc, (uchar*)end); + mbl= cs->cset->ctype(cs, &ctype, doc, end); if (true_word_char(ctype, *doc)) mwc= 0; else if (!misc_word_char(*doc) || mwc) @@ -238,7 +239,7 @@ uchar ft_simple_get_word(CHARSET_INFO *c if (skip_stopwords == FALSE || (length >= ft_min_word_len && length < ft_max_word_len && - !is_stopword((char*) word->pos, word->len))) + !is_stopword(word->pos, word->len))) { *start= doc; DBUG_RETURN(1); @@ -257,8 +258,9 @@ void ft_parse_init(TREE *wtree, CHARSET_ static int ft_add_word(MYSQL_FTPARSER_PARAM *param, - char *word, int word_len, - MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info __attribute__((unused))) + const uchar *word, size_t word_len, + MYSQL_FTPARSER_BOOLEAN_INFO *boolean_info + __attribute__((unused))) { TREE *wtree; FT_WORD w; @@ -286,10 +288,10 @@ static int ft_add_word(MYSQL_FTPARSER_PA static int ft_parse_internal(MYSQL_FTPARSER_PARAM *param, - char *doc_arg, int doc_len) + const uchar *doc_arg, size_t doc_len) { - uchar *doc= (uchar*) doc_arg; - uchar *end= doc + doc_len; + const uchar *doc= doc_arg; + const uchar *end= doc + doc_len; MY_FT_PARSER_PARAM *ft_param=param->mysql_ftparam; TREE *wtree= ft_param->wtree; FT_WORD w; @@ -302,7 +304,7 @@ static int ft_parse_internal(MYSQL_FTPAR } -int ft_parse(TREE *wtree, uchar *doc, int doclen, +int ft_parse(TREE *wtree, const uchar *doc, int doclen, struct st_mysql_ftparser *parser, MYSQL_FTPARSER_PARAM *param, MEM_ROOT *mem_root) { @@ -377,7 +379,7 @@ MYSQL_FTPARSER_PARAM *ftparser_call_init mysql_add_word != 0 - parser is initialized, or no initialization needed. */ info->ftparser_param[ftparser_nr].mysql_add_word= - (int (*)(struct st_mysql_ftparser_param *, char *, int, + (int (*)(struct st_mysql_ftparser_param *, const uchar *, size_t, MYSQL_FTPARSER_BOOLEAN_INFO *)) 1; if (parser->init && parser->init(&info->ftparser_param[ftparser_nr])) return 0; === modified file 'storage/myisam/ft_stopwords.c' --- a/storage/myisam/ft_stopwords.c 2008-04-28 16:24:05 +0000 +++ b/storage/myisam/ft_stopwords.c 2009-11-29 23:08:56 +0000 @@ -66,7 +66,8 @@ int ft_init_stopwords() { File fd; uint len; - uchar *buffer, *start, *end; + uchar *buffer; + const uchar *start, *end; FT_WORD w; int error=-1; @@ -109,7 +110,7 @@ err0: } -int is_stopword(char *word, uint len) +int is_stopword(const uchar *word, size_t len) { FT_STOPWORD sw; sw.pos=word; === modified file 'storage/myisam/ftdefs.h' --- a/storage/myisam/ftdefs.h 2009-01-26 06:35:15 +0000 +++ b/storage/myisam/ftdefs.h 2009-11-29 23:08:56 +0000 @@ -96,18 +96,18 @@ #define FTB_RQUOT (ft_boolean_syntax[11]) typedef struct st_ft_word { - uchar * pos; + const uchar *pos; uint len; double weight; } FT_WORD; -int is_stopword(char *word, uint len); +int is_stopword(const uchar *word, size_t len); uint _ft_make_key(MI_INFO *, uint , uchar *, FT_WORD *, my_off_t); -uchar ft_get_word(CHARSET_INFO *, uchar **, uchar *, FT_WORD *, +uchar ft_get_word(CHARSET_INFO *, const uchar **, const uchar *, FT_WORD *, MYSQL_FTPARSER_BOOLEAN_INFO *); -uchar ft_simple_get_word(CHARSET_INFO *, uchar **, const uchar *, +uchar ft_simple_get_word(CHARSET_INFO *, const uchar **, const uchar *, FT_WORD *, my_bool); typedef struct _st_ft_seg_iterator { @@ -121,7 +121,7 @@ void _mi_ft_segiterator_dummy_init(const uint _mi_ft_segiterator(FT_SEG_ITERATOR *); void ft_parse_init(TREE *, CHARSET_INFO *); -int ft_parse(TREE *, uchar *, int, struct st_mysql_ftparser *parser, +int ft_parse(TREE *, const uchar *, int, struct st_mysql_ftparser *parser, MYSQL_FTPARSER_PARAM *, MEM_ROOT *); FT_WORD * ft_linearize(TREE *, MEM_ROOT *); FT_WORD * _mi_ft_parserecord(MI_INFO *, uint, const uchar *, MEM_ROOT *); === modified file 'storage/myisam/ha_myisam.cc' --- a/storage/myisam/ha_myisam.cc 2009-10-06 14:53:46 +0000 +++ b/storage/myisam/ha_myisam.cc 2009-11-29 23:08:56 +0000 @@ -497,10 +497,9 @@ int check_definition(MI_KEYDEF *t1_keyin extern "C" { -volatile int *killed_ptr(HA_CHECK *param) +int killed_ptr(HA_CHECK *param) { - /* In theory Unsafe conversion, but should be ok for now */ - return (int*) &(((THD *)(param->thd))->killed); + return thd_killed((THD*)param->thd); } void mi_check_print_error(HA_CHECK *param, const char *fmt,...) === modified file 'storage/myisam/mi_check.c' --- a/storage/myisam/mi_check.c 2009-10-15 21:38:29 +0000 +++ b/storage/myisam/mi_check.c 2009-11-29 23:08:56 +0000 @@ -166,7 +166,7 @@ int chk_del(HA_CHECK *param, register MI empty=0; for (i= info->state->del ; i > 0L && next_link != HA_OFFSET_ERROR ; i--) { - if (*killed_ptr(param)) + if (killed_ptr(param)) DBUG_RETURN(1); if (test_flag & T_VERBOSE) printf(" %9s",llstr(next_link,buff)); @@ -261,7 +261,7 @@ static int check_k_link(HA_CHECK *param, records= (ha_rows) (info->state->key_file_length / block_size); while (next_link != HA_OFFSET_ERROR && records > 0) { - if (*killed_ptr(param)) + if (killed_ptr(param)) DBUG_RETURN(1); if (param->testflag & T_VERBOSE) printf("%16s",llstr(next_link,llbuff)); @@ -778,7 +778,7 @@ static int chk_index(HA_CHECK *param, MI } for ( ;; ) { - if (*killed_ptr(param)) + if (killed_ptr(param)) goto err; memcpy((char*) info->lastkey,(char*) key,key_length); info->lastkey_length=key_length; @@ -990,7 +990,7 @@ int chk_data_link(HA_CHECK *param, MI_IN bzero((char*) key_checksum, info->s->base.keys * sizeof(key_checksum[0])); while (pos < info->state->data_file_length) { - if (*killed_ptr(param)) + if (killed_ptr(param)) goto err2; switch (info->s->data_file_type) { case BLOCK_RECORD: @@ -3247,7 +3247,7 @@ static int sort_get_next_record(MI_SORT_ char llbuff[22],llbuff2[22]; DBUG_ENTER("sort_get_next_record"); - if (*killed_ptr(param)) + if (killed_ptr(param)) DBUG_RETURN(1); switch (share->data_file_type) { === modified file 'storage/myisam/myisam_ftdump.c' --- a/storage/myisam/myisam_ftdump.c 2007-05-10 09:59:39 +0000 +++ b/storage/myisam/myisam_ftdump.c 2009-11-29 23:08:56 +0000 @@ -113,7 +113,7 @@ int main(int argc,char *argv[]) subkeys=ft_sintXkorr(info->lastkey+keylen+1); if (subkeys >= 0) - weight=*(float*)&subkeys; + weight= *(float*) (char*) &subkeys; #ifdef HAVE_SNPRINTF snprintf(buf,MAX_LEN,"%.*s",(int) keylen,info->lastkey+1); === modified file 'storage/myisam/myisamchk.c' --- a/storage/myisam/myisamchk.c 2009-09-19 21:21:29 +0000 +++ b/storage/myisam/myisamchk.c 2009-11-29 23:08:56 +0000 @@ -1745,11 +1745,9 @@ err: sorting */ -static int not_killed= 0; - -volatile int *killed_ptr(HA_CHECK *param __attribute__((unused))) +int killed_ptr(HA_CHECK *param __attribute__((unused))) { - return &not_killed; /* always NULL */ + return 0; } /* print warnings and errors */ === modified file 'storage/myisam/myisamdef.h' --- a/storage/myisam/myisamdef.h 2009-10-06 06:57:22 +0000 +++ b/storage/myisam/myisamdef.h 2009-11-29 23:08:56 +0000 @@ -725,7 +725,7 @@ my_bool mi_dynmap_file(MI_INFO *info, my void mi_remap_file(MI_INFO *info, my_off_t size); /* Functions needed by mi_check */ -volatile int *killed_ptr(HA_CHECK *param); +int killed_ptr(HA_CHECK *param); void mi_check_print_error _VARARGS((HA_CHECK *param, const char *fmt, ...)); void mi_check_print_warning _VARARGS((HA_CHECK *param, const char *fmt, ...)); void mi_check_print_info _VARARGS((HA_CHECK *param, const char *fmt, ...)); === modified file 'storage/myisam/myisamlog.c' --- a/storage/myisam/myisamlog.c 2008-02-18 22:35:17 +0000 +++ b/storage/myisam/myisamlog.c 2009-11-29 23:08:56 +0000 @@ -385,7 +385,7 @@ static int examine_log(char * file_name, file_info.name=0; file_info.show_name=0; file_info.record=0; - if (read_string(&cache,(uchar**) &file_info.name, + if (read_string(&cache,(uchar**) (char*) &file_info.name, (uint) mi_uint2korr(head))) goto err; { === modified file 'storage/myisam/sort.c' --- a/storage/myisam/sort.c 2009-10-15 21:38:29 +0000 +++ b/storage/myisam/sort.c 2009-11-29 23:08:56 +0000 @@ -900,7 +900,6 @@ merge_buffers(MI_SORT_PARAM *info, uint uchar *strpos; BUFFPEK *buffpek,**refpek; QUEUE queue; - volatile int *killed= killed_ptr(info->sort_info->param); DBUG_ENTER("merge_buffers"); count=error=0; @@ -933,10 +932,6 @@ merge_buffers(MI_SORT_PARAM *info, uint { for (;;) { - if (*killed) - { - error=1; goto err; - } buffpek=(BUFFPEK*) queue_top(&queue); if (to_file) { @@ -956,6 +951,12 @@ merge_buffers(MI_SORT_PARAM *info, uint buffpek->key+=sort_length; if (! --buffpek->mem_count) { + /* It's enough to check for killedptr before a slow operation */ + if (killed_ptr(info->sort_info->param)) + { + error=1; + goto err; + } if (!(error=(int) info->read_to_buffer(from_file,buffpek,sort_length))) { uchar *base= buffpek->base; === modified file 'storage/xtradb/fil/fil0fil.c' --- a/storage/xtradb/fil/fil0fil.c 2009-11-13 21:26:08 +0000 +++ b/storage/xtradb/fil/fil0fil.c 2009-11-29 23:08:56 +0000 @@ -3139,7 +3139,7 @@ skip_info: if (mach_read_from_4(page + FIL_PAGE_OFFSET) || !offset) { mach_write_to_4(page + FIL_PAGE_ARCH_LOG_NO_OR_SPACE_ID, id); - for (i = 0; i < n_index; i++) { + for (i = 0; (ulint) i < n_index; i++) { if (offset / UNIV_PAGE_SIZE == root_page[i]) { /* this is index root page */ mach_write_to_4(page + FIL_PAGE_DATA + PAGE_BTR_SEG_LEAF === modified file 'storage/xtradb/trx/trx0i_s.c' --- a/storage/xtradb/trx/trx0i_s.c 2009-09-07 10:22:53 +0000 +++ b/storage/xtradb/trx/trx0i_s.c 2009-11-29 23:08:56 +0000 @@ -28,11 +28,11 @@ table cache" for later retrieval. Created July 17, 2007 Vasil Dimov *******************************************************/ -#include <mysql/plugin.h> #include "mysql_addons.h" #include "univ.i" +#include <mysql/plugin.h> #include "buf0buf.h" #include "dict0dict.h" #include "ha0storage.h"

1 0

[Maria-developers] Updated (by Knielsen): Store in binlog text of statements that caused RBR events (47)
by worklog-noreply＠askmonty.org 30 Nov '09

30 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Store in binlog text of statements that caused RBR events CREATION DATE..: Sat, 15 Aug 2009, 23:48 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: Knielsen CATEGORY.......: Server-RawIdeaBin TASK ID........: 47 (http://askmonty.org/worklog/?tid=47) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Knielsen - Mon, 30 Nov 2009, 11:21)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.18210 2009-11-30 11:21:33.000000000 +0200 +++ /tmp/wklog.47.new.18210 2009-11-30 11:21:33.000000000 +0200 @@ -28,7 +28,14 @@ Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query -text in its data part. Example: +text in its data part. + +The numeric code for this event must be assigned carefully. It should be +coordinated with MySQL/Sun, otherwise we can get into a situation where MySQL +uses the same numeric code for one event that MariaDB uses for +ANNOTATE_ROWS_EVENT, which would make merging the two impossible. + +Example: ... ************************ -=-=(Alexi - Mon, 30 Nov 2009, 10:33)=-=- Low Level Design modified. --- /tmp/wklog.47.old.16188 2009-11-30 10:33:44.000000000 +0200 +++ /tmp/wklog.47.new.16188 2009-11-30 10:33:44.000000000 +0200 @@ -2,6 +2,7 @@ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. When sending events to a slave, master should simply skip Annotate_rows events (they are not needed for replication). + [ ??? Multi-master - currently not clear ] 2. When sending events to mysqlbinlog (remote case), master must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 13:00)=-=- Low Level Design modified. --- /tmp/wklog.47.old.32047 2009-11-29 13:00:21.000000000 +0200 +++ /tmp/wklog.47.new.32047 2009-11-29 13:00:21.000000000 +0200 @@ -1 +1,7 @@ +mysql_binlog_send() [sql/sql_repl.cc] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +1. When sending events to a slave, master should simply skip + Annotate_rows events (they are not needed for replication). +2. When sending events to mysqlbinlog (remote case), master + must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 09:50)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.24993 2009-11-29 07:50:36.000000000 +0000 +++ /tmp/wklog.47.new.24993 2009-11-29 07:50:36.000000000 +0000 @@ -4,3 +4,96 @@ > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). +New server option +~~~~~~~~~~~~~~~~~ + --binlog-annotate-row-events + +Setting this option makes RBR (row-) events in the binary log to be +preceded by Annotate rows events (see below). The corresponding +'binlog_annotate_row_events' system variable is dynamic and has both +global and session values. Default global value is OFF. + +Note. Session values are usefull to make it possible to annotate only + some selected statements: + + ... + SET SESSION binlog_annotate_row_events=ON; + ... statements to be annotated ... + SET SESSION binlog_annotate_row_events=OFF; + ... statements not to be annotated ... + +New binlog event type +~~~~~~~~~~~~~~~~~~~~~ + Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] + +Describes the query which caused the corresponding row event. In binary log, +precedes each Table_map_log_event. Contains empty post-header and the query +text in its data part. Example: + + ... + ************************ + ANNOTATE_ROWS_EVENT [51] + ************************ + 000000C7 | 54 1B 12 4B | time_when = 1259477844 + 000000CB | 33 | event_type = 51 + 000000CC | 64 00 00 00 | server_id = 100 + 000000D0 | 2C 00 00 00 | event_len = 44 + 000000D4 | F3 00 00 00 | log_pos = 000000F3 + 000000D8 | 00 00 | flags = <none> + ------------------------ + 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" + 000000DE | 72 74 20 69 | + 000000E2 | 6E 74 6F 20 | + 000000E6 | 74 31 20 76 | + 000000EA | 61 6C 75 65 | + 000000EE | 73 20 28 31 | + 000000F2 | 29 | + ************************ + TABLE_MAP_EVENT [19] + ************************ + 000000F3 | 54 1B 12 4B | time_when = 1259477844 + 000000F7 | 13 | event_type = 19 + 000000F8 | 64 00 00 00 | server_id = 100 + 000000FC | 29 00 00 00 | event_len = 41 + 00000100 | 1C 01 00 00 | log_pos = 0000011C + 00000104 | 00 00 | flags = <none> + ------------------------ + ... + ************************ + WRITE_ROWS_EVENT [23] + ************************ + 0000011C | 54 1B 12 4B | time_when = 1259477844 + 00000120 | 17 | event_type = 23 + 00000121 | 64 00 00 00 | server_id = 100 + 00000125 | 22 00 00 00 | event_len = 34 + 00000129 | 3E 01 00 00 | log_pos = 0000013E + 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F + ------------------------ + 0000012F | 0F 00 00 00 | table_id = 15 + ... + +mysqlbinlog output +~~~~~~~~~~~~~~~~~~ +Something like this: + + ... + # at 199 + # at 243 + # at 284 + #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values +(1)` + #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped +to number 15 + #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 +flags: STMT_END_F + + BINLOG ' + VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= + VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= + VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== + '/*!*/; + ### INSERT INTO test.t1 + ### SET + ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ + ... + -=-=(Knielsen - Fri, 27 Nov 2009, 13:30)=-=- Observers changed: Knielsen -=-=(Psergey - Sun, 16 Aug 2009, 11:08)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.12485 2009-08-16 11:08:33.000000000 +0300 +++ /tmp/wklog.47.new.12485 2009-08-16 11:08:33.000000000 +0300 @@ -1 +1,6 @@ +First suggestion: + +> I think for this we would actually need a new binlog event type +> (Comment_log_event?). Unless we want to log an empty statement Query_log_event +> containing only a comment (a bit of a hack). -=-=(Psergey - Sun, 16 Aug 2009, 00:02)=-=- Dependency created: 39 now depends on 47 DESCRIPTION: Store in binlog (and show in mysqlbinlog output) texts of statements that caused RBR events This is needed for (list from Monty): - Easier to understand why updates happened - Would make it easier to find out where in application things went wrong (as you can search for exact strings) - Allow one to filter things based on comments in the statement. The cost of this can be that the binlog will be approximately 2x in size (especially insert of big blob's would be a bit painful), so this should be an optional feature. HIGH-LEVEL SPECIFICATION: First suggestion: > I think for this we would actually need a new binlog event type > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). New server option ~~~~~~~~~~~~~~~~~ --binlog-annotate-row-events Setting this option makes RBR (row-) events in the binary log to be preceded by Annotate rows events (see below). The corresponding 'binlog_annotate_row_events' system variable is dynamic and has both global and session values. Default global value is OFF. Note. Session values are usefull to make it possible to annotate only some selected statements: ... SET SESSION binlog_annotate_row_events=ON; ... statements to be annotated ... SET SESSION binlog_annotate_row_events=OFF; ... statements not to be annotated ... New binlog event type ~~~~~~~~~~~~~~~~~~~~~ Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query text in its data part. The numeric code for this event must be assigned carefully. It should be coordinated with MySQL/Sun, otherwise we can get into a situation where MySQL uses the same numeric code for one event that MariaDB uses for ANNOTATE_ROWS_EVENT, which would make merging the two impossible. Example: ... ************************ ANNOTATE_ROWS_EVENT [51] ************************ 000000C7 | 54 1B 12 4B | time_when = 1259477844 000000CB | 33 | event_type = 51 000000CC | 64 00 00 00 | server_id = 100 000000D0 | 2C 00 00 00 | event_len = 44 000000D4 | F3 00 00 00 | log_pos = 000000F3 000000D8 | 00 00 | flags = <none> ------------------------ 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" 000000DE | 72 74 20 69 | 000000E2 | 6E 74 6F 20 | 000000E6 | 74 31 20 76 | 000000EA | 61 6C 75 65 | 000000EE | 73 20 28 31 | 000000F2 | 29 | ************************ TABLE_MAP_EVENT [19] ************************ 000000F3 | 54 1B 12 4B | time_when = 1259477844 000000F7 | 13 | event_type = 19 000000F8 | 64 00 00 00 | server_id = 100 000000FC | 29 00 00 00 | event_len = 41 00000100 | 1C 01 00 00 | log_pos = 0000011C 00000104 | 00 00 | flags = <none> ------------------------ ... ************************ WRITE_ROWS_EVENT [23] ************************ 0000011C | 54 1B 12 4B | time_when = 1259477844 00000120 | 17 | event_type = 23 00000121 | 64 00 00 00 | server_id = 100 00000125 | 22 00 00 00 | event_len = 34 00000129 | 3E 01 00 00 | log_pos = 0000013E 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F ------------------------ 0000012F | 0F 00 00 00 | table_id = 15 ... mysqlbinlog output ~~~~~~~~~~~~~~~~~~ Something like this: ... # at 199 # at 243 # at 284 #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values (1)` #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped to number 15 #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 flags: STMT_END_F BINLOG ' VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== '/*!*/; ### INSERT INTO test.t1 ### SET ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ ... LOW-LEVEL DESIGN: mysql_binlog_send() [sql/sql_repl.cc] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. When sending events to a slave, master should simply skip Annotate_rows events (they are not needed for replication). [ ??? Multi-master - currently not clear ] 2. When sending events to mysqlbinlog (remote case), master must send Annotate_rows events as well. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Knielsen): Store in binlog text of statements that caused RBR events (47)
by worklog-noreply＠askmonty.org 30 Nov '09

30 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Store in binlog text of statements that caused RBR events CREATION DATE..: Sat, 15 Aug 2009, 23:48 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: Knielsen CATEGORY.......: Server-RawIdeaBin TASK ID........: 47 (http://askmonty.org/worklog/?tid=47) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Knielsen - Mon, 30 Nov 2009, 11:21)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.18210 2009-11-30 11:21:33.000000000 +0200 +++ /tmp/wklog.47.new.18210 2009-11-30 11:21:33.000000000 +0200 @@ -28,7 +28,14 @@ Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query -text in its data part. Example: +text in its data part. + +The numeric code for this event must be assigned carefully. It should be +coordinated with MySQL/Sun, otherwise we can get into a situation where MySQL +uses the same numeric code for one event that MariaDB uses for +ANNOTATE_ROWS_EVENT, which would make merging the two impossible. + +Example: ... ************************ -=-=(Alexi - Mon, 30 Nov 2009, 10:33)=-=- Low Level Design modified. --- /tmp/wklog.47.old.16188 2009-11-30 10:33:44.000000000 +0200 +++ /tmp/wklog.47.new.16188 2009-11-30 10:33:44.000000000 +0200 @@ -2,6 +2,7 @@ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. When sending events to a slave, master should simply skip Annotate_rows events (they are not needed for replication). + [ ??? Multi-master - currently not clear ] 2. When sending events to mysqlbinlog (remote case), master must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 13:00)=-=- Low Level Design modified. --- /tmp/wklog.47.old.32047 2009-11-29 13:00:21.000000000 +0200 +++ /tmp/wklog.47.new.32047 2009-11-29 13:00:21.000000000 +0200 @@ -1 +1,7 @@ +mysql_binlog_send() [sql/sql_repl.cc] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +1. When sending events to a slave, master should simply skip + Annotate_rows events (they are not needed for replication). +2. When sending events to mysqlbinlog (remote case), master + must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 09:50)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.24993 2009-11-29 07:50:36.000000000 +0000 +++ /tmp/wklog.47.new.24993 2009-11-29 07:50:36.000000000 +0000 @@ -4,3 +4,96 @@ > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). +New server option +~~~~~~~~~~~~~~~~~ + --binlog-annotate-row-events + +Setting this option makes RBR (row-) events in the binary log to be +preceded by Annotate rows events (see below). The corresponding +'binlog_annotate_row_events' system variable is dynamic and has both +global and session values. Default global value is OFF. + +Note. Session values are usefull to make it possible to annotate only + some selected statements: + + ... + SET SESSION binlog_annotate_row_events=ON; + ... statements to be annotated ... + SET SESSION binlog_annotate_row_events=OFF; + ... statements not to be annotated ... + +New binlog event type +~~~~~~~~~~~~~~~~~~~~~ + Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] + +Describes the query which caused the corresponding row event. In binary log, +precedes each Table_map_log_event. Contains empty post-header and the query +text in its data part. Example: + + ... + ************************ + ANNOTATE_ROWS_EVENT [51] + ************************ + 000000C7 | 54 1B 12 4B | time_when = 1259477844 + 000000CB | 33 | event_type = 51 + 000000CC | 64 00 00 00 | server_id = 100 + 000000D0 | 2C 00 00 00 | event_len = 44 + 000000D4 | F3 00 00 00 | log_pos = 000000F3 + 000000D8 | 00 00 | flags = <none> + ------------------------ + 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" + 000000DE | 72 74 20 69 | + 000000E2 | 6E 74 6F 20 | + 000000E6 | 74 31 20 76 | + 000000EA | 61 6C 75 65 | + 000000EE | 73 20 28 31 | + 000000F2 | 29 | + ************************ + TABLE_MAP_EVENT [19] + ************************ + 000000F3 | 54 1B 12 4B | time_when = 1259477844 + 000000F7 | 13 | event_type = 19 + 000000F8 | 64 00 00 00 | server_id = 100 + 000000FC | 29 00 00 00 | event_len = 41 + 00000100 | 1C 01 00 00 | log_pos = 0000011C + 00000104 | 00 00 | flags = <none> + ------------------------ + ... + ************************ + WRITE_ROWS_EVENT [23] + ************************ + 0000011C | 54 1B 12 4B | time_when = 1259477844 + 00000120 | 17 | event_type = 23 + 00000121 | 64 00 00 00 | server_id = 100 + 00000125 | 22 00 00 00 | event_len = 34 + 00000129 | 3E 01 00 00 | log_pos = 0000013E + 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F + ------------------------ + 0000012F | 0F 00 00 00 | table_id = 15 + ... + +mysqlbinlog output +~~~~~~~~~~~~~~~~~~ +Something like this: + + ... + # at 199 + # at 243 + # at 284 + #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values +(1)` + #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped +to number 15 + #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 +flags: STMT_END_F + + BINLOG ' + VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= + VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= + VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== + '/*!*/; + ### INSERT INTO test.t1 + ### SET + ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ + ... + -=-=(Knielsen - Fri, 27 Nov 2009, 13:30)=-=- Observers changed: Knielsen -=-=(Psergey - Sun, 16 Aug 2009, 11:08)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.12485 2009-08-16 11:08:33.000000000 +0300 +++ /tmp/wklog.47.new.12485 2009-08-16 11:08:33.000000000 +0300 @@ -1 +1,6 @@ +First suggestion: + +> I think for this we would actually need a new binlog event type +> (Comment_log_event?). Unless we want to log an empty statement Query_log_event +> containing only a comment (a bit of a hack). -=-=(Psergey - Sun, 16 Aug 2009, 00:02)=-=- Dependency created: 39 now depends on 47 DESCRIPTION: Store in binlog (and show in mysqlbinlog output) texts of statements that caused RBR events This is needed for (list from Monty): - Easier to understand why updates happened - Would make it easier to find out where in application things went wrong (as you can search for exact strings) - Allow one to filter things based on comments in the statement. The cost of this can be that the binlog will be approximately 2x in size (especially insert of big blob's would be a bit painful), so this should be an optional feature. HIGH-LEVEL SPECIFICATION: First suggestion: > I think for this we would actually need a new binlog event type > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). New server option ~~~~~~~~~~~~~~~~~ --binlog-annotate-row-events Setting this option makes RBR (row-) events in the binary log to be preceded by Annotate rows events (see below). The corresponding 'binlog_annotate_row_events' system variable is dynamic and has both global and session values. Default global value is OFF. Note. Session values are usefull to make it possible to annotate only some selected statements: ... SET SESSION binlog_annotate_row_events=ON; ... statements to be annotated ... SET SESSION binlog_annotate_row_events=OFF; ... statements not to be annotated ... New binlog event type ~~~~~~~~~~~~~~~~~~~~~ Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query text in its data part. The numeric code for this event must be assigned carefully. It should be coordinated with MySQL/Sun, otherwise we can get into a situation where MySQL uses the same numeric code for one event that MariaDB uses for ANNOTATE_ROWS_EVENT, which would make merging the two impossible. Example: ... ************************ ANNOTATE_ROWS_EVENT [51] ************************ 000000C7 | 54 1B 12 4B | time_when = 1259477844 000000CB | 33 | event_type = 51 000000CC | 64 00 00 00 | server_id = 100 000000D0 | 2C 00 00 00 | event_len = 44 000000D4 | F3 00 00 00 | log_pos = 000000F3 000000D8 | 00 00 | flags = <none> ------------------------ 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" 000000DE | 72 74 20 69 | 000000E2 | 6E 74 6F 20 | 000000E6 | 74 31 20 76 | 000000EA | 61 6C 75 65 | 000000EE | 73 20 28 31 | 000000F2 | 29 | ************************ TABLE_MAP_EVENT [19] ************************ 000000F3 | 54 1B 12 4B | time_when = 1259477844 000000F7 | 13 | event_type = 19 000000F8 | 64 00 00 00 | server_id = 100 000000FC | 29 00 00 00 | event_len = 41 00000100 | 1C 01 00 00 | log_pos = 0000011C 00000104 | 00 00 | flags = <none> ------------------------ ... ************************ WRITE_ROWS_EVENT [23] ************************ 0000011C | 54 1B 12 4B | time_when = 1259477844 00000120 | 17 | event_type = 23 00000121 | 64 00 00 00 | server_id = 100 00000125 | 22 00 00 00 | event_len = 34 00000129 | 3E 01 00 00 | log_pos = 0000013E 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F ------------------------ 0000012F | 0F 00 00 00 | table_id = 15 ... mysqlbinlog output ~~~~~~~~~~~~~~~~~~ Something like this: ... # at 199 # at 243 # at 284 #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values (1)` #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped to number 15 #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 flags: STMT_END_F BINLOG ' VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== '/*!*/; ### INSERT INTO test.t1 ### SET ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ ... LOW-LEVEL DESIGN: mysql_binlog_send() [sql/sql_repl.cc] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. When sending events to a slave, master should simply skip Annotate_rows events (they are not needed for replication). [ ??? Multi-master - currently not clear ] 2. When sending events to mysqlbinlog (remote case), master must send Annotate_rows events as well. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Knielsen): Store in binlog text of statements that caused RBR events (47)
by worklog-noreply＠askmonty.org 30 Nov '09

30 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Store in binlog text of statements that caused RBR events CREATION DATE..: Sat, 15 Aug 2009, 23:48 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: Knielsen CATEGORY.......: Server-RawIdeaBin TASK ID........: 47 (http://askmonty.org/worklog/?tid=47) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Knielsen - Mon, 30 Nov 2009, 11:21)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.18210 2009-11-30 11:21:33.000000000 +0200 +++ /tmp/wklog.47.new.18210 2009-11-30 11:21:33.000000000 +0200 @@ -28,7 +28,14 @@ Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query -text in its data part. Example: +text in its data part. + +The numeric code for this event must be assigned carefully. It should be +coordinated with MySQL/Sun, otherwise we can get into a situation where MySQL +uses the same numeric code for one event that MariaDB uses for +ANNOTATE_ROWS_EVENT, which would make merging the two impossible. + +Example: ... ************************ -=-=(Alexi - Mon, 30 Nov 2009, 10:33)=-=- Low Level Design modified. --- /tmp/wklog.47.old.16188 2009-11-30 10:33:44.000000000 +0200 +++ /tmp/wklog.47.new.16188 2009-11-30 10:33:44.000000000 +0200 @@ -2,6 +2,7 @@ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. When sending events to a slave, master should simply skip Annotate_rows events (they are not needed for replication). + [ ??? Multi-master - currently not clear ] 2. When sending events to mysqlbinlog (remote case), master must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 13:00)=-=- Low Level Design modified. --- /tmp/wklog.47.old.32047 2009-11-29 13:00:21.000000000 +0200 +++ /tmp/wklog.47.new.32047 2009-11-29 13:00:21.000000000 +0200 @@ -1 +1,7 @@ +mysql_binlog_send() [sql/sql_repl.cc] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +1. When sending events to a slave, master should simply skip + Annotate_rows events (they are not needed for replication). +2. When sending events to mysqlbinlog (remote case), master + must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 09:50)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.24993 2009-11-29 07:50:36.000000000 +0000 +++ /tmp/wklog.47.new.24993 2009-11-29 07:50:36.000000000 +0000 @@ -4,3 +4,96 @@ > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). +New server option +~~~~~~~~~~~~~~~~~ + --binlog-annotate-row-events + +Setting this option makes RBR (row-) events in the binary log to be +preceded by Annotate rows events (see below). The corresponding +'binlog_annotate_row_events' system variable is dynamic and has both +global and session values. Default global value is OFF. + +Note. Session values are usefull to make it possible to annotate only + some selected statements: + + ... + SET SESSION binlog_annotate_row_events=ON; + ... statements to be annotated ... + SET SESSION binlog_annotate_row_events=OFF; + ... statements not to be annotated ... + +New binlog event type +~~~~~~~~~~~~~~~~~~~~~ + Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] + +Describes the query which caused the corresponding row event. In binary log, +precedes each Table_map_log_event. Contains empty post-header and the query +text in its data part. Example: + + ... + ************************ + ANNOTATE_ROWS_EVENT [51] + ************************ + 000000C7 | 54 1B 12 4B | time_when = 1259477844 + 000000CB | 33 | event_type = 51 + 000000CC | 64 00 00 00 | server_id = 100 + 000000D0 | 2C 00 00 00 | event_len = 44 + 000000D4 | F3 00 00 00 | log_pos = 000000F3 + 000000D8 | 00 00 | flags = <none> + ------------------------ + 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" + 000000DE | 72 74 20 69 | + 000000E2 | 6E 74 6F 20 | + 000000E6 | 74 31 20 76 | + 000000EA | 61 6C 75 65 | + 000000EE | 73 20 28 31 | + 000000F2 | 29 | + ************************ + TABLE_MAP_EVENT [19] + ************************ + 000000F3 | 54 1B 12 4B | time_when = 1259477844 + 000000F7 | 13 | event_type = 19 + 000000F8 | 64 00 00 00 | server_id = 100 + 000000FC | 29 00 00 00 | event_len = 41 + 00000100 | 1C 01 00 00 | log_pos = 0000011C + 00000104 | 00 00 | flags = <none> + ------------------------ + ... + ************************ + WRITE_ROWS_EVENT [23] + ************************ + 0000011C | 54 1B 12 4B | time_when = 1259477844 + 00000120 | 17 | event_type = 23 + 00000121 | 64 00 00 00 | server_id = 100 + 00000125 | 22 00 00 00 | event_len = 34 + 00000129 | 3E 01 00 00 | log_pos = 0000013E + 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F + ------------------------ + 0000012F | 0F 00 00 00 | table_id = 15 + ... + +mysqlbinlog output +~~~~~~~~~~~~~~~~~~ +Something like this: + + ... + # at 199 + # at 243 + # at 284 + #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values +(1)` + #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped +to number 15 + #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 +flags: STMT_END_F + + BINLOG ' + VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= + VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= + VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== + '/*!*/; + ### INSERT INTO test.t1 + ### SET + ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ + ... + -=-=(Knielsen - Fri, 27 Nov 2009, 13:30)=-=- Observers changed: Knielsen -=-=(Psergey - Sun, 16 Aug 2009, 11:08)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.12485 2009-08-16 11:08:33.000000000 +0300 +++ /tmp/wklog.47.new.12485 2009-08-16 11:08:33.000000000 +0300 @@ -1 +1,6 @@ +First suggestion: + +> I think for this we would actually need a new binlog event type +> (Comment_log_event?). Unless we want to log an empty statement Query_log_event +> containing only a comment (a bit of a hack). -=-=(Psergey - Sun, 16 Aug 2009, 00:02)=-=- Dependency created: 39 now depends on 47 DESCRIPTION: Store in binlog (and show in mysqlbinlog output) texts of statements that caused RBR events This is needed for (list from Monty): - Easier to understand why updates happened - Would make it easier to find out where in application things went wrong (as you can search for exact strings) - Allow one to filter things based on comments in the statement. The cost of this can be that the binlog will be approximately 2x in size (especially insert of big blob's would be a bit painful), so this should be an optional feature. HIGH-LEVEL SPECIFICATION: First suggestion: > I think for this we would actually need a new binlog event type > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). New server option ~~~~~~~~~~~~~~~~~ --binlog-annotate-row-events Setting this option makes RBR (row-) events in the binary log to be preceded by Annotate rows events (see below). The corresponding 'binlog_annotate_row_events' system variable is dynamic and has both global and session values. Default global value is OFF. Note. Session values are usefull to make it possible to annotate only some selected statements: ... SET SESSION binlog_annotate_row_events=ON; ... statements to be annotated ... SET SESSION binlog_annotate_row_events=OFF; ... statements not to be annotated ... New binlog event type ~~~~~~~~~~~~~~~~~~~~~ Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query text in its data part. The numeric code for this event must be assigned carefully. It should be coordinated with MySQL/Sun, otherwise we can get into a situation where MySQL uses the same numeric code for one event that MariaDB uses for ANNOTATE_ROWS_EVENT, which would make merging the two impossible. Example: ... ************************ ANNOTATE_ROWS_EVENT [51] ************************ 000000C7 | 54 1B 12 4B | time_when = 1259477844 000000CB | 33 | event_type = 51 000000CC | 64 00 00 00 | server_id = 100 000000D0 | 2C 00 00 00 | event_len = 44 000000D4 | F3 00 00 00 | log_pos = 000000F3 000000D8 | 00 00 | flags = <none> ------------------------ 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" 000000DE | 72 74 20 69 | 000000E2 | 6E 74 6F 20 | 000000E6 | 74 31 20 76 | 000000EA | 61 6C 75 65 | 000000EE | 73 20 28 31 | 000000F2 | 29 | ************************ TABLE_MAP_EVENT [19] ************************ 000000F3 | 54 1B 12 4B | time_when = 1259477844 000000F7 | 13 | event_type = 19 000000F8 | 64 00 00 00 | server_id = 100 000000FC | 29 00 00 00 | event_len = 41 00000100 | 1C 01 00 00 | log_pos = 0000011C 00000104 | 00 00 | flags = <none> ------------------------ ... ************************ WRITE_ROWS_EVENT [23] ************************ 0000011C | 54 1B 12 4B | time_when = 1259477844 00000120 | 17 | event_type = 23 00000121 | 64 00 00 00 | server_id = 100 00000125 | 22 00 00 00 | event_len = 34 00000129 | 3E 01 00 00 | log_pos = 0000013E 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F ------------------------ 0000012F | 0F 00 00 00 | table_id = 15 ... mysqlbinlog output ~~~~~~~~~~~~~~~~~~ Something like this: ... # at 199 # at 243 # at 284 #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values (1)` #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped to number 15 #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 flags: STMT_END_F BINLOG ' VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== '/*!*/; ### INSERT INTO test.t1 ### SET ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ ... LOW-LEVEL DESIGN: mysql_binlog_send() [sql/sql_repl.cc] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. When sending events to a slave, master should simply skip Annotate_rows events (they are not needed for replication). [ ??? Multi-master - currently not clear ] 2. When sending events to mysqlbinlog (remote case), master must send Annotate_rows events as well. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Alexi): Store in binlog text of statements that caused RBR events (47)
by worklog-noreply＠askmonty.org 30 Nov '09

30 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Store in binlog text of statements that caused RBR events CREATION DATE..: Sat, 15 Aug 2009, 23:48 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: Knielsen CATEGORY.......: Server-RawIdeaBin TASK ID........: 47 (http://askmonty.org/worklog/?tid=47) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Alexi - Mon, 30 Nov 2009, 10:33)=-=- Low Level Design modified. --- /tmp/wklog.47.old.16188 2009-11-30 10:33:44.000000000 +0200 +++ /tmp/wklog.47.new.16188 2009-11-30 10:33:44.000000000 +0200 @@ -2,6 +2,7 @@ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. When sending events to a slave, master should simply skip Annotate_rows events (they are not needed for replication). + [ ??? Multi-master - currently not clear ] 2. When sending events to mysqlbinlog (remote case), master must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 13:00)=-=- Low Level Design modified. --- /tmp/wklog.47.old.32047 2009-11-29 13:00:21.000000000 +0200 +++ /tmp/wklog.47.new.32047 2009-11-29 13:00:21.000000000 +0200 @@ -1 +1,7 @@ +mysql_binlog_send() [sql/sql_repl.cc] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +1. When sending events to a slave, master should simply skip + Annotate_rows events (they are not needed for replication). +2. When sending events to mysqlbinlog (remote case), master + must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 09:50)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.24993 2009-11-29 07:50:36.000000000 +0000 +++ /tmp/wklog.47.new.24993 2009-11-29 07:50:36.000000000 +0000 @@ -4,3 +4,96 @@ > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). +New server option +~~~~~~~~~~~~~~~~~ + --binlog-annotate-row-events + +Setting this option makes RBR (row-) events in the binary log to be +preceded by Annotate rows events (see below). The corresponding +'binlog_annotate_row_events' system variable is dynamic and has both +global and session values. Default global value is OFF. + +Note. Session values are usefull to make it possible to annotate only + some selected statements: + + ... + SET SESSION binlog_annotate_row_events=ON; + ... statements to be annotated ... + SET SESSION binlog_annotate_row_events=OFF; + ... statements not to be annotated ... + +New binlog event type +~~~~~~~~~~~~~~~~~~~~~ + Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] + +Describes the query which caused the corresponding row event. In binary log, +precedes each Table_map_log_event. Contains empty post-header and the query +text in its data part. Example: + + ... + ************************ + ANNOTATE_ROWS_EVENT [51] + ************************ + 000000C7 | 54 1B 12 4B | time_when = 1259477844 + 000000CB | 33 | event_type = 51 + 000000CC | 64 00 00 00 | server_id = 100 + 000000D0 | 2C 00 00 00 | event_len = 44 + 000000D4 | F3 00 00 00 | log_pos = 000000F3 + 000000D8 | 00 00 | flags = <none> + ------------------------ + 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" + 000000DE | 72 74 20 69 | + 000000E2 | 6E 74 6F 20 | + 000000E6 | 74 31 20 76 | + 000000EA | 61 6C 75 65 | + 000000EE | 73 20 28 31 | + 000000F2 | 29 | + ************************ + TABLE_MAP_EVENT [19] + ************************ + 000000F3 | 54 1B 12 4B | time_when = 1259477844 + 000000F7 | 13 | event_type = 19 + 000000F8 | 64 00 00 00 | server_id = 100 + 000000FC | 29 00 00 00 | event_len = 41 + 00000100 | 1C 01 00 00 | log_pos = 0000011C + 00000104 | 00 00 | flags = <none> + ------------------------ + ... + ************************ + WRITE_ROWS_EVENT [23] + ************************ + 0000011C | 54 1B 12 4B | time_when = 1259477844 + 00000120 | 17 | event_type = 23 + 00000121 | 64 00 00 00 | server_id = 100 + 00000125 | 22 00 00 00 | event_len = 34 + 00000129 | 3E 01 00 00 | log_pos = 0000013E + 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F + ------------------------ + 0000012F | 0F 00 00 00 | table_id = 15 + ... + +mysqlbinlog output +~~~~~~~~~~~~~~~~~~ +Something like this: + + ... + # at 199 + # at 243 + # at 284 + #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values +(1)` + #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped +to number 15 + #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 +flags: STMT_END_F + + BINLOG ' + VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= + VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= + VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== + '/*!*/; + ### INSERT INTO test.t1 + ### SET + ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ + ... + -=-=(Knielsen - Fri, 27 Nov 2009, 13:30)=-=- Observers changed: Knielsen -=-=(Psergey - Sun, 16 Aug 2009, 11:08)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.12485 2009-08-16 11:08:33.000000000 +0300 +++ /tmp/wklog.47.new.12485 2009-08-16 11:08:33.000000000 +0300 @@ -1 +1,6 @@ +First suggestion: + +> I think for this we would actually need a new binlog event type +> (Comment_log_event?). Unless we want to log an empty statement Query_log_event +> containing only a comment (a bit of a hack). -=-=(Psergey - Sun, 16 Aug 2009, 00:02)=-=- Dependency created: 39 now depends on 47 DESCRIPTION: Store in binlog (and show in mysqlbinlog output) texts of statements that caused RBR events This is needed for (list from Monty): - Easier to understand why updates happened - Would make it easier to find out where in application things went wrong (as you can search for exact strings) - Allow one to filter things based on comments in the statement. The cost of this can be that the binlog will be approximately 2x in size (especially insert of big blob's would be a bit painful), so this should be an optional feature. HIGH-LEVEL SPECIFICATION: First suggestion: > I think for this we would actually need a new binlog event type > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). New server option ~~~~~~~~~~~~~~~~~ --binlog-annotate-row-events Setting this option makes RBR (row-) events in the binary log to be preceded by Annotate rows events (see below). The corresponding 'binlog_annotate_row_events' system variable is dynamic and has both global and session values. Default global value is OFF. Note. Session values are usefull to make it possible to annotate only some selected statements: ... SET SESSION binlog_annotate_row_events=ON; ... statements to be annotated ... SET SESSION binlog_annotate_row_events=OFF; ... statements not to be annotated ... New binlog event type ~~~~~~~~~~~~~~~~~~~~~ Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query text in its data part. Example: ... ************************ ANNOTATE_ROWS_EVENT [51] ************************ 000000C7 | 54 1B 12 4B | time_when = 1259477844 000000CB | 33 | event_type = 51 000000CC | 64 00 00 00 | server_id = 100 000000D0 | 2C 00 00 00 | event_len = 44 000000D4 | F3 00 00 00 | log_pos = 000000F3 000000D8 | 00 00 | flags = <none> ------------------------ 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" 000000DE | 72 74 20 69 | 000000E2 | 6E 74 6F 20 | 000000E6 | 74 31 20 76 | 000000EA | 61 6C 75 65 | 000000EE | 73 20 28 31 | 000000F2 | 29 | ************************ TABLE_MAP_EVENT [19] ************************ 000000F3 | 54 1B 12 4B | time_when = 1259477844 000000F7 | 13 | event_type = 19 000000F8 | 64 00 00 00 | server_id = 100 000000FC | 29 00 00 00 | event_len = 41 00000100 | 1C 01 00 00 | log_pos = 0000011C 00000104 | 00 00 | flags = <none> ------------------------ ... ************************ WRITE_ROWS_EVENT [23] ************************ 0000011C | 54 1B 12 4B | time_when = 1259477844 00000120 | 17 | event_type = 23 00000121 | 64 00 00 00 | server_id = 100 00000125 | 22 00 00 00 | event_len = 34 00000129 | 3E 01 00 00 | log_pos = 0000013E 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F ------------------------ 0000012F | 0F 00 00 00 | table_id = 15 ... mysqlbinlog output ~~~~~~~~~~~~~~~~~~ Something like this: ... # at 199 # at 243 # at 284 #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values (1)` #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped to number 15 #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 flags: STMT_END_F BINLOG ' VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== '/*!*/; ### INSERT INTO test.t1 ### SET ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ ... LOW-LEVEL DESIGN: mysql_binlog_send() [sql/sql_repl.cc] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. When sending events to a slave, master should simply skip Annotate_rows events (they are not needed for replication). [ ??? Multi-master - currently not clear ] 2. When sending events to mysqlbinlog (remote case), master must send Annotate_rows events as well. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Alexi): Store in binlog text of statements that caused RBR events (47)
by worklog-noreply＠askmonty.org 30 Nov '09

30 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Store in binlog text of statements that caused RBR events CREATION DATE..: Sat, 15 Aug 2009, 23:48 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: Knielsen CATEGORY.......: Server-RawIdeaBin TASK ID........: 47 (http://askmonty.org/worklog/?tid=47) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Alexi - Mon, 30 Nov 2009, 10:33)=-=- Low Level Design modified. --- /tmp/wklog.47.old.16188 2009-11-30 10:33:44.000000000 +0200 +++ /tmp/wklog.47.new.16188 2009-11-30 10:33:44.000000000 +0200 @@ -2,6 +2,7 @@ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. When sending events to a slave, master should simply skip Annotate_rows events (they are not needed for replication). + [ ??? Multi-master - currently not clear ] 2. When sending events to mysqlbinlog (remote case), master must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 13:00)=-=- Low Level Design modified. --- /tmp/wklog.47.old.32047 2009-11-29 13:00:21.000000000 +0200 +++ /tmp/wklog.47.new.32047 2009-11-29 13:00:21.000000000 +0200 @@ -1 +1,7 @@ +mysql_binlog_send() [sql/sql_repl.cc] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +1. When sending events to a slave, master should simply skip + Annotate_rows events (they are not needed for replication). +2. When sending events to mysqlbinlog (remote case), master + must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 09:50)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.24993 2009-11-29 07:50:36.000000000 +0000 +++ /tmp/wklog.47.new.24993 2009-11-29 07:50:36.000000000 +0000 @@ -4,3 +4,96 @@ > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). +New server option +~~~~~~~~~~~~~~~~~ + --binlog-annotate-row-events + +Setting this option makes RBR (row-) events in the binary log to be +preceded by Annotate rows events (see below). The corresponding +'binlog_annotate_row_events' system variable is dynamic and has both +global and session values. Default global value is OFF. + +Note. Session values are usefull to make it possible to annotate only + some selected statements: + + ... + SET SESSION binlog_annotate_row_events=ON; + ... statements to be annotated ... + SET SESSION binlog_annotate_row_events=OFF; + ... statements not to be annotated ... + +New binlog event type +~~~~~~~~~~~~~~~~~~~~~ + Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] + +Describes the query which caused the corresponding row event. In binary log, +precedes each Table_map_log_event. Contains empty post-header and the query +text in its data part. Example: + + ... + ************************ + ANNOTATE_ROWS_EVENT [51] + ************************ + 000000C7 | 54 1B 12 4B | time_when = 1259477844 + 000000CB | 33 | event_type = 51 + 000000CC | 64 00 00 00 | server_id = 100 + 000000D0 | 2C 00 00 00 | event_len = 44 + 000000D4 | F3 00 00 00 | log_pos = 000000F3 + 000000D8 | 00 00 | flags = <none> + ------------------------ + 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" + 000000DE | 72 74 20 69 | + 000000E2 | 6E 74 6F 20 | + 000000E6 | 74 31 20 76 | + 000000EA | 61 6C 75 65 | + 000000EE | 73 20 28 31 | + 000000F2 | 29 | + ************************ + TABLE_MAP_EVENT [19] + ************************ + 000000F3 | 54 1B 12 4B | time_when = 1259477844 + 000000F7 | 13 | event_type = 19 + 000000F8 | 64 00 00 00 | server_id = 100 + 000000FC | 29 00 00 00 | event_len = 41 + 00000100 | 1C 01 00 00 | log_pos = 0000011C + 00000104 | 00 00 | flags = <none> + ------------------------ + ... + ************************ + WRITE_ROWS_EVENT [23] + ************************ + 0000011C | 54 1B 12 4B | time_when = 1259477844 + 00000120 | 17 | event_type = 23 + 00000121 | 64 00 00 00 | server_id = 100 + 00000125 | 22 00 00 00 | event_len = 34 + 00000129 | 3E 01 00 00 | log_pos = 0000013E + 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F + ------------------------ + 0000012F | 0F 00 00 00 | table_id = 15 + ... + +mysqlbinlog output +~~~~~~~~~~~~~~~~~~ +Something like this: + + ... + # at 199 + # at 243 + # at 284 + #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values +(1)` + #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped +to number 15 + #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 +flags: STMT_END_F + + BINLOG ' + VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= + VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= + VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== + '/*!*/; + ### INSERT INTO test.t1 + ### SET + ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ + ... + -=-=(Knielsen - Fri, 27 Nov 2009, 13:30)=-=- Observers changed: Knielsen -=-=(Psergey - Sun, 16 Aug 2009, 11:08)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.12485 2009-08-16 11:08:33.000000000 +0300 +++ /tmp/wklog.47.new.12485 2009-08-16 11:08:33.000000000 +0300 @@ -1 +1,6 @@ +First suggestion: + +> I think for this we would actually need a new binlog event type +> (Comment_log_event?). Unless we want to log an empty statement Query_log_event +> containing only a comment (a bit of a hack). -=-=(Psergey - Sun, 16 Aug 2009, 00:02)=-=- Dependency created: 39 now depends on 47 DESCRIPTION: Store in binlog (and show in mysqlbinlog output) texts of statements that caused RBR events This is needed for (list from Monty): - Easier to understand why updates happened - Would make it easier to find out where in application things went wrong (as you can search for exact strings) - Allow one to filter things based on comments in the statement. The cost of this can be that the binlog will be approximately 2x in size (especially insert of big blob's would be a bit painful), so this should be an optional feature. HIGH-LEVEL SPECIFICATION: First suggestion: > I think for this we would actually need a new binlog event type > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). New server option ~~~~~~~~~~~~~~~~~ --binlog-annotate-row-events Setting this option makes RBR (row-) events in the binary log to be preceded by Annotate rows events (see below). The corresponding 'binlog_annotate_row_events' system variable is dynamic and has both global and session values. Default global value is OFF. Note. Session values are usefull to make it possible to annotate only some selected statements: ... SET SESSION binlog_annotate_row_events=ON; ... statements to be annotated ... SET SESSION binlog_annotate_row_events=OFF; ... statements not to be annotated ... New binlog event type ~~~~~~~~~~~~~~~~~~~~~ Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query text in its data part. Example: ... ************************ ANNOTATE_ROWS_EVENT [51] ************************ 000000C7 | 54 1B 12 4B | time_when = 1259477844 000000CB | 33 | event_type = 51 000000CC | 64 00 00 00 | server_id = 100 000000D0 | 2C 00 00 00 | event_len = 44 000000D4 | F3 00 00 00 | log_pos = 000000F3 000000D8 | 00 00 | flags = <none> ------------------------ 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" 000000DE | 72 74 20 69 | 000000E2 | 6E 74 6F 20 | 000000E6 | 74 31 20 76 | 000000EA | 61 6C 75 65 | 000000EE | 73 20 28 31 | 000000F2 | 29 | ************************ TABLE_MAP_EVENT [19] ************************ 000000F3 | 54 1B 12 4B | time_when = 1259477844 000000F7 | 13 | event_type = 19 000000F8 | 64 00 00 00 | server_id = 100 000000FC | 29 00 00 00 | event_len = 41 00000100 | 1C 01 00 00 | log_pos = 0000011C 00000104 | 00 00 | flags = <none> ------------------------ ... ************************ WRITE_ROWS_EVENT [23] ************************ 0000011C | 54 1B 12 4B | time_when = 1259477844 00000120 | 17 | event_type = 23 00000121 | 64 00 00 00 | server_id = 100 00000125 | 22 00 00 00 | event_len = 34 00000129 | 3E 01 00 00 | log_pos = 0000013E 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F ------------------------ 0000012F | 0F 00 00 00 | table_id = 15 ... mysqlbinlog output ~~~~~~~~~~~~~~~~~~ Something like this: ... # at 199 # at 243 # at 284 #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values (1)` #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped to number 15 #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 flags: STMT_END_F BINLOG ' VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== '/*!*/; ### INSERT INTO test.t1 ### SET ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ ... LOW-LEVEL DESIGN: mysql_binlog_send() [sql/sql_repl.cc] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. When sending events to a slave, master should simply skip Annotate_rows events (they are not needed for replication). [ ??? Multi-master - currently not clear ] 2. When sending events to mysqlbinlog (remote case), master must send Annotate_rows events as well. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Alexi): Store in binlog text of statements that caused RBR events (47)
by worklog-noreply＠askmonty.org 30 Nov '09

30 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Store in binlog text of statements that caused RBR events CREATION DATE..: Sat, 15 Aug 2009, 23:48 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: Knielsen CATEGORY.......: Server-RawIdeaBin TASK ID........: 47 (http://askmonty.org/worklog/?tid=47) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Alexi - Mon, 30 Nov 2009, 10:33)=-=- Low Level Design modified. --- /tmp/wklog.47.old.16188 2009-11-30 10:33:44.000000000 +0200 +++ /tmp/wklog.47.new.16188 2009-11-30 10:33:44.000000000 +0200 @@ -2,6 +2,7 @@ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. When sending events to a slave, master should simply skip Annotate_rows events (they are not needed for replication). + [ ??? Multi-master - currently not clear ] 2. When sending events to mysqlbinlog (remote case), master must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 13:00)=-=- Low Level Design modified. --- /tmp/wklog.47.old.32047 2009-11-29 13:00:21.000000000 +0200 +++ /tmp/wklog.47.new.32047 2009-11-29 13:00:21.000000000 +0200 @@ -1 +1,7 @@ +mysql_binlog_send() [sql/sql_repl.cc] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +1. When sending events to a slave, master should simply skip + Annotate_rows events (they are not needed for replication). +2. When sending events to mysqlbinlog (remote case), master + must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 09:50)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.24993 2009-11-29 07:50:36.000000000 +0000 +++ /tmp/wklog.47.new.24993 2009-11-29 07:50:36.000000000 +0000 @@ -4,3 +4,96 @@ > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). +New server option +~~~~~~~~~~~~~~~~~ + --binlog-annotate-row-events + +Setting this option makes RBR (row-) events in the binary log to be +preceded by Annotate rows events (see below). The corresponding +'binlog_annotate_row_events' system variable is dynamic and has both +global and session values. Default global value is OFF. + +Note. Session values are usefull to make it possible to annotate only + some selected statements: + + ... + SET SESSION binlog_annotate_row_events=ON; + ... statements to be annotated ... + SET SESSION binlog_annotate_row_events=OFF; + ... statements not to be annotated ... + +New binlog event type +~~~~~~~~~~~~~~~~~~~~~ + Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] + +Describes the query which caused the corresponding row event. In binary log, +precedes each Table_map_log_event. Contains empty post-header and the query +text in its data part. Example: + + ... + ************************ + ANNOTATE_ROWS_EVENT [51] + ************************ + 000000C7 | 54 1B 12 4B | time_when = 1259477844 + 000000CB | 33 | event_type = 51 + 000000CC | 64 00 00 00 | server_id = 100 + 000000D0 | 2C 00 00 00 | event_len = 44 + 000000D4 | F3 00 00 00 | log_pos = 000000F3 + 000000D8 | 00 00 | flags = <none> + ------------------------ + 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" + 000000DE | 72 74 20 69 | + 000000E2 | 6E 74 6F 20 | + 000000E6 | 74 31 20 76 | + 000000EA | 61 6C 75 65 | + 000000EE | 73 20 28 31 | + 000000F2 | 29 | + ************************ + TABLE_MAP_EVENT [19] + ************************ + 000000F3 | 54 1B 12 4B | time_when = 1259477844 + 000000F7 | 13 | event_type = 19 + 000000F8 | 64 00 00 00 | server_id = 100 + 000000FC | 29 00 00 00 | event_len = 41 + 00000100 | 1C 01 00 00 | log_pos = 0000011C + 00000104 | 00 00 | flags = <none> + ------------------------ + ... + ************************ + WRITE_ROWS_EVENT [23] + ************************ + 0000011C | 54 1B 12 4B | time_when = 1259477844 + 00000120 | 17 | event_type = 23 + 00000121 | 64 00 00 00 | server_id = 100 + 00000125 | 22 00 00 00 | event_len = 34 + 00000129 | 3E 01 00 00 | log_pos = 0000013E + 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F + ------------------------ + 0000012F | 0F 00 00 00 | table_id = 15 + ... + +mysqlbinlog output +~~~~~~~~~~~~~~~~~~ +Something like this: + + ... + # at 199 + # at 243 + # at 284 + #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values +(1)` + #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped +to number 15 + #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 +flags: STMT_END_F + + BINLOG ' + VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= + VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= + VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== + '/*!*/; + ### INSERT INTO test.t1 + ### SET + ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ + ... + -=-=(Knielsen - Fri, 27 Nov 2009, 13:30)=-=- Observers changed: Knielsen -=-=(Psergey - Sun, 16 Aug 2009, 11:08)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.12485 2009-08-16 11:08:33.000000000 +0300 +++ /tmp/wklog.47.new.12485 2009-08-16 11:08:33.000000000 +0300 @@ -1 +1,6 @@ +First suggestion: + +> I think for this we would actually need a new binlog event type +> (Comment_log_event?). Unless we want to log an empty statement Query_log_event +> containing only a comment (a bit of a hack). -=-=(Psergey - Sun, 16 Aug 2009, 00:02)=-=- Dependency created: 39 now depends on 47 DESCRIPTION: Store in binlog (and show in mysqlbinlog output) texts of statements that caused RBR events This is needed for (list from Monty): - Easier to understand why updates happened - Would make it easier to find out where in application things went wrong (as you can search for exact strings) - Allow one to filter things based on comments in the statement. The cost of this can be that the binlog will be approximately 2x in size (especially insert of big blob's would be a bit painful), so this should be an optional feature. HIGH-LEVEL SPECIFICATION: First suggestion: > I think for this we would actually need a new binlog event type > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). New server option ~~~~~~~~~~~~~~~~~ --binlog-annotate-row-events Setting this option makes RBR (row-) events in the binary log to be preceded by Annotate rows events (see below). The corresponding 'binlog_annotate_row_events' system variable is dynamic and has both global and session values. Default global value is OFF. Note. Session values are usefull to make it possible to annotate only some selected statements: ... SET SESSION binlog_annotate_row_events=ON; ... statements to be annotated ... SET SESSION binlog_annotate_row_events=OFF; ... statements not to be annotated ... New binlog event type ~~~~~~~~~~~~~~~~~~~~~ Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query text in its data part. Example: ... ************************ ANNOTATE_ROWS_EVENT [51] ************************ 000000C7 | 54 1B 12 4B | time_when = 1259477844 000000CB | 33 | event_type = 51 000000CC | 64 00 00 00 | server_id = 100 000000D0 | 2C 00 00 00 | event_len = 44 000000D4 | F3 00 00 00 | log_pos = 000000F3 000000D8 | 00 00 | flags = <none> ------------------------ 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" 000000DE | 72 74 20 69 | 000000E2 | 6E 74 6F 20 | 000000E6 | 74 31 20 76 | 000000EA | 61 6C 75 65 | 000000EE | 73 20 28 31 | 000000F2 | 29 | ************************ TABLE_MAP_EVENT [19] ************************ 000000F3 | 54 1B 12 4B | time_when = 1259477844 000000F7 | 13 | event_type = 19 000000F8 | 64 00 00 00 | server_id = 100 000000FC | 29 00 00 00 | event_len = 41 00000100 | 1C 01 00 00 | log_pos = 0000011C 00000104 | 00 00 | flags = <none> ------------------------ ... ************************ WRITE_ROWS_EVENT [23] ************************ 0000011C | 54 1B 12 4B | time_when = 1259477844 00000120 | 17 | event_type = 23 00000121 | 64 00 00 00 | server_id = 100 00000125 | 22 00 00 00 | event_len = 34 00000129 | 3E 01 00 00 | log_pos = 0000013E 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F ------------------------ 0000012F | 0F 00 00 00 | table_id = 15 ... mysqlbinlog output ~~~~~~~~~~~~~~~~~~ Something like this: ... # at 199 # at 243 # at 284 #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values (1)` #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped to number 15 #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 flags: STMT_END_F BINLOG ' VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== '/*!*/; ### INSERT INTO test.t1 ### SET ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ ... LOW-LEVEL DESIGN: mysql_binlog_send() [sql/sql_repl.cc] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. When sending events to a slave, master should simply skip Annotate_rows events (they are not needed for replication). [ ??? Multi-master - currently not clear ] 2. When sending events to mysqlbinlog (remote case), master must send Annotate_rows events as well. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (monty:2771)
by Michael Widenius 30 Nov '09

30 Nov '09

#At lp:maria based on revid:monty@askmonty.org-20091129230856-358dn8f5bqx1d67g 2771 Michael Widenius 2009-11-30 Ignore temporary directory modified: .bzrignore === modified file '.bzrignore' --- a/.bzrignore 2009-11-06 17:22:32 +0000 +++ b/.bzrignore 2009-11-29 23:16:14 +0000 @@ -1922,3 +1922,4 @@ libmysqld/examples/mysqltest.cc extra/libevent/event-config.h libmysqld/opt_table_elimination.cc libmysqld/ha_federatedx.cc +tmp

1 0

[Maria-developers] Updated (by Alexi): Store in binlog text of statements that caused RBR events (47)
by worklog-noreply＠askmonty.org 29 Nov '09

29 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Store in binlog text of statements that caused RBR events CREATION DATE..: Sat, 15 Aug 2009, 23:48 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: Knielsen CATEGORY.......: Server-RawIdeaBin TASK ID........: 47 (http://askmonty.org/worklog/?tid=47) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Alexi - Sun, 29 Nov 2009, 13:00)=-=- Low Level Design modified. --- /tmp/wklog.47.old.32047 2009-11-29 13:00:21.000000000 +0200 +++ /tmp/wklog.47.new.32047 2009-11-29 13:00:21.000000000 +0200 @@ -1 +1,7 @@ +mysql_binlog_send() [sql/sql_repl.cc] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +1. When sending events to a slave, master should simply skip + Annotate_rows events (they are not needed for replication). +2. When sending events to mysqlbinlog (remote case), master + must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 09:50)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.24993 2009-11-29 07:50:36.000000000 +0000 +++ /tmp/wklog.47.new.24993 2009-11-29 07:50:36.000000000 +0000 @@ -4,3 +4,96 @@ > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). +New server option +~~~~~~~~~~~~~~~~~ + --binlog-annotate-row-events + +Setting this option makes RBR (row-) events in the binary log to be +preceded by Annotate rows events (see below). The corresponding +'binlog_annotate_row_events' system variable is dynamic and has both +global and session values. Default global value is OFF. + +Note. Session values are usefull to make it possible to annotate only + some selected statements: + + ... + SET SESSION binlog_annotate_row_events=ON; + ... statements to be annotated ... + SET SESSION binlog_annotate_row_events=OFF; + ... statements not to be annotated ... + +New binlog event type +~~~~~~~~~~~~~~~~~~~~~ + Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] + +Describes the query which caused the corresponding row event. In binary log, +precedes each Table_map_log_event. Contains empty post-header and the query +text in its data part. Example: + + ... + ************************ + ANNOTATE_ROWS_EVENT [51] + ************************ + 000000C7 | 54 1B 12 4B | time_when = 1259477844 + 000000CB | 33 | event_type = 51 + 000000CC | 64 00 00 00 | server_id = 100 + 000000D0 | 2C 00 00 00 | event_len = 44 + 000000D4 | F3 00 00 00 | log_pos = 000000F3 + 000000D8 | 00 00 | flags = <none> + ------------------------ + 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" + 000000DE | 72 74 20 69 | + 000000E2 | 6E 74 6F 20 | + 000000E6 | 74 31 20 76 | + 000000EA | 61 6C 75 65 | + 000000EE | 73 20 28 31 | + 000000F2 | 29 | + ************************ + TABLE_MAP_EVENT [19] + ************************ + 000000F3 | 54 1B 12 4B | time_when = 1259477844 + 000000F7 | 13 | event_type = 19 + 000000F8 | 64 00 00 00 | server_id = 100 + 000000FC | 29 00 00 00 | event_len = 41 + 00000100 | 1C 01 00 00 | log_pos = 0000011C + 00000104 | 00 00 | flags = <none> + ------------------------ + ... + ************************ + WRITE_ROWS_EVENT [23] + ************************ + 0000011C | 54 1B 12 4B | time_when = 1259477844 + 00000120 | 17 | event_type = 23 + 00000121 | 64 00 00 00 | server_id = 100 + 00000125 | 22 00 00 00 | event_len = 34 + 00000129 | 3E 01 00 00 | log_pos = 0000013E + 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F + ------------------------ + 0000012F | 0F 00 00 00 | table_id = 15 + ... + +mysqlbinlog output +~~~~~~~~~~~~~~~~~~ +Something like this: + + ... + # at 199 + # at 243 + # at 284 + #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values +(1)` + #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped +to number 15 + #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 +flags: STMT_END_F + + BINLOG ' + VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= + VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= + VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== + '/*!*/; + ### INSERT INTO test.t1 + ### SET + ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ + ... + -=-=(Knielsen - Fri, 27 Nov 2009, 13:30)=-=- Observers changed: Knielsen -=-=(Psergey - Sun, 16 Aug 2009, 11:08)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.12485 2009-08-16 11:08:33.000000000 +0300 +++ /tmp/wklog.47.new.12485 2009-08-16 11:08:33.000000000 +0300 @@ -1 +1,6 @@ +First suggestion: + +> I think for this we would actually need a new binlog event type +> (Comment_log_event?). Unless we want to log an empty statement Query_log_event +> containing only a comment (a bit of a hack). -=-=(Psergey - Sun, 16 Aug 2009, 00:02)=-=- Dependency created: 39 now depends on 47 DESCRIPTION: Store in binlog (and show in mysqlbinlog output) texts of statements that caused RBR events This is needed for (list from Monty): - Easier to understand why updates happened - Would make it easier to find out where in application things went wrong (as you can search for exact strings) - Allow one to filter things based on comments in the statement. The cost of this can be that the binlog will be approximately 2x in size (especially insert of big blob's would be a bit painful), so this should be an optional feature. HIGH-LEVEL SPECIFICATION: First suggestion: > I think for this we would actually need a new binlog event type > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). New server option ~~~~~~~~~~~~~~~~~ --binlog-annotate-row-events Setting this option makes RBR (row-) events in the binary log to be preceded by Annotate rows events (see below). The corresponding 'binlog_annotate_row_events' system variable is dynamic and has both global and session values. Default global value is OFF. Note. Session values are usefull to make it possible to annotate only some selected statements: ... SET SESSION binlog_annotate_row_events=ON; ... statements to be annotated ... SET SESSION binlog_annotate_row_events=OFF; ... statements not to be annotated ... New binlog event type ~~~~~~~~~~~~~~~~~~~~~ Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query text in its data part. Example: ... ************************ ANNOTATE_ROWS_EVENT [51] ************************ 000000C7 | 54 1B 12 4B | time_when = 1259477844 000000CB | 33 | event_type = 51 000000CC | 64 00 00 00 | server_id = 100 000000D0 | 2C 00 00 00 | event_len = 44 000000D4 | F3 00 00 00 | log_pos = 000000F3 000000D8 | 00 00 | flags = <none> ------------------------ 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" 000000DE | 72 74 20 69 | 000000E2 | 6E 74 6F 20 | 000000E6 | 74 31 20 76 | 000000EA | 61 6C 75 65 | 000000EE | 73 20 28 31 | 000000F2 | 29 | ************************ TABLE_MAP_EVENT [19] ************************ 000000F3 | 54 1B 12 4B | time_when = 1259477844 000000F7 | 13 | event_type = 19 000000F8 | 64 00 00 00 | server_id = 100 000000FC | 29 00 00 00 | event_len = 41 00000100 | 1C 01 00 00 | log_pos = 0000011C 00000104 | 00 00 | flags = <none> ------------------------ ... ************************ WRITE_ROWS_EVENT [23] ************************ 0000011C | 54 1B 12 4B | time_when = 1259477844 00000120 | 17 | event_type = 23 00000121 | 64 00 00 00 | server_id = 100 00000125 | 22 00 00 00 | event_len = 34 00000129 | 3E 01 00 00 | log_pos = 0000013E 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F ------------------------ 0000012F | 0F 00 00 00 | table_id = 15 ... mysqlbinlog output ~~~~~~~~~~~~~~~~~~ Something like this: ... # at 199 # at 243 # at 284 #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values (1)` #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped to number 15 #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 flags: STMT_END_F BINLOG ' VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== '/*!*/; ### INSERT INTO test.t1 ### SET ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ ... LOW-LEVEL DESIGN: mysql_binlog_send() [sql/sql_repl.cc] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. When sending events to a slave, master should simply skip Annotate_rows events (they are not needed for replication). 2. When sending events to mysqlbinlog (remote case), master must send Annotate_rows events as well. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Alexi): Store in binlog text of statements that caused RBR events (47)
by worklog-noreply＠askmonty.org 29 Nov '09

29 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Store in binlog text of statements that caused RBR events CREATION DATE..: Sat, 15 Aug 2009, 23:48 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: Knielsen CATEGORY.......: Server-RawIdeaBin TASK ID........: 47 (http://askmonty.org/worklog/?tid=47) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Alexi - Sun, 29 Nov 2009, 13:00)=-=- Low Level Design modified. --- /tmp/wklog.47.old.32047 2009-11-29 13:00:21.000000000 +0200 +++ /tmp/wklog.47.new.32047 2009-11-29 13:00:21.000000000 +0200 @@ -1 +1,7 @@ +mysql_binlog_send() [sql/sql_repl.cc] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +1. When sending events to a slave, master should simply skip + Annotate_rows events (they are not needed for replication). +2. When sending events to mysqlbinlog (remote case), master + must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 09:50)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.24993 2009-11-29 07:50:36.000000000 +0000 +++ /tmp/wklog.47.new.24993 2009-11-29 07:50:36.000000000 +0000 @@ -4,3 +4,96 @@ > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). +New server option +~~~~~~~~~~~~~~~~~ + --binlog-annotate-row-events + +Setting this option makes RBR (row-) events in the binary log to be +preceded by Annotate rows events (see below). The corresponding +'binlog_annotate_row_events' system variable is dynamic and has both +global and session values. Default global value is OFF. + +Note. Session values are usefull to make it possible to annotate only + some selected statements: + + ... + SET SESSION binlog_annotate_row_events=ON; + ... statements to be annotated ... + SET SESSION binlog_annotate_row_events=OFF; + ... statements not to be annotated ... + +New binlog event type +~~~~~~~~~~~~~~~~~~~~~ + Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] + +Describes the query which caused the corresponding row event. In binary log, +precedes each Table_map_log_event. Contains empty post-header and the query +text in its data part. Example: + + ... + ************************ + ANNOTATE_ROWS_EVENT [51] + ************************ + 000000C7 | 54 1B 12 4B | time_when = 1259477844 + 000000CB | 33 | event_type = 51 + 000000CC | 64 00 00 00 | server_id = 100 + 000000D0 | 2C 00 00 00 | event_len = 44 + 000000D4 | F3 00 00 00 | log_pos = 000000F3 + 000000D8 | 00 00 | flags = <none> + ------------------------ + 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" + 000000DE | 72 74 20 69 | + 000000E2 | 6E 74 6F 20 | + 000000E6 | 74 31 20 76 | + 000000EA | 61 6C 75 65 | + 000000EE | 73 20 28 31 | + 000000F2 | 29 | + ************************ + TABLE_MAP_EVENT [19] + ************************ + 000000F3 | 54 1B 12 4B | time_when = 1259477844 + 000000F7 | 13 | event_type = 19 + 000000F8 | 64 00 00 00 | server_id = 100 + 000000FC | 29 00 00 00 | event_len = 41 + 00000100 | 1C 01 00 00 | log_pos = 0000011C + 00000104 | 00 00 | flags = <none> + ------------------------ + ... + ************************ + WRITE_ROWS_EVENT [23] + ************************ + 0000011C | 54 1B 12 4B | time_when = 1259477844 + 00000120 | 17 | event_type = 23 + 00000121 | 64 00 00 00 | server_id = 100 + 00000125 | 22 00 00 00 | event_len = 34 + 00000129 | 3E 01 00 00 | log_pos = 0000013E + 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F + ------------------------ + 0000012F | 0F 00 00 00 | table_id = 15 + ... + +mysqlbinlog output +~~~~~~~~~~~~~~~~~~ +Something like this: + + ... + # at 199 + # at 243 + # at 284 + #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values +(1)` + #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped +to number 15 + #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 +flags: STMT_END_F + + BINLOG ' + VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= + VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= + VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== + '/*!*/; + ### INSERT INTO test.t1 + ### SET + ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ + ... + -=-=(Knielsen - Fri, 27 Nov 2009, 13:30)=-=- Observers changed: Knielsen -=-=(Psergey - Sun, 16 Aug 2009, 11:08)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.12485 2009-08-16 11:08:33.000000000 +0300 +++ /tmp/wklog.47.new.12485 2009-08-16 11:08:33.000000000 +0300 @@ -1 +1,6 @@ +First suggestion: + +> I think for this we would actually need a new binlog event type +> (Comment_log_event?). Unless we want to log an empty statement Query_log_event +> containing only a comment (a bit of a hack). -=-=(Psergey - Sun, 16 Aug 2009, 00:02)=-=- Dependency created: 39 now depends on 47 DESCRIPTION: Store in binlog (and show in mysqlbinlog output) texts of statements that caused RBR events This is needed for (list from Monty): - Easier to understand why updates happened - Would make it easier to find out where in application things went wrong (as you can search for exact strings) - Allow one to filter things based on comments in the statement. The cost of this can be that the binlog will be approximately 2x in size (especially insert of big blob's would be a bit painful), so this should be an optional feature. HIGH-LEVEL SPECIFICATION: First suggestion: > I think for this we would actually need a new binlog event type > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). New server option ~~~~~~~~~~~~~~~~~ --binlog-annotate-row-events Setting this option makes RBR (row-) events in the binary log to be preceded by Annotate rows events (see below). The corresponding 'binlog_annotate_row_events' system variable is dynamic and has both global and session values. Default global value is OFF. Note. Session values are usefull to make it possible to annotate only some selected statements: ... SET SESSION binlog_annotate_row_events=ON; ... statements to be annotated ... SET SESSION binlog_annotate_row_events=OFF; ... statements not to be annotated ... New binlog event type ~~~~~~~~~~~~~~~~~~~~~ Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query text in its data part. Example: ... ************************ ANNOTATE_ROWS_EVENT [51] ************************ 000000C7 | 54 1B 12 4B | time_when = 1259477844 000000CB | 33 | event_type = 51 000000CC | 64 00 00 00 | server_id = 100 000000D0 | 2C 00 00 00 | event_len = 44 000000D4 | F3 00 00 00 | log_pos = 000000F3 000000D8 | 00 00 | flags = <none> ------------------------ 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" 000000DE | 72 74 20 69 | 000000E2 | 6E 74 6F 20 | 000000E6 | 74 31 20 76 | 000000EA | 61 6C 75 65 | 000000EE | 73 20 28 31 | 000000F2 | 29 | ************************ TABLE_MAP_EVENT [19] ************************ 000000F3 | 54 1B 12 4B | time_when = 1259477844 000000F7 | 13 | event_type = 19 000000F8 | 64 00 00 00 | server_id = 100 000000FC | 29 00 00 00 | event_len = 41 00000100 | 1C 01 00 00 | log_pos = 0000011C 00000104 | 00 00 | flags = <none> ------------------------ ... ************************ WRITE_ROWS_EVENT [23] ************************ 0000011C | 54 1B 12 4B | time_when = 1259477844 00000120 | 17 | event_type = 23 00000121 | 64 00 00 00 | server_id = 100 00000125 | 22 00 00 00 | event_len = 34 00000129 | 3E 01 00 00 | log_pos = 0000013E 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F ------------------------ 0000012F | 0F 00 00 00 | table_id = 15 ... mysqlbinlog output ~~~~~~~~~~~~~~~~~~ Something like this: ... # at 199 # at 243 # at 284 #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values (1)` #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped to number 15 #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 flags: STMT_END_F BINLOG ' VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== '/*!*/; ### INSERT INTO test.t1 ### SET ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ ... LOW-LEVEL DESIGN: mysql_binlog_send() [sql/sql_repl.cc] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. When sending events to a slave, master should simply skip Annotate_rows events (they are not needed for replication). 2. When sending events to mysqlbinlog (remote case), master must send Annotate_rows events as well. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Alexi): Store in binlog text of statements that caused RBR events (47)
by worklog-noreply＠askmonty.org 29 Nov '09

29 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Store in binlog text of statements that caused RBR events CREATION DATE..: Sat, 15 Aug 2009, 23:48 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: Knielsen CATEGORY.......: Server-RawIdeaBin TASK ID........: 47 (http://askmonty.org/worklog/?tid=47) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Alexi - Sun, 29 Nov 2009, 13:00)=-=- Low Level Design modified. --- /tmp/wklog.47.old.32047 2009-11-29 13:00:21.000000000 +0200 +++ /tmp/wklog.47.new.32047 2009-11-29 13:00:21.000000000 +0200 @@ -1 +1,7 @@ +mysql_binlog_send() [sql/sql_repl.cc] +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +1. When sending events to a slave, master should simply skip + Annotate_rows events (they are not needed for replication). +2. When sending events to mysqlbinlog (remote case), master + must send Annotate_rows events as well. -=-=(Alexi - Sun, 29 Nov 2009, 09:50)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.24993 2009-11-29 07:50:36.000000000 +0000 +++ /tmp/wklog.47.new.24993 2009-11-29 07:50:36.000000000 +0000 @@ -4,3 +4,96 @@ > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). +New server option +~~~~~~~~~~~~~~~~~ + --binlog-annotate-row-events + +Setting this option makes RBR (row-) events in the binary log to be +preceded by Annotate rows events (see below). The corresponding +'binlog_annotate_row_events' system variable is dynamic and has both +global and session values. Default global value is OFF. + +Note. Session values are usefull to make it possible to annotate only + some selected statements: + + ... + SET SESSION binlog_annotate_row_events=ON; + ... statements to be annotated ... + SET SESSION binlog_annotate_row_events=OFF; + ... statements not to be annotated ... + +New binlog event type +~~~~~~~~~~~~~~~~~~~~~ + Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] + +Describes the query which caused the corresponding row event. In binary log, +precedes each Table_map_log_event. Contains empty post-header and the query +text in its data part. Example: + + ... + ************************ + ANNOTATE_ROWS_EVENT [51] + ************************ + 000000C7 | 54 1B 12 4B | time_when = 1259477844 + 000000CB | 33 | event_type = 51 + 000000CC | 64 00 00 00 | server_id = 100 + 000000D0 | 2C 00 00 00 | event_len = 44 + 000000D4 | F3 00 00 00 | log_pos = 000000F3 + 000000D8 | 00 00 | flags = <none> + ------------------------ + 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" + 000000DE | 72 74 20 69 | + 000000E2 | 6E 74 6F 20 | + 000000E6 | 74 31 20 76 | + 000000EA | 61 6C 75 65 | + 000000EE | 73 20 28 31 | + 000000F2 | 29 | + ************************ + TABLE_MAP_EVENT [19] + ************************ + 000000F3 | 54 1B 12 4B | time_when = 1259477844 + 000000F7 | 13 | event_type = 19 + 000000F8 | 64 00 00 00 | server_id = 100 + 000000FC | 29 00 00 00 | event_len = 41 + 00000100 | 1C 01 00 00 | log_pos = 0000011C + 00000104 | 00 00 | flags = <none> + ------------------------ + ... + ************************ + WRITE_ROWS_EVENT [23] + ************************ + 0000011C | 54 1B 12 4B | time_when = 1259477844 + 00000120 | 17 | event_type = 23 + 00000121 | 64 00 00 00 | server_id = 100 + 00000125 | 22 00 00 00 | event_len = 34 + 00000129 | 3E 01 00 00 | log_pos = 0000013E + 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F + ------------------------ + 0000012F | 0F 00 00 00 | table_id = 15 + ... + +mysqlbinlog output +~~~~~~~~~~~~~~~~~~ +Something like this: + + ... + # at 199 + # at 243 + # at 284 + #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values +(1)` + #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped +to number 15 + #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 +flags: STMT_END_F + + BINLOG ' + VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= + VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= + VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== + '/*!*/; + ### INSERT INTO test.t1 + ### SET + ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ + ... + -=-=(Knielsen - Fri, 27 Nov 2009, 13:30)=-=- Observers changed: Knielsen -=-=(Psergey - Sun, 16 Aug 2009, 11:08)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.12485 2009-08-16 11:08:33.000000000 +0300 +++ /tmp/wklog.47.new.12485 2009-08-16 11:08:33.000000000 +0300 @@ -1 +1,6 @@ +First suggestion: + +> I think for this we would actually need a new binlog event type +> (Comment_log_event?). Unless we want to log an empty statement Query_log_event +> containing only a comment (a bit of a hack). -=-=(Psergey - Sun, 16 Aug 2009, 00:02)=-=- Dependency created: 39 now depends on 47 DESCRIPTION: Store in binlog (and show in mysqlbinlog output) texts of statements that caused RBR events This is needed for (list from Monty): - Easier to understand why updates happened - Would make it easier to find out where in application things went wrong (as you can search for exact strings) - Allow one to filter things based on comments in the statement. The cost of this can be that the binlog will be approximately 2x in size (especially insert of big blob's would be a bit painful), so this should be an optional feature. HIGH-LEVEL SPECIFICATION: First suggestion: > I think for this we would actually need a new binlog event type > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). New server option ~~~~~~~~~~~~~~~~~ --binlog-annotate-row-events Setting this option makes RBR (row-) events in the binary log to be preceded by Annotate rows events (see below). The corresponding 'binlog_annotate_row_events' system variable is dynamic and has both global and session values. Default global value is OFF. Note. Session values are usefull to make it possible to annotate only some selected statements: ... SET SESSION binlog_annotate_row_events=ON; ... statements to be annotated ... SET SESSION binlog_annotate_row_events=OFF; ... statements not to be annotated ... New binlog event type ~~~~~~~~~~~~~~~~~~~~~ Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query text in its data part. Example: ... ************************ ANNOTATE_ROWS_EVENT [51] ************************ 000000C7 | 54 1B 12 4B | time_when = 1259477844 000000CB | 33 | event_type = 51 000000CC | 64 00 00 00 | server_id = 100 000000D0 | 2C 00 00 00 | event_len = 44 000000D4 | F3 00 00 00 | log_pos = 000000F3 000000D8 | 00 00 | flags = <none> ------------------------ 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" 000000DE | 72 74 20 69 | 000000E2 | 6E 74 6F 20 | 000000E6 | 74 31 20 76 | 000000EA | 61 6C 75 65 | 000000EE | 73 20 28 31 | 000000F2 | 29 | ************************ TABLE_MAP_EVENT [19] ************************ 000000F3 | 54 1B 12 4B | time_when = 1259477844 000000F7 | 13 | event_type = 19 000000F8 | 64 00 00 00 | server_id = 100 000000FC | 29 00 00 00 | event_len = 41 00000100 | 1C 01 00 00 | log_pos = 0000011C 00000104 | 00 00 | flags = <none> ------------------------ ... ************************ WRITE_ROWS_EVENT [23] ************************ 0000011C | 54 1B 12 4B | time_when = 1259477844 00000120 | 17 | event_type = 23 00000121 | 64 00 00 00 | server_id = 100 00000125 | 22 00 00 00 | event_len = 34 00000129 | 3E 01 00 00 | log_pos = 0000013E 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F ------------------------ 0000012F | 0F 00 00 00 | table_id = 15 ... mysqlbinlog output ~~~~~~~~~~~~~~~~~~ Something like this: ... # at 199 # at 243 # at 284 #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values (1)` #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped to number 15 #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 flags: STMT_END_F BINLOG ' VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== '/*!*/; ### INSERT INTO test.t1 ### SET ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ ... LOW-LEVEL DESIGN: mysql_binlog_send() [sql/sql_repl.cc] ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1. When sending events to a slave, master should simply skip Annotate_rows events (they are not needed for replication). 2. When sending events to mysqlbinlog (remote case), master must send Annotate_rows events as well. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Alexi): Store in binlog text of statements that caused RBR events (47)
by worklog-noreply＠askmonty.org 29 Nov '09

29 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Store in binlog text of statements that caused RBR events CREATION DATE..: Sat, 15 Aug 2009, 23:48 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: Knielsen CATEGORY.......: Server-RawIdeaBin TASK ID........: 47 (http://askmonty.org/worklog/?tid=47) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Alexi - Sun, 29 Nov 2009, 09:50)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.24993 2009-11-29 07:50:36.000000000 +0000 +++ /tmp/wklog.47.new.24993 2009-11-29 07:50:36.000000000 +0000 @@ -4,3 +4,96 @@ > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). +New server option +~~~~~~~~~~~~~~~~~ + --binlog-annotate-row-events + +Setting this option makes RBR (row-) events in the binary log to be +preceded by Annotate rows events (see below). The corresponding +'binlog_annotate_row_events' system variable is dynamic and has both +global and session values. Default global value is OFF. + +Note. Session values are usefull to make it possible to annotate only + some selected statements: + + ... + SET SESSION binlog_annotate_row_events=ON; + ... statements to be annotated ... + SET SESSION binlog_annotate_row_events=OFF; + ... statements not to be annotated ... + +New binlog event type +~~~~~~~~~~~~~~~~~~~~~ + Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] + +Describes the query which caused the corresponding row event. In binary log, +precedes each Table_map_log_event. Contains empty post-header and the query +text in its data part. Example: + + ... + ************************ + ANNOTATE_ROWS_EVENT [51] + ************************ + 000000C7 | 54 1B 12 4B | time_when = 1259477844 + 000000CB | 33 | event_type = 51 + 000000CC | 64 00 00 00 | server_id = 100 + 000000D0 | 2C 00 00 00 | event_len = 44 + 000000D4 | F3 00 00 00 | log_pos = 000000F3 + 000000D8 | 00 00 | flags = <none> + ------------------------ + 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" + 000000DE | 72 74 20 69 | + 000000E2 | 6E 74 6F 20 | + 000000E6 | 74 31 20 76 | + 000000EA | 61 6C 75 65 | + 000000EE | 73 20 28 31 | + 000000F2 | 29 | + ************************ + TABLE_MAP_EVENT [19] + ************************ + 000000F3 | 54 1B 12 4B | time_when = 1259477844 + 000000F7 | 13 | event_type = 19 + 000000F8 | 64 00 00 00 | server_id = 100 + 000000FC | 29 00 00 00 | event_len = 41 + 00000100 | 1C 01 00 00 | log_pos = 0000011C + 00000104 | 00 00 | flags = <none> + ------------------------ + ... + ************************ + WRITE_ROWS_EVENT [23] + ************************ + 0000011C | 54 1B 12 4B | time_when = 1259477844 + 00000120 | 17 | event_type = 23 + 00000121 | 64 00 00 00 | server_id = 100 + 00000125 | 22 00 00 00 | event_len = 34 + 00000129 | 3E 01 00 00 | log_pos = 0000013E + 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F + ------------------------ + 0000012F | 0F 00 00 00 | table_id = 15 + ... + +mysqlbinlog output +~~~~~~~~~~~~~~~~~~ +Something like this: + + ... + # at 199 + # at 243 + # at 284 + #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values +(1)` + #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped +to number 15 + #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 +flags: STMT_END_F + + BINLOG ' + VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= + VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= + VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== + '/*!*/; + ### INSERT INTO test.t1 + ### SET + ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ + ... + -=-=(Knielsen - Fri, 27 Nov 2009, 13:30)=-=- Observers changed: Knielsen -=-=(Psergey - Sun, 16 Aug 2009, 11:08)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.12485 2009-08-16 11:08:33.000000000 +0300 +++ /tmp/wklog.47.new.12485 2009-08-16 11:08:33.000000000 +0300 @@ -1 +1,6 @@ +First suggestion: + +> I think for this we would actually need a new binlog event type +> (Comment_log_event?). Unless we want to log an empty statement Query_log_event +> containing only a comment (a bit of a hack). -=-=(Psergey - Sun, 16 Aug 2009, 00:02)=-=- Dependency created: 39 now depends on 47 DESCRIPTION: Store in binlog (and show in mysqlbinlog output) texts of statements that caused RBR events This is needed for (list from Monty): - Easier to understand why updates happened - Would make it easier to find out where in application things went wrong (as you can search for exact strings) - Allow one to filter things based on comments in the statement. The cost of this can be that the binlog will be approximately 2x in size (especially insert of big blob's would be a bit painful), so this should be an optional feature. HIGH-LEVEL SPECIFICATION: First suggestion: > I think for this we would actually need a new binlog event type > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). New server option ~~~~~~~~~~~~~~~~~ --binlog-annotate-row-events Setting this option makes RBR (row-) events in the binary log to be preceded by Annotate rows events (see below). The corresponding 'binlog_annotate_row_events' system variable is dynamic and has both global and session values. Default global value is OFF. Note. Session values are usefull to make it possible to annotate only some selected statements: ... SET SESSION binlog_annotate_row_events=ON; ... statements to be annotated ... SET SESSION binlog_annotate_row_events=OFF; ... statements not to be annotated ... New binlog event type ~~~~~~~~~~~~~~~~~~~~~ Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query text in its data part. Example: ... ************************ ANNOTATE_ROWS_EVENT [51] ************************ 000000C7 | 54 1B 12 4B | time_when = 1259477844 000000CB | 33 | event_type = 51 000000CC | 64 00 00 00 | server_id = 100 000000D0 | 2C 00 00 00 | event_len = 44 000000D4 | F3 00 00 00 | log_pos = 000000F3 000000D8 | 00 00 | flags = <none> ------------------------ 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" 000000DE | 72 74 20 69 | 000000E2 | 6E 74 6F 20 | 000000E6 | 74 31 20 76 | 000000EA | 61 6C 75 65 | 000000EE | 73 20 28 31 | 000000F2 | 29 | ************************ TABLE_MAP_EVENT [19] ************************ 000000F3 | 54 1B 12 4B | time_when = 1259477844 000000F7 | 13 | event_type = 19 000000F8 | 64 00 00 00 | server_id = 100 000000FC | 29 00 00 00 | event_len = 41 00000100 | 1C 01 00 00 | log_pos = 0000011C 00000104 | 00 00 | flags = <none> ------------------------ ... ************************ WRITE_ROWS_EVENT [23] ************************ 0000011C | 54 1B 12 4B | time_when = 1259477844 00000120 | 17 | event_type = 23 00000121 | 64 00 00 00 | server_id = 100 00000125 | 22 00 00 00 | event_len = 34 00000129 | 3E 01 00 00 | log_pos = 0000013E 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F ------------------------ 0000012F | 0F 00 00 00 | table_id = 15 ... mysqlbinlog output ~~~~~~~~~~~~~~~~~~ Something like this: ... # at 199 # at 243 # at 284 #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values (1)` #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped to number 15 #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 flags: STMT_END_F BINLOG ' VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== '/*!*/; ### INSERT INTO test.t1 ### SET ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ ... ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Alexi): Store in binlog text of statements that caused RBR events (47)
by worklog-noreply＠askmonty.org 29 Nov '09

29 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Store in binlog text of statements that caused RBR events CREATION DATE..: Sat, 15 Aug 2009, 23:48 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: Knielsen CATEGORY.......: Server-RawIdeaBin TASK ID........: 47 (http://askmonty.org/worklog/?tid=47) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Alexi - Sun, 29 Nov 2009, 09:50)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.24993 2009-11-29 07:50:36.000000000 +0000 +++ /tmp/wklog.47.new.24993 2009-11-29 07:50:36.000000000 +0000 @@ -4,3 +4,96 @@ > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). +New server option +~~~~~~~~~~~~~~~~~ + --binlog-annotate-row-events + +Setting this option makes RBR (row-) events in the binary log to be +preceded by Annotate rows events (see below). The corresponding +'binlog_annotate_row_events' system variable is dynamic and has both +global and session values. Default global value is OFF. + +Note. Session values are usefull to make it possible to annotate only + some selected statements: + + ... + SET SESSION binlog_annotate_row_events=ON; + ... statements to be annotated ... + SET SESSION binlog_annotate_row_events=OFF; + ... statements not to be annotated ... + +New binlog event type +~~~~~~~~~~~~~~~~~~~~~ + Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] + +Describes the query which caused the corresponding row event. In binary log, +precedes each Table_map_log_event. Contains empty post-header and the query +text in its data part. Example: + + ... + ************************ + ANNOTATE_ROWS_EVENT [51] + ************************ + 000000C7 | 54 1B 12 4B | time_when = 1259477844 + 000000CB | 33 | event_type = 51 + 000000CC | 64 00 00 00 | server_id = 100 + 000000D0 | 2C 00 00 00 | event_len = 44 + 000000D4 | F3 00 00 00 | log_pos = 000000F3 + 000000D8 | 00 00 | flags = <none> + ------------------------ + 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" + 000000DE | 72 74 20 69 | + 000000E2 | 6E 74 6F 20 | + 000000E6 | 74 31 20 76 | + 000000EA | 61 6C 75 65 | + 000000EE | 73 20 28 31 | + 000000F2 | 29 | + ************************ + TABLE_MAP_EVENT [19] + ************************ + 000000F3 | 54 1B 12 4B | time_when = 1259477844 + 000000F7 | 13 | event_type = 19 + 000000F8 | 64 00 00 00 | server_id = 100 + 000000FC | 29 00 00 00 | event_len = 41 + 00000100 | 1C 01 00 00 | log_pos = 0000011C + 00000104 | 00 00 | flags = <none> + ------------------------ + ... + ************************ + WRITE_ROWS_EVENT [23] + ************************ + 0000011C | 54 1B 12 4B | time_when = 1259477844 + 00000120 | 17 | event_type = 23 + 00000121 | 64 00 00 00 | server_id = 100 + 00000125 | 22 00 00 00 | event_len = 34 + 00000129 | 3E 01 00 00 | log_pos = 0000013E + 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F + ------------------------ + 0000012F | 0F 00 00 00 | table_id = 15 + ... + +mysqlbinlog output +~~~~~~~~~~~~~~~~~~ +Something like this: + + ... + # at 199 + # at 243 + # at 284 + #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values +(1)` + #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped +to number 15 + #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 +flags: STMT_END_F + + BINLOG ' + VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= + VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= + VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== + '/*!*/; + ### INSERT INTO test.t1 + ### SET + ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ + ... + -=-=(Knielsen - Fri, 27 Nov 2009, 13:30)=-=- Observers changed: Knielsen -=-=(Psergey - Sun, 16 Aug 2009, 11:08)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.12485 2009-08-16 11:08:33.000000000 +0300 +++ /tmp/wklog.47.new.12485 2009-08-16 11:08:33.000000000 +0300 @@ -1 +1,6 @@ +First suggestion: + +> I think for this we would actually need a new binlog event type +> (Comment_log_event?). Unless we want to log an empty statement Query_log_event +> containing only a comment (a bit of a hack). -=-=(Psergey - Sun, 16 Aug 2009, 00:02)=-=- Dependency created: 39 now depends on 47 DESCRIPTION: Store in binlog (and show in mysqlbinlog output) texts of statements that caused RBR events This is needed for (list from Monty): - Easier to understand why updates happened - Would make it easier to find out where in application things went wrong (as you can search for exact strings) - Allow one to filter things based on comments in the statement. The cost of this can be that the binlog will be approximately 2x in size (especially insert of big blob's would be a bit painful), so this should be an optional feature. HIGH-LEVEL SPECIFICATION: First suggestion: > I think for this we would actually need a new binlog event type > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). New server option ~~~~~~~~~~~~~~~~~ --binlog-annotate-row-events Setting this option makes RBR (row-) events in the binary log to be preceded by Annotate rows events (see below). The corresponding 'binlog_annotate_row_events' system variable is dynamic and has both global and session values. Default global value is OFF. Note. Session values are usefull to make it possible to annotate only some selected statements: ... SET SESSION binlog_annotate_row_events=ON; ... statements to be annotated ... SET SESSION binlog_annotate_row_events=OFF; ... statements not to be annotated ... New binlog event type ~~~~~~~~~~~~~~~~~~~~~ Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query text in its data part. Example: ... ************************ ANNOTATE_ROWS_EVENT [51] ************************ 000000C7 | 54 1B 12 4B | time_when = 1259477844 000000CB | 33 | event_type = 51 000000CC | 64 00 00 00 | server_id = 100 000000D0 | 2C 00 00 00 | event_len = 44 000000D4 | F3 00 00 00 | log_pos = 000000F3 000000D8 | 00 00 | flags = <none> ------------------------ 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" 000000DE | 72 74 20 69 | 000000E2 | 6E 74 6F 20 | 000000E6 | 74 31 20 76 | 000000EA | 61 6C 75 65 | 000000EE | 73 20 28 31 | 000000F2 | 29 | ************************ TABLE_MAP_EVENT [19] ************************ 000000F3 | 54 1B 12 4B | time_when = 1259477844 000000F7 | 13 | event_type = 19 000000F8 | 64 00 00 00 | server_id = 100 000000FC | 29 00 00 00 | event_len = 41 00000100 | 1C 01 00 00 | log_pos = 0000011C 00000104 | 00 00 | flags = <none> ------------------------ ... ************************ WRITE_ROWS_EVENT [23] ************************ 0000011C | 54 1B 12 4B | time_when = 1259477844 00000120 | 17 | event_type = 23 00000121 | 64 00 00 00 | server_id = 100 00000125 | 22 00 00 00 | event_len = 34 00000129 | 3E 01 00 00 | log_pos = 0000013E 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F ------------------------ 0000012F | 0F 00 00 00 | table_id = 15 ... mysqlbinlog output ~~~~~~~~~~~~~~~~~~ Something like this: ... # at 199 # at 243 # at 284 #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values (1)` #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped to number 15 #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 flags: STMT_END_F BINLOG ' VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== '/*!*/; ### INSERT INTO test.t1 ### SET ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ ... ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Alexi): Store in binlog text of statements that caused RBR events (47)
by worklog-noreply＠askmonty.org 29 Nov '09

29 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Store in binlog text of statements that caused RBR events CREATION DATE..: Sat, 15 Aug 2009, 23:48 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: Knielsen CATEGORY.......: Server-RawIdeaBin TASK ID........: 47 (http://askmonty.org/worklog/?tid=47) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Alexi - Sun, 29 Nov 2009, 09:50)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.24993 2009-11-29 07:50:36.000000000 +0000 +++ /tmp/wklog.47.new.24993 2009-11-29 07:50:36.000000000 +0000 @@ -4,3 +4,96 @@ > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). +New server option +~~~~~~~~~~~~~~~~~ + --binlog-annotate-row-events + +Setting this option makes RBR (row-) events in the binary log to be +preceded by Annotate rows events (see below). The corresponding +'binlog_annotate_row_events' system variable is dynamic and has both +global and session values. Default global value is OFF. + +Note. Session values are usefull to make it possible to annotate only + some selected statements: + + ... + SET SESSION binlog_annotate_row_events=ON; + ... statements to be annotated ... + SET SESSION binlog_annotate_row_events=OFF; + ... statements not to be annotated ... + +New binlog event type +~~~~~~~~~~~~~~~~~~~~~ + Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] + +Describes the query which caused the corresponding row event. In binary log, +precedes each Table_map_log_event. Contains empty post-header and the query +text in its data part. Example: + + ... + ************************ + ANNOTATE_ROWS_EVENT [51] + ************************ + 000000C7 | 54 1B 12 4B | time_when = 1259477844 + 000000CB | 33 | event_type = 51 + 000000CC | 64 00 00 00 | server_id = 100 + 000000D0 | 2C 00 00 00 | event_len = 44 + 000000D4 | F3 00 00 00 | log_pos = 000000F3 + 000000D8 | 00 00 | flags = <none> + ------------------------ + 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" + 000000DE | 72 74 20 69 | + 000000E2 | 6E 74 6F 20 | + 000000E6 | 74 31 20 76 | + 000000EA | 61 6C 75 65 | + 000000EE | 73 20 28 31 | + 000000F2 | 29 | + ************************ + TABLE_MAP_EVENT [19] + ************************ + 000000F3 | 54 1B 12 4B | time_when = 1259477844 + 000000F7 | 13 | event_type = 19 + 000000F8 | 64 00 00 00 | server_id = 100 + 000000FC | 29 00 00 00 | event_len = 41 + 00000100 | 1C 01 00 00 | log_pos = 0000011C + 00000104 | 00 00 | flags = <none> + ------------------------ + ... + ************************ + WRITE_ROWS_EVENT [23] + ************************ + 0000011C | 54 1B 12 4B | time_when = 1259477844 + 00000120 | 17 | event_type = 23 + 00000121 | 64 00 00 00 | server_id = 100 + 00000125 | 22 00 00 00 | event_len = 34 + 00000129 | 3E 01 00 00 | log_pos = 0000013E + 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F + ------------------------ + 0000012F | 0F 00 00 00 | table_id = 15 + ... + +mysqlbinlog output +~~~~~~~~~~~~~~~~~~ +Something like this: + + ... + # at 199 + # at 243 + # at 284 + #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values +(1)` + #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped +to number 15 + #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 +flags: STMT_END_F + + BINLOG ' + VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= + VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= + VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== + '/*!*/; + ### INSERT INTO test.t1 + ### SET + ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ + ... + -=-=(Knielsen - Fri, 27 Nov 2009, 13:30)=-=- Observers changed: Knielsen -=-=(Psergey - Sun, 16 Aug 2009, 11:08)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.12485 2009-08-16 11:08:33.000000000 +0300 +++ /tmp/wklog.47.new.12485 2009-08-16 11:08:33.000000000 +0300 @@ -1 +1,6 @@ +First suggestion: + +> I think for this we would actually need a new binlog event type +> (Comment_log_event?). Unless we want to log an empty statement Query_log_event +> containing only a comment (a bit of a hack). -=-=(Psergey - Sun, 16 Aug 2009, 00:02)=-=- Dependency created: 39 now depends on 47 DESCRIPTION: Store in binlog (and show in mysqlbinlog output) texts of statements that caused RBR events This is needed for (list from Monty): - Easier to understand why updates happened - Would make it easier to find out where in application things went wrong (as you can search for exact strings) - Allow one to filter things based on comments in the statement. The cost of this can be that the binlog will be approximately 2x in size (especially insert of big blob's would be a bit painful), so this should be an optional feature. HIGH-LEVEL SPECIFICATION: First suggestion: > I think for this we would actually need a new binlog event type > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). New server option ~~~~~~~~~~~~~~~~~ --binlog-annotate-row-events Setting this option makes RBR (row-) events in the binary log to be preceded by Annotate rows events (see below). The corresponding 'binlog_annotate_row_events' system variable is dynamic and has both global and session values. Default global value is OFF. Note. Session values are usefull to make it possible to annotate only some selected statements: ... SET SESSION binlog_annotate_row_events=ON; ... statements to be annotated ... SET SESSION binlog_annotate_row_events=OFF; ... statements not to be annotated ... New binlog event type ~~~~~~~~~~~~~~~~~~~~~ Annotate_rows_log_event [ ANNOTATE_ROWS_EVENT ] Describes the query which caused the corresponding row event. In binary log, precedes each Table_map_log_event. Contains empty post-header and the query text in its data part. Example: ... ************************ ANNOTATE_ROWS_EVENT [51] ************************ 000000C7 | 54 1B 12 4B | time_when = 1259477844 000000CB | 33 | event_type = 51 000000CC | 64 00 00 00 | server_id = 100 000000D0 | 2C 00 00 00 | event_len = 44 000000D4 | F3 00 00 00 | log_pos = 000000F3 000000D8 | 00 00 | flags = <none> ------------------------ 000000DA | 69 6E 73 65 | query = "insert into t1 values (1)" 000000DE | 72 74 20 69 | 000000E2 | 6E 74 6F 20 | 000000E6 | 74 31 20 76 | 000000EA | 61 6C 75 65 | 000000EE | 73 20 28 31 | 000000F2 | 29 | ************************ TABLE_MAP_EVENT [19] ************************ 000000F3 | 54 1B 12 4B | time_when = 1259477844 000000F7 | 13 | event_type = 19 000000F8 | 64 00 00 00 | server_id = 100 000000FC | 29 00 00 00 | event_len = 41 00000100 | 1C 01 00 00 | log_pos = 0000011C 00000104 | 00 00 | flags = <none> ------------------------ ... ************************ WRITE_ROWS_EVENT [23] ************************ 0000011C | 54 1B 12 4B | time_when = 1259477844 00000120 | 17 | event_type = 23 00000121 | 64 00 00 00 | server_id = 100 00000125 | 22 00 00 00 | event_len = 34 00000129 | 3E 01 00 00 | log_pos = 0000013E 0000012D | 10 00 | flags = LOG_EVENT_UPDATE_TABLE_MAP_VERSION_F ------------------------ 0000012F | 0F 00 00 00 | table_id = 15 ... mysqlbinlog output ~~~~~~~~~~~~~~~~~~ Something like this: ... # at 199 # at 243 # at 284 #091129 9:57:24 server id 100 end_log_pos 243 Query: `insert into t1 values (1)` #091129 9:57:24 server id 100 end_log_pos 284 Table_map: `test`.`t1` mapped to number 15 #091129 9:57:24 server id 100 end_log_pos 318 Write_rows: table id 15 flags: STMT_END_F BINLOG ' VBsSSzNkAAAALAAAAPMAAAAAAGluc2VydCBpbnRvIHQxIHZhbHVlcyAoMSk= VBsSSxNkAAAAKQAAABwBAAAAAA8AAAAAAAAABHRlc3QAAnQxAAEDAAE= VBsSSxdkAAAAIgAAAD4BAAAQAA8AAAAAAAEAAf/+AQAAAA== '/*!*/; ### INSERT INTO test.t1 ### SET ### @1=1 /* INT meta=0 nullable=1 is_null=0 */ ... ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] bzr commit into Mariadb 5.2, with Maria 2.0:maria/5.2 branch (sanja:2725)
by sanja＠askmonty.org 27 Nov '09

27 Nov '09

#At lp:maria/5.2 2725 sanja(a)askmonty.org 2009-07-07 Group commit and small optimisation of flush log (in case of page without CRC and sector protection) added. (for review) modified: mysql-test/suite/maria/r/maria3.result storage/maria/ha_maria.cc storage/maria/ma_init.c storage/maria/ma_loghandler.c storage/maria/ma_loghandler.h per-file messages: mysql-test/suite/maria/r/maria3.result New variables added storage/maria/ha_maria.cc New variables for group commit and routines to control them added. storage/maria/ma_init.c Service thread stop (if it is present) storage/maria/ma_loghandler.c 2 types of group commit added. If page is not protetcted by CRC or sector protection (i.e. we do not change header when add data to the page) we do not copy beginning of the page during forcing the buffer where the page is last to close. storage/maria/ma_loghandler.h Routines to control group commit added. === modified file 'mysql-test/suite/maria/r/maria3.result' --- a/mysql-test/suite/maria/r/maria3.result 2009-02-19 09:01:25 +0000 +++ b/mysql-test/suite/maria/r/maria3.result 2009-07-07 00:37:23 +0000 @@ -264,6 +264,8 @@ Variable_name Value maria_block_size 8192 maria_checkpoint_interval 30 maria_force_start_after_recovery_failures 0 +maria_group_commit none +maria_group_commit_rate 800 maria_log_file_size 4294959104 maria_log_purge_type immediate maria_max_sort_file_size 9223372036853727232 @@ -285,6 +287,7 @@ Maria_pagecache_read_requests # Maria_pagecache_reads # Maria_pagecache_write_requests # Maria_pagecache_writes # +Maria_transaction_log_syncs # create table t1 (b char(0)); insert into t1 values(NULL),(""); select length(b) from t1; === modified file 'storage/maria/ha_maria.cc' --- a/storage/maria/ha_maria.cc 2009-05-19 09:28:05 +0000 +++ b/storage/maria/ha_maria.cc 2009-07-07 00:37:23 +0000 @@ -101,22 +101,40 @@ TYPELIB maria_translog_purge_type_typeli array_elements(maria_translog_purge_type_names) - 1, "", maria_translog_purge_type_names, NULL }; + +/* transactional log directory sync */ const char *maria_sync_log_dir_names[]= { "NEVER", "NEWFILE", "ALWAYS", NullS }; - TYPELIB maria_sync_log_dir_typelib= { array_elements(maria_sync_log_dir_names) - 1, "", maria_sync_log_dir_names, NULL }; +/* transactional log group commit */ +const char *maria_group_commit_names[]= +{ + "none", "hard", "soft", NullS +}; +TYPELIB maria_group_commit_typelib= +{ + array_elements(maria_group_commit_names) - 1, "", + maria_group_commit_names, NULL +}; + /** Interval between background checkpoints in seconds */ static ulong checkpoint_interval; static void update_checkpoint_interval(MYSQL_THD thd, struct st_mysql_sys_var *var, void *var_ptr, const void *save); +static void update_maria_group_commit(MYSQL_THD thd, + struct st_mysql_sys_var *var, + void *var_ptr, const void *save); +static void update_maria_group_commit_rate(MYSQL_THD thd, + struct st_mysql_sys_var *var, + void *var_ptr, const void *save); /** After that many consecutive recovery failures, remove logs */ static ulong force_start_after_recovery_failures; static void update_log_file_size(MYSQL_THD thd, @@ -163,6 +181,24 @@ static MYSQL_SYSVAR_ULONG(log_file_size, NULL, update_log_file_size, TRANSLOG_FILE_SIZE, TRANSLOG_MIN_FILE_SIZE, 0xffffffffL, TRANSLOG_PAGE_SIZE); +static MYSQL_SYSVAR_ENUM(group_commit, maria_group_commit, + PLUGIN_VAR_RQCMDARG, + "Specifies maria group commit mode. " + "Possible values are \"none\" (no group commit), " + "\"hard\" (with waiting to actual commit), " + "\"soft\" (no wait for commit (DANGEROUS!!!))", + NULL, update_maria_group_commit, + TRANSLOG_GCOMMIT_NONE, &maria_group_commit_typelib); + +static MYSQL_SYSVAR_ULONG(group_commit_rate, maria_group_commit_rate, + PLUGIN_VAR_RQCMDARG, + "Number of commits per 100 seconds. (in other words one commit for" + "every 100/maria_group_commit_rate second). 0 stands for no waiting" + "for other threads to come and do a commit in \"hard\" mode and no" + " sync()/commit at all in \"soft\" mode. Option has only an effect" + "if maria_group_commit is used", + NULL, update_maria_group_commit_rate, 800, 0, UINT_MAX, 1); + static MYSQL_SYSVAR_ENUM(log_purge_type, log_purge_type, PLUGIN_VAR_RQCMDARG, "Specifies how maria transactional log will be purged. " @@ -3254,6 +3290,8 @@ static struct st_mysql_sys_var* system_v MYSQL_SYSVAR(block_size), MYSQL_SYSVAR(checkpoint_interval), MYSQL_SYSVAR(force_start_after_recovery_failures), + MYSQL_SYSVAR(group_commit), + MYSQL_SYSVAR(group_commit_rate), MYSQL_SYSVAR(page_checksum), MYSQL_SYSVAR(log_dir_path), MYSQL_SYSVAR(log_file_size), @@ -3284,6 +3322,97 @@ static void update_checkpoint_interval(M } /** + @brief Updates group commit mode +*/ + +static void update_maria_group_commit(MYSQL_THD thd, + struct st_mysql_sys_var *var, + void *var_ptr, const void *save) +{ + ulong value= (ulong)*((long *)var_ptr); + DBUG_ENTER("update_maria_group_commit"); + DBUG_PRINT("enter", ("old value: %lu new value %lu rate %lu", + value, (ulong)(*(long *)save), + maria_group_commit_rate)); + /* old value */ + switch (value) { + case TRANSLOG_GCOMMIT_NONE: + break; + case TRANSLOG_GCOMMIT_HARD: + translog_hard_group_commit(FALSE); + break; + case TRANSLOG_GCOMMIT_SOFT: + translog_soft_sync(FALSE); + if (maria_group_commit_rate) + translog_soft_sync_end(); + break; + default: + DBUG_ASSERT(0); /* impossible */ + } + value= *(ulong *)var_ptr= (ulong)(*(long *)save); + translog_sync(); + /* new value */ + switch (value) { + case TRANSLOG_GCOMMIT_NONE: + break; + case TRANSLOG_GCOMMIT_HARD: + translog_hard_group_commit(TRUE); + break; + case TRANSLOG_GCOMMIT_SOFT: + translog_soft_sync(TRUE); + /* variable change made under global lock so we can just read it */ + if (maria_group_commit_rate) + translog_soft_sync_start(); + break; + default: + DBUG_ASSERT(0); /* impossible */ + } + DBUG_VOID_RETURN; +} + +/** + @brief Updates group commit rate +*/ + +static void update_maria_group_commit_rate(MYSQL_THD thd, + struct st_mysql_sys_var *var, + void *var_ptr, const void *save) +{ + ulong new_value= (ulong)*((long *)save); + ulong *value_ptr= (ulong*) var_ptr; + DBUG_ENTER("update_maria_group_commit_rate"); + DBUG_PRINT("enter", ("old value: %lu new value %lu group commit %lu", + *value_ptr, new_value, maria_group_commit)); + if (new_value && + ((TRANSLOG_RATE_BASE * 1000000000ULL / new_value + + TRANSLOG_RATE_BASE / 2) / + TRANSLOG_RATE_BASE) == 0) + new_value= 0; /* protection against too small value */ + + /* variable change made under global lock so we can just read it */ + switch (maria_group_commit) { + case TRANSLOG_GCOMMIT_NONE: + *value_ptr= new_value; + translog_set_group_commit_rate(new_value); + break; + case TRANSLOG_GCOMMIT_HARD: + *value_ptr= new_value; + translog_set_group_commit_rate(new_value); + break; + case TRANSLOG_GCOMMIT_SOFT: + if (*value_ptr) + translog_soft_sync_end(); + translog_set_group_commit_rate(new_value); + if ((*value_ptr= new_value)) + translog_soft_sync_start(); + break; + default: + DBUG_ASSERT(0); /* impossible */ + } + DBUG_VOID_RETURN; +} + +/** @brief Updates the transaction log file limit. */ @@ -3305,6 +3434,7 @@ static SHOW_VAR status_variables[]= { {"Maria_pagecache_reads", (char*) &maria_pagecache_var.global_cache_read, SHOW_LONGLONG}, {"Maria_pagecache_write_requests", (char*) &maria_pagecache_var.global_cache_w_requests, SHOW_LONGLONG}, {"Maria_pagecache_writes", (char*) &maria_pagecache_var.global_cache_write, SHOW_LONGLONG}, + {"Maria_transaction_log_syncs", (char*) &translog_syncs, SHOW_LONGLONG}, {NullS, NullS, SHOW_LONG} }; === modified file 'storage/maria/ma_init.c' --- a/storage/maria/ma_init.c 2008-10-09 20:03:54 +0000 +++ b/storage/maria/ma_init.c 2009-07-07 00:37:23 +0000 @@ -82,6 +82,11 @@ void maria_end(void) maria_inited= maria_multi_threaded= FALSE; ft_free_stopwords(); ma_checkpoint_end(); + if (translog_status == TRANSLOG_OK) + { + translog_soft_sync_end(); + translog_sync(); + } if ((trid= trnman_get_max_trid()) > max_trid_in_control_file) { /* === modified file 'storage/maria/ma_loghandler.c' --- a/storage/maria/ma_loghandler.c 2009-05-19 09:28:05 +0000 +++ b/storage/maria/ma_loghandler.c 2009-07-07 00:37:23 +0000 @@ -18,6 +18,7 @@ #include "ma_blockrec.h" /* for some constants and in-write hooks */ #include "ma_key_recover.h" /* For some in-write hooks */ #include "ma_checkpoint.h" +#include "ma_servicethread.h" /* On Windows, neither my_open() nor my_sync() work for directories. @@ -47,6 +48,15 @@ #include <m_ctype.h> #endif +/** @brief protects checkpoint_in_progress */ +static pthread_mutex_t LOCK_soft_sync; +/** @brief for killing the background checkpoint thread */ +static pthread_cond_t COND_soft_sync; +/** @brief control structure for checkpoint background thread */ +static MA_SERVICE_THREAD_CONTROL soft_sync_control= + {THREAD_DEAD, FALSE, &LOCK_soft_sync, &COND_soft_sync}; + + /* transaction log file descriptor */ typedef struct st_translog_file { @@ -124,10 +134,20 @@ struct st_translog_buffer /* Previous buffer offset to detect it flush finish */ TRANSLOG_ADDRESS prev_buffer_offset; /* + If the buffer was forced to close it save value of its horizon + otherwise LSN_IMPOSSIBLE + */ + TRANSLOG_ADDRESS pre_force_close_horizon; + /* How much is written (or will be written when copy_to_buffer_in_progress become 0) to this buffer */ translog_size_t size; + /* + How much data was skipped during moving page from previous buffer + to this one (it is optimisation of forcing buffer to finish + */ + uint skipped_data; /* File handler for this buffer */ TRANSLOG_FILE *file; /* Threads which are waiting for buffer filling/freeing */ @@ -304,6 +324,7 @@ struct st_translog_descriptor */ pthread_mutex_t log_flush_lock; pthread_cond_t log_flush_cond; + pthread_cond_t new_goal_cond; /* Protects changing of headers of finished files (max_lsn) */ pthread_mutex_t file_header_lock; @@ -344,13 +365,39 @@ static struct st_translog_descriptor log ulong log_purge_type= TRANSLOG_PURGE_IMMIDIATE; ulong log_file_size= TRANSLOG_FILE_SIZE; +/* sync() of log files directory mode */ ulong sync_log_dir= TRANSLOG_SYNC_DIR_NEWFILE; +ulong maria_group_commit= TRANSLOG_GCOMMIT_NONE; +ulong maria_group_commit_rate= 0; /* Marker for end of log */ static uchar end_of_log= 0; #define END_OF_LOG &end_of_log +/** + Switch for "soft" sync (no real sync() but periodical sync by service + thread) +*/ +static volatile my_bool soft_sync= FALSE; +/** + Switch for "hard" group commit mode +*/ +static volatile my_bool hard_group_commit= FALSE; +/** + File numbers interval which have to be sync() +*/ +static uint32 soft_sync_min= 0; +static uint32 soft_sync_max= 0; +/** + stores interval in nanoseconds/TRANSLOG_RATE_BASE (to + fit into uint32) +*/ +static uint32 group_commit_wait= 0; enum enum_translog_status translog_status= TRANSLOG_UNINITED; +ulonglong translog_syncs= 0; /* Number of sync()s */ + +/* time of last flush */ +static ulonglong flush_start= 0; /* chunk types */ #define TRANSLOG_CHUNK_LSN 0x00 /* 0 chunk refer as LSN (head or tail */ @@ -980,12 +1027,17 @@ static TRANSLOG_FILE *get_logfile_by_num static TRANSLOG_FILE *get_current_logfile() { TRANSLOG_FILE *file; + DBUG_ENTER("get_current_logfile"); rw_rdlock(&log_descriptor.open_files_lock); + DBUG_PRINT("info", ("max_file: %lu min_file: %lu open_files: %lu", + (ulong) log_descriptor.max_file, + (ulong) log_descriptor.min_file, + (ulong) log_descriptor.open_files.elements)); DBUG_ASSERT(log_descriptor.max_file - log_descriptor.min_file + 1 == log_descriptor.open_files.elements); file= *dynamic_element(&log_descriptor.open_files, 0, TRANSLOG_FILE **); rw_unlock(&log_descriptor.open_files_lock); - return (file); + DBUG_RETURN(file); } uchar NEAR maria_trans_file_magic[]= @@ -1069,6 +1121,7 @@ static my_bool translog_write_file_heade static my_bool translog_max_lsn_to_header(File file, LSN lsn) { uchar lsn_buff[LSN_STORE_SIZE]; + my_bool rc; DBUG_ENTER("translog_max_lsn_to_header"); DBUG_PRINT("enter", ("File descriptor: %ld " "lsn: (%lu,0x%lx)", @@ -1077,11 +1130,13 @@ static my_bool translog_max_lsn_to_heade lsn_store(lsn_buff, lsn); - DBUG_RETURN(my_pwrite(file, lsn_buff, - LSN_STORE_SIZE, - (LOG_HEADER_DATA_SIZE - LSN_STORE_SIZE), - log_write_flags) != 0 || - my_sync(file, MYF(MY_WME)) != 0); + if (!(rc= (my_pwrite(file, lsn_buff, + LSN_STORE_SIZE, + (LOG_HEADER_DATA_SIZE - LSN_STORE_SIZE), + log_write_flags) != 0 || + my_sync(file, MYF(MY_WME)) != 0))) + translog_syncs++; + DBUG_RETURN(rc); } @@ -1423,7 +1478,9 @@ LSN translog_get_file_max_lsn_stored(uin static my_bool translog_buffer_init(struct st_translog_buffer *buffer, int num) { DBUG_ENTER("translog_buffer_init"); - buffer->prev_last_lsn= buffer->last_lsn= LSN_IMPOSSIBLE; + buffer->pre_force_close_horizon= + buffer->prev_last_lsn= buffer->last_lsn= + LSN_IMPOSSIBLE; DBUG_PRINT("info", ("last_lsn and prev_last_lsn set to 0 buffer: 0x%lx", (ulong) buffer)); @@ -1435,6 +1492,7 @@ static my_bool translog_buffer_init(stru memset(buffer->buffer, TRANSLOG_FILLER, TRANSLOG_WRITE_BUFFER); /* Buffer size */ buffer->size= 0; + buffer->skipped_data= 0; /* cond of thread which is waiting for buffer filling */ if (pthread_cond_init(&buffer->waiting_filling_buffer, 0)) DBUG_RETURN(1); @@ -1489,7 +1547,10 @@ static my_bool translog_close_log_file(T TODO: sync only we have changed the log */ if (!file->is_sync) + { rc= my_sync(file->handler.file, MYF(MY_WME)); + translog_syncs++; + } rc|= my_close(file->handler.file, MYF(MY_WME)); my_free(file, MYF(0)); return test(rc); @@ -2044,7 +2105,8 @@ static void translog_start_buffer(struct (ulong) LSN_OFFSET(log_descriptor.horizon), (ulong) LSN_OFFSET(log_descriptor.horizon))); DBUG_ASSERT(buffer_no == buffer->buffer_no); - buffer->prev_last_lsn= buffer->last_lsn= LSN_IMPOSSIBLE; + buffer->pre_force_close_horizon= + buffer->prev_last_lsn= buffer->last_lsn= LSN_IMPOSSIBLE; DBUG_PRINT("info", ("last_lsn and prev_last_lsn set to 0 buffer: 0x%lx", (ulong) buffer)); buffer->offset= log_descriptor.horizon; @@ -2052,6 +2114,7 @@ static void translog_start_buffer(struct buffer->file= get_current_logfile(); buffer->overlay= 0; buffer->size= 0; + buffer->skipped_data= 0; translog_cursor_init(cursor, buffer, buffer_no); DBUG_PRINT("info", ("file: #%ld (%d) init cursor #%u: 0x%lx " "chaser: %d Size: %lu (%lu)", @@ -2523,6 +2586,7 @@ static my_bool translog_buffer_flush(str TRANSLOG_ADDRESS offset= buffer->offset; TRANSLOG_FILE *file= buffer->file; uint8 ver= buffer->ver; + uint skipped_data; DBUG_ENTER("translog_buffer_flush"); DBUG_PRINT("enter", ("Buffer: #%u 0x%lx file: %d offset: (%lu,0x%lx) size: %lu", @@ -2557,6 +2621,8 @@ static my_bool translog_buffer_flush(str disk */ file= buffer->file; + skipped_data= buffer->skipped_data; + DBUG_ASSERT(skipped_data < TRANSLOG_PAGE_SIZE); for (i= 0, pg= LSN_OFFSET(buffer->offset) / TRANSLOG_PAGE_SIZE; i < buffer->size; i+= TRANSLOG_PAGE_SIZE, pg++) @@ -2573,13 +2639,16 @@ static my_bool translog_buffer_flush(str DBUG_ASSERT(i + TRANSLOG_PAGE_SIZE <= buffer->size); if (translog_status != TRANSLOG_OK && translog_status != TRANSLOG_SHUTDOWN) DBUG_RETURN(1); - if (pagecache_inject(log_descriptor.pagecache, + if (pagecache_write_part(log_descriptor.pagecache, &file->handler, pg, 3, buffer->buffer + i, PAGECACHE_PLAIN_PAGE, PAGECACHE_LOCK_LEFT_UNLOCKED, - PAGECACHE_PIN_LEFT_UNPINNED, 0, - LSN_IMPOSSIBLE)) + PAGECACHE_PIN_LEFT_UNPINNED, + PAGECACHE_WRITE_DONE, 0, + LSN_IMPOSSIBLE, + skipped_data, + TRANSLOG_PAGE_SIZE - skipped_data)) { DBUG_PRINT("error", ("Can't write page (%lu,0x%lx) to pagecache, error: %d", @@ -2589,10 +2658,12 @@ static my_bool translog_buffer_flush(str translog_stop_writing(); DBUG_RETURN(1); } + skipped_data= 0; } file->is_sync= 0; - if (my_pwrite(file->handler.file, buffer->buffer, - buffer->size, LSN_OFFSET(buffer->offset), + if (my_pwrite(file->handler.file, buffer->buffer + buffer->skipped_data, + buffer->size - buffer->skipped_data, + LSN_OFFSET(buffer->offset) + buffer->skipped_data, log_write_flags)) { DBUG_PRINT("error", ("Can't write buffer (%lu,0x%lx) size %lu " @@ -2985,6 +3056,7 @@ restart: uchar *from, *table= NULL; int is_last_unfinished_page; uint last_protected_sector= 0; + uint skipped_data= curr_buffer->skipped_data; TRANSLOG_FILE file_copy; uint8 ver= curr_buffer->ver; translog_wait_for_writers(curr_buffer); @@ -2997,7 +3069,25 @@ restart: } DBUG_ASSERT(LSN_FILE_NO(addr) == LSN_FILE_NO(curr_buffer->offset)); from= curr_buffer->buffer + (addr - curr_buffer->offset); - memcpy(buffer, from, TRANSLOG_PAGE_SIZE); + if (skipped_data > (addr - curr_buffer->offset)) + { + /* + We read page part of which is not present in buffer, + so we should read absent part from file (page cache actually) + */ + file= get_logfile_by_number(file_no); + DBUG_ASSERT(file != NULL); + buffer= pagecache_read(log_descriptor.pagecache, &file->handler, + LSN_OFFSET(addr) / TRANSLOG_PAGE_SIZE, + 3, buffer, + PAGECACHE_PLAIN_PAGE, + PAGECACHE_LOCK_LEFT_UNLOCKED, + NULL); + } + else + skipped_data= 0; /* Read after skipped in buffer data */ + memcpy(buffer + skipped_data, from + skipped_data, + TRANSLOG_PAGE_SIZE - skipped_data); /* We can use copy then in translog_page_validator() because it do not put it permanently somewhere. @@ -3291,6 +3381,7 @@ static my_bool translog_truncate_log(TRA uint32 next_page_offset, page_rest; uint32 i; File fd; + int rc; TRANSLOG_VALIDATOR_DATA data; char path[FN_REFLEN]; uchar page_buff[TRANSLOG_PAGE_SIZE]; @@ -3316,14 +3407,19 @@ static my_bool translog_truncate_log(TRA TRANSLOG_PAGE_SIZE); page_rest= next_page_offset - LSN_OFFSET(addr); memset(page_buff, TRANSLOG_FILLER, page_rest); - if ((fd= open_logfile_by_number_no_cache(LSN_FILE_NO(addr))) < 0 || - ((my_chsize(fd, next_page_offset, TRANSLOG_FILLER, MYF(MY_WME)) || - (page_rest && my_pwrite(fd, page_buff, page_rest, LSN_OFFSET(addr), - log_write_flags)) || - my_sync(fd, MYF(MY_WME))) | - my_close(fd, MYF(MY_WME))) || - (sync_log_dir >= TRANSLOG_SYNC_DIR_ALWAYS && - sync_dir(log_descriptor.directory_fd, MYF(MY_WME | MY_IGNORE_BADFD)))) + rc= ((fd= open_logfile_by_number_no_cache(LSN_FILE_NO(addr))) < 0 || + ((my_chsize(fd, next_page_offset, TRANSLOG_FILLER, MYF(MY_WME)) || + (page_rest && my_pwrite(fd, page_buff, page_rest, LSN_OFFSET(addr), + log_write_flags)) || + my_sync(fd, MYF(MY_WME))))); + translog_syncs++; + rc|= (fd > 0 && my_close(fd, MYF(MY_WME))); + if (sync_log_dir >= TRANSLOG_SYNC_DIR_ALWAYS) + { + rc|= sync_dir(log_descriptor.directory_fd, MYF(MY_WME | MY_IGNORE_BADFD)); + translog_syncs++; + } + if (rc) DBUG_RETURN(1); /* fix the horizon */ @@ -3511,6 +3607,7 @@ my_bool translog_init_with_table(const c pthread_mutex_init(&log_descriptor.dirty_buffer_mask_lock, MY_MUTEX_INIT_FAST) || pthread_cond_init(&log_descriptor.log_flush_cond, 0) || + pthread_cond_init(&log_descriptor.new_goal_cond, 0) || my_rwlock_init(&log_descriptor.open_files_lock, NULL) || my_init_dynamic_array(&log_descriptor.open_files, @@ -3912,7 +4009,6 @@ my_bool translog_init_with_table(const c log_descriptor.flushed= log_descriptor.horizon; log_descriptor.in_buffers_only= log_descriptor.bc.buffer->offset; log_descriptor.max_lsn= LSN_IMPOSSIBLE; /* set to 0 */ - log_descriptor.previous_flush_horizon= log_descriptor.horizon; /* Now 'flushed' is set to 'horizon' value, but 'horizon' is (potentially) address of the next LSN and we want indicate that all LSNs that are @@ -3995,6 +4091,10 @@ my_bool translog_init_with_table(const c It is beginning of the log => there is no LSNs in the log => There is no harm in leaving it "as-is". */ + log_descriptor.previous_flush_horizon= log_descriptor.horizon; + DBUG_PRINT("info", ("previous_flush_horizon: (%lu,0x%lx)", + LSN_IN_PARTS(log_descriptor. + previous_flush_horizon))); DBUG_RETURN(0); } file_no--; @@ -4070,6 +4170,9 @@ my_bool translog_init_with_table(const c translog_free_record_header(&rec); } } + log_descriptor.previous_flush_horizon= log_descriptor.horizon; + DBUG_PRINT("info", ("previous_flush_horizon: (%lu,0x%lx)", + LSN_IN_PARTS(log_descriptor.previous_flush_horizon))); DBUG_RETURN(0); err: ma_message_no_user(0, "log initialization failed"); @@ -4157,6 +4260,7 @@ void translog_destroy() pthread_mutex_destroy(&log_descriptor.log_flush_lock); pthread_mutex_destroy(&log_descriptor.dirty_buffer_mask_lock); pthread_cond_destroy(&log_descriptor.log_flush_cond); + pthread_cond_destroy(&log_descriptor.new_goal_cond); rwlock_destroy(&log_descriptor.open_files_lock); delete_dynamic(&log_descriptor.open_files); delete_dynamic(&log_descriptor.unfinished_files); @@ -6885,11 +6989,11 @@ int translog_read_record_header_from_buf { translog_size_t res; DBUG_ENTER("translog_read_record_header_from_buffer"); + DBUG_PRINT("info", ("page byte: 0x%x offset: %u", + (uint) page[page_offset], (uint) page_offset)); DBUG_ASSERT(translog_is_LSN_chunk(page[page_offset])); DBUG_ASSERT(translog_status == TRANSLOG_OK || translog_status == TRANSLOG_READONLY); - DBUG_PRINT("info", ("page byte: 0x%x offset: %u", - (uint) page[page_offset], (uint) page_offset)); buff->type= (page[page_offset] & TRANSLOG_REC_TYPE); buff->short_trid= uint2korr(page + page_offset + 1); DBUG_PRINT("info", ("Type %u, Short TrID %u, LSN (%lu,0x%lx)", @@ -7356,27 +7460,27 @@ static void translog_force_current_buffe "Buffer addr: (%lu,0x%lx) " "Page addr: (%lu,0x%lx) " "size: %lu (%lu) Pg: %u left: %u in progress %u", - (uint) log_descriptor.bc.buffer_no, - (ulong) log_descriptor.bc.buffer, - LSN_IN_PARTS(log_descriptor.bc.buffer->offset), + (uint) old_buffer_no, + (ulong) old_buffer, + LSN_IN_PARTS(old_buffer->offset), (ulong) LSN_FILE_NO(log_descriptor.horizon), (ulong) (LSN_OFFSET(log_descriptor.horizon) - log_descriptor.bc.current_page_fill), - (ulong) log_descriptor.bc.buffer->size, + (ulong) old_buffer->size, (ulong) (log_descriptor.bc.ptr -log_descriptor.bc. buffer->buffer), (uint) log_descriptor.bc.current_page_fill, (uint) left, - (uint) log_descriptor.bc.buffer-> + (uint) old_buffer-> copy_to_buffer_in_progress)); translog_lock_assert_owner(); LINT_INIT(current_page_fill); - new_buff_beginning= log_descriptor.bc.buffer->offset; - new_buff_beginning+= log_descriptor.bc.buffer->size; /* increase offset */ + new_buff_beginning= old_buffer->offset; + new_buff_beginning+= old_buffer->size; /* increase offset */ DBUG_ASSERT(log_descriptor.bc.ptr !=NULL); DBUG_ASSERT(LSN_FILE_NO(log_descriptor.horizon) == - LSN_FILE_NO(log_descriptor.bc.buffer->offset)); + LSN_FILE_NO(old_buffer->offset)); translog_check_cursor(&log_descriptor.bc); DBUG_ASSERT(left < TRANSLOG_PAGE_SIZE); if (left) @@ -7387,18 +7491,20 @@ static void translog_force_current_buffe */ DBUG_PRINT("info", ("left: %u", (uint) left)); + old_buffer->pre_force_close_horizon= + old_buffer->offset + old_buffer->size; /* decrease offset */ new_buff_beginning-= log_descriptor.bc.current_page_fill; current_page_fill= log_descriptor.bc.current_page_fill; memset(log_descriptor.bc.ptr, TRANSLOG_FILLER, left); - log_descriptor.bc.buffer->size+= left; + old_buffer->size+= left; DBUG_PRINT("info", ("Finish Page buffer #%u: 0x%lx " "Size: %lu", - (uint) log_descriptor.bc.buffer->buffer_no, - (ulong) log_descriptor.bc.buffer, - (ulong) log_descriptor.bc.buffer->size)); - DBUG_ASSERT(log_descriptor.bc.buffer->buffer_no == + (uint) old_buffer->buffer_no, + (ulong) old_buffer, + (ulong) old_buffer->size)); + DBUG_ASSERT(old_buffer->buffer_no == log_descriptor.bc.buffer_no); } else @@ -7509,11 +7615,21 @@ static void translog_force_current_buffe if (left) { - /* - TODO: do not copy beginning of the page if we have no CRC or sector - checks on - */ - memcpy(new_buffer->buffer, data, current_page_fill); + if (log_descriptor.flags & + (TRANSLOG_PAGE_CRC | TRANSLOG_SECTOR_PROTECTION)) + memcpy(new_buffer->buffer, data, current_page_fill); + else + { + /* + This page header does not change if we add more data to the page so + we can not copy it and will not overwrite later + */ + new_buffer->skipped_data= current_page_fill; +#ifndef DBUG_OFF + memset(new_buffer->buffer, 0xa5, current_page_fill); +#endif + DBUG_ASSERT(new_buffer->skipped_data < TRANSLOG_PAGE_SIZE); + } } old_buffer->next_buffer_offset= new_buffer->offset; translog_buffer_lock(new_buffer); @@ -7561,6 +7677,7 @@ void translog_flush_set_new_goal_and_wai { log_descriptor.next_pass_max_lsn= lsn; log_descriptor.max_lsn_requester= pthread_self(); + pthread_cond_broadcast(&log_descriptor.new_goal_cond); } while (flush_no == log_descriptor.flush_no) { @@ -7572,67 +7689,79 @@ void translog_flush_set_new_goal_and_wai /** - @brief Flush the log up to given LSN (included) + @brief sync() range of files (inclusive) and directory (by request) - @param lsn log record serial number up to which (inclusive) - the log has to be flushed + @param min min internal file number to flush + @param max max internal file number to flush + @param sync_dir need sync directory - @return Operation status + return Operation status @retval 0 OK @retval 1 Error - */ -my_bool translog_flush(TRANSLOG_ADDRESS lsn) +static my_bool translog_sync_files(uint32 min, uint32 max, + my_bool sync_dir) { - LSN sent_to_disk= LSN_IMPOSSIBLE; - TRANSLOG_ADDRESS flush_horizon; - uint fn, i; - dirty_buffer_mask_t dirty_buffer_mask; - uint8 last_buffer_no, start_buffer_no; + uint fn; my_bool rc= 0; - DBUG_ENTER("translog_flush"); - DBUG_PRINT("enter", ("Flush up to LSN: (%lu,0x%lx)", LSN_IN_PARTS(lsn))); - DBUG_ASSERT(translog_status == TRANSLOG_OK || - translog_status == TRANSLOG_READONLY); - LINT_INIT(sent_to_disk); - - pthread_mutex_lock(&log_descriptor.log_flush_lock); - DBUG_PRINT("info", ("Everything is flushed up to (%lu,0x%lx)", - LSN_IN_PARTS(log_descriptor.flushed))); - if (cmp_translog_addr(log_descriptor.flushed, lsn) >= 0) + ulonglong flush_interval; + DBUG_ENTER("translog_sync_files"); + DBUG_PRINT("info", ("min: %lu max: %lu sync dir: %d", + (ulong) min, (ulong) max, (int) sync_dir)); + DBUG_ASSERT(min <= max); + + flush_interval= group_commit_wait; + if (flush_interval) + flush_start= my_micro_time(); + for (fn= min; fn <= max; fn++) { - pthread_mutex_unlock(&log_descriptor.log_flush_lock); - DBUG_RETURN(0); - } - if (log_descriptor.flush_in_progress) - { - translog_flush_set_new_goal_and_wait(lsn); - if (!pthread_equal(log_descriptor.max_lsn_requester, pthread_self())) + TRANSLOG_FILE *file= get_logfile_by_number(fn); + DBUG_ASSERT(file != NULL); + if (!file->is_sync) { - /* fix lsn if it was horizon */ - if (cmp_translog_addr(lsn, log_descriptor.bc.buffer->last_lsn) > 0) - lsn= BUFFER_MAX_LSN(log_descriptor.bc.buffer); - translog_flush_wait_for_end(lsn); - pthread_mutex_unlock(&log_descriptor.log_flush_lock); - DBUG_RETURN(0); + if (my_sync(file->handler.file, MYF(MY_WME))) + { + rc= 1; + translog_stop_writing(); + DBUG_RETURN(rc); + } + translog_syncs++; + file->is_sync= 1; } - log_descriptor.next_pass_max_lsn= LSN_IMPOSSIBLE; } - log_descriptor.flush_in_progress= 1; - flush_horizon= log_descriptor.previous_flush_horizon; - DBUG_PRINT("info", ("flush_in_progress is set")); - pthread_mutex_unlock(&log_descriptor.log_flush_lock); - translog_lock(); - if (log_descriptor.is_everything_flushed) + if (sync_dir) { - DBUG_PRINT("info", ("everything is flushed")); - rc= (translog_status == TRANSLOG_READONLY); - translog_unlock(); - goto out; + if (!(rc= sync_dir(log_descriptor.directory_fd, + MYF(MY_WME | MY_IGNORE_BADFD)))) + translog_syncs++; } + DBUG_RETURN(rc); +} + + +/* + @brief Flushes buffers with LSNs in them less or equal address <lsn> + + @param lsn address up to which all LSNs should be flushed, + can be reset to real last LSN address + @parem sent_to_disk returns 'sent to disk' position + @param flush_horizon returns horizon of the flush + + @note About terminology see comment to translog_flush(). +*/ + +void translog_flush_buffers(TRANSLOG_ADDRESS *lsn, + TRANSLOG_ADDRESS *sent_to_disk, + TRANSLOG_ADDRESS *flush_horizon) +{ + dirty_buffer_mask_t dirty_buffer_mask; + uint i; + uint8 last_buffer_no, start_buffer_no; + DBUG_ENTER("translog_flush_buffers"); + /* We will recheck information when will lock buffers one by one so we can use unprotected read here (this is just for @@ -7656,15 +7785,15 @@ my_bool translog_flush(TRANSLOG_ADDRESS /* if LSN up to which we have to flush bigger then maximum LSN of previous buffer and at least one LSN was saved in the current buffer (last_lsn != - LSN_IMPOSSIBLE) then we better finish the current buffer. + LSN_IMPOSSIBLE) then we have to close the current buffer. */ - if (cmp_translog_addr(lsn, log_descriptor.bc.buffer->prev_last_lsn) > 0 && + if (cmp_translog_addr(*lsn, log_descriptor.bc.buffer->prev_last_lsn) > 0 && log_descriptor.bc.buffer->last_lsn != LSN_IMPOSSIBLE) { struct st_translog_buffer *buffer= log_descriptor.bc.buffer; - lsn= log_descriptor.bc.buffer->last_lsn; /* fix lsn if it was horizon */ + *lsn= log_descriptor.bc.buffer->last_lsn; /* fix lsn if it was horizon */ DBUG_PRINT("info", ("LSN to flush fixed to last lsn: (%lu,0x%lx)", - LSN_IN_PARTS(log_descriptor.bc.buffer->last_lsn))); + LSN_IN_PARTS(log_descriptor.bc.buffer->last_lsn))); last_buffer_no= log_descriptor.bc.buffer_no; log_descriptor.is_everything_flushed= 1; translog_force_current_buffer_to_finish(); @@ -7676,8 +7805,10 @@ my_bool translog_flush(TRANSLOG_ADDRESS TRANSLOG_BUFFERS_NO); translog_unlock(); } - sent_to_disk= translog_get_sent_to_disk(); - if (cmp_translog_addr(lsn, sent_to_disk) > 0) + + /* flush buffers */ + *sent_to_disk= translog_get_sent_to_disk(); + if (cmp_translog_addr(*lsn, *sent_to_disk) > 0) { DBUG_PRINT("info", ("Start buffer #: %u last buffer #: %u", @@ -7697,53 +7828,237 @@ my_bool translog_flush(TRANSLOG_ADDRESS LSN_IN_PARTS(buffer->last_lsn), (buffer->file ? "dirty" : "closed"))); - if (buffer->prev_last_lsn <= lsn && + if (buffer->prev_last_lsn <= *lsn && buffer->file != NULL) { - DBUG_ASSERT(flush_horizon <= buffer->offset + buffer->size); - flush_horizon= buffer->offset + buffer->size; + DBUG_ASSERT(*flush_horizon <= buffer->offset + buffer->size); + *flush_horizon= (buffer->pre_force_close_horizon != LSN_IMPOSSIBLE ? + buffer->pre_force_close_horizon : + buffer->offset + buffer->size); + /* pre_force_close_horizon is reset during new buffer start */ + DBUG_PRINT("info", ("flush_horizon: (%lu,0x%lx)", + LSN_IN_PARTS(*flush_horizon))); + DBUG_ASSERT(*flush_horizon <= log_descriptor.horizon); + translog_buffer_flush(buffer); } translog_buffer_unlock(buffer); i= (i + 1) % TRANSLOG_BUFFERS_NO; } while (i != last_buffer_no); - sent_to_disk= translog_get_sent_to_disk(); + *sent_to_disk= translog_get_sent_to_disk(); + } + + DBUG_VOID_RETURN; +} + +/** + @brief Flush the log up to given LSN (included) + + @param lsn log record serial number up to which (inclusive) + the log has to be flushed + + @return Operation status + @retval 0 OK + @retval 1 Error + + @note + + - Non group commit logic: Commits made in passes. Thread which started + flush first is performing actual flush, other threads sets new goal (LSN) + of the next pass (if it is maximum) and waits for the pass end or just + wait for the pass end. + + - If hard group commit enabled and rate set to zero: + The first thread sends all changed buffers to disk. This is repeated + as long as there are new LSNs added. The process can not loop + forever because we have limited number of threads and they will wait + for the data to be synced. + Pseudo code: + + do + send changed buffers to disk + while new_goal + sync + + - If hard group commit switched ON and less than rate microseconds has + passed from last sync, then after buffers have been sent to disk + wait until rate microseconds has passed since last sync, do sync and return. + This ensures that if we call sync infrequently we don't do any waits. + + - If soft group commit enabled everything works as with 'non group commit' + but the thread doesn't do any real sync(). If rate is not zero the + sync() will be performed by a service thread with the given rate + when needed (new LSN appears). + + @note Terminology: + 'sent to disk' means written to disk but not sync()ed, + 'flushed' mean sent to disk and synced(). +*/ + +my_bool translog_flush(TRANSLOG_ADDRESS lsn) +{ + struct timespec abstime; + ulonglong flush_interval; + ulonglong time_spent; + LSN sent_to_disk= LSN_IMPOSSIBLE; + TRANSLOG_ADDRESS flush_horizon; + my_bool rc= 0; + my_bool hgroup_commit_at_start; + DBUG_ENTER("translog_flush"); + DBUG_PRINT("enter", ("Flush up to LSN: (%lu,0x%lx)", LSN_IN_PARTS(lsn))); + DBUG_ASSERT(translog_status == TRANSLOG_OK || + translog_status == TRANSLOG_READONLY); + LINT_INIT(sent_to_disk); + + pthread_mutex_lock(&log_descriptor.log_flush_lock); + DBUG_PRINT("info", ("Everything is flushed up to (%lu,0x%lx)", + LSN_IN_PARTS(log_descriptor.flushed))); + if (cmp_translog_addr(log_descriptor.flushed, lsn) >= 0) + + + { + pthread_mutex_unlock(&log_descriptor.log_flush_lock); + DBUG_RETURN(0); } + if (log_descriptor.flush_in_progress) + { + translog_lock(); + /* fix lsn if it was horizon */ + if (cmp_translog_addr(lsn, log_descriptor.bc.buffer->last_lsn) > 0) + lsn= BUFFER_MAX_LSN(log_descriptor.bc.buffer); + translog_unlock(); + translog_flush_set_new_goal_and_wait(lsn); + if (!pthread_equal(log_descriptor.max_lsn_requester, pthread_self())) + { + /* + translog_flush_wait_for_end() release log_flush_lock while is + waiting then acquire it again + */ + translog_flush_wait_for_end(lsn); + pthread_mutex_unlock(&log_descriptor.log_flush_lock); + DBUG_RETURN(0); + } + log_descriptor.next_pass_max_lsn= LSN_IMPOSSIBLE; + } + log_descriptor.flush_in_progress= 1; + flush_horizon= log_descriptor.previous_flush_horizon; + DBUG_PRINT("info", ("flush_in_progress is set, flush_horizon: (%lu,0x%lx)", + LSN_IN_PARTS(flush_horizon))); + pthread_mutex_unlock(&log_descriptor.log_flush_lock); + + hgroup_commit_at_start= hard_group_commit; + if (hgroup_commit_at_start) + flush_interval= group_commit_wait * TRANSLOG_RATE_BASE; - /* sync files from previous flush till current one */ - for (fn= LSN_FILE_NO(log_descriptor.flushed); fn <= LSN_FILE_NO(lsn); fn++) + translog_lock(); + if (log_descriptor.is_everything_flushed) { - TRANSLOG_FILE *file= get_logfile_by_number(fn); - DBUG_ASSERT(file != NULL); - if (!file->is_sync) + DBUG_PRINT("info", ("everything is flushed")); + translog_unlock(); + pthread_mutex_lock(&log_descriptor.log_flush_lock); + goto out; + } + + for (;;) + { + /* Following function flushes buffers and makes translog_unlock() */ + translog_flush_buffers(&lsn, &sent_to_disk, &flush_horizon); + + if (!hgroup_commit_at_start) + break; /* flush pass is ended */ + +retest: + if (flush_interval != 0 && + (my_micro_time() - flush_start) >= flush_interval) + break; /* flush pass is ended */ + + pthread_mutex_lock(&log_descriptor.log_flush_lock); + if (log_descriptor.next_pass_max_lsn != LSN_IMPOSSIBLE) + { + /* take next goal */ + lsn= log_descriptor.next_pass_max_lsn; + log_descriptor.next_pass_max_lsn= LSN_IMPOSSIBLE; + /* prevent other thread from continue */ + log_descriptor.max_lsn_requester= pthread_self(); + DBUG_PRINT("info", ("flush took next goal: (%lu,0x%lx)", + LSN_IN_PARTS(lsn))); + } + else { - if (my_sync(file->handler.file, MYF(MY_WME))) + if (flush_interval == 0 || + (time_spent= (my_micro_time() - flush_start)) >= flush_interval) { - rc= 1; - translog_stop_writing(); - sent_to_disk= LSN_IMPOSSIBLE; - goto out; + pthread_mutex_unlock(&log_descriptor.log_flush_lock); + break; } - file->is_sync= 1; + DBUG_PRINT("info", ("flush waits: %llu interval: %llu spent: %llu", + flush_interval - time_spent, + flush_interval, time_spent)); + /* wait time or next goal */ + set_timespec_nsec(abstime, flush_interval - time_spent); + pthread_cond_timedwait(&log_descriptor.new_goal_cond, + &log_descriptor.log_flush_lock, + &abstime); + pthread_mutex_unlock(&log_descriptor.log_flush_lock); + DBUG_PRINT("info", ("retest conditions")); + goto retest; } + pthread_mutex_unlock(&log_descriptor.log_flush_lock); + + /* next flush pass */ + DBUG_PRINT("info", ("next flush pass")); + translog_lock(); } - if (sync_log_dir >= TRANSLOG_SYNC_DIR_ALWAYS && - (LSN_FILE_NO(log_descriptor.previous_flush_horizon) != - LSN_FILE_NO(flush_horizon) || - ((LSN_OFFSET(log_descriptor.previous_flush_horizon) - 1) / - TRANSLOG_PAGE_SIZE) != - ((LSN_OFFSET(flush_horizon) - 1) / TRANSLOG_PAGE_SIZE))) - rc|= sync_dir(log_descriptor.directory_fd, MYF(MY_WME | MY_IGNORE_BADFD)); + /* + sync() files from previous flush till current one + */ + if (!soft_sync || hgroup_commit_at_start) + { + if ((rc= + translog_sync_files(LSN_FILE_NO(log_descriptor.flushed), + LSN_FILE_NO(lsn), + sync_log_dir >= TRANSLOG_SYNC_DIR_ALWAYS && + (LSN_FILE_NO(log_descriptor. + previous_flush_horizon) != + LSN_FILE_NO(flush_horizon) || + (LSN_OFFSET(log_descriptor. + previous_flush_horizon) / + TRANSLOG_PAGE_SIZE) != + (LSN_OFFSET(flush_horizon) / + TRANSLOG_PAGE_SIZE))))) + { + sent_to_disk= LSN_IMPOSSIBLE; + pthread_mutex_lock(&log_descriptor.log_flush_lock); + goto out; + } + /* keep values for soft sync() and forced sync() actual */ + { + uint32 fileno= LSN_FILE_NO(lsn); + my_atomic_rwlock_wrlock(&soft_sync_rwl); + my_atomic_store32(&soft_sync_min, fileno); + my_atomic_store32(&soft_sync_max, fileno); + my_atomic_rwlock_wrunlock(&soft_sync_rwl); + } + } + else + { + my_atomic_rwlock_wrlock(&soft_sync_rwl); + my_atomic_store32(&soft_sync_max, LSN_FILE_NO(lsn)); + my_atomic_rwlock_wrunlock(&soft_sync_rwl); + } + + DBUG_ASSERT(flush_horizon <= log_descriptor.horizon); + + pthread_mutex_lock(&log_descriptor.log_flush_lock); log_descriptor.previous_flush_horizon= flush_horizon; out: - pthread_mutex_lock(&log_descriptor.log_flush_lock); if (sent_to_disk != LSN_IMPOSSIBLE) log_descriptor.flushed= sent_to_disk; log_descriptor.flush_in_progress= 0; log_descriptor.flush_no++; DBUG_PRINT("info", ("flush_in_progress is dropped")); - pthread_mutex_unlock(&log_descriptor.log_flush_lock);\ + pthread_mutex_unlock(&log_descriptor.log_flush_lock); pthread_cond_broadcast(&log_descriptor.log_flush_cond); DBUG_RETURN(rc); } @@ -8113,6 +8428,8 @@ LSN translog_first_theoretical_lsn() my_bool translog_purge(TRANSLOG_ADDRESS low) { uint32 last_need_file= LSN_FILE_NO(low); + uint32 min_unsync; + int soft; TRANSLOG_ADDRESS horizon= translog_get_horizon(); int rc= 0; DBUG_ENTER("translog_purge"); @@ -8120,12 +8437,23 @@ my_bool translog_purge(TRANSLOG_ADDRESS DBUG_ASSERT(translog_status == TRANSLOG_OK || translog_status == TRANSLOG_READONLY); + soft= soft_sync; + DBUG_PRINT("info", ("min_unsync: %lu", (ulong) min_unsync)); + if (soft && min_unsync < last_need_file) + { + last_need_file= min_unsync; + DBUG_PRINT("info", ("last_need_file set to %lu", (ulong)last_need_file)); + } + pthread_mutex_lock(&log_descriptor.purger_lock); + DBUG_PRINT("info", ("last_lsn_checked file: %lu:", + (ulong) log_descriptor.last_lsn_checked)); if (LSN_FILE_NO(log_descriptor.last_lsn_checked) < last_need_file) { uint32 i; uint32 min_file= translog_first_file(horizon, 1); DBUG_ASSERT(min_file != 0); /* log is already started */ + DBUG_PRINT("info", ("min_file: %lu:",(ulong) min_file)); for(i= min_file; i < last_need_file && rc == 0; i++) { LSN lsn= translog_get_file_max_lsn_stored(i); @@ -8356,6 +8684,153 @@ my_bool translog_log_debug_info(TRN *trn } + +/** + Sets soft sync mode + + @param mode TRUE if we need switch soft sync on else off +*/ + +void translog_soft_sync(my_bool mode) +{ + soft_sync= mode; +} + + +/** + Sets hard group commit + + @param mode TRUE if we need switch hard group commit on else off +*/ + +void translog_hard_group_commit(my_bool mode) +{ + hard_group_commit= mode; +} + + +/** + @brief forced log sync (used when we are switching modes) +*/ + +void translog_sync() +{ + uint32 max= get_current_logfile()->number; + uint32 min; + DBUG_ENTER("ma_translog_sync"); + + my_atomic_rwlock_rdlock(&soft_sync_rwl); + min= my_atomic_load32(&soft_sync_min); + my_atomic_rwlock_rdunlock(&soft_sync_rwl); + if (!min) + min= max; + + translog_sync_files(min, max, sync_log_dir >= TRANSLOG_SYNC_DIR_ALWAYS); + + DBUG_VOID_RETURN; +} + + +/** + @brief set rate for group commit + + @param rate rate to set. + + @note internally it stores interval in nanoseconds/TRANSLOG_RATE_BASE (to + fit into uint32) +*/ + +void translog_set_group_commit_rate(uint32 rate) +{ + DBUG_ENTER("translog_set_group_commit_rate"); + ulonglong wait_time; + if (rate) + { + wait_time= ((TRANSLOG_RATE_BASE * 1000000000ULL / rate + + TRANSLOG_RATE_BASE / 2) / + TRANSLOG_RATE_BASE); + if (wait_time == 0) + wait_time= 1; /* protection from getting special value */ + } + else + wait_time= 0; + group_commit_wait= wait_time; + DBUG_PRINT("info", ("rate: %lu wait: %llu", + (ulong)rate, (ulonglong)wait_time)); + DBUG_VOID_RETURN; +} + + +/** + @brief syncing service thread +*/ + +static pthread_handler_t +ma_soft_sync_background( void *arg __attribute__((unused))) +{ + + my_thread_init(); + { + DBUG_ENTER("ma_soft_sync_background"); + for(;;) + { + ulonglong prev_loop= my_micro_time(); + ulonglong time, sleep; + uint32 min, max; + my_atomic_rwlock_rdlock(&soft_sync_rwl); + min= my_atomic_load32(&soft_sync_min); + max= my_atomic_load32(&soft_sync_max); + my_atomic_store32(&soft_sync_min, max); + my_atomic_rwlock_rdunlock(&soft_sync_rwl); + + sleep= group_commit_wait * TRANSLOG_RATE_BASE; + translog_sync_files(min, max, FALSE); + time= my_micro_time() - prev_loop; + if (time > sleep) + sleep= 0; + else + sleep-= time; + if (my_service_thread_sleep(&soft_sync_control, sleep)) + break; + } + my_service_thread_signal_end(&soft_sync_control); + my_thread_end(); + DBUG_RETURN(0); + } +} + + +/** + @brief Starts syncing thread +*/ + +int translog_soft_sync_start(void) +{ + pthread_t th; + int res= 0; + DBUG_ENTER("translog_soft_sync_start"); + if (!(res= ma_service_thread_control_init(&soft_sync_control))) + if (!(res= pthread_create(&th, NULL, ma_soft_sync_background, NULL))) + soft_sync_control.status= THREAD_RUNNING; + DBUG_RETURN(res); +} + + +/** + @brief Stops syncing thread +*/ + +void translog_soft_sync_end(void) +{ + DBUG_ENTER("translog_soft_sync_end"); + if (soft_sync_control.inited) + { + ma_service_thread_control_end(&soft_sync_control); + } + DBUG_VOID_RETURN; +} + + #ifdef MARIA_DUMP_LOG #include <my_getopt.h> extern void translog_example_table_init(); === modified file 'storage/maria/ma_loghandler.h' --- a/storage/maria/ma_loghandler.h 2009-01-15 22:25:53 +0000 +++ b/storage/maria/ma_loghandler.h 2009-07-07 00:37:23 +0000 @@ -342,6 +342,14 @@ enum enum_translog_status TRANSLOG_SHUTDOWN /* going to shutdown the loghandler */ }; extern enum enum_translog_status translog_status; +extern ulonglong translog_syncs; /* Number of sync()s */ + +void translog_soft_sync(my_bool mode); +void translog_hard_group_commit(my_bool mode); +int translog_soft_sync_start(void); +void translog_soft_sync_end(void); +void translog_sync(); +void translog_set_group_commit_rate(uint32 rate); /* all the rest added because of recovery; should we make @@ -441,6 +449,18 @@ extern LOG_DESC log_record_type_descript typedef enum { + TRANSLOG_GCOMMIT_NONE, + TRANSLOG_GCOMMIT_HARD, + TRANSLOG_GCOMMIT_SOFT +} enum_maria_group_commit; +extern ulong maria_group_commit; +extern ulong maria_group_commit_rate; +/** + group commit interval is TRANSLOG_RATE_BASE/<rate> seconds +*/ +#define TRANSLOG_RATE_BASE 100 +typedef enum +{ TRANSLOG_PURGE_IMMIDIATE, TRANSLOG_PURGE_EXTERNAL, TRANSLOG_PURGE_ONDEMAND

2 1

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (knielsen:2776)
by knielsen＠knielsen-hq.org 27 Nov '09

27 Nov '09

#At lp:maria 2776 knielsen(a)knielsen-hq.org 2009-11-27 After-merge fixes for MySQL 5.1.41 merge into MariaDB: more fixes for Buildbot problems. modified: mysql-test/mysql-test-run.pl mysql-test/suite/rpl/r/rpl_temporary_errors.result mysql-test/suite/rpl/t/rpl_temporary_errors.test sql/sql_select.cc per-file messages: mysql-test/mysql-test-run.pl Manually apply similar patch to the one in Bug#47983. mysql-test/suite/rpl/r/rpl_temporary_errors.result Fix wrong failure with warning in error log due to per-test suppressions not being active during server shutdown. mysql-test/suite/rpl/t/rpl_temporary_errors.test Fix wrong failure with warning in error log due to per-test suppressions not being active during server shutdown. sql/sql_select.cc Manually cherry-pick fix for Bug#45989. For some reason, that fix was reverted in MySQL 5.1, even though it is necessary to plug the memory leak. === modified file 'mysql-test/mysql-test-run.pl' --- a/mysql-test/mysql-test-run.pl 2009-11-26 08:02:33 +0000 +++ b/mysql-test/mysql-test-run.pl 2009-11-27 13:20:59 +0000 @@ -3994,6 +3994,11 @@ sub extract_warning_lines ($) { qr/Slave SQL thread retried transaction/, qr/Slave $additional info$/, qr/Incorrect information in file/, + qr/Slave I\/O: Get master SERVER_ID failed with error:.*/, + qr/Slave I\/O: Get master clock failed with error:.*/, + qr/Slave I\/O: Get master COLLATION_SERVER failed with error:.*/, + qr/Slave I\/O: Get master TIME_ZONE failed with error:.*/, + qr/Slave I\/O: error reconnecting to master '.*' - retry-time: [1-3] retries/, ); my $match_count= 0; === modified file 'mysql-test/suite/rpl/r/rpl_temporary_errors.result' --- a/mysql-test/suite/rpl/r/rpl_temporary_errors.result 2008-10-13 18:33:08 +0000 +++ b/mysql-test/suite/rpl/r/rpl_temporary_errors.result 2009-11-27 13:20:59 +0000 @@ -82,3 +82,4 @@ Last_SQL_Error DROP TABLE t1; **** On Master **** DROP TABLE t1; +STOP SLAVE; === modified file 'mysql-test/suite/rpl/t/rpl_temporary_errors.test' --- a/mysql-test/suite/rpl/t/rpl_temporary_errors.test 2008-10-13 18:33:08 +0000 +++ b/mysql-test/suite/rpl/t/rpl_temporary_errors.test 2009-11-27 13:20:59 +0000 @@ -31,3 +31,12 @@ DROP TABLE t1; --echo **** On Master **** connection master; DROP TABLE t1; + +# We must wait for the slave to stop. +# Otherwise the warnings in the error log about deadlock may be written to +# the error log only during shutdown, and currently the suppression of +# "Deadlock found" set in this test case is not effective during server +# shutdown. +connection slave; +STOP SLAVE; +--source include/wait_for_slave_to_stop.inc === modified file 'sql/sql_select.cc' --- a/sql/sql_select.cc 2009-11-16 20:54:33 +0000 +++ b/sql/sql_select.cc 2009-11-27 13:20:59 +0000 @@ -2311,7 +2311,7 @@ JOIN::destroy() tab->cleanup(); } tmp_join->tmp_join= 0; - tmp_table_param.copy_field= 0; + tmp_table_param.cleanup(); DBUG_RETURN(tmp_join->destroy()); } cond_equal= 0;

1 0

[Maria-developers] Updated (by Knielsen): Store in binlog text of statements that caused RBR events (47)
by worklog-noreply＠askmonty.org 27 Nov '09

27 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Store in binlog text of statements that caused RBR events CREATION DATE..: Sat, 15 Aug 2009, 23:48 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: Knielsen CATEGORY.......: Server-RawIdeaBin TASK ID........: 47 (http://askmonty.org/worklog/?tid=47) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Knielsen - Fri, 27 Nov 2009, 13:30)=-=- Observers changed: Knielsen -=-=(Psergey - Sun, 16 Aug 2009, 11:08)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.12485 2009-08-16 11:08:33.000000000 +0300 +++ /tmp/wklog.47.new.12485 2009-08-16 11:08:33.000000000 +0300 @@ -1 +1,6 @@ +First suggestion: + +> I think for this we would actually need a new binlog event type +> (Comment_log_event?). Unless we want to log an empty statement Query_log_event +> containing only a comment (a bit of a hack). -=-=(Psergey - Sun, 16 Aug 2009, 00:02)=-=- Dependency created: 39 now depends on 47 DESCRIPTION: Store in binlog (and show in mysqlbinlog output) texts of statements that caused RBR events This is needed for (list from Monty): - Easier to understand why updates happened - Would make it easier to find out where in application things went wrong (as you can search for exact strings) - Allow one to filter things based on comments in the statement. The cost of this can be that the binlog will be approximately 2x in size (especially insert of big blob's would be a bit painful), so this should be an optional feature. HIGH-LEVEL SPECIFICATION: First suggestion: > I think for this we would actually need a new binlog event type > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Knielsen): Store in binlog text of statements that caused RBR events (47)
by worklog-noreply＠askmonty.org 27 Nov '09

27 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Store in binlog text of statements that caused RBR events CREATION DATE..: Sat, 15 Aug 2009, 23:48 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: Knielsen CATEGORY.......: Server-RawIdeaBin TASK ID........: 47 (http://askmonty.org/worklog/?tid=47) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Knielsen - Fri, 27 Nov 2009, 13:30)=-=- Observers changed: Knielsen -=-=(Psergey - Sun, 16 Aug 2009, 11:08)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.12485 2009-08-16 11:08:33.000000000 +0300 +++ /tmp/wklog.47.new.12485 2009-08-16 11:08:33.000000000 +0300 @@ -1 +1,6 @@ +First suggestion: + +> I think for this we would actually need a new binlog event type +> (Comment_log_event?). Unless we want to log an empty statement Query_log_event +> containing only a comment (a bit of a hack). -=-=(Psergey - Sun, 16 Aug 2009, 00:02)=-=- Dependency created: 39 now depends on 47 DESCRIPTION: Store in binlog (and show in mysqlbinlog output) texts of statements that caused RBR events This is needed for (list from Monty): - Easier to understand why updates happened - Would make it easier to find out where in application things went wrong (as you can search for exact strings) - Allow one to filter things based on comments in the statement. The cost of this can be that the binlog will be approximately 2x in size (especially insert of big blob's would be a bit painful), so this should be an optional feature. HIGH-LEVEL SPECIFICATION: First suggestion: > I think for this we would actually need a new binlog event type > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Knielsen): Store in binlog text of statements that caused RBR events (47)
by worklog-noreply＠askmonty.org 27 Nov '09

27 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Store in binlog text of statements that caused RBR events CREATION DATE..: Sat, 15 Aug 2009, 23:48 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: Knielsen CATEGORY.......: Server-RawIdeaBin TASK ID........: 47 (http://askmonty.org/worklog/?tid=47) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Knielsen - Fri, 27 Nov 2009, 13:30)=-=- Observers changed: Knielsen -=-=(Psergey - Sun, 16 Aug 2009, 11:08)=-=- High-Level Specification modified. --- /tmp/wklog.47.old.12485 2009-08-16 11:08:33.000000000 +0300 +++ /tmp/wklog.47.new.12485 2009-08-16 11:08:33.000000000 +0300 @@ -1 +1,6 @@ +First suggestion: + +> I think for this we would actually need a new binlog event type +> (Comment_log_event?). Unless we want to log an empty statement Query_log_event +> containing only a comment (a bit of a hack). -=-=(Psergey - Sun, 16 Aug 2009, 00:02)=-=- Dependency created: 39 now depends on 47 DESCRIPTION: Store in binlog (and show in mysqlbinlog output) texts of statements that caused RBR events This is needed for (list from Monty): - Easier to understand why updates happened - Would make it easier to find out where in application things went wrong (as you can search for exact strings) - Allow one to filter things based on comments in the statement. The cost of this can be that the binlog will be approximately 2x in size (especially insert of big blob's would be a bit painful), so this should be an optional feature. HIGH-LEVEL SPECIFICATION: First suggestion: > I think for this we would actually need a new binlog event type > (Comment_log_event?). Unless we want to log an empty statement Query_log_event > containing only a comment (a bit of a hack). ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Timour): Subquery optimization: Efficient NOT IN execution with NULLs (68)
by worklog-noreply＠askmonty.org 27 Nov '09

27 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Subquery optimization: Efficient NOT IN execution with NULLs CREATION DATE..: Fri, 27 Nov 2009, 13:22 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: CATEGORY.......: Server-RawIdeaBin TASK ID........: 68 (http://askmonty.org/worklog/?tid=68) VERSION........: Benchmarks-3.0 STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Timour - Fri, 27 Nov 2009, 13:23)=-=- High-Level Specification modified. --- /tmp/wklog.68.old.17140 2009-11-27 11:23:17.000000000 +0000 +++ /tmp/wklog.68.new.17140 2009-11-27 11:23:17.000000000 +0000 @@ -1 +1,72 @@ +This a copy of the initial algorithm proposed by Igor: +====================================================== +For each left side tuple (v_1,...,v_n) we have to find the following set +of rowids for the temp table containing N rows as the result of +materialization of the subquery: + + R= INTERSECT (rowid{a_i=v_i} UNION rowid{a_i is null} where i runs +trough all indexes from [1..n] such that v_i is not null. + +Bear in mind the following specifics of this intersection: + (1) For each i: rowid{a_i=v_i} and rowid{a_i is null} are disjoint + (2) For each i: rowid{a_i is null} is the same for each tuple + +Due to (2) it makes sense to build rowid{a_i is null} only once. +A good representation for such sets would be bitmaps: +- it requires minimum memory: not more than N*n bits in total +- search of an element in a set is extremely cheap + +Taken all above into account I could suggest the following algorithm to +build R: + + Using indexes (read about them below) for each column participating in the + intersection, + merge ordered sets rowid{a_i=v_i} in the following manner. + If a rowid r has been encountered maximum in k sets +rowid{a_i1=v_i1},...,rowid(a_ik=v_ik), + then it has to be checked against all rowid{a_i=v_i} such that i is +not in {i1,...,ik}. + As soon as we fail to find r in one of these sets we discard it. + If r has been found in all of them then r belongs to the set R. + +Here we use the property (1): any r from rowid{a_i=v_i} UNION rowid{a_i +is null} is either +belongs to rowid{a_i=v_i} or to rowid{a_i is null}. From this we can +infer that for any r from R +indexes a_i can be uniquely divided into two groups: one contains +indexes a_i where r belongs to +the sets rowid{a_i=v_i}, the other contains indexes a_j such that r +belongs to rowid{a_j is null}. + +Now let's talk how to get elements from rowid{a_i=v_i} in a sorted order +needed for the merge procedure. We could use BTREE indexes for temp +table. But they are rather expensive and +take a lot of memory as the are implemented with RB trees. +I would suggest creating for each column from the temporary table just +an array of rowids sorted by the value from column a. +Index lookup in such an array is cheap. It's also rather cheap to check +that the next rowid refers to a row with a different value in column a. +The array can be created on demand. + +Other consideration that may be taken into account: +1. If columns a_j1,...,a_jm do not contain null values in the temporary +table at all, create for them only one index array (and of course do not +create any bitmaps for them). +2. Consider the ratio d(a_i)=N'/ V(a_i), where N' is the number of rows +where a_i is not null and V(a_i) is the number of distinct values for +a_i excluding nulls. + If d(a_i) is close to 1 then do not create any index array: check +whether there is a match running through the records that have been +filtered in. Anyway if d(a_i) is close to 1 then a intersection with +rowid{a_i=v_i} would not reduce the number of remaining rowids +significantly. + If additionally N-N' is small do not create a bitmap for this column +either. +3. If for a column a_i d(a_i) is not close to 1, but N-N' is small a +sorted array of rowids from the set rowid{a_i is null} can be used +instead of a bitmap. +4. We always have a match if R0= INTERSECT rowid{a_i is null} is not +empty. Here i runs through all indexes from [1..n] such that v_i is not +null. For a given subset of columns this fact has to be checked only +once. It can be easily done with bitmap intersection. DESCRIPTION: The goal of this task is to implement efficient execution of NOT IN subquery predicates of the form: <oe_1,...,oe_n> NOT IN <non_correlated subquery> when either some oe_i, or some subqury result column contains NULLs. The problem with such predicates is that it is possible to use index lookups only when neither argument of the predicate contains NULLs. If some argument contains a NULL, then due to NULL semantics, it plays the role of a wildcard. If we were to use regular index lookups, then we would get 'no match' for some outer tuple (thus the predicate evaluates to FALSE), while the SQL semantics means 'partial match', and the predicate should evaluate to NULL. This task implements an efficient algorithm to compute such 'parial matches', where a NULL matches any value. HIGH-LEVEL SPECIFICATION: This a copy of the initial algorithm proposed by Igor: ====================================================== For each left side tuple (v_1,...,v_n) we have to find the following set of rowids for the temp table containing N rows as the result of materialization of the subquery: R= INTERSECT (rowid{a_i=v_i} UNION rowid{a_i is null} where i runs trough all indexes from [1..n] such that v_i is not null. Bear in mind the following specifics of this intersection: (1) For each i: rowid{a_i=v_i} and rowid{a_i is null} are disjoint (2) For each i: rowid{a_i is null} is the same for each tuple Due to (2) it makes sense to build rowid{a_i is null} only once. A good representation for such sets would be bitmaps: - it requires minimum memory: not more than N*n bits in total - search of an element in a set is extremely cheap Taken all above into account I could suggest the following algorithm to build R: Using indexes (read about them below) for each column participating in the intersection, merge ordered sets rowid{a_i=v_i} in the following manner. If a rowid r has been encountered maximum in k sets rowid{a_i1=v_i1},...,rowid(a_ik=v_ik), then it has to be checked against all rowid{a_i=v_i} such that i is not in {i1,...,ik}. As soon as we fail to find r in one of these sets we discard it. If r has been found in all of them then r belongs to the set R. Here we use the property (1): any r from rowid{a_i=v_i} UNION rowid{a_i is null} is either belongs to rowid{a_i=v_i} or to rowid{a_i is null}. From this we can infer that for any r from R indexes a_i can be uniquely divided into two groups: one contains indexes a_i where r belongs to the sets rowid{a_i=v_i}, the other contains indexes a_j such that r belongs to rowid{a_j is null}. Now let's talk how to get elements from rowid{a_i=v_i} in a sorted order needed for the merge procedure. We could use BTREE indexes for temp table. But they are rather expensive and take a lot of memory as the are implemented with RB trees. I would suggest creating for each column from the temporary table just an array of rowids sorted by the value from column a. Index lookup in such an array is cheap. It's also rather cheap to check that the next rowid refers to a row with a different value in column a. The array can be created on demand. Other consideration that may be taken into account: 1. If columns a_j1,...,a_jm do not contain null values in the temporary table at all, create for them only one index array (and of course do not create any bitmaps for them). 2. Consider the ratio d(a_i)=N'/ V(a_i), where N' is the number of rows where a_i is not null and V(a_i) is the number of distinct values for a_i excluding nulls. If d(a_i) is close to 1 then do not create any index array: check whether there is a match running through the records that have been filtered in. Anyway if d(a_i) is close to 1 then a intersection with rowid{a_i=v_i} would not reduce the number of remaining rowids significantly. If additionally N-N' is small do not create a bitmap for this column either. 3. If for a column a_i d(a_i) is not close to 1, but N-N' is small a sorted array of rowids from the set rowid{a_i is null} can be used instead of a bitmap. 4. We always have a match if R0= INTERSECT rowid{a_i is null} is not empty. Here i runs through all indexes from [1..n] such that v_i is not null. For a given subset of columns this fact has to be checked only once. It can be easily done with bitmap intersection. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Timour): Subquery optimization: Efficient NOT IN execution with NULLs (68)
by worklog-noreply＠askmonty.org 27 Nov '09

27 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Subquery optimization: Efficient NOT IN execution with NULLs CREATION DATE..: Fri, 27 Nov 2009, 13:22 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: CATEGORY.......: Server-RawIdeaBin TASK ID........: 68 (http://askmonty.org/worklog/?tid=68) VERSION........: Benchmarks-3.0 STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Timour - Fri, 27 Nov 2009, 13:23)=-=- High-Level Specification modified. --- /tmp/wklog.68.old.17140 2009-11-27 11:23:17.000000000 +0000 +++ /tmp/wklog.68.new.17140 2009-11-27 11:23:17.000000000 +0000 @@ -1 +1,72 @@ +This a copy of the initial algorithm proposed by Igor: +====================================================== +For each left side tuple (v_1,...,v_n) we have to find the following set +of rowids for the temp table containing N rows as the result of +materialization of the subquery: + + R= INTERSECT (rowid{a_i=v_i} UNION rowid{a_i is null} where i runs +trough all indexes from [1..n] such that v_i is not null. + +Bear in mind the following specifics of this intersection: + (1) For each i: rowid{a_i=v_i} and rowid{a_i is null} are disjoint + (2) For each i: rowid{a_i is null} is the same for each tuple + +Due to (2) it makes sense to build rowid{a_i is null} only once. +A good representation for such sets would be bitmaps: +- it requires minimum memory: not more than N*n bits in total +- search of an element in a set is extremely cheap + +Taken all above into account I could suggest the following algorithm to +build R: + + Using indexes (read about them below) for each column participating in the + intersection, + merge ordered sets rowid{a_i=v_i} in the following manner. + If a rowid r has been encountered maximum in k sets +rowid{a_i1=v_i1},...,rowid(a_ik=v_ik), + then it has to be checked against all rowid{a_i=v_i} such that i is +not in {i1,...,ik}. + As soon as we fail to find r in one of these sets we discard it. + If r has been found in all of them then r belongs to the set R. + +Here we use the property (1): any r from rowid{a_i=v_i} UNION rowid{a_i +is null} is either +belongs to rowid{a_i=v_i} or to rowid{a_i is null}. From this we can +infer that for any r from R +indexes a_i can be uniquely divided into two groups: one contains +indexes a_i where r belongs to +the sets rowid{a_i=v_i}, the other contains indexes a_j such that r +belongs to rowid{a_j is null}. + +Now let's talk how to get elements from rowid{a_i=v_i} in a sorted order +needed for the merge procedure. We could use BTREE indexes for temp +table. But they are rather expensive and +take a lot of memory as the are implemented with RB trees. +I would suggest creating for each column from the temporary table just +an array of rowids sorted by the value from column a. +Index lookup in such an array is cheap. It's also rather cheap to check +that the next rowid refers to a row with a different value in column a. +The array can be created on demand. + +Other consideration that may be taken into account: +1. If columns a_j1,...,a_jm do not contain null values in the temporary +table at all, create for them only one index array (and of course do not +create any bitmaps for them). +2. Consider the ratio d(a_i)=N'/ V(a_i), where N' is the number of rows +where a_i is not null and V(a_i) is the number of distinct values for +a_i excluding nulls. + If d(a_i) is close to 1 then do not create any index array: check +whether there is a match running through the records that have been +filtered in. Anyway if d(a_i) is close to 1 then a intersection with +rowid{a_i=v_i} would not reduce the number of remaining rowids +significantly. + If additionally N-N' is small do not create a bitmap for this column +either. +3. If for a column a_i d(a_i) is not close to 1, but N-N' is small a +sorted array of rowids from the set rowid{a_i is null} can be used +instead of a bitmap. +4. We always have a match if R0= INTERSECT rowid{a_i is null} is not +empty. Here i runs through all indexes from [1..n] such that v_i is not +null. For a given subset of columns this fact has to be checked only +once. It can be easily done with bitmap intersection. DESCRIPTION: The goal of this task is to implement efficient execution of NOT IN subquery predicates of the form: <oe_1,...,oe_n> NOT IN <non_correlated subquery> when either some oe_i, or some subqury result column contains NULLs. The problem with such predicates is that it is possible to use index lookups only when neither argument of the predicate contains NULLs. If some argument contains a NULL, then due to NULL semantics, it plays the role of a wildcard. If we were to use regular index lookups, then we would get 'no match' for some outer tuple (thus the predicate evaluates to FALSE), while the SQL semantics means 'partial match', and the predicate should evaluate to NULL. This task implements an efficient algorithm to compute such 'parial matches', where a NULL matches any value. HIGH-LEVEL SPECIFICATION: This a copy of the initial algorithm proposed by Igor: ====================================================== For each left side tuple (v_1,...,v_n) we have to find the following set of rowids for the temp table containing N rows as the result of materialization of the subquery: R= INTERSECT (rowid{a_i=v_i} UNION rowid{a_i is null} where i runs trough all indexes from [1..n] such that v_i is not null. Bear in mind the following specifics of this intersection: (1) For each i: rowid{a_i=v_i} and rowid{a_i is null} are disjoint (2) For each i: rowid{a_i is null} is the same for each tuple Due to (2) it makes sense to build rowid{a_i is null} only once. A good representation for such sets would be bitmaps: - it requires minimum memory: not more than N*n bits in total - search of an element in a set is extremely cheap Taken all above into account I could suggest the following algorithm to build R: Using indexes (read about them below) for each column participating in the intersection, merge ordered sets rowid{a_i=v_i} in the following manner. If a rowid r has been encountered maximum in k sets rowid{a_i1=v_i1},...,rowid(a_ik=v_ik), then it has to be checked against all rowid{a_i=v_i} such that i is not in {i1,...,ik}. As soon as we fail to find r in one of these sets we discard it. If r has been found in all of them then r belongs to the set R. Here we use the property (1): any r from rowid{a_i=v_i} UNION rowid{a_i is null} is either belongs to rowid{a_i=v_i} or to rowid{a_i is null}. From this we can infer that for any r from R indexes a_i can be uniquely divided into two groups: one contains indexes a_i where r belongs to the sets rowid{a_i=v_i}, the other contains indexes a_j such that r belongs to rowid{a_j is null}. Now let's talk how to get elements from rowid{a_i=v_i} in a sorted order needed for the merge procedure. We could use BTREE indexes for temp table. But they are rather expensive and take a lot of memory as the are implemented with RB trees. I would suggest creating for each column from the temporary table just an array of rowids sorted by the value from column a. Index lookup in such an array is cheap. It's also rather cheap to check that the next rowid refers to a row with a different value in column a. The array can be created on demand. Other consideration that may be taken into account: 1. If columns a_j1,...,a_jm do not contain null values in the temporary table at all, create for them only one index array (and of course do not create any bitmaps for them). 2. Consider the ratio d(a_i)=N'/ V(a_i), where N' is the number of rows where a_i is not null and V(a_i) is the number of distinct values for a_i excluding nulls. If d(a_i) is close to 1 then do not create any index array: check whether there is a match running through the records that have been filtered in. Anyway if d(a_i) is close to 1 then a intersection with rowid{a_i=v_i} would not reduce the number of remaining rowids significantly. If additionally N-N' is small do not create a bitmap for this column either. 3. If for a column a_i d(a_i) is not close to 1, but N-N' is small a sorted array of rowids from the set rowid{a_i is null} can be used instead of a bitmap. 4. We always have a match if R0= INTERSECT rowid{a_i is null} is not empty. Here i runs through all indexes from [1..n] such that v_i is not null. For a given subset of columns this fact has to be checked only once. It can be easily done with bitmap intersection. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] New (by Timour): Subquery optimization: Efficient NOT IN execution with NULLs (68)
by worklog-noreply＠askmonty.org 27 Nov '09

27 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Subquery optimization: Efficient NOT IN execution with NULLs CREATION DATE..: Fri, 27 Nov 2009, 13:22 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: CATEGORY.......: Server-RawIdeaBin TASK ID........: 68 (http://askmonty.org/worklog/?tid=68) VERSION........: Benchmarks-3.0 STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: DESCRIPTION: The goal of this task is to implement efficient execution of NOT IN subquery predicates of the form: <oe_1,...,oe_n> NOT IN <non_correlated subquery> when either some oe_i, or some subqury result column contains NULLs. The problem with such predicates is that it is possible to use index lookups only when neither argument of the predicate contains NULLs. If some argument contains a NULL, then due to NULL semantics, it plays the role of a wildcard. If we were to use regular index lookups, then we would get 'no match' for some outer tuple (thus the predicate evaluates to FALSE), while the SQL semantics means 'partial match', and the predicate should evaluate to NULL. This task implements an efficient algorithm to compute such 'parial matches', where a NULL matches any value. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] New (by Timour): Subquery optimization: Efficient NOT IN execution with NULLs (68)
by worklog-noreply＠askmonty.org 27 Nov '09

27 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Subquery optimization: Efficient NOT IN execution with NULLs CREATION DATE..: Fri, 27 Nov 2009, 13:22 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: CATEGORY.......: Server-RawIdeaBin TASK ID........: 68 (http://askmonty.org/worklog/?tid=68) VERSION........: Benchmarks-3.0 STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: DESCRIPTION: The goal of this task is to implement efficient execution of NOT IN subquery predicates of the form: <oe_1,...,oe_n> NOT IN <non_correlated subquery> when either some oe_i, or some subqury result column contains NULLs. The problem with such predicates is that it is possible to use index lookups only when neither argument of the predicate contains NULLs. If some argument contains a NULL, then due to NULL semantics, it plays the role of a wildcard. If we were to use regular index lookups, then we would get 'no match' for some outer tuple (thus the predicate evaluates to FALSE), while the SQL semantics means 'partial match', and the predicate should evaluate to NULL. This task implements an efficient algorithm to compute such 'parial matches', where a NULL matches any value. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Monty): Provide key cache statistics (58)
by worklog-noreply＠askmonty.org 27 Nov '09

27 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Provide key cache statistics CREATION DATE..: Thu, 22 Oct 2009, 12:28 SUPERVISOR.....: Bothorsen IMPLEMENTOR....: COPIES TO......: CATEGORY.......: Server-RawIdeaBin TASK ID........: 58 (http://askmonty.org/worklog/?tid=58) VERSION........: WorkLog-3.4 STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Monty - Fri, 27 Nov 2009, 12:36)=-=- High Level Description modified. --- /tmp/wklog.58.old.15108 2009-11-27 10:36:13.000000000 +0000 +++ /tmp/wklog.58.new.15108 2009-11-27 10:36:13.000000000 +0000 @@ -1,8 +1,7 @@ -Provide the key cache -statistics available for the default cache (key blocks used, -unflushed, etc.) for all named caches. +Provide the same key cache statistics for all MyISAM key caches as we provide +for the default key cache. -Monty answers in an email: +Background: Monty answers in an email to customer: This is a much simpler task and something that is important to get done. We would have to introduce a 'show keycache statistics' command @@ -11,3 +10,44 @@ The raw coding is probably 1-2 days, but on top if this we need testing, a test environment and building of biniaries for you. Henrik/Bo and Igor will come back to you with a more exact estimate. + +----------------- +Currently we have: + +show status like "key%"; ++------------------------+-------+ +| Variable_name | Value | ++------------------------+-------+ +| Key_blocks_not_flushed | 0 | +| Key_blocks_unused | 13389 | +| Key_blocks_used | 7 | +| Key_read_requests | 22 | +| Key_reads | 7 | +| Key_write_requests | 0 | +| Key_writes | 0 | ++------------------------+-------+ + +Sergei suggested we introduce for each new key cache a set of variables prefixed +with the keycache name: + +show status like "keycache1.%"; +Which should show: + ++----------------------------------+-------+ +| Variable_name | Value | ++----------------------------------+-------+ +| keycache1.Key_blocks_not_flushed | 0 | +| keycache1.Key_blocks_unused | 13389 | +| keycache1.Key_blocks_used | 7 | +| keycache1.Key_read_requests | 22 | +| keycache1.Key_reads | 7 | +| keycache1.Key_write_requests | 0 | +| keycache1.Key_writes | 0 | ++----------------------------------+-------+ + +The task would thus to be to automatically introduce new variables when we +create a new key cache and automatically remove these variables when the cache +is deleted. + +The other option would be to add an information schema where we populate the +information schema with data from all existing key caches. DESCRIPTION: Provide the same key cache statistics for all MyISAM key caches as we provide for the default key cache. Background: Monty answers in an email to customer: This is a much simpler task and something that is important to get done. We would have to introduce a 'show keycache statistics' command that would iterate over all keycaches and provide the statistics. The raw coding is probably 1-2 days, but on top if this we need testing, a test environment and building of biniaries for you. Henrik/Bo and Igor will come back to you with a more exact estimate. ----------------- Currently we have: show status like "key%"; +------------------------+-------+ | Variable_name | Value | +------------------------+-------+ | Key_blocks_not_flushed | 0 | | Key_blocks_unused | 13389 | | Key_blocks_used | 7 | | Key_read_requests | 22 | | Key_reads | 7 | | Key_write_requests | 0 | | Key_writes | 0 | +------------------------+-------+ Sergei suggested we introduce for each new key cache a set of variables prefixed with the keycache name: show status like "keycache1.%"; Which should show: +----------------------------------+-------+ | Variable_name | Value | +----------------------------------+-------+ | keycache1.Key_blocks_not_flushed | 0 | | keycache1.Key_blocks_unused | 13389 | | keycache1.Key_blocks_used | 7 | | keycache1.Key_read_requests | 22 | | keycache1.Key_reads | 7 | | keycache1.Key_write_requests | 0 | | keycache1.Key_writes | 0 | +----------------------------------+-------+ The task would thus to be to automatically introduce new variables when we create a new key cache and automatically remove these variables when the cache is deleted. The other option would be to add an information schema where we populate the information schema with data from all existing key caches. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Test failures in PBXT after merging with MySQL-5.1.41
by Kristian Nielsen 27 Nov '09

27 Nov '09

Hi Paul, Vladimir, We are working on merging latest MySQL (5.1.41) into MariaDB. After the merge, I see some test failures in the PBXT test suite. See here: https://askmonty.org/buildbot/builders/hardy-amd64-fulltest/builds/118/step… Or if you prefer, see here for a search for these failures in the build history database: http://askmonty.org/buildbot/reports/cross_reference#branch=5.1-merge&revis… >From a quick look, it seems to me most/all of these are just the need to adjust result files for changes/bugfixes in MySQL. But before spending time on fixing, I though it was better to coordinate with you, to not duplicate effort. So perhaps you already did the similar changes for the PBXT testsuite against MySQL 5.1.41? Is there maybe a newer version of PBXT that we should merge? Or alternatively, do you want me to fix the testsuite, and then send you a patch for merging into PBXT to update to 5.1.41? - Kristian.

3 11

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (monty:2769)
by Michael Widenius 26 Nov '09

26 Nov '09

#At lp:maria based on revid:monty@askmonty.org-20091126155705-5g7fmkbcpdu6ad6j 2769 Michael Widenius 2009-11-26 Added protection around usage of thd->mysys_var (May be changed to 0 by scheduler) modified: sql/mysqld.cc sql/scheduler.cc sql/sql_class.cc sql/sql_class.h sql/sql_insert.cc sql/sql_show.cc === modified file 'sql/mysqld.cc' --- a/sql/mysqld.cc 2009-10-31 19:22:50 +0000 +++ b/sql/mysqld.cc 2009-11-26 20:19:33 +0000 @@ -957,6 +957,7 @@ static void close_connections(void) tmp->killed= THD::KILL_CONNECTION; thread_scheduler.post_kill_notification(tmp); + pthread_mutex_lock(&tmp->LOCK_thd_data); if (tmp->mysys_var) { tmp->mysys_var->abort=1; @@ -979,6 +980,7 @@ static void close_connections(void) } pthread_mutex_unlock(&tmp->mysys_var->mutex); } + pthread_mutex_unlock(&tmp->LOCK_thd_data); } (void) pthread_mutex_unlock(&LOCK_thread_count); // For unlink from list === modified file 'sql/scheduler.cc' --- a/sql/scheduler.cc 2009-09-07 20:50:10 +0000 +++ b/sql/scheduler.cc 2009-11-26 20:19:33 +0000 @@ -235,9 +235,7 @@ void thd_scheduler::thread_detach() if (thread_attached) { THD* thd = (THD*)list.data; - pthread_mutex_lock(&thd->LOCK_thd_data); - thd->mysys_var= NULL; - pthread_mutex_unlock(&thd->LOCK_thd_data); + thd->reset_globals(); thread_attached= FALSE; #ifndef DBUG_OFF /* === modified file 'sql/sql_class.cc' --- a/sql/sql_class.cc 2009-09-15 10:46:35 +0000 +++ b/sql/sql_class.cc 2009-11-26 20:19:33 +0000 @@ -1166,6 +1166,19 @@ bool THD::store_globals() } +/** + Untie THD from current thread + + Used when using --thread-handling=pool-of-threads +*/ + +void THD::reset_globals() +{ + pthread_mutex_lock(&LOCK_thd_data); + mysys_var= 0; + pthread_mutex_unlock(&LOCK_thd_data); +} + /* Cleanup after query. === modified file 'sql/sql_class.h' --- a/sql/sql_class.h 2009-09-15 10:46:35 +0000 +++ b/sql/sql_class.h 2009-11-26 20:19:33 +0000 @@ -1906,6 +1906,7 @@ public: void cleanup(void); void cleanup_after_query(); bool store_globals(); + void reset_globals(); #ifdef SIGNAL_WITH_VIO_CLOSE inline void set_active_vio(Vio* vio) { === modified file 'sql/sql_insert.cc' --- a/sql/sql_insert.cc 2009-11-07 10:34:19 +0000 +++ b/sql/sql_insert.cc 2009-11-26 20:19:33 +0000 @@ -2279,6 +2279,7 @@ void kill_delayed_threads(void) while ((di= it++)) { di->thd.killed= THD::KILL_CONNECTION; + pthread_mutex_lock(&di->thd.LOCK_thd_data); if (di->thd.mysys_var) { pthread_mutex_lock(&di->thd.mysys_var->mutex); @@ -2297,6 +2298,7 @@ void kill_delayed_threads(void) } pthread_mutex_unlock(&di->thd.mysys_var->mutex); } + pthread_mutex_unlock(&di->thd.LOCK_thd_data); } VOID(pthread_mutex_unlock(&LOCK_delayed_insert)); // For unlink from list } === modified file 'sql/sql_show.cc' --- a/sql/sql_show.cc 2009-11-02 09:30:21 +0000 +++ b/sql/sql_show.cc 2009-11-26 20:19:33 +0000 @@ -1747,6 +1747,7 @@ void mysqld_list_processes(THD *thd,cons if ((thd_info->db=tmp->db)) // Safe test thd_info->db=thd->strdup(thd_info->db); thd_info->command=(int) tmp->command; + pthread_mutex_lock(&tmp->LOCK_thd_data); if ((mysys_var= tmp->mysys_var)) pthread_mutex_lock(&mysys_var->mutex); thd_info->proc_info= (char*) (tmp->killed == THD::KILL_CONNECTION? "Killed" : 0); @@ -1766,6 +1767,7 @@ void mysqld_list_processes(THD *thd,cons #endif if (mysys_var) pthread_mutex_unlock(&mysys_var->mutex); + pthread_mutex_unlock(&tmp->LOCK_thd_data); thd_info->start_time= tmp->start_time; thd_info->query=0;

1 0

[Maria-developers] Updated (by Psergey): MRR backport (67)
by worklog-noreply＠askmonty.org 26 Nov '09

26 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: MRR backport CREATION DATE..: Thu, 26 Nov 2009, 15:19 SUPERVISOR.....: Monty IMPLEMENTOR....: Psergey COPIES TO......: CATEGORY.......: Server-Sprint TASK ID........: 67 (http://askmonty.org/worklog/?tid=67) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Psergey - Thu, 26 Nov 2009, 20:21)=-=- High-Level Specification modified. --- /tmp/wklog.67.old.9329 2009-11-26 20:21:28.000000000 +0200 +++ /tmp/wklog.67.new.9329 2009-11-26 20:21:28.000000000 +0200 @@ -65,17 +65,19 @@ 2.5 Make MRR code more of a module ---------------------------------- -Some code in handler.cc can be moved to separate file. -But changes in opt_range.cc can't. -TODO: Sort out how much we really can do here. Initial guess is not much as the -code consists of: +It is not possible to make MRR to be a totally separate module, as its code +consists of : - Default MRR implementation in handler.cc - Changes in opt_range.cc to use MRR instead of multiple records_in_range() - calls. These rely on opt_range.cc's internal structures like SEL_ARG trees and + calls. These rely on opt_range.cc's internal stuctures like SEL_ARG trees and so there is not much point in moving them out. -- DS-MRR implementations which are spread over storage engines. -and the only modularization we see is to move #1 into a separate file which -won't achieve much. +- DS-MRR impelementations which are spread over storage engines. + +We'll try to modularize what we can: +- Move out default MRR implementation from handler.cc +- Move possible parts out of opt_range.cc into a separate file. + + 2.6 Improve the cost model -------------------------- -=-=(Psergey - Thu, 26 Nov 2009, 19:06)=-=- High-Level Specification modified. --- /tmp/wklog.67.old.6449 2009-11-26 19:06:04.000000000 +0200 +++ /tmp/wklog.67.new.6449 2009-11-26 19:06:04.000000000 +0200 @@ -1,4 +1,3 @@ - <contents> 1. Requirements 2. Required actions @@ -44,6 +43,7 @@ http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=index_condi… http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=mrr +http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=icp 2.2 Backport DS-MRR code to MariaDB 5.2 --------------------------------------- -=-=(Psergey - Thu, 26 Nov 2009, 18:15)=-=- High-Level Specification modified. --- /tmp/wklog.67.old.4161 2009-11-26 18:15:36.000000000 +0200 +++ /tmp/wklog.67.new.4161 2009-11-26 18:15:36.000000000 +0200 @@ -1,3 +1,17 @@ + +<contents> +1. Requirements +2. Required actions +2.1 Fix DS-MRR/InnoDB bugs +2.2 Backport DS-MRR code to MariaDB 5.2 +2.3 Introduce control variables +2.4 Other backport issues +2.5 Make MRR code more of a module +2.6 Improve the cost model +2.7 Let DS-MRR support clustered primary keys +</contents> + + 1. Requirements =============== @@ -63,4 +77,28 @@ and the only modularization we see is to move #1 into a separate file which won't achieve much. +2.6 Improve the cost model +-------------------------- +At the moment DS-MRR cost formula re-uses non-MRR scan costs, which uses +records_in_range() calls, followed by index_only_read_time() or read_time() +calls to produce the estimate for read cost. + + We should change this (TODO sort out how exactly) + +Note: this means that the query plans will change from MariaDB 5.2. + +2.7 Let DS-MRR support clustered primary keys +--------------------------------------------- +At the moment DS-MRR is not supported for clustered primary keys. It is not +needed when MRR is used for range access, because range access is done over +an ordered list of ranges, but it is useful for BKA. + +TODO: + it's useful for BKA because BKA makes MRR scans over un-orderered + non-disjoint lists of ranges. Then we can sort these and do ordered scans. + There is still no use for DS-MRR over clustered primary key for range + access, where the ranges are disjoint and ordered. + How about postponing this item until BKA is backported? + + -=-=(Guest - Thu, 26 Nov 2009, 16:52)=-=- High-Level Specification modified. --- /tmp/wklog.67.old.694 2009-11-26 14:52:53.000000000 +0000 +++ /tmp/wklog.67.new.694 2009-11-26 14:52:53.000000000 +0000 @@ -1 +1,66 @@ +1. Requirements +=============== + +We need the following: + +1. Latest MRR interface support, including extensions to support ICP when + using BKA. +2. Let DS-MRR support clustered primary keys (needed when using BKA). +3. Remove conditions used for key access from the condition pushed to index + (ATM this manifests itself as "Using index condition" appearing where there + was no "Using where". TODO: example of this?) +4. Introduce a separate @@optimizer_switch flag for turning on/out ICP (atm it + is switched on/off by @@engine_condition_pushdown) +5. Introduce a separate @@mrr_buffer_size variable to control MRR buffer size + for range+MRR scans. ATM it is controlled by @@read_rnd_size flag and that + makes it unobvious for a number of users. +6. Rename multi_range_read_info_const() to look like it is not a part of MRR + interface. +8. Try to make MRR to be more of a module +7. Improve MRR's cost model. + +2. Required actions +=================== + +Roughly in the order in which it will be done: + +2.1 Fix DS-MRR/InnoDB bugs +-------------------------- +We need to fix the bugs listed here: + +http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=index_condition_pushdown +http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=mrr + +2.2 Backport DS-MRR code to MariaDB 5.2 +--------------------------------------- +The easiest way seems to be to to manually move the needed code from mysql-6.0 +(or whatever it's called now) to MariaDB. + +2.3 Introduce control variables +------------------------------- +Act on items #4 and #5 from the requirements. Should be easy as +@@optimizer_switch is supported in 5.1 codebase. + +2.4 Other backport issues +------------------------- +* Figure out what to do with NDB/MRR. 5.1 codebase has "old" NDB/MRR + implementation. mysql-6.0 (and NDB's branch) have the updated NDB/MRR + but merging it into 5.1 can be very labor-intensive. + Will it be ok to disable NDB/MRR altogether? + + +2.5 Make MRR code more of a module +---------------------------------- +Some code in handler.cc can be moved to separate file. +But changes in opt_range.cc can't. +TODO: Sort out how much we really can do here. Initial guess is not much as the +code consists of: +- Default MRR implementation in handler.cc +- Changes in opt_range.cc to use MRR instead of multiple records_in_range() + calls. These rely on opt_range.cc's internal structures like SEL_ARG trees and + so there is not much point in moving them out. +- DS-MRR implementations which are spread over storage engines. +and the only modularization we see is to move #1 into a separate file which +won't achieve much. + DESCRIPTION: Backport DS-MRR into MariaDB-5.2 codebase, also adding certain extra features to make it more usable. HIGH-LEVEL SPECIFICATION: <contents> 1. Requirements 2. Required actions 2.1 Fix DS-MRR/InnoDB bugs 2.2 Backport DS-MRR code to MariaDB 5.2 2.3 Introduce control variables 2.4 Other backport issues 2.5 Make MRR code more of a module 2.6 Improve the cost model 2.7 Let DS-MRR support clustered primary keys </contents> 1. Requirements =============== We need the following: 1. Latest MRR interface support, including extensions to support ICP when using BKA. 2. Let DS-MRR support clustered primary keys (needed when using BKA). 3. Remove conditions used for key access from the condition pushed to index (ATM this manifests itself as "Using index condition" appearing where there was no "Using where". TODO: example of this?) 4. Introduce a separate @@optimizer_switch flag for turning on/out ICP (atm it is switched on/off by @@engine_condition_pushdown) 5. Introduce a separate @@mrr_buffer_size variable to control MRR buffer size for range+MRR scans. ATM it is controlled by @@read_rnd_size flag and that makes it unobvious for a number of users. 6. Rename multi_range_read_info_const() to look like it is not a part of MRR interface. 8. Try to make MRR to be more of a module 7. Improve MRR's cost model. 2. Required actions =================== Roughly in the order in which it will be done: 2.1 Fix DS-MRR/InnoDB bugs -------------------------- We need to fix the bugs listed here: http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=index_condi… http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=mrr http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=icp 2.2 Backport DS-MRR code to MariaDB 5.2 --------------------------------------- The easiest way seems to be to to manually move the needed code from mysql-6.0 (or whatever it's called now) to MariaDB. 2.3 Introduce control variables ------------------------------- Act on items #4 and #5 from the requirements. Should be easy as @@optimizer_switch is supported in 5.1 codebase. 2.4 Other backport issues ------------------------- * Figure out what to do with NDB/MRR. 5.1 codebase has "old" NDB/MRR implementation. mysql-6.0 (and NDB's branch) have the updated NDB/MRR but merging it into 5.1 can be very labor-intensive. Will it be ok to disable NDB/MRR altogether? 2.5 Make MRR code more of a module ---------------------------------- It is not possible to make MRR to be a totally separate module, as its code consists of : - Default MRR implementation in handler.cc - Changes in opt_range.cc to use MRR instead of multiple records_in_range() calls. These rely on opt_range.cc's internal stuctures like SEL_ARG trees and so there is not much point in moving them out. - DS-MRR impelementations which are spread over storage engines. We'll try to modularize what we can: - Move out default MRR implementation from handler.cc - Move possible parts out of opt_range.cc into a separate file. 2.6 Improve the cost model -------------------------- At the moment DS-MRR cost formula re-uses non-MRR scan costs, which uses records_in_range() calls, followed by index_only_read_time() or read_time() calls to produce the estimate for read cost. We should change this (TODO sort out how exactly) Note: this means that the query plans will change from MariaDB 5.2. 2.7 Let DS-MRR support clustered primary keys --------------------------------------------- At the moment DS-MRR is not supported for clustered primary keys. It is not needed when MRR is used for range access, because range access is done over an ordered list of ranges, but it is useful for BKA. TODO: it's useful for BKA because BKA makes MRR scans over un-orderered non-disjoint lists of ranges. Then we can sort these and do ordered scans. There is still no use for DS-MRR over clustered primary key for range access, where the ranges are disjoint and ordered. How about postponing this item until BKA is backported? ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Psergey): MRR backport (67)
by worklog-noreply＠askmonty.org 26 Nov '09

26 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: MRR backport CREATION DATE..: Thu, 26 Nov 2009, 15:19 SUPERVISOR.....: Monty IMPLEMENTOR....: Psergey COPIES TO......: CATEGORY.......: Server-Sprint TASK ID........: 67 (http://askmonty.org/worklog/?tid=67) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Psergey - Thu, 26 Nov 2009, 20:21)=-=- High-Level Specification modified. --- /tmp/wklog.67.old.9329 2009-11-26 20:21:28.000000000 +0200 +++ /tmp/wklog.67.new.9329 2009-11-26 20:21:28.000000000 +0200 @@ -65,17 +65,19 @@ 2.5 Make MRR code more of a module ---------------------------------- -Some code in handler.cc can be moved to separate file. -But changes in opt_range.cc can't. -TODO: Sort out how much we really can do here. Initial guess is not much as the -code consists of: +It is not possible to make MRR to be a totally separate module, as its code +consists of : - Default MRR implementation in handler.cc - Changes in opt_range.cc to use MRR instead of multiple records_in_range() - calls. These rely on opt_range.cc's internal structures like SEL_ARG trees and + calls. These rely on opt_range.cc's internal stuctures like SEL_ARG trees and so there is not much point in moving them out. -- DS-MRR implementations which are spread over storage engines. -and the only modularization we see is to move #1 into a separate file which -won't achieve much. +- DS-MRR impelementations which are spread over storage engines. + +We'll try to modularize what we can: +- Move out default MRR implementation from handler.cc +- Move possible parts out of opt_range.cc into a separate file. + + 2.6 Improve the cost model -------------------------- -=-=(Psergey - Thu, 26 Nov 2009, 19:06)=-=- High-Level Specification modified. --- /tmp/wklog.67.old.6449 2009-11-26 19:06:04.000000000 +0200 +++ /tmp/wklog.67.new.6449 2009-11-26 19:06:04.000000000 +0200 @@ -1,4 +1,3 @@ - <contents> 1. Requirements 2. Required actions @@ -44,6 +43,7 @@ http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=index_condi… http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=mrr +http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=icp 2.2 Backport DS-MRR code to MariaDB 5.2 --------------------------------------- -=-=(Psergey - Thu, 26 Nov 2009, 18:15)=-=- High-Level Specification modified. --- /tmp/wklog.67.old.4161 2009-11-26 18:15:36.000000000 +0200 +++ /tmp/wklog.67.new.4161 2009-11-26 18:15:36.000000000 +0200 @@ -1,3 +1,17 @@ + +<contents> +1. Requirements +2. Required actions +2.1 Fix DS-MRR/InnoDB bugs +2.2 Backport DS-MRR code to MariaDB 5.2 +2.3 Introduce control variables +2.4 Other backport issues +2.5 Make MRR code more of a module +2.6 Improve the cost model +2.7 Let DS-MRR support clustered primary keys +</contents> + + 1. Requirements =============== @@ -63,4 +77,28 @@ and the only modularization we see is to move #1 into a separate file which won't achieve much. +2.6 Improve the cost model +-------------------------- +At the moment DS-MRR cost formula re-uses non-MRR scan costs, which uses +records_in_range() calls, followed by index_only_read_time() or read_time() +calls to produce the estimate for read cost. + + We should change this (TODO sort out how exactly) + +Note: this means that the query plans will change from MariaDB 5.2. + +2.7 Let DS-MRR support clustered primary keys +--------------------------------------------- +At the moment DS-MRR is not supported for clustered primary keys. It is not +needed when MRR is used for range access, because range access is done over +an ordered list of ranges, but it is useful for BKA. + +TODO: + it's useful for BKA because BKA makes MRR scans over un-orderered + non-disjoint lists of ranges. Then we can sort these and do ordered scans. + There is still no use for DS-MRR over clustered primary key for range + access, where the ranges are disjoint and ordered. + How about postponing this item until BKA is backported? + + -=-=(Guest - Thu, 26 Nov 2009, 16:52)=-=- High-Level Specification modified. --- /tmp/wklog.67.old.694 2009-11-26 14:52:53.000000000 +0000 +++ /tmp/wklog.67.new.694 2009-11-26 14:52:53.000000000 +0000 @@ -1 +1,66 @@ +1. Requirements +=============== + +We need the following: + +1. Latest MRR interface support, including extensions to support ICP when + using BKA. +2. Let DS-MRR support clustered primary keys (needed when using BKA). +3. Remove conditions used for key access from the condition pushed to index + (ATM this manifests itself as "Using index condition" appearing where there + was no "Using where". TODO: example of this?) +4. Introduce a separate @@optimizer_switch flag for turning on/out ICP (atm it + is switched on/off by @@engine_condition_pushdown) +5. Introduce a separate @@mrr_buffer_size variable to control MRR buffer size + for range+MRR scans. ATM it is controlled by @@read_rnd_size flag and that + makes it unobvious for a number of users. +6. Rename multi_range_read_info_const() to look like it is not a part of MRR + interface. +8. Try to make MRR to be more of a module +7. Improve MRR's cost model. + +2. Required actions +=================== + +Roughly in the order in which it will be done: + +2.1 Fix DS-MRR/InnoDB bugs +-------------------------- +We need to fix the bugs listed here: + +http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=index_condition_pushdown +http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=mrr + +2.2 Backport DS-MRR code to MariaDB 5.2 +--------------------------------------- +The easiest way seems to be to to manually move the needed code from mysql-6.0 +(or whatever it's called now) to MariaDB. + +2.3 Introduce control variables +------------------------------- +Act on items #4 and #5 from the requirements. Should be easy as +@@optimizer_switch is supported in 5.1 codebase. + +2.4 Other backport issues +------------------------- +* Figure out what to do with NDB/MRR. 5.1 codebase has "old" NDB/MRR + implementation. mysql-6.0 (and NDB's branch) have the updated NDB/MRR + but merging it into 5.1 can be very labor-intensive. + Will it be ok to disable NDB/MRR altogether? + + +2.5 Make MRR code more of a module +---------------------------------- +Some code in handler.cc can be moved to separate file. +But changes in opt_range.cc can't. +TODO: Sort out how much we really can do here. Initial guess is not much as the +code consists of: +- Default MRR implementation in handler.cc +- Changes in opt_range.cc to use MRR instead of multiple records_in_range() + calls. These rely on opt_range.cc's internal structures like SEL_ARG trees and + so there is not much point in moving them out. +- DS-MRR implementations which are spread over storage engines. +and the only modularization we see is to move #1 into a separate file which +won't achieve much. + DESCRIPTION: Backport DS-MRR into MariaDB-5.2 codebase, also adding certain extra features to make it more usable. HIGH-LEVEL SPECIFICATION: <contents> 1. Requirements 2. Required actions 2.1 Fix DS-MRR/InnoDB bugs 2.2 Backport DS-MRR code to MariaDB 5.2 2.3 Introduce control variables 2.4 Other backport issues 2.5 Make MRR code more of a module 2.6 Improve the cost model 2.7 Let DS-MRR support clustered primary keys </contents> 1. Requirements =============== We need the following: 1. Latest MRR interface support, including extensions to support ICP when using BKA. 2. Let DS-MRR support clustered primary keys (needed when using BKA). 3. Remove conditions used for key access from the condition pushed to index (ATM this manifests itself as "Using index condition" appearing where there was no "Using where". TODO: example of this?) 4. Introduce a separate @@optimizer_switch flag for turning on/out ICP (atm it is switched on/off by @@engine_condition_pushdown) 5. Introduce a separate @@mrr_buffer_size variable to control MRR buffer size for range+MRR scans. ATM it is controlled by @@read_rnd_size flag and that makes it unobvious for a number of users. 6. Rename multi_range_read_info_const() to look like it is not a part of MRR interface. 8. Try to make MRR to be more of a module 7. Improve MRR's cost model. 2. Required actions =================== Roughly in the order in which it will be done: 2.1 Fix DS-MRR/InnoDB bugs -------------------------- We need to fix the bugs listed here: http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=index_condi… http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=mrr http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=icp 2.2 Backport DS-MRR code to MariaDB 5.2 --------------------------------------- The easiest way seems to be to to manually move the needed code from mysql-6.0 (or whatever it's called now) to MariaDB. 2.3 Introduce control variables ------------------------------- Act on items #4 and #5 from the requirements. Should be easy as @@optimizer_switch is supported in 5.1 codebase. 2.4 Other backport issues ------------------------- * Figure out what to do with NDB/MRR. 5.1 codebase has "old" NDB/MRR implementation. mysql-6.0 (and NDB's branch) have the updated NDB/MRR but merging it into 5.1 can be very labor-intensive. Will it be ok to disable NDB/MRR altogether? 2.5 Make MRR code more of a module ---------------------------------- It is not possible to make MRR to be a totally separate module, as its code consists of : - Default MRR implementation in handler.cc - Changes in opt_range.cc to use MRR instead of multiple records_in_range() calls. These rely on opt_range.cc's internal stuctures like SEL_ARG trees and so there is not much point in moving them out. - DS-MRR impelementations which are spread over storage engines. We'll try to modularize what we can: - Move out default MRR implementation from handler.cc - Move possible parts out of opt_range.cc into a separate file. 2.6 Improve the cost model -------------------------- At the moment DS-MRR cost formula re-uses non-MRR scan costs, which uses records_in_range() calls, followed by index_only_read_time() or read_time() calls to produce the estimate for read cost. We should change this (TODO sort out how exactly) Note: this means that the query plans will change from MariaDB 5.2. 2.7 Let DS-MRR support clustered primary keys --------------------------------------------- At the moment DS-MRR is not supported for clustered primary keys. It is not needed when MRR is used for range access, because range access is done over an ordered list of ranges, but it is useful for BKA. TODO: it's useful for BKA because BKA makes MRR scans over un-orderered non-disjoint lists of ranges. Then we can sort these and do ordered scans. There is still no use for DS-MRR over clustered primary key for range access, where the ranges are disjoint and ordered. How about postponing this item until BKA is backported? ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Psergey): MRR backport (67)
by worklog-noreply＠askmonty.org 26 Nov '09

26 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: MRR backport CREATION DATE..: Thu, 26 Nov 2009, 15:19 SUPERVISOR.....: Monty IMPLEMENTOR....: Psergey COPIES TO......: CATEGORY.......: Server-Sprint TASK ID........: 67 (http://askmonty.org/worklog/?tid=67) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Psergey - Thu, 26 Nov 2009, 19:06)=-=- High-Level Specification modified. --- /tmp/wklog.67.old.6449 2009-11-26 19:06:04.000000000 +0200 +++ /tmp/wklog.67.new.6449 2009-11-26 19:06:04.000000000 +0200 @@ -1,4 +1,3 @@ - <contents> 1. Requirements 2. Required actions @@ -44,6 +43,7 @@ http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=index_condi… http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=mrr +http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=icp 2.2 Backport DS-MRR code to MariaDB 5.2 --------------------------------------- -=-=(Psergey - Thu, 26 Nov 2009, 18:15)=-=- High-Level Specification modified. --- /tmp/wklog.67.old.4161 2009-11-26 18:15:36.000000000 +0200 +++ /tmp/wklog.67.new.4161 2009-11-26 18:15:36.000000000 +0200 @@ -1,3 +1,17 @@ + +<contents> +1. Requirements +2. Required actions +2.1 Fix DS-MRR/InnoDB bugs +2.2 Backport DS-MRR code to MariaDB 5.2 +2.3 Introduce control variables +2.4 Other backport issues +2.5 Make MRR code more of a module +2.6 Improve the cost model +2.7 Let DS-MRR support clustered primary keys +</contents> + + 1. Requirements =============== @@ -63,4 +77,28 @@ and the only modularization we see is to move #1 into a separate file which won't achieve much. +2.6 Improve the cost model +-------------------------- +At the moment DS-MRR cost formula re-uses non-MRR scan costs, which uses +records_in_range() calls, followed by index_only_read_time() or read_time() +calls to produce the estimate for read cost. + + We should change this (TODO sort out how exactly) + +Note: this means that the query plans will change from MariaDB 5.2. + +2.7 Let DS-MRR support clustered primary keys +--------------------------------------------- +At the moment DS-MRR is not supported for clustered primary keys. It is not +needed when MRR is used for range access, because range access is done over +an ordered list of ranges, but it is useful for BKA. + +TODO: + it's useful for BKA because BKA makes MRR scans over un-orderered + non-disjoint lists of ranges. Then we can sort these and do ordered scans. + There is still no use for DS-MRR over clustered primary key for range + access, where the ranges are disjoint and ordered. + How about postponing this item until BKA is backported? + + -=-=(Guest - Thu, 26 Nov 2009, 16:52)=-=- High-Level Specification modified. --- /tmp/wklog.67.old.694 2009-11-26 14:52:53.000000000 +0000 +++ /tmp/wklog.67.new.694 2009-11-26 14:52:53.000000000 +0000 @@ -1 +1,66 @@ +1. Requirements +=============== + +We need the following: + +1. Latest MRR interface support, including extensions to support ICP when + using BKA. +2. Let DS-MRR support clustered primary keys (needed when using BKA). +3. Remove conditions used for key access from the condition pushed to index + (ATM this manifests itself as "Using index condition" appearing where there + was no "Using where". TODO: example of this?) +4. Introduce a separate @@optimizer_switch flag for turning on/out ICP (atm it + is switched on/off by @@engine_condition_pushdown) +5. Introduce a separate @@mrr_buffer_size variable to control MRR buffer size + for range+MRR scans. ATM it is controlled by @@read_rnd_size flag and that + makes it unobvious for a number of users. +6. Rename multi_range_read_info_const() to look like it is not a part of MRR + interface. +8. Try to make MRR to be more of a module +7. Improve MRR's cost model. + +2. Required actions +=================== + +Roughly in the order in which it will be done: + +2.1 Fix DS-MRR/InnoDB bugs +-------------------------- +We need to fix the bugs listed here: + +http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=index_condition_pushdown +http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=mrr + +2.2 Backport DS-MRR code to MariaDB 5.2 +--------------------------------------- +The easiest way seems to be to to manually move the needed code from mysql-6.0 +(or whatever it's called now) to MariaDB. + +2.3 Introduce control variables +------------------------------- +Act on items #4 and #5 from the requirements. Should be easy as +@@optimizer_switch is supported in 5.1 codebase. + +2.4 Other backport issues +------------------------- +* Figure out what to do with NDB/MRR. 5.1 codebase has "old" NDB/MRR + implementation. mysql-6.0 (and NDB's branch) have the updated NDB/MRR + but merging it into 5.1 can be very labor-intensive. + Will it be ok to disable NDB/MRR altogether? + + +2.5 Make MRR code more of a module +---------------------------------- +Some code in handler.cc can be moved to separate file. +But changes in opt_range.cc can't. +TODO: Sort out how much we really can do here. Initial guess is not much as the +code consists of: +- Default MRR implementation in handler.cc +- Changes in opt_range.cc to use MRR instead of multiple records_in_range() + calls. These rely on opt_range.cc's internal structures like SEL_ARG trees and + so there is not much point in moving them out. +- DS-MRR implementations which are spread over storage engines. +and the only modularization we see is to move #1 into a separate file which +won't achieve much. + DESCRIPTION: Backport DS-MRR into MariaDB-5.2 codebase, also adding certain extra features to make it more usable. HIGH-LEVEL SPECIFICATION: <contents> 1. Requirements 2. Required actions 2.1 Fix DS-MRR/InnoDB bugs 2.2 Backport DS-MRR code to MariaDB 5.2 2.3 Introduce control variables 2.4 Other backport issues 2.5 Make MRR code more of a module 2.6 Improve the cost model 2.7 Let DS-MRR support clustered primary keys </contents> 1. Requirements =============== We need the following: 1. Latest MRR interface support, including extensions to support ICP when using BKA. 2. Let DS-MRR support clustered primary keys (needed when using BKA). 3. Remove conditions used for key access from the condition pushed to index (ATM this manifests itself as "Using index condition" appearing where there was no "Using where". TODO: example of this?) 4. Introduce a separate @@optimizer_switch flag for turning on/out ICP (atm it is switched on/off by @@engine_condition_pushdown) 5. Introduce a separate @@mrr_buffer_size variable to control MRR buffer size for range+MRR scans. ATM it is controlled by @@read_rnd_size flag and that makes it unobvious for a number of users. 6. Rename multi_range_read_info_const() to look like it is not a part of MRR interface. 8. Try to make MRR to be more of a module 7. Improve MRR's cost model. 2. Required actions =================== Roughly in the order in which it will be done: 2.1 Fix DS-MRR/InnoDB bugs -------------------------- We need to fix the bugs listed here: http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=index_condi… http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=mrr http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=icp 2.2 Backport DS-MRR code to MariaDB 5.2 --------------------------------------- The easiest way seems to be to to manually move the needed code from mysql-6.0 (or whatever it's called now) to MariaDB. 2.3 Introduce control variables ------------------------------- Act on items #4 and #5 from the requirements. Should be easy as @@optimizer_switch is supported in 5.1 codebase. 2.4 Other backport issues ------------------------- * Figure out what to do with NDB/MRR. 5.1 codebase has "old" NDB/MRR implementation. mysql-6.0 (and NDB's branch) have the updated NDB/MRR but merging it into 5.1 can be very labor-intensive. Will it be ok to disable NDB/MRR altogether? 2.5 Make MRR code more of a module ---------------------------------- Some code in handler.cc can be moved to separate file. But changes in opt_range.cc can't. TODO: Sort out how much we really can do here. Initial guess is not much as the code consists of: - Default MRR implementation in handler.cc - Changes in opt_range.cc to use MRR instead of multiple records_in_range() calls. These rely on opt_range.cc's internal structures like SEL_ARG trees and so there is not much point in moving them out. - DS-MRR implementations which are spread over storage engines. and the only modularization we see is to move #1 into a separate file which won't achieve much. 2.6 Improve the cost model -------------------------- At the moment DS-MRR cost formula re-uses non-MRR scan costs, which uses records_in_range() calls, followed by index_only_read_time() or read_time() calls to produce the estimate for read cost. We should change this (TODO sort out how exactly) Note: this means that the query plans will change from MariaDB 5.2. 2.7 Let DS-MRR support clustered primary keys --------------------------------------------- At the moment DS-MRR is not supported for clustered primary keys. It is not needed when MRR is used for range access, because range access is done over an ordered list of ranges, but it is useful for BKA. TODO: it's useful for BKA because BKA makes MRR scans over un-orderered non-disjoint lists of ranges. Then we can sort these and do ordered scans. There is still no use for DS-MRR over clustered primary key for range access, where the ranges are disjoint and ordered. How about postponing this item until BKA is backported? ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Psergey): MRR backport (67)
by worklog-noreply＠askmonty.org 26 Nov '09

26 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: MRR backport CREATION DATE..: Thu, 26 Nov 2009, 15:19 SUPERVISOR.....: Monty IMPLEMENTOR....: Psergey COPIES TO......: CATEGORY.......: Server-Sprint TASK ID........: 67 (http://askmonty.org/worklog/?tid=67) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Psergey - Thu, 26 Nov 2009, 19:06)=-=- High-Level Specification modified. --- /tmp/wklog.67.old.6449 2009-11-26 19:06:04.000000000 +0200 +++ /tmp/wklog.67.new.6449 2009-11-26 19:06:04.000000000 +0200 @@ -1,4 +1,3 @@ - <contents> 1. Requirements 2. Required actions @@ -44,6 +43,7 @@ http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=index_condi… http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=mrr +http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=icp 2.2 Backport DS-MRR code to MariaDB 5.2 --------------------------------------- -=-=(Psergey - Thu, 26 Nov 2009, 18:15)=-=- High-Level Specification modified. --- /tmp/wklog.67.old.4161 2009-11-26 18:15:36.000000000 +0200 +++ /tmp/wklog.67.new.4161 2009-11-26 18:15:36.000000000 +0200 @@ -1,3 +1,17 @@ + +<contents> +1. Requirements +2. Required actions +2.1 Fix DS-MRR/InnoDB bugs +2.2 Backport DS-MRR code to MariaDB 5.2 +2.3 Introduce control variables +2.4 Other backport issues +2.5 Make MRR code more of a module +2.6 Improve the cost model +2.7 Let DS-MRR support clustered primary keys +</contents> + + 1. Requirements =============== @@ -63,4 +77,28 @@ and the only modularization we see is to move #1 into a separate file which won't achieve much. +2.6 Improve the cost model +-------------------------- +At the moment DS-MRR cost formula re-uses non-MRR scan costs, which uses +records_in_range() calls, followed by index_only_read_time() or read_time() +calls to produce the estimate for read cost. + + We should change this (TODO sort out how exactly) + +Note: this means that the query plans will change from MariaDB 5.2. + +2.7 Let DS-MRR support clustered primary keys +--------------------------------------------- +At the moment DS-MRR is not supported for clustered primary keys. It is not +needed when MRR is used for range access, because range access is done over +an ordered list of ranges, but it is useful for BKA. + +TODO: + it's useful for BKA because BKA makes MRR scans over un-orderered + non-disjoint lists of ranges. Then we can sort these and do ordered scans. + There is still no use for DS-MRR over clustered primary key for range + access, where the ranges are disjoint and ordered. + How about postponing this item until BKA is backported? + + -=-=(Guest - Thu, 26 Nov 2009, 16:52)=-=- High-Level Specification modified. --- /tmp/wklog.67.old.694 2009-11-26 14:52:53.000000000 +0000 +++ /tmp/wklog.67.new.694 2009-11-26 14:52:53.000000000 +0000 @@ -1 +1,66 @@ +1. Requirements +=============== + +We need the following: + +1. Latest MRR interface support, including extensions to support ICP when + using BKA. +2. Let DS-MRR support clustered primary keys (needed when using BKA). +3. Remove conditions used for key access from the condition pushed to index + (ATM this manifests itself as "Using index condition" appearing where there + was no "Using where". TODO: example of this?) +4. Introduce a separate @@optimizer_switch flag for turning on/out ICP (atm it + is switched on/off by @@engine_condition_pushdown) +5. Introduce a separate @@mrr_buffer_size variable to control MRR buffer size + for range+MRR scans. ATM it is controlled by @@read_rnd_size flag and that + makes it unobvious for a number of users. +6. Rename multi_range_read_info_const() to look like it is not a part of MRR + interface. +8. Try to make MRR to be more of a module +7. Improve MRR's cost model. + +2. Required actions +=================== + +Roughly in the order in which it will be done: + +2.1 Fix DS-MRR/InnoDB bugs +-------------------------- +We need to fix the bugs listed here: + +http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=index_condition_pushdown +http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=mrr + +2.2 Backport DS-MRR code to MariaDB 5.2 +--------------------------------------- +The easiest way seems to be to to manually move the needed code from mysql-6.0 +(or whatever it's called now) to MariaDB. + +2.3 Introduce control variables +------------------------------- +Act on items #4 and #5 from the requirements. Should be easy as +@@optimizer_switch is supported in 5.1 codebase. + +2.4 Other backport issues +------------------------- +* Figure out what to do with NDB/MRR. 5.1 codebase has "old" NDB/MRR + implementation. mysql-6.0 (and NDB's branch) have the updated NDB/MRR + but merging it into 5.1 can be very labor-intensive. + Will it be ok to disable NDB/MRR altogether? + + +2.5 Make MRR code more of a module +---------------------------------- +Some code in handler.cc can be moved to separate file. +But changes in opt_range.cc can't. +TODO: Sort out how much we really can do here. Initial guess is not much as the +code consists of: +- Default MRR implementation in handler.cc +- Changes in opt_range.cc to use MRR instead of multiple records_in_range() + calls. These rely on opt_range.cc's internal structures like SEL_ARG trees and + so there is not much point in moving them out. +- DS-MRR implementations which are spread over storage engines. +and the only modularization we see is to move #1 into a separate file which +won't achieve much. + DESCRIPTION: Backport DS-MRR into MariaDB-5.2 codebase, also adding certain extra features to make it more usable. HIGH-LEVEL SPECIFICATION: <contents> 1. Requirements 2. Required actions 2.1 Fix DS-MRR/InnoDB bugs 2.2 Backport DS-MRR code to MariaDB 5.2 2.3 Introduce control variables 2.4 Other backport issues 2.5 Make MRR code more of a module 2.6 Improve the cost model 2.7 Let DS-MRR support clustered primary keys </contents> 1. Requirements =============== We need the following: 1. Latest MRR interface support, including extensions to support ICP when using BKA. 2. Let DS-MRR support clustered primary keys (needed when using BKA). 3. Remove conditions used for key access from the condition pushed to index (ATM this manifests itself as "Using index condition" appearing where there was no "Using where". TODO: example of this?) 4. Introduce a separate @@optimizer_switch flag for turning on/out ICP (atm it is switched on/off by @@engine_condition_pushdown) 5. Introduce a separate @@mrr_buffer_size variable to control MRR buffer size for range+MRR scans. ATM it is controlled by @@read_rnd_size flag and that makes it unobvious for a number of users. 6. Rename multi_range_read_info_const() to look like it is not a part of MRR interface. 8. Try to make MRR to be more of a module 7. Improve MRR's cost model. 2. Required actions =================== Roughly in the order in which it will be done: 2.1 Fix DS-MRR/InnoDB bugs -------------------------- We need to fix the bugs listed here: http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=index_condi… http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=mrr http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=icp 2.2 Backport DS-MRR code to MariaDB 5.2 --------------------------------------- The easiest way seems to be to to manually move the needed code from mysql-6.0 (or whatever it's called now) to MariaDB. 2.3 Introduce control variables ------------------------------- Act on items #4 and #5 from the requirements. Should be easy as @@optimizer_switch is supported in 5.1 codebase. 2.4 Other backport issues ------------------------- * Figure out what to do with NDB/MRR. 5.1 codebase has "old" NDB/MRR implementation. mysql-6.0 (and NDB's branch) have the updated NDB/MRR but merging it into 5.1 can be very labor-intensive. Will it be ok to disable NDB/MRR altogether? 2.5 Make MRR code more of a module ---------------------------------- Some code in handler.cc can be moved to separate file. But changes in opt_range.cc can't. TODO: Sort out how much we really can do here. Initial guess is not much as the code consists of: - Default MRR implementation in handler.cc - Changes in opt_range.cc to use MRR instead of multiple records_in_range() calls. These rely on opt_range.cc's internal structures like SEL_ARG trees and so there is not much point in moving them out. - DS-MRR implementations which are spread over storage engines. and the only modularization we see is to move #1 into a separate file which won't achieve much. 2.6 Improve the cost model -------------------------- At the moment DS-MRR cost formula re-uses non-MRR scan costs, which uses records_in_range() calls, followed by index_only_read_time() or read_time() calls to produce the estimate for read cost. We should change this (TODO sort out how exactly) Note: this means that the query plans will change from MariaDB 5.2. 2.7 Let DS-MRR support clustered primary keys --------------------------------------------- At the moment DS-MRR is not supported for clustered primary keys. It is not needed when MRR is used for range access, because range access is done over an ordered list of ranges, but it is useful for BKA. TODO: it's useful for BKA because BKA makes MRR scans over un-orderered non-disjoint lists of ranges. Then we can sort these and do ordered scans. There is still no use for DS-MRR over clustered primary key for range access, where the ranges are disjoint and ordered. How about postponing this item until BKA is backported? ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Psergey): MRR backport (67)
by worklog-noreply＠askmonty.org 26 Nov '09

26 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: MRR backport CREATION DATE..: Thu, 26 Nov 2009, 15:19 SUPERVISOR.....: Monty IMPLEMENTOR....: Psergey COPIES TO......: CATEGORY.......: Server-Sprint TASK ID........: 67 (http://askmonty.org/worklog/?tid=67) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Psergey - Thu, 26 Nov 2009, 18:15)=-=- High-Level Specification modified. --- /tmp/wklog.67.old.4161 2009-11-26 18:15:36.000000000 +0200 +++ /tmp/wklog.67.new.4161 2009-11-26 18:15:36.000000000 +0200 @@ -1,3 +1,17 @@ + +<contents> +1. Requirements +2. Required actions +2.1 Fix DS-MRR/InnoDB bugs +2.2 Backport DS-MRR code to MariaDB 5.2 +2.3 Introduce control variables +2.4 Other backport issues +2.5 Make MRR code more of a module +2.6 Improve the cost model +2.7 Let DS-MRR support clustered primary keys +</contents> + + 1. Requirements =============== @@ -63,4 +77,28 @@ and the only modularization we see is to move #1 into a separate file which won't achieve much. +2.6 Improve the cost model +-------------------------- +At the moment DS-MRR cost formula re-uses non-MRR scan costs, which uses +records_in_range() calls, followed by index_only_read_time() or read_time() +calls to produce the estimate for read cost. + + We should change this (TODO sort out how exactly) + +Note: this means that the query plans will change from MariaDB 5.2. + +2.7 Let DS-MRR support clustered primary keys +--------------------------------------------- +At the moment DS-MRR is not supported for clustered primary keys. It is not +needed when MRR is used for range access, because range access is done over +an ordered list of ranges, but it is useful for BKA. + +TODO: + it's useful for BKA because BKA makes MRR scans over un-orderered + non-disjoint lists of ranges. Then we can sort these and do ordered scans. + There is still no use for DS-MRR over clustered primary key for range + access, where the ranges are disjoint and ordered. + How about postponing this item until BKA is backported? + + -=-=(Guest - Thu, 26 Nov 2009, 16:52)=-=- High-Level Specification modified. --- /tmp/wklog.67.old.694 2009-11-26 14:52:53.000000000 +0000 +++ /tmp/wklog.67.new.694 2009-11-26 14:52:53.000000000 +0000 @@ -1 +1,66 @@ +1. Requirements +=============== + +We need the following: + +1. Latest MRR interface support, including extensions to support ICP when + using BKA. +2. Let DS-MRR support clustered primary keys (needed when using BKA). +3. Remove conditions used for key access from the condition pushed to index + (ATM this manifests itself as "Using index condition" appearing where there + was no "Using where". TODO: example of this?) +4. Introduce a separate @@optimizer_switch flag for turning on/out ICP (atm it + is switched on/off by @@engine_condition_pushdown) +5. Introduce a separate @@mrr_buffer_size variable to control MRR buffer size + for range+MRR scans. ATM it is controlled by @@read_rnd_size flag and that + makes it unobvious for a number of users. +6. Rename multi_range_read_info_const() to look like it is not a part of MRR + interface. +8. Try to make MRR to be more of a module +7. Improve MRR's cost model. + +2. Required actions +=================== + +Roughly in the order in which it will be done: + +2.1 Fix DS-MRR/InnoDB bugs +-------------------------- +We need to fix the bugs listed here: + +http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=index_condition_pushdown +http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=mrr + +2.2 Backport DS-MRR code to MariaDB 5.2 +--------------------------------------- +The easiest way seems to be to to manually move the needed code from mysql-6.0 +(or whatever it's called now) to MariaDB. + +2.3 Introduce control variables +------------------------------- +Act on items #4 and #5 from the requirements. Should be easy as +@@optimizer_switch is supported in 5.1 codebase. + +2.4 Other backport issues +------------------------- +* Figure out what to do with NDB/MRR. 5.1 codebase has "old" NDB/MRR + implementation. mysql-6.0 (and NDB's branch) have the updated NDB/MRR + but merging it into 5.1 can be very labor-intensive. + Will it be ok to disable NDB/MRR altogether? + + +2.5 Make MRR code more of a module +---------------------------------- +Some code in handler.cc can be moved to separate file. +But changes in opt_range.cc can't. +TODO: Sort out how much we really can do here. Initial guess is not much as the +code consists of: +- Default MRR implementation in handler.cc +- Changes in opt_range.cc to use MRR instead of multiple records_in_range() + calls. These rely on opt_range.cc's internal structures like SEL_ARG trees and + so there is not much point in moving them out. +- DS-MRR implementations which are spread over storage engines. +and the only modularization we see is to move #1 into a separate file which +won't achieve much. + DESCRIPTION: Backport DS-MRR into MariaDB-5.2 codebase, also adding certain extra features to make it more usable. HIGH-LEVEL SPECIFICATION: <contents> 1. Requirements 2. Required actions 2.1 Fix DS-MRR/InnoDB bugs 2.2 Backport DS-MRR code to MariaDB 5.2 2.3 Introduce control variables 2.4 Other backport issues 2.5 Make MRR code more of a module 2.6 Improve the cost model 2.7 Let DS-MRR support clustered primary keys </contents> 1. Requirements =============== We need the following: 1. Latest MRR interface support, including extensions to support ICP when using BKA. 2. Let DS-MRR support clustered primary keys (needed when using BKA). 3. Remove conditions used for key access from the condition pushed to index (ATM this manifests itself as "Using index condition" appearing where there was no "Using where". TODO: example of this?) 4. Introduce a separate @@optimizer_switch flag for turning on/out ICP (atm it is switched on/off by @@engine_condition_pushdown) 5. Introduce a separate @@mrr_buffer_size variable to control MRR buffer size for range+MRR scans. ATM it is controlled by @@read_rnd_size flag and that makes it unobvious for a number of users. 6. Rename multi_range_read_info_const() to look like it is not a part of MRR interface. 8. Try to make MRR to be more of a module 7. Improve MRR's cost model. 2. Required actions =================== Roughly in the order in which it will be done: 2.1 Fix DS-MRR/InnoDB bugs -------------------------- We need to fix the bugs listed here: http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=index_condi… http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=mrr 2.2 Backport DS-MRR code to MariaDB 5.2 --------------------------------------- The easiest way seems to be to to manually move the needed code from mysql-6.0 (or whatever it's called now) to MariaDB. 2.3 Introduce control variables ------------------------------- Act on items #4 and #5 from the requirements. Should be easy as @@optimizer_switch is supported in 5.1 codebase. 2.4 Other backport issues ------------------------- * Figure out what to do with NDB/MRR. 5.1 codebase has "old" NDB/MRR implementation. mysql-6.0 (and NDB's branch) have the updated NDB/MRR but merging it into 5.1 can be very labor-intensive. Will it be ok to disable NDB/MRR altogether? 2.5 Make MRR code more of a module ---------------------------------- Some code in handler.cc can be moved to separate file. But changes in opt_range.cc can't. TODO: Sort out how much we really can do here. Initial guess is not much as the code consists of: - Default MRR implementation in handler.cc - Changes in opt_range.cc to use MRR instead of multiple records_in_range() calls. These rely on opt_range.cc's internal structures like SEL_ARG trees and so there is not much point in moving them out. - DS-MRR implementations which are spread over storage engines. and the only modularization we see is to move #1 into a separate file which won't achieve much. 2.6 Improve the cost model -------------------------- At the moment DS-MRR cost formula re-uses non-MRR scan costs, which uses records_in_range() calls, followed by index_only_read_time() or read_time() calls to produce the estimate for read cost. We should change this (TODO sort out how exactly) Note: this means that the query plans will change from MariaDB 5.2. 2.7 Let DS-MRR support clustered primary keys --------------------------------------------- At the moment DS-MRR is not supported for clustered primary keys. It is not needed when MRR is used for range access, because range access is done over an ordered list of ranges, but it is useful for BKA. TODO: it's useful for BKA because BKA makes MRR scans over un-orderered non-disjoint lists of ranges. Then we can sort these and do ordered scans. There is still no use for DS-MRR over clustered primary key for range access, where the ranges are disjoint and ordered. How about postponing this item until BKA is backported? ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Psergey): MRR backport (67)
by worklog-noreply＠askmonty.org 26 Nov '09

26 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: MRR backport CREATION DATE..: Thu, 26 Nov 2009, 15:19 SUPERVISOR.....: Monty IMPLEMENTOR....: Psergey COPIES TO......: CATEGORY.......: Server-Sprint TASK ID........: 67 (http://askmonty.org/worklog/?tid=67) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Psergey - Thu, 26 Nov 2009, 18:15)=-=- High-Level Specification modified. --- /tmp/wklog.67.old.4161 2009-11-26 18:15:36.000000000 +0200 +++ /tmp/wklog.67.new.4161 2009-11-26 18:15:36.000000000 +0200 @@ -1,3 +1,17 @@ + +<contents> +1. Requirements +2. Required actions +2.1 Fix DS-MRR/InnoDB bugs +2.2 Backport DS-MRR code to MariaDB 5.2 +2.3 Introduce control variables +2.4 Other backport issues +2.5 Make MRR code more of a module +2.6 Improve the cost model +2.7 Let DS-MRR support clustered primary keys +</contents> + + 1. Requirements =============== @@ -63,4 +77,28 @@ and the only modularization we see is to move #1 into a separate file which won't achieve much. +2.6 Improve the cost model +-------------------------- +At the moment DS-MRR cost formula re-uses non-MRR scan costs, which uses +records_in_range() calls, followed by index_only_read_time() or read_time() +calls to produce the estimate for read cost. + + We should change this (TODO sort out how exactly) + +Note: this means that the query plans will change from MariaDB 5.2. + +2.7 Let DS-MRR support clustered primary keys +--------------------------------------------- +At the moment DS-MRR is not supported for clustered primary keys. It is not +needed when MRR is used for range access, because range access is done over +an ordered list of ranges, but it is useful for BKA. + +TODO: + it's useful for BKA because BKA makes MRR scans over un-orderered + non-disjoint lists of ranges. Then we can sort these and do ordered scans. + There is still no use for DS-MRR over clustered primary key for range + access, where the ranges are disjoint and ordered. + How about postponing this item until BKA is backported? + + -=-=(Guest - Thu, 26 Nov 2009, 16:52)=-=- High-Level Specification modified. --- /tmp/wklog.67.old.694 2009-11-26 14:52:53.000000000 +0000 +++ /tmp/wklog.67.new.694 2009-11-26 14:52:53.000000000 +0000 @@ -1 +1,66 @@ +1. Requirements +=============== + +We need the following: + +1. Latest MRR interface support, including extensions to support ICP when + using BKA. +2. Let DS-MRR support clustered primary keys (needed when using BKA). +3. Remove conditions used for key access from the condition pushed to index + (ATM this manifests itself as "Using index condition" appearing where there + was no "Using where". TODO: example of this?) +4. Introduce a separate @@optimizer_switch flag for turning on/out ICP (atm it + is switched on/off by @@engine_condition_pushdown) +5. Introduce a separate @@mrr_buffer_size variable to control MRR buffer size + for range+MRR scans. ATM it is controlled by @@read_rnd_size flag and that + makes it unobvious for a number of users. +6. Rename multi_range_read_info_const() to look like it is not a part of MRR + interface. +8. Try to make MRR to be more of a module +7. Improve MRR's cost model. + +2. Required actions +=================== + +Roughly in the order in which it will be done: + +2.1 Fix DS-MRR/InnoDB bugs +-------------------------- +We need to fix the bugs listed here: + +http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=index_condition_pushdown +http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=mrr + +2.2 Backport DS-MRR code to MariaDB 5.2 +--------------------------------------- +The easiest way seems to be to to manually move the needed code from mysql-6.0 +(or whatever it's called now) to MariaDB. + +2.3 Introduce control variables +------------------------------- +Act on items #4 and #5 from the requirements. Should be easy as +@@optimizer_switch is supported in 5.1 codebase. + +2.4 Other backport issues +------------------------- +* Figure out what to do with NDB/MRR. 5.1 codebase has "old" NDB/MRR + implementation. mysql-6.0 (and NDB's branch) have the updated NDB/MRR + but merging it into 5.1 can be very labor-intensive. + Will it be ok to disable NDB/MRR altogether? + + +2.5 Make MRR code more of a module +---------------------------------- +Some code in handler.cc can be moved to separate file. +But changes in opt_range.cc can't. +TODO: Sort out how much we really can do here. Initial guess is not much as the +code consists of: +- Default MRR implementation in handler.cc +- Changes in opt_range.cc to use MRR instead of multiple records_in_range() + calls. These rely on opt_range.cc's internal structures like SEL_ARG trees and + so there is not much point in moving them out. +- DS-MRR implementations which are spread over storage engines. +and the only modularization we see is to move #1 into a separate file which +won't achieve much. + DESCRIPTION: Backport DS-MRR into MariaDB-5.2 codebase, also adding certain extra features to make it more usable. HIGH-LEVEL SPECIFICATION: <contents> 1. Requirements 2. Required actions 2.1 Fix DS-MRR/InnoDB bugs 2.2 Backport DS-MRR code to MariaDB 5.2 2.3 Introduce control variables 2.4 Other backport issues 2.5 Make MRR code more of a module 2.6 Improve the cost model 2.7 Let DS-MRR support clustered primary keys </contents> 1. Requirements =============== We need the following: 1. Latest MRR interface support, including extensions to support ICP when using BKA. 2. Let DS-MRR support clustered primary keys (needed when using BKA). 3. Remove conditions used for key access from the condition pushed to index (ATM this manifests itself as "Using index condition" appearing where there was no "Using where". TODO: example of this?) 4. Introduce a separate @@optimizer_switch flag for turning on/out ICP (atm it is switched on/off by @@engine_condition_pushdown) 5. Introduce a separate @@mrr_buffer_size variable to control MRR buffer size for range+MRR scans. ATM it is controlled by @@read_rnd_size flag and that makes it unobvious for a number of users. 6. Rename multi_range_read_info_const() to look like it is not a part of MRR interface. 8. Try to make MRR to be more of a module 7. Improve MRR's cost model. 2. Required actions =================== Roughly in the order in which it will be done: 2.1 Fix DS-MRR/InnoDB bugs -------------------------- We need to fix the bugs listed here: http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=index_condi… http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=mrr 2.2 Backport DS-MRR code to MariaDB 5.2 --------------------------------------- The easiest way seems to be to to manually move the needed code from mysql-6.0 (or whatever it's called now) to MariaDB. 2.3 Introduce control variables ------------------------------- Act on items #4 and #5 from the requirements. Should be easy as @@optimizer_switch is supported in 5.1 codebase. 2.4 Other backport issues ------------------------- * Figure out what to do with NDB/MRR. 5.1 codebase has "old" NDB/MRR implementation. mysql-6.0 (and NDB's branch) have the updated NDB/MRR but merging it into 5.1 can be very labor-intensive. Will it be ok to disable NDB/MRR altogether? 2.5 Make MRR code more of a module ---------------------------------- Some code in handler.cc can be moved to separate file. But changes in opt_range.cc can't. TODO: Sort out how much we really can do here. Initial guess is not much as the code consists of: - Default MRR implementation in handler.cc - Changes in opt_range.cc to use MRR instead of multiple records_in_range() calls. These rely on opt_range.cc's internal structures like SEL_ARG trees and so there is not much point in moving them out. - DS-MRR implementations which are spread over storage engines. and the only modularization we see is to move #1 into a separate file which won't achieve much. 2.6 Improve the cost model -------------------------- At the moment DS-MRR cost formula re-uses non-MRR scan costs, which uses records_in_range() calls, followed by index_only_read_time() or read_time() calls to produce the estimate for read cost. We should change this (TODO sort out how exactly) Note: this means that the query plans will change from MariaDB 5.2. 2.7 Let DS-MRR support clustered primary keys --------------------------------------------- At the moment DS-MRR is not supported for clustered primary keys. It is not needed when MRR is used for range access, because range access is done over an ordered list of ranges, but it is useful for BKA. TODO: it's useful for BKA because BKA makes MRR scans over un-orderered non-disjoint lists of ranges. Then we can sort these and do ordered scans. There is still no use for DS-MRR over clustered primary key for range access, where the ranges are disjoint and ordered. How about postponing this item until BKA is backported? ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (monty:2768) Bug#48357
by Michael Widenius 26 Nov '09

26 Nov '09

#At lp:maria based on revid:monty@askmonty.org-20091125231823-ausq10s23qyih61x 2768 Michael Widenius 2009-11-26 Fix for Bug #48357 SHOW BINLOG EVENTS: Wrong offset or I/O error modified: sql/log_event.cc per-file messages: sql/log_event.cc gcc 4.4.1 assumes that variables that you cast away will not change (strict-aliasing) The symptom was that mysql-test-run binlog.binglog_database got errors in the log === modified file 'sql/log_event.cc' --- a/sql/log_event.cc 2009-10-15 21:38:29 +0000 +++ b/sql/log_event.cc 2009-11-26 15:57:05 +0000 @@ -2133,7 +2133,7 @@ void Query_log_event::pack_info(Protocol /** Utility function for the next method (Query_log_event::write()) . */ -static void write_str_with_code_and_len(char **dst, const char *src, +static void write_str_with_code_and_len(uchar **dst, const char *src, int len, uint code) { /* @@ -2143,7 +2143,7 @@ static void write_str_with_code_and_len( DBUG_ASSERT(len <= 255); DBUG_ASSERT(src); *((*dst)++)= code; - *((*dst)++)= (uchar) len; + *((*dst)++)= len; bmove(*dst, src, len); (*dst)+= len; } @@ -2229,7 +2229,7 @@ bool Query_log_event::write(IO_CACHE* fi } if (catalog_len) // i.e. this var is inited (false for 4.0 events) { - write_str_with_code_and_len((char **)(&start), + write_str_with_code_and_len(&start, catalog, catalog_len, Q_CATALOG_NZ_CODE); /* In 5.0.x where x<4 masters we used to store the end zero here. This was @@ -2267,7 +2267,7 @@ bool Query_log_event::write(IO_CACHE* fi { /* In the TZ sys table, column Name is of length 64 so this should be ok */ DBUG_ASSERT(time_zone_len <= MAX_TIME_ZONE_NAME_LENGTH); - write_str_with_code_and_len((char **)(&start), + write_str_with_code_and_len(&start, time_zone_str, time_zone_len, Q_TIME_ZONE_CODE); } if (lc_time_names_number)

1 0

[Maria-developers] [Branch ~maria-captains/maria/5.1] Rev 2768: Apply the strict aliasing patch from http://lists.mysql.com/commits/89843
by noreply＠launchpad.net 26 Nov '09

26 Nov '09

------------------------------------------------------------ revno: 2768 committer: Bo Thorsen <bo.thorsen(a)canonical.com> branch nick: trunk-local timestamp: Thu 2009-11-26 16:32:49 +0100 message: Apply the strict aliasing patch from http://lists.mysql.com/commits/89843 modified: sql/log_event.cc -- lp:maria https://code.launchpad.net/~maria-captains/maria/5.1 Your team Maria developers is subscribed to branch lp:maria. To unsubscribe from this branch go to https://code.launchpad.net/~maria-captains/maria/5.1/+edit-subscription.

1 0

[Maria-developers] Updated (by Guest): MRR backport (67)
by worklog-noreply＠askmonty.org 26 Nov '09

26 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: MRR backport CREATION DATE..: Thu, 26 Nov 2009, 15:19 SUPERVISOR.....: Monty IMPLEMENTOR....: Psergey COPIES TO......: CATEGORY.......: Server-Sprint TASK ID........: 67 (http://askmonty.org/worklog/?tid=67) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Guest - Thu, 26 Nov 2009, 16:52)=-=- High-Level Specification modified. --- /tmp/wklog.67.old.694 2009-11-26 14:52:53.000000000 +0000 +++ /tmp/wklog.67.new.694 2009-11-26 14:52:53.000000000 +0000 @@ -1 +1,66 @@ +1. Requirements +=============== + +We need the following: + +1. Latest MRR interface support, including extensions to support ICP when + using BKA. +2. Let DS-MRR support clustered primary keys (needed when using BKA). +3. Remove conditions used for key access from the condition pushed to index + (ATM this manifests itself as "Using index condition" appearing where there + was no "Using where". TODO: example of this?) +4. Introduce a separate @@optimizer_switch flag for turning on/out ICP (atm it + is switched on/off by @@engine_condition_pushdown) +5. Introduce a separate @@mrr_buffer_size variable to control MRR buffer size + for range+MRR scans. ATM it is controlled by @@read_rnd_size flag and that + makes it unobvious for a number of users. +6. Rename multi_range_read_info_const() to look like it is not a part of MRR + interface. +8. Try to make MRR to be more of a module +7. Improve MRR's cost model. + +2. Required actions +=================== + +Roughly in the order in which it will be done: + +2.1 Fix DS-MRR/InnoDB bugs +-------------------------- +We need to fix the bugs listed here: + +http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=index_condition_pushdown +http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=mrr + +2.2 Backport DS-MRR code to MariaDB 5.2 +--------------------------------------- +The easiest way seems to be to to manually move the needed code from mysql-6.0 +(or whatever it's called now) to MariaDB. + +2.3 Introduce control variables +------------------------------- +Act on items #4 and #5 from the requirements. Should be easy as +@@optimizer_switch is supported in 5.1 codebase. + +2.4 Other backport issues +------------------------- +* Figure out what to do with NDB/MRR. 5.1 codebase has "old" NDB/MRR + implementation. mysql-6.0 (and NDB's branch) have the updated NDB/MRR + but merging it into 5.1 can be very labor-intensive. + Will it be ok to disable NDB/MRR altogether? + + +2.5 Make MRR code more of a module +---------------------------------- +Some code in handler.cc can be moved to separate file. +But changes in opt_range.cc can't. +TODO: Sort out how much we really can do here. Initial guess is not much as the +code consists of: +- Default MRR implementation in handler.cc +- Changes in opt_range.cc to use MRR instead of multiple records_in_range() + calls. These rely on opt_range.cc's internal structures like SEL_ARG trees and + so there is not much point in moving them out. +- DS-MRR implementations which are spread over storage engines. +and the only modularization we see is to move #1 into a separate file which +won't achieve much. + DESCRIPTION: Backport DS-MRR into MariaDB-5.2 codebase, also adding certain extra features to make it more usable. HIGH-LEVEL SPECIFICATION: 1. Requirements =============== We need the following: 1. Latest MRR interface support, including extensions to support ICP when using BKA. 2. Let DS-MRR support clustered primary keys (needed when using BKA). 3. Remove conditions used for key access from the condition pushed to index (ATM this manifests itself as "Using index condition" appearing where there was no "Using where". TODO: example of this?) 4. Introduce a separate @@optimizer_switch flag for turning on/out ICP (atm it is switched on/off by @@engine_condition_pushdown) 5. Introduce a separate @@mrr_buffer_size variable to control MRR buffer size for range+MRR scans. ATM it is controlled by @@read_rnd_size flag and that makes it unobvious for a number of users. 6. Rename multi_range_read_info_const() to look like it is not a part of MRR interface. 8. Try to make MRR to be more of a module 7. Improve MRR's cost model. 2. Required actions =================== Roughly in the order in which it will be done: 2.1 Fix DS-MRR/InnoDB bugs -------------------------- We need to fix the bugs listed here: http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=index_condi… http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=mrr 2.2 Backport DS-MRR code to MariaDB 5.2 --------------------------------------- The easiest way seems to be to to manually move the needed code from mysql-6.0 (or whatever it's called now) to MariaDB. 2.3 Introduce control variables ------------------------------- Act on items #4 and #5 from the requirements. Should be easy as @@optimizer_switch is supported in 5.1 codebase. 2.4 Other backport issues ------------------------- * Figure out what to do with NDB/MRR. 5.1 codebase has "old" NDB/MRR implementation. mysql-6.0 (and NDB's branch) have the updated NDB/MRR but merging it into 5.1 can be very labor-intensive. Will it be ok to disable NDB/MRR altogether? 2.5 Make MRR code more of a module ---------------------------------- Some code in handler.cc can be moved to separate file. But changes in opt_range.cc can't. TODO: Sort out how much we really can do here. Initial guess is not much as the code consists of: - Default MRR implementation in handler.cc - Changes in opt_range.cc to use MRR instead of multiple records_in_range() calls. These rely on opt_range.cc's internal structures like SEL_ARG trees and so there is not much point in moving them out. - DS-MRR implementations which are spread over storage engines. and the only modularization we see is to move #1 into a separate file which won't achieve much. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Guest): MRR backport (67)
by worklog-noreply＠askmonty.org 26 Nov '09

26 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: MRR backport CREATION DATE..: Thu, 26 Nov 2009, 15:19 SUPERVISOR.....: Monty IMPLEMENTOR....: Psergey COPIES TO......: CATEGORY.......: Server-Sprint TASK ID........: 67 (http://askmonty.org/worklog/?tid=67) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Guest - Thu, 26 Nov 2009, 16:52)=-=- High-Level Specification modified. --- /tmp/wklog.67.old.694 2009-11-26 14:52:53.000000000 +0000 +++ /tmp/wklog.67.new.694 2009-11-26 14:52:53.000000000 +0000 @@ -1 +1,66 @@ +1. Requirements +=============== + +We need the following: + +1. Latest MRR interface support, including extensions to support ICP when + using BKA. +2. Let DS-MRR support clustered primary keys (needed when using BKA). +3. Remove conditions used for key access from the condition pushed to index + (ATM this manifests itself as "Using index condition" appearing where there + was no "Using where". TODO: example of this?) +4. Introduce a separate @@optimizer_switch flag for turning on/out ICP (atm it + is switched on/off by @@engine_condition_pushdown) +5. Introduce a separate @@mrr_buffer_size variable to control MRR buffer size + for range+MRR scans. ATM it is controlled by @@read_rnd_size flag and that + makes it unobvious for a number of users. +6. Rename multi_range_read_info_const() to look like it is not a part of MRR + interface. +8. Try to make MRR to be more of a module +7. Improve MRR's cost model. + +2. Required actions +=================== + +Roughly in the order in which it will be done: + +2.1 Fix DS-MRR/InnoDB bugs +-------------------------- +We need to fix the bugs listed here: + +http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=index_condition_pushdown +http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=mrr + +2.2 Backport DS-MRR code to MariaDB 5.2 +--------------------------------------- +The easiest way seems to be to to manually move the needed code from mysql-6.0 +(or whatever it's called now) to MariaDB. + +2.3 Introduce control variables +------------------------------- +Act on items #4 and #5 from the requirements. Should be easy as +@@optimizer_switch is supported in 5.1 codebase. + +2.4 Other backport issues +------------------------- +* Figure out what to do with NDB/MRR. 5.1 codebase has "old" NDB/MRR + implementation. mysql-6.0 (and NDB's branch) have the updated NDB/MRR + but merging it into 5.1 can be very labor-intensive. + Will it be ok to disable NDB/MRR altogether? + + +2.5 Make MRR code more of a module +---------------------------------- +Some code in handler.cc can be moved to separate file. +But changes in opt_range.cc can't. +TODO: Sort out how much we really can do here. Initial guess is not much as the +code consists of: +- Default MRR implementation in handler.cc +- Changes in opt_range.cc to use MRR instead of multiple records_in_range() + calls. These rely on opt_range.cc's internal structures like SEL_ARG trees and + so there is not much point in moving them out. +- DS-MRR implementations which are spread over storage engines. +and the only modularization we see is to move #1 into a separate file which +won't achieve much. + DESCRIPTION: Backport DS-MRR into MariaDB-5.2 codebase, also adding certain extra features to make it more usable. HIGH-LEVEL SPECIFICATION: 1. Requirements =============== We need the following: 1. Latest MRR interface support, including extensions to support ICP when using BKA. 2. Let DS-MRR support clustered primary keys (needed when using BKA). 3. Remove conditions used for key access from the condition pushed to index (ATM this manifests itself as "Using index condition" appearing where there was no "Using where". TODO: example of this?) 4. Introduce a separate @@optimizer_switch flag for turning on/out ICP (atm it is switched on/off by @@engine_condition_pushdown) 5. Introduce a separate @@mrr_buffer_size variable to control MRR buffer size for range+MRR scans. ATM it is controlled by @@read_rnd_size flag and that makes it unobvious for a number of users. 6. Rename multi_range_read_info_const() to look like it is not a part of MRR interface. 8. Try to make MRR to be more of a module 7. Improve MRR's cost model. 2. Required actions =================== Roughly in the order in which it will be done: 2.1 Fix DS-MRR/InnoDB bugs -------------------------- We need to fix the bugs listed here: http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=index_condi… http://bugs.mysql.com/search.php?cmd=display&status=Active&tags=mrr 2.2 Backport DS-MRR code to MariaDB 5.2 --------------------------------------- The easiest way seems to be to to manually move the needed code from mysql-6.0 (or whatever it's called now) to MariaDB. 2.3 Introduce control variables ------------------------------- Act on items #4 and #5 from the requirements. Should be easy as @@optimizer_switch is supported in 5.1 codebase. 2.4 Other backport issues ------------------------- * Figure out what to do with NDB/MRR. 5.1 codebase has "old" NDB/MRR implementation. mysql-6.0 (and NDB's branch) have the updated NDB/MRR but merging it into 5.1 can be very labor-intensive. Will it be ok to disable NDB/MRR altogether? 2.5 Make MRR code more of a module ---------------------------------- Some code in handler.cc can be moved to separate file. But changes in opt_range.cc can't. TODO: Sort out how much we really can do here. Initial guess is not much as the code consists of: - Default MRR implementation in handler.cc - Changes in opt_range.cc to use MRR instead of multiple records_in_range() calls. These rely on opt_range.cc's internal structures like SEL_ARG trees and so there is not much point in moving them out. - DS-MRR implementations which are spread over storage engines. and the only modularization we see is to move #1 into a separate file which won't achieve much. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] New (by Psergey): MRR backport (67)
by worklog-noreply＠askmonty.org 26 Nov '09

26 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: MRR backport CREATION DATE..: Thu, 26 Nov 2009, 15:19 SUPERVISOR.....: Monty IMPLEMENTOR....: Psergey COPIES TO......: CATEGORY.......: Server-Sprint TASK ID........: 67 (http://askmonty.org/worklog/?tid=67) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: DESCRIPTION: Backport DS-MRR into MariaDB-5.2 codebase, also adding certain extra features to make it more usable. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] New (by Psergey): MRR backport (67)
by worklog-noreply＠askmonty.org 26 Nov '09

26 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: MRR backport CREATION DATE..: Thu, 26 Nov 2009, 15:19 SUPERVISOR.....: Monty IMPLEMENTOR....: Psergey COPIES TO......: CATEGORY.......: Server-Sprint TASK ID........: 67 (http://askmonty.org/worklog/?tid=67) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: DESCRIPTION: Backport DS-MRR into MariaDB-5.2 codebase, also adding certain extra features to make it more usable. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Help with memory leak in EXPLAIN
by Kristian Nielsen 26 Nov '09

26 Nov '09

Hi Sergey, I encountered a memory leak while working on fixing the MySQL 5.1.41 merge. The memory leak is repeatable with the following test case: ------------------------------ cut here ------------------------------ # # Bug#45989 memory leak after explain encounters an error in the query # CREATE TABLE t1(a LONGTEXT); INSERT INTO t1 VALUES (repeat('a',@@global.max_allowed_packet)); INSERT INTO t1 VALUES (repeat('b',@@global.max_allowed_packet)); --error ER_BAD_FIELD_ERROR EXPLAIN SELECT DISTINCT 1 FROM t1, (SELECT DISTINCTROW a AS away FROM t1 GROUP BY a WITH ROLLUP) as d1 WHERE t1.a = d1.a; DROP TABLE t1; ------------------------------ cut here ------------------------------ perl mysql-test-run.pl --valgrind kn.test Errors/warnings were found in logfiles during server shutdown after running the following sequence(s) of tests: main.kn ==14709== 1,048,584 bytes in 1 blocks are possibly lost in loss record 7 of 7 ==14709== at 0x4C22FAB: malloc (vg_replace_malloc.c:207) ==14709== by 0xB1DDD0: my_malloc (my_malloc.c:37) ==14709== by 0x6691C1: String::real_alloc(unsigned) (sql_string.cc:43) ==14709== by 0x5D7BFF: String::alloc(unsigned) (sql_string.h:210) ==14709== by 0x669835: String::copy(String const&) (sql_string.cc:192) ==14709== by 0x7E8BB0: do_save_blob(Copy_field*) (field_conv.cc:296) ==14709== by 0x7E6922: do_copy_null(Copy_field*) (field_conv.cc:207) ==14709== by 0x6EB28B: copy_fields(TMP_TABLE_PARAM*) (sql_select.cc:15374) ==14709== by 0x6F7F1B: end_write_group(JOIN*, st_join_table*, bool) (sql_select.cc:12715) ==14709== by 0x6F4415: evaluate_join_record(JOIN*, st_join_table*, int) (sql_select.cc:11514) ==14709== by 0x6F464E: sub_select(JOIN*, st_join_table*, bool) (sql_select.cc:11399) ==14709== by 0x705B1E: do_select(JOIN*, List<Item>*, st_table*, Procedure*) (sql_select.cc:11155) ==14709== by 0x7161C2: JOIN::exec() (sql_select.cc:1809) ==14709== by 0x7127B6: mysql_select(THD*, Item***, TABLE_LIST*, unsigned, List<Item>&, Item*, unsigned, st_order*, st_order*, Item*, st_order*, unsigned long long, select_result*, st_select_lex_unit*, st_select_lex*) (sql_select.cc:2425) ==14709== by 0x82EEEB: mysql_derived_filling(THD*, st_lex*, TABLE_LIST*) (sql_derived.cc:294) ==14709== by 0x82EC8B: mysql_handle_derived(st_lex*, bool (*)(THD*, st_lex*, TABLE_LIST*)) (sql_derived.cc:56) This bug is also in main lp:maria MariaDB trunk, and also in current MySQL (there is a MySQL bug, http://bugs.mysql.com/bug.php?id=45989) Could you take a look, and see if it is something you/we could fix, or if you have any other suggestions for how to deal with this? - Kristian.

2 1

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (knielsen:2775)
by knielsen＠knielsen-hq.org 26 Nov '09

26 Nov '09

#At lp:maria 2775 knielsen(a)knielsen-hq.org 2009-11-26 After-merge fix for MySQL 5.1.41 merge into MariaDB: Fix incorrect merge breaking the build. modified: unittest/mysys/Makefile.am === modified file 'unittest/mysys/Makefile.am' --- a/unittest/mysys/Makefile.am 2009-11-16 20:49:51 +0000 +++ b/unittest/mysys/Makefile.am 2009-11-26 08:44:38 +0000 @@ -18,15 +18,13 @@ INCLUDES = @ZLIB_INCLUDES@ -I$(top_buil noinst_HEADERS = thr_template.c -noinst_PROGRAMS = bitmap-t base64-t my_atomic-t lf-t waiting_threads-t - LDADD = $(top_builddir)/unittest/mytap/libmytap.a \ $(top_builddir)/mysys/libmysys.a \ $(top_builddir)/dbug/libdbug.a \ $(top_builddir)/strings/libmystrings.a EXTRA_DIST = CMakeLists.txt -noinst_PROGRAMS = bitmap-t base64-t +noinst_PROGRAMS = bitmap-t base64-t lf-t waiting_threads-t if NEED_THREAD # my_atomic-t is used to check thread functions, so it is safe to

1 0

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (knielsen:2774)
by knielsen＠knielsen-hq.org 26 Nov '09

26 Nov '09

#At lp:maria 2774 knielsen(a)knielsen-hq.org 2009-11-26 After-merge fixes for merge of MySQL 5.1.41 into MariaDB: more post-shutdown warning fixes. modified: mysql-test/mysql-test-run.pl === modified file 'mysql-test/mysql-test-run.pl' --- a/mysql-test/mysql-test-run.pl 2009-11-26 07:45:54 +0000 +++ b/mysql-test/mysql-test-run.pl 2009-11-26 08:02:33 +0000 @@ -3992,7 +3992,8 @@ sub extract_warning_lines ($) { qr/Setting lower_case_table_names=2/, qr/deprecated/, qr/Slave SQL thread retried transaction/, - qw/Slave $additional info$/, + qr/Slave $additional info$/, + qr/Incorrect information in file/, ); my $match_count= 0;

1 0

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (knielsen:2773)
by knielsen＠knielsen-hq.org 26 Nov '09

26 Nov '09

#At lp:maria 2773 knielsen(a)knielsen-hq.org 2009-11-26 After-merge fix for MySQL 5.1.41 merge into MariaDB: Another warning antipattern also needed during shutdown. modified: mysql-test/mysql-test-run.pl === modified file 'mysql-test/mysql-test-run.pl' --- a/mysql-test/mysql-test-run.pl 2009-11-25 11:06:01 +0000 +++ b/mysql-test/mysql-test-run.pl 2009-11-26 07:45:54 +0000 @@ -3989,6 +3989,7 @@ sub extract_warning_lines ($) { qr/unknown option '--loose-/, qr/unknown variable 'loose-/, qr/Now setting lower_case_table_names to [02]/, + qr/Setting lower_case_table_names=2/, qr/deprecated/, qr/Slave SQL thread retried transaction/, qw/Slave $additional info$/,

1 0

[Maria-developers] [Branch ~maria-captains/maria/5.1] Rev 2767: Fixed LPBUG#485443 --with-fast-mutexes and without safe mutexes (debug build) maria do not builds
by noreply＠launchpad.net 26 Nov '09

26 Nov '09

------------------------------------------------------------ revno: 2767 committer: Michael Widenius <monty(a)askmonty.org> branch nick: maria-5.1 timestamp: Thu 2009-11-26 01:18:23 +0200 message: Fixed LPBUG#485443 --with-fast-mutexes and without safe mutexes (debug build) maria do not builds Added 'mariadb_SERVER' as extra config group for MariaDB embedded server modified: client/mysql.cc mysys/thr_mutex.c -- lp:maria https://code.launchpad.net/~maria-captains/maria/5.1 Your team Maria developers is subscribed to branch lp:maria. To unsubscribe from this branch go to https://code.launchpad.net/~maria-captains/maria/5.1/+edit-subscription.

1 0

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (monty:2767)
by Michael Widenius 26 Nov '09

26 Nov '09

#At lp:maria based on revid:monty@askmonty.org-20091125073628-7fekfnr2hmkbjos1 2767 Michael Widenius 2009-11-26 Fixed LPBUG#485443 --with-fast-mutexes and without safe mutexes (debug build) maria do not builds Added 'mariadb_SERVER' as extra config group for MariaDB embedded server modified: client/mysql.cc mysys/thr_mutex.c per-file messages: client/mysql.cc Cleanup Added 'mariadb_SERVER' as extra config group for MariaDB embedded server mysys/thr_mutex.c Fixed LPBUG#485443 --with-fast-mutexes and without safe mutexes (debug build) maria do not builds === modified file 'client/mysql.cc' --- a/client/mysql.cc 2009-11-06 17:22:32 +0000 +++ b/client/mysql.cc 2009-11-25 23:18:23 +0000 @@ -83,7 +83,7 @@ extern "C" { #include <term.h> #endif #endif -#endif +#endif /* defined(HAVE_CURSES_H) && defined(HAVE_TERM_H) */ #undef bcmp // Fix problem with new readline #if defined(__WIN__) @@ -92,7 +92,6 @@ extern "C" { #include <readline/readline.h> #define HAVE_READLINE #endif - //int vidattr(long unsigned int attrs); // Was missing in sun curses } #if !defined(HAVE_VIDATTR) @@ -1024,7 +1023,7 @@ static const char *load_default_groups[] static int embedded_server_arg_count= 0; static char *embedded_server_args[MAX_SERVER_ARGS]; static const char *embedded_server_groups[]= -{ "server", "embedded", "mysql_SERVER", 0 }; +{ "server", "embedded", "mysql_SERVER", "mariadb_SERVER", 0 }; #ifdef HAVE_READLINE /* === modified file 'mysys/thr_mutex.c' --- a/mysys/thr_mutex.c 2009-06-22 08:06:35 +0000 +++ b/mysys/thr_mutex.c 2009-11-25 23:18:23 +0000 @@ -36,6 +36,7 @@ #undef pthread_mutex_init #undef pthread_mutex_lock #undef pthread_mutex_unlock +#undef pthread_mutex_trylock #undef pthread_mutex_destroy #undef pthread_cond_wait #undef pthread_cond_timedwait @@ -838,31 +839,9 @@ static void print_deadlock_warning(safe_ DBUG_VOID_RETURN; } +#elif defined(MY_PTHREAD_FASTMUTEX) -#endif /* THREAD && SAFE_MUTEX */ - -#if defined(THREAD) && defined(MY_PTHREAD_FASTMUTEX) && !defined(SAFE_MUTEX) - -#include "mysys_priv.h" -#include "my_static.h" -#include <m_string.h> - -#include <m_ctype.h> -#include <hash.h> -#include <myisampack.h> -#include <mysys_err.h> -#include <my_sys.h> - -#undef pthread_mutex_t -#undef pthread_mutex_init -#undef pthread_mutex_lock -#undef pthread_mutex_trylock -#undef pthread_mutex_unlock -#undef pthread_mutex_destroy -#undef pthread_cond_wait -#undef pthread_cond_timedwait - -ulong mutex_delay(ulong delayloops) +static ulong mutex_delay(ulong delayloops) { ulong i; volatile ulong j; @@ -943,6 +922,6 @@ void fastmutex_global_init(void) cpu_count= sysconf(_SC_NPROCESSORS_CONF); #endif } - -#endif /* SAFE_MUTEX_DEFINED */ + +#endif /* defined(MY_PTHREAD_FASTMUTEX) */ #endif /* THREAD */

1 0

[Maria-developers] New (by Sanja): Subquery optimization: Avoid recalculating subquery if external fields values found in subquery cache (66)
by worklog-noreply＠askmonty.org 25 Nov '09

25 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Subquery optimization: Avoid recalculating subquery if external fields values found in subquery cache CREATION DATE..: Wed, 25 Nov 2009, 22:25 SUPERVISOR.....: Monty IMPLEMENTOR....: Sanja COPIES TO......: CATEGORY.......: Server-BackLog TASK ID........: 66 (http://askmonty.org/worklog/?tid=66) VERSION........: Server-5.2 STATUS.........: Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: DESCRIPTION: Collect all outer items/references (left part of the subquiery and outer references inside the subquery) in key string. Compare the string (which represents certain value set of the references) against values in hash table and return cached result of subquery if the reference values combination has already been used. For example in the following subquery: (L1, L2) IN (SELECT A, B FROM T WHERE T.F1>OTER_FIELD) set of references to look into the subquery cache is (L1, L2, OTER_FIELD). The subquery cache should be implemented as simple LRU connected to the subquery. Size of the subquery cache (in number of results (but maybe in used memory amount)) is limited by session variable (query parameter?). ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] New (by Sanja): Subquery optimization: Avoid recalculating subquery if external fields values found in subquery cache (66)
by worklog-noreply＠askmonty.org 25 Nov '09

25 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Subquery optimization: Avoid recalculating subquery if external fields values found in subquery cache CREATION DATE..: Wed, 25 Nov 2009, 22:25 SUPERVISOR.....: Monty IMPLEMENTOR....: Sanja COPIES TO......: CATEGORY.......: Server-BackLog TASK ID........: 66 (http://askmonty.org/worklog/?tid=66) VERSION........: Server-5.2 STATUS.........: Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: DESCRIPTION: Collect all outer items/references (left part of the subquiery and outer references inside the subquery) in key string. Compare the string (which represents certain value set of the references) against values in hash table and return cached result of subquery if the reference values combination has already been used. For example in the following subquery: (L1, L2) IN (SELECT A, B FROM T WHERE T.F1>OTER_FIELD) set of references to look into the subquery cache is (L1, L2, OTER_FIELD). The subquery cache should be implemented as simple LRU connected to the subquery. Size of the subquery cache (in number of results (but maybe in used memory amount)) is limited by session variable (query parameter?). ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] WL47: Second suggestion.
by Alexi1952 25 Nov '09

25 Nov '09

Having to start with WL47 which already contains the 'First suggestion', I'd like to discuss: Second suggestion ~~~~~~~~~~~~~~~~~ Add query text to the RBR events data part (together with adding query_len field to their header). Reason (compared with the First suggestion): - this makes binary log bit smaller (no need to store headers of new events). - almost all changes needed in this case are localized in the RBR-event classes (e.g. replication, being, roughly speaking, based on calling the apply_event() and other Log_event methods, shouldn't care about whether there is a query_text in the event or not). We can recognize whether an RBR-event in a binary log contain query text or not by looking at it header length in the Format description event (the post_header member): containing the additional query_len field, the header of the 'extended' event have a diferent size. Because we are able to recognize RBR-events format _only_ by the Format description event (at the beginning of the binary log), we can not allow to mix in one binary log both 'standard' and 'extended' formats (which is possible within the First suggestion). If possibility of such a mix is important then we should prefer the First suggestion. Any remarks ? Which way to go ?

1 0

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (knielsen:2772)
by knielsen＠knielsen-hq.org 25 Nov '09

25 Nov '09

#At lp:maria 2772 knielsen(a)knielsen-hq.org 2009-11-25 After-merge fixes following Merge of Mysql 5.1.41 into MariaDB: Fix path for SSL stuff when running testsuite from installed package. modified: mysql-test/lib/My/ConfigFactory.pm === modified file 'mysql-test/lib/My/ConfigFactory.pm' --- a/mysql-test/lib/My/ConfigFactory.pm 2009-11-16 20:49:51 +0000 +++ b/mysql-test/lib/My/ConfigFactory.pm 2009-11-25 13:00:28 +0000 @@ -142,8 +142,7 @@ sub fix_secure_file_priv { sub fix_std_data { my ($self, $config, $group_name, $group)= @_; - my $basedir= $self->get_basedir($group); - return "$basedir/mysql-test/std_data"; + return "$::opt_vardir/std_data"; } sub ssl_supported {

1 0

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (knielsen:2771)
by knielsen＠knielsen-hq.org 25 Nov '09

25 Nov '09

#At lp:maria 2771 knielsen(a)knielsen-hq.org 2009-11-25 After-merge fix for merge of MySQL 5.1.41 into MariaDB: Valgrind fixes. modified: mysql-test/mysql-test-run.pl sql/log_event.cc per-file messages: mysql-test/mysql-test-run.pl Fix Valgrind warnings: add more post-shutdown warning suppressions, and revert bad previous change. sql/log_event.cc Manually apply fix for Bug#48340 (basically missing initialisation of thd->lex->local_file in Load_log_event::do_apply_event()) === modified file 'mysql-test/mysql-test-run.pl' --- a/mysql-test/mysql-test-run.pl 2009-11-24 09:34:42 +0000 +++ b/mysql-test/mysql-test-run.pl 2009-11-25 11:06:01 +0000 @@ -223,7 +223,7 @@ my $opt_strace_client; our $opt_user = "root"; my $opt_valgrind= 0; -my $opt_valgrind_mysqld= 0; +our $opt_valgrind_mysqld= 0; my $opt_valgrind_mysqltest= 0; my @default_valgrind_args= ("--show-reachable=yes"); my @valgrind_args; @@ -3990,6 +3990,8 @@ sub extract_warning_lines ($) { qr/unknown variable 'loose-/, qr/Now setting lower_case_table_names to [02]/, qr/deprecated/, + qr/Slave SQL thread retried transaction/, + qw/Slave $additional info$/, ); my $match_count= 0; === modified file 'sql/log_event.cc' --- a/sql/log_event.cc 2009-11-16 20:49:51 +0000 +++ b/sql/log_event.cc 2009-11-25 11:06:01 +0000 @@ -4509,6 +4509,7 @@ int Load_log_event::do_apply_event(NET* as the present method does not call mysql_parse(). */ lex_start(thd); + thd->lex->local_file= local_fname; mysql_reset_thd_for_next_command(thd); if (!use_rli_only_for_errors)

1 0

[Maria-developers] Updated (by Guest): Add a mysqlbinlog option to filter updates to certain tables (40)
by worklog-noreply＠askmonty.org 25 Nov '09

25 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Add a mysqlbinlog option to filter updates to certain tables CREATION DATE..: Mon, 10 Aug 2009, 13:25 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: Psergey CATEGORY.......: Server-Sprint TASK ID........: 40 (http://askmonty.org/worklog/?tid=40) VERSION........: Server-9.x STATUS.........: Assigned PRIORITY.......: 60 WORKED HOURS...: 32 ESTIMATE.......: 32 (hours remain) ORIG. ESTIMATE.: 48 PROGRESS NOTES: -=-=(Guest - Wed, 25 Nov 2009, 11:41)=-=- Status updated. --- /tmp/wklog.40.old.5760 2009-11-25 11:41:09.000000000 +0200 +++ /tmp/wklog.40.new.5760 2009-11-25 11:41:09.000000000 +0200 @@ -1 +1 @@ -Un-Assigned +Assigned -=-=(Guest - Wed, 25 Nov 2009, 11:41)=-=- Category updated. --- /tmp/wklog.40.old.5737 2009-11-25 11:41:03.000000000 +0200 +++ /tmp/wklog.40.new.5737 2009-11-25 11:41:03.000000000 +0200 @@ -1 +1 @@ -Server-RawIdeaBin +Server-Sprint -=-=(Bothorsen - Tue, 17 Nov 2009, 17:20)=-=- Alex is closer to a working patch now. Worked 14 hours and estimate 32 hours remain (original estimate unchanged). -=-=(Bothorsen - Thu, 12 Nov 2009, 13:13)=-=- Work hours by Alexi and Bo + estimated time for the task. Worked 16 hours and estimate 46 hours remain (original estimate increased by 14 hours). -=-=(Alexi - Sun, 08 Nov 2009, 15:18)=-=- Low Level Design modified. --- /tmp/wklog.40.old.15787 2009-11-08 15:18:11.000000000 +0200 +++ /tmp/wklog.40.new.15787 2009-11-08 15:18:11.000000000 +0200 @@ -62,7 +62,7 @@ it considers the query to extent to the end of the event. 2. For 'db' (current db) the trailing zero is redundant since the length is already known. -3. db_len = 0 means that this is the current db. +3. In tables_info, db_len = 0 means that this is the current db. When reading Query events from binary log, we can recognize its format by its post-header length: in extended case the post-header includes 4 @@ -75,6 +75,77 @@ + #define Q_QUERY_LEN_OFFSET Q_STATUS_VARS_LEN_OFFSET + 2 + #define Q_QUERY_TABLES_INFO_LEN_OFFSET Q_QUERY_LEN_OFFSET + 2 + +*********************************************************************** +HELP NEEDED +*********************************************************************** +The QUERY_HEADER_LEN is used in the definition of MAX_LOG_EVENT_HEADER: + +log_event.h +~~~~~~~~~~~ +#define MAX_LOG_EVENT_HEADER ( /* in order of Query_log_event::write */ \ + LOG_EVENT_HEADER_LEN + /* write_header */ \ + QUERY_HEADER_LEN + /* write_data */ \ + EXECUTE_LOAD_QUERY_EXTRA_HEADER_LEN + /*write_post_header_for_derived */ \ + MAX_SIZE_LOG_EVENT_STATUS + /* status */ \ + NAME_LEN + 1) + +which is used only for setting + + thd->variables.max_allowed_packet + mysql->net.max_packet_size + +Looks like (but I am not quite sure) that QUERY_HEADER_LEN can simply +(without making any other changes) be substituted in this definition by +QUERY_HEADER_LEN_EXT. + +Below I list all places where MAX_LOG_EVENT_HEADER is used: + +slave.cc +~~~~~~~~ +static int init_slave_thread(...) +{ ... + /* + Adding MAX_LOG_EVENT_HEADER_LEN to the max_allowed_packet on all + slave threads, since a replication event can become this much larger + than the corresponding packet (query) sent from client to master. + */ + thd->variables.max_allowed_packet= global_system_variables.max_allowed_packet + + MAX_LOG_EVENT_HEADER; /* note, incr over the global not session var */ + ... +} +pthread_handler_t handle_slave_io(...) +{ ... + /* + Adding MAX_LOG_EVENT_HEADER_LEN to the max_packet_size on the I/O + thread, since a replication event can become this much larger than + the corresponding packet (query) sent from client to master. + */ + mysql->net.max_packet_size= thd->net.max_packet_size+= MAX_LOG_EVENT_HEADER; + ... +} + +sql_repl.cc +~~~~~~~~~~~ +void mysql_binlog_send(...) +{ ... + /* + Adding MAX_LOG_EVENT_HEADER_LEN, since a binlog event can become + this larger than the corresponding packet (query) sent + from client to master. + */ + thd->variables.max_allowed_packet+= MAX_LOG_EVENT_HEADER; + ... +} +bool mysql_show_binlog_events(...) +{ ... + /* + to account binlog event header size + */ + thd->variables.max_allowed_packet+= MAX_LOG_EVENT_HEADER; + ... +} + 3. Changes in log events ************************ @@ -84,7 +155,7 @@ This setting is done in Format description event constructor which creates the event for writing to binary log: - if (binlog_with_tables_info) + if (opt_binlog_with_tables_info) post_header_len[QUERY_EVENT - 1] = QUERY_HEADER_LEN_EXT; else post_header_len[QUERY_EVENT - 1] = QUERY_HEADER_LEN; @@ -99,12 +170,12 @@ following manner: switch (binlog_ver) { - case 4: /* MySQL 5.0 and higher */ + #ifndef MYSQL_CLIENT + case 4: /* MySQL 5.0 and higher */ ... -+ #else -+ <error> + break; + #endif + case 1: case 3: ... @@ -132,7 +203,7 @@ -------------------------------- [Creates the event for binlogging] -In case of binlog_with_tables_info = TRUE, set additionally query_len, +In case of opt_binlog_with_tables_info = TRUE, set additionally query_len, tables_info_len, and tables_info members (the constructor is to have an additional 'tables_info' argument). @@ -140,7 +211,7 @@ ---------------- [Writes the event to binlog] -In case of binlog_with_tables_info = TRUE, write additional members +In case of opt_binlog_with_tables_info = TRUE, write additional members (query_len, tables_info_len, and tables_info) to binary log. Also write corresponding whole event length to the common-header. -=-=(Alexi - Sun, 08 Nov 2009, 10:40)=-=- Low Level Design modified. --- /tmp/wklog.40.old.5055 2009-11-08 08:40:02.000000000 +0000 +++ /tmp/wklog.40.new.5055 2009-11-08 08:40:02.000000000 +0000 @@ -3,6 +3,7 @@ 1. Adding --binlog-with-tables-info option ****************************************** +GLOBAL, read-only option. When set, Query events are to be written in the extended binary format which contains tables_info. When not set, Query events -=-=(Alexi - Thu, 05 Nov 2009, 12:37)=-=- Low Level Design modified. --- /tmp/wklog.40.old.11441 2009-11-05 12:37:16.000000000 +0200 +++ /tmp/wklog.40.new.11441 2009-11-05 12:37:16.000000000 +0200 @@ -1,9 +1,18 @@ OPTION: 2.5 Extend Query Events With Tables Info ================================================ -1. Query_log_event Binary Format -******************************** -Changes to be done: +1. Adding --binlog-with-tables-info option +****************************************** + +When set, Query events are to be written in the extended binary +format which contains tables_info. When not set, Query events +are to be written in usual format (without any changes). + +2. Query event extended binary format +************************************* + +When --binlog-with-tables-info is set, Query events are writen +to binary log in the following (extended) format. Query_log_event binary format --------------------------------- @@ -24,12 +33,12 @@ error_code 2 status_vars_len 2 + query_len 2 (see Note 1) -+ tables_info_len 2 (see Note 2) ++ tables_info_len 2 --------------------------------- BODY: status_vars status_vars_len - db db_len + 1 -+ db db_len (see Note 3) ++ db db_len (see Note 2) query query_len + tables_info @@ -37,7 +46,7 @@ --------------------------------- Name Size (bytes) --------------------------------- - db_len 1 (see Note 4) + db_len 1 (see Note 3) db db_len table_name_len 1 table_name table_name_len @@ -48,19 +57,99 @@ table_name table_name_len NOTES -1. Currently Query_log_event format doesn't include 'query_len' because +1. In usual format, Query_log_event doesn't include 'query_len' because it considers the query to extent to the end of the event. -2. If tables_info is not included in the event (--binlog-with-tables-info - option), tables_info_len = 0. -3. The trailing zero is redundant since the length is already known. -4. In case of db = current db, db_len = 0 and db = empty, because - current db is already included in the current event format. +2. For 'db' (current db) the trailing zero is redundant since the length + is already known. +3. db_len = 0 means that this is the current db. + +When reading Query events from binary log, we can recognize its format +by its post-header length: in extended case the post-header includes 4 +additional bytes. + + #define QUERY_HEADER_LEN (QUERY_HEADER_MINIMAL_LEN + 4) ++ #define QUERY_HEADER_LEN_EXT (QUERY_HEADER_LEN + 4) + ... + #define Q_STATUS_VARS_LEN_OFFSET 11 ++ #define Q_QUERY_LEN_OFFSET Q_STATUS_VARS_LEN_OFFSET + 2 ++ #define Q_QUERY_TABLES_INFO_LEN_OFFSET Q_QUERY_LEN_OFFSET + 2 + +3. Changes in log events +************************ + +3.1. Format description event +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +Changes needed here concern setting post-header length for Query events. +This setting is done in Format description event constructor which creates +the event for writing to binary log: + + if (binlog_with_tables_info) + post_header_len[QUERY_EVENT - 1] = QUERY_HEADER_LEN_EXT; + else + post_header_len[QUERY_EVENT - 1] = QUERY_HEADER_LEN; + +This change is to be done only for case binlog_ver = 4. + +NOTE. The refered above constructor is allowed to be invoked in a client +context for creating "artificial" Format description events in case of +MySQL < 5.0 (e.g. see mysqlbinlog code). To avoid compilation problems +(because of 'binlog_with_tables_info') and taking into account the +"MySQL < 5.0" restriction, we have to #ifdef out the above code in +following manner: + + switch (binlog_ver) { + case 4: /* MySQL 5.0 and higher */ ++ #ifndef MYSQL_CLIENT + ... ++ #else ++ <error> ++ #endif + case 1: + case 3: + ... + } + +3.2. Query event +~~~~~~~~~~~~~~~~ +Changes needed here include adding tables_info and tables_info_len +members (member for query length already exists) and modifying the +following function-members: + +Query_log_event(buf) constructor +-------------------------------- +[Parses binary format written to the 'buf'] + +Getting post-header length from the Format description event (passed +to the constructor as an argument), define whether buf contains an +extended or usual Query event and parse the buf contents accordingly. + +NOTE. Defining Query event format here should be done with taking into +account that this constructor can be called within a Query-derived +event with the event_type argument != QUERY_EVENT. + +Query_log_event(thd) constructor +-------------------------------- +[Creates the event for binlogging] + +In case of binlog_with_tables_info = TRUE, set additionally query_len, +tables_info_len, and tables_info members (the constructor is to have +an additional 'tables_info' argument). + +write() function +---------------- +[Writes the event to binlog] + +In case of binlog_with_tables_info = TRUE, write additional members +(query_len, tables_info_len, and tables_info) to binary log. Also +write corresponding whole event length to the common-header. + +<To be continued> -2. Where to get tables info from? +4. Where to get tables info from? ********************************* -2.1. Case study: CREATE TABLE -****************************** +4.1. Case study: CREATE TABLE +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ *** CREATE TABLE table [SELECT ...] @@ -129,4 +218,4 @@ } } -To be continued +<To be continued> -=-=(Alexi - Wed, 04 Nov 2009, 10:21)=-=- Low Level Design modified. --- /tmp/wklog.40.old.6734 2009-11-04 10:21:20.000000000 +0200 +++ /tmp/wklog.40.new.6734 2009-11-04 10:21:20.000000000 +0200 @@ -21,9 +21,9 @@ slave_proxy_id 4 exec_time 4 db_len 1 -+ query_len 2 (see Note 1) error_code 2 status_vars_len 2 ++ query_len 2 (see Note 1) + tables_info_len 2 (see Note 2) --------------------------------- BODY: -=-=(Alexi - Tue, 03 Nov 2009, 11:19)=-=- Low Level Design modified. --- /tmp/wklog.40.old.7187 2009-11-03 11:19:22.000000000 +0200 +++ /tmp/wklog.40.new.7187 2009-11-03 11:19:22.000000000 +0200 @@ -1 +1,132 @@ +OPTION: 2.5 Extend Query Events With Tables Info +================================================ +1. Query_log_event Binary Format +******************************** +Changes to be done: + + Query_log_event binary format + --------------------------------- + Name Size (bytes) + --------------------------------- + COMMON HEADER: + timestamp 4 + type 1 + server_id 4 + total_size 4 + master_position 4 + flags 2 + --------------------------------- + POST HEADER: + slave_proxy_id 4 + exec_time 4 + db_len 1 ++ query_len 2 (see Note 1) + error_code 2 + status_vars_len 2 ++ tables_info_len 2 (see Note 2) + --------------------------------- + BODY: + status_vars status_vars_len +- db db_len + 1 ++ db db_len (see Note 3) + query query_len ++ tables_info + + tables_info binary format + --------------------------------- + Name Size (bytes) + --------------------------------- + db_len 1 (see Note 4) + db db_len + table_name_len 1 + table_name table_name_len + ... + db_len 1 + db db_len + table_name_len 1 + table_name table_name_len + +NOTES +1. Currently Query_log_event format doesn't include 'query_len' because + it considers the query to extent to the end of the event. +2. If tables_info is not included in the event (--binlog-with-tables-info + option), tables_info_len = 0. +3. The trailing zero is redundant since the length is already known. +4. In case of db = current db, db_len = 0 and db = empty, because + current db is already included in the current event format. + +2. Where to get tables info from? +********************************* + +2.1. Case study: CREATE TABLE +****************************** + +*** CREATE TABLE table [SELECT ...] + + bool mysql_create_table_no_lock( + THD *thd, + const char *db, + const char *table_name, ...) + { + ... + // ------------------------------------- + // WL40: To be included in tables_info: + // * db, table_name + // * thd->lex->query_tables (tables refered to in + // the select-part; empty if no select-part) + // ------------------------------------- + write_bin_log(thd, TRUE, thd->query, thd->query_length); + } + +*** CREATE TABLE table LIKE src-table + + bool mysql_create_like_table( + ... + TABLE_LIST *table, + TABLE_LIST *src_table, + ...) + { + ... + if (thd->current_stmt_binlog_row_based) + { // RBR: In this case we don't replicate temp tables + if (!(create_info->options & HA_LEX_CREATE_TMP_TABLE)) + { + if (src_table->table->s->tmp_table) + { // CREATE normal-table LIKE temp-table: + + // Generate new query without LIKE-part + store_create_info(thd, table, &query, create_info, FALSE); + + // ------------------------------------- + // WL40: To include to tables_info: + // * table (src_table is not included) + // ------------------------------------- + write_bin_log(thd, TRUE, query.ptr(), query.length()); + } + else + { // CREATE normal-table LIKE normal-table + + // ------------------------------------- + // WL40: To include to log_tables_info: + // * table + // * src_table + // ------------------------------------- + write_bin_log(thd, TRUE, thd->query, thd->query_length); + } + } + // CREATE temp-table LIKE ... + // This case is not replicated + } + else + { // SBR: + // ------------------------------------- + // WL40: To include to tables_info: + // * table + // * src_table + // ------------------------------------- + write_bin_log(thd, TRUE, thd->query, thd->query_length); + } + } + +To be continued -=-=(Alexi - Mon, 02 Nov 2009, 11:34)=-=- Worked 2 hours on option 2.5 Worked 2 hours and estimate 48 hours remain (original estimate increased by 50 hours). ------------------------------------------------------------ -=-=(View All Progress Notes, 16 total)=-=- http://askmonty.org/worklog/index.pl?tid=40&nolimit=1 DESCRIPTION: Replication slave can be set to filter updates to certain tables with --replicate-[wild-]{do,ignore}-table options. This task is about adding similar functionality to mysqlbinlog. HIGH-LEVEL SPECIFICATION: 1. Context ---------- (See http://askmonty.org/wiki/index.php/Scratch/ReplicationOptions for global overview) At the moment, the server has these replication slave options: --replicate-do-table=db.tbl --replicate-ignore-table=db.tbl --replicate-wild-do-table=pattern.pattern --replicate-wild-ignore-table=pattern.pattern They affect both RBR and SBR events. SBR events are checked after the statement has been parsed, the server iterates over list of used tables and checks them againist --replicate instructions. What is interesting is that this scheme still allows to update the ignored table through a VIEW. 2. Table filtering in mysqlbinlog --------------------------------- Per-table filtering of RBR events is easy (as it is relatively easy to extract the name of the table that the event applies to). Per-table filtering of SBR events is hard, as generally it is not apparent which tables the statement refers to. This opens possible options: 2.1 Put the parser into mysqlbinlog ----------------------------------- Once we have a full parser in mysqlbinlog, we'll be able to check which tables are used by a statement, and will allow to show behaviour identical to those that one obtains when using --replicate-* slave options. (It is not clear how much effort is needed to put the parser into mysqlbinlog. Any guesses?) 2.2 Use dumb regexp match ------------------------- Use a really dumb approach. A query is considered to be modifying table X if it matches an expression CREATE TABLE $tablename DROP $tablename UPDATE ...$tablename ... SET // here '...' can't contain the word 'SET' DELETE ...$tablename ... WHERE // same as above ALTER TABLE $tablename .. etc (go get from the grammar) .. The advantage over doing the same in awk is that mysqlbinlog will also process RBR statements, and together with that will provide a working solution for those who are careful with their table names not mixing with string constants and such. (TODO: string constants are of particular concern as they come from [potentially hostile] users, unlike e.g. table aliases which come from [not hostile] developers. Remove also all string constants before attempting to do match?) 2.3 Have the master put annotations ----------------------------------- We could add a master option so that it injects into query a mark that tells which tables the query will affect, e.g. for the query UPDATE t1 LEFT JOIN db3.t2 ON ... WHERE ... the binlog will have /* !mysqlbinlog: updates t1,db3.t2 */ UPDATE t1 LEFT JOIN ... and further processing in mysqlbinlog will be trivial. 2.4 Implement server functionality to ignore certain tables ----------------------------------------------------------- We could add a general facility in the server to ignore certain tables: SET SESSION ignored_tables = "db1.t1,db2.t2"; This would work similar to --replicate-ignore-table, but in a general way not restricted to the slave SQL thread. It would then be trivial for mysqlbinlog to add such statements at the start of the output, or probably the user could just do it manually with no need for additional options for mysqlbinlog. It might be useful to integrate this with the code that already handles --replicate-ignore-db and similar slave options. 2.5 Extend Query Events With Tables Info ---------------------------------------- We could extend query events structure with a tables info - a list of tables which the query refers to: <current query event structure> tables_info_len dbase_len dbase table_len table ... dbase_len dbase table_len table Note. In case of <dbase> = current data base, we can set dbase_len = 0 and dbase = empty because current query event structure already includes current data base name. Note. Possibly it is reasonable also to add a --binlog-with-tables-info option which defines whether tables info must be included to the query events. LOW-LEVEL DESIGN: OPTION: 2.5 Extend Query Events With Tables Info ================================================ 1. Adding --binlog-with-tables-info option ****************************************** GLOBAL, read-only option. When set, Query events are to be written in the extended binary format which contains tables_info. When not set, Query events are to be written in usual format (without any changes). 2. Query event extended binary format ************************************* When --binlog-with-tables-info is set, Query events are writen to binary log in the following (extended) format. Query_log_event binary format --------------------------------- Name Size (bytes) --------------------------------- COMMON HEADER: timestamp 4 type 1 server_id 4 total_size 4 master_position 4 flags 2 --------------------------------- POST HEADER: slave_proxy_id 4 exec_time 4 db_len 1 error_code 2 status_vars_len 2 + query_len 2 (see Note 1) + tables_info_len 2 --------------------------------- BODY: status_vars status_vars_len - db db_len + 1 + db db_len (see Note 2) query query_len + tables_info tables_info binary format --------------------------------- Name Size (bytes) --------------------------------- db_len 1 (see Note 3) db db_len table_name_len 1 table_name table_name_len ... db_len 1 db db_len table_name_len 1 table_name table_name_len NOTES 1. In usual format, Query_log_event doesn't include 'query_len' because it considers the query to extent to the end of the event. 2. For 'db' (current db) the trailing zero is redundant since the length is already known. 3. In tables_info, db_len = 0 means that this is the current db. When reading Query events from binary log, we can recognize its format by its post-header length: in extended case the post-header includes 4 additional bytes. #define QUERY_HEADER_LEN (QUERY_HEADER_MINIMAL_LEN + 4) + #define QUERY_HEADER_LEN_EXT (QUERY_HEADER_LEN + 4) ... #define Q_STATUS_VARS_LEN_OFFSET 11 + #define Q_QUERY_LEN_OFFSET Q_STATUS_VARS_LEN_OFFSET + 2 + #define Q_QUERY_TABLES_INFO_LEN_OFFSET Q_QUERY_LEN_OFFSET + 2 *********************************************************************** HELP NEEDED *********************************************************************** The QUERY_HEADER_LEN is used in the definition of MAX_LOG_EVENT_HEADER: log_event.h ~~~~~~~~~~~ #define MAX_LOG_EVENT_HEADER ( /* in order of Query_log_event::write */ \ LOG_EVENT_HEADER_LEN + /* write_header */ \ QUERY_HEADER_LEN + /* write_data */ \ EXECUTE_LOAD_QUERY_EXTRA_HEADER_LEN + /*write_post_header_for_derived */ \ MAX_SIZE_LOG_EVENT_STATUS + /* status */ \ NAME_LEN + 1) which is used only for setting thd->variables.max_allowed_packet mysql->net.max_packet_size Looks like (but I am not quite sure) that QUERY_HEADER_LEN can simply (without making any other changes) be substituted in this definition by QUERY_HEADER_LEN_EXT. Below I list all places where MAX_LOG_EVENT_HEADER is used: slave.cc ~~~~~~~~ static int init_slave_thread(...) { ... /* Adding MAX_LOG_EVENT_HEADER_LEN to the max_allowed_packet on all slave threads, since a replication event can become this much larger than the corresponding packet (query) sent from client to master. */ thd->variables.max_allowed_packet= global_system_variables.max_allowed_packet + MAX_LOG_EVENT_HEADER; /* note, incr over the global not session var */ ... } pthread_handler_t handle_slave_io(...) { ... /* Adding MAX_LOG_EVENT_HEADER_LEN to the max_packet_size on the I/O thread, since a replication event can become this much larger than the corresponding packet (query) sent from client to master. */ mysql->net.max_packet_size= thd->net.max_packet_size+= MAX_LOG_EVENT_HEADER; ... } sql_repl.cc ~~~~~~~~~~~ void mysql_binlog_send(...) { ... /* Adding MAX_LOG_EVENT_HEADER_LEN, since a binlog event can become this larger than the corresponding packet (query) sent from client to master. */ thd->variables.max_allowed_packet+= MAX_LOG_EVENT_HEADER; ... } bool mysql_show_binlog_events(...) { ... /* to account binlog event header size */ thd->variables.max_allowed_packet+= MAX_LOG_EVENT_HEADER; ... } 3. Changes in log events ************************ 3.1. Format description event ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Changes needed here concern setting post-header length for Query events. This setting is done in Format description event constructor which creates the event for writing to binary log: if (opt_binlog_with_tables_info) post_header_len[QUERY_EVENT - 1] = QUERY_HEADER_LEN_EXT; else post_header_len[QUERY_EVENT - 1] = QUERY_HEADER_LEN; This change is to be done only for case binlog_ver = 4. NOTE. The refered above constructor is allowed to be invoked in a client context for creating "artificial" Format description events in case of MySQL < 5.0 (e.g. see mysqlbinlog code). To avoid compilation problems (because of 'binlog_with_tables_info') and taking into account the "MySQL < 5.0" restriction, we have to #ifdef out the above code in following manner: switch (binlog_ver) { + #ifndef MYSQL_CLIENT case 4: /* MySQL 5.0 and higher */ ... break; + #endif case 1: case 3: ... } 3.2. Query event ~~~~~~~~~~~~~~~~ Changes needed here include adding tables_info and tables_info_len members (member for query length already exists) and modifying the following function-members: Query_log_event(buf) constructor -------------------------------- [Parses binary format written to the 'buf'] Getting post-header length from the Format description event (passed to the constructor as an argument), define whether buf contains an extended or usual Query event and parse the buf contents accordingly. NOTE. Defining Query event format here should be done with taking into account that this constructor can be called within a Query-derived event with the event_type argument != QUERY_EVENT. Query_log_event(thd) constructor -------------------------------- [Creates the event for binlogging] In case of opt_binlog_with_tables_info = TRUE, set additionally query_len, tables_info_len, and tables_info members (the constructor is to have an additional 'tables_info' argument). write() function ---------------- [Writes the event to binlog] In case of opt_binlog_with_tables_info = TRUE, write additional members (query_len, tables_info_len, and tables_info) to binary log. Also write corresponding whole event length to the common-header. <To be continued> 4. Where to get tables info from? ********************************* 4.1. Case study: CREATE TABLE ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ *** CREATE TABLE table [SELECT ...] bool mysql_create_table_no_lock( THD *thd, const char *db, const char *table_name, ...) { ... // ------------------------------------- // WL40: To be included in tables_info: // * db, table_name // * thd->lex->query_tables (tables refered to in // the select-part; empty if no select-part) // ------------------------------------- write_bin_log(thd, TRUE, thd->query, thd->query_length); } *** CREATE TABLE table LIKE src-table bool mysql_create_like_table( ... TABLE_LIST *table, TABLE_LIST *src_table, ...) { ... if (thd->current_stmt_binlog_row_based) { // RBR: In this case we don't replicate temp tables if (!(create_info->options & HA_LEX_CREATE_TMP_TABLE)) { if (src_table->table->s->tmp_table) { // CREATE normal-table LIKE temp-table: // Generate new query without LIKE-part store_create_info(thd, table, &query, create_info, FALSE); // ------------------------------------- // WL40: To include to tables_info: // * table (src_table is not included) // ------------------------------------- write_bin_log(thd, TRUE, query.ptr(), query.length()); } else { // CREATE normal-table LIKE normal-table // ------------------------------------- // WL40: To include to log_tables_info: // * table // * src_table // ------------------------------------- write_bin_log(thd, TRUE, thd->query, thd->query_length); } } // CREATE temp-table LIKE ... // This case is not replicated } else { // SBR: // ------------------------------------- // WL40: To include to tables_info: // * table // * src_table // ------------------------------------- write_bin_log(thd, TRUE, thd->query, thd->query_length); } } <To be continued> ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Guest): Add a mysqlbinlog option to filter updates to certain tables (40)
by worklog-noreply＠askmonty.org 25 Nov '09

25 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Add a mysqlbinlog option to filter updates to certain tables CREATION DATE..: Mon, 10 Aug 2009, 13:25 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: Psergey CATEGORY.......: Server-Sprint TASK ID........: 40 (http://askmonty.org/worklog/?tid=40) VERSION........: Server-9.x STATUS.........: Assigned PRIORITY.......: 60 WORKED HOURS...: 32 ESTIMATE.......: 32 (hours remain) ORIG. ESTIMATE.: 48 PROGRESS NOTES: -=-=(Guest - Wed, 25 Nov 2009, 11:41)=-=- Status updated. --- /tmp/wklog.40.old.5760 2009-11-25 11:41:09.000000000 +0200 +++ /tmp/wklog.40.new.5760 2009-11-25 11:41:09.000000000 +0200 @@ -1 +1 @@ -Un-Assigned +Assigned -=-=(Guest - Wed, 25 Nov 2009, 11:41)=-=- Category updated. --- /tmp/wklog.40.old.5737 2009-11-25 11:41:03.000000000 +0200 +++ /tmp/wklog.40.new.5737 2009-11-25 11:41:03.000000000 +0200 @@ -1 +1 @@ -Server-RawIdeaBin +Server-Sprint -=-=(Bothorsen - Tue, 17 Nov 2009, 17:20)=-=- Alex is closer to a working patch now. Worked 14 hours and estimate 32 hours remain (original estimate unchanged). -=-=(Bothorsen - Thu, 12 Nov 2009, 13:13)=-=- Work hours by Alexi and Bo + estimated time for the task. Worked 16 hours and estimate 46 hours remain (original estimate increased by 14 hours). -=-=(Alexi - Sun, 08 Nov 2009, 15:18)=-=- Low Level Design modified. --- /tmp/wklog.40.old.15787 2009-11-08 15:18:11.000000000 +0200 +++ /tmp/wklog.40.new.15787 2009-11-08 15:18:11.000000000 +0200 @@ -62,7 +62,7 @@ it considers the query to extent to the end of the event. 2. For 'db' (current db) the trailing zero is redundant since the length is already known. -3. db_len = 0 means that this is the current db. +3. In tables_info, db_len = 0 means that this is the current db. When reading Query events from binary log, we can recognize its format by its post-header length: in extended case the post-header includes 4 @@ -75,6 +75,77 @@ + #define Q_QUERY_LEN_OFFSET Q_STATUS_VARS_LEN_OFFSET + 2 + #define Q_QUERY_TABLES_INFO_LEN_OFFSET Q_QUERY_LEN_OFFSET + 2 + +*********************************************************************** +HELP NEEDED +*********************************************************************** +The QUERY_HEADER_LEN is used in the definition of MAX_LOG_EVENT_HEADER: + +log_event.h +~~~~~~~~~~~ +#define MAX_LOG_EVENT_HEADER ( /* in order of Query_log_event::write */ \ + LOG_EVENT_HEADER_LEN + /* write_header */ \ + QUERY_HEADER_LEN + /* write_data */ \ + EXECUTE_LOAD_QUERY_EXTRA_HEADER_LEN + /*write_post_header_for_derived */ \ + MAX_SIZE_LOG_EVENT_STATUS + /* status */ \ + NAME_LEN + 1) + +which is used only for setting + + thd->variables.max_allowed_packet + mysql->net.max_packet_size + +Looks like (but I am not quite sure) that QUERY_HEADER_LEN can simply +(without making any other changes) be substituted in this definition by +QUERY_HEADER_LEN_EXT. + +Below I list all places where MAX_LOG_EVENT_HEADER is used: + +slave.cc +~~~~~~~~ +static int init_slave_thread(...) +{ ... + /* + Adding MAX_LOG_EVENT_HEADER_LEN to the max_allowed_packet on all + slave threads, since a replication event can become this much larger + than the corresponding packet (query) sent from client to master. + */ + thd->variables.max_allowed_packet= global_system_variables.max_allowed_packet + + MAX_LOG_EVENT_HEADER; /* note, incr over the global not session var */ + ... +} +pthread_handler_t handle_slave_io(...) +{ ... + /* + Adding MAX_LOG_EVENT_HEADER_LEN to the max_packet_size on the I/O + thread, since a replication event can become this much larger than + the corresponding packet (query) sent from client to master. + */ + mysql->net.max_packet_size= thd->net.max_packet_size+= MAX_LOG_EVENT_HEADER; + ... +} + +sql_repl.cc +~~~~~~~~~~~ +void mysql_binlog_send(...) +{ ... + /* + Adding MAX_LOG_EVENT_HEADER_LEN, since a binlog event can become + this larger than the corresponding packet (query) sent + from client to master. + */ + thd->variables.max_allowed_packet+= MAX_LOG_EVENT_HEADER; + ... +} +bool mysql_show_binlog_events(...) +{ ... + /* + to account binlog event header size + */ + thd->variables.max_allowed_packet+= MAX_LOG_EVENT_HEADER; + ... +} + 3. Changes in log events ************************ @@ -84,7 +155,7 @@ This setting is done in Format description event constructor which creates the event for writing to binary log: - if (binlog_with_tables_info) + if (opt_binlog_with_tables_info) post_header_len[QUERY_EVENT - 1] = QUERY_HEADER_LEN_EXT; else post_header_len[QUERY_EVENT - 1] = QUERY_HEADER_LEN; @@ -99,12 +170,12 @@ following manner: switch (binlog_ver) { - case 4: /* MySQL 5.0 and higher */ + #ifndef MYSQL_CLIENT + case 4: /* MySQL 5.0 and higher */ ... -+ #else -+ <error> + break; + #endif + case 1: case 3: ... @@ -132,7 +203,7 @@ -------------------------------- [Creates the event for binlogging] -In case of binlog_with_tables_info = TRUE, set additionally query_len, +In case of opt_binlog_with_tables_info = TRUE, set additionally query_len, tables_info_len, and tables_info members (the constructor is to have an additional 'tables_info' argument). @@ -140,7 +211,7 @@ ---------------- [Writes the event to binlog] -In case of binlog_with_tables_info = TRUE, write additional members +In case of opt_binlog_with_tables_info = TRUE, write additional members (query_len, tables_info_len, and tables_info) to binary log. Also write corresponding whole event length to the common-header. -=-=(Alexi - Sun, 08 Nov 2009, 10:40)=-=- Low Level Design modified. --- /tmp/wklog.40.old.5055 2009-11-08 08:40:02.000000000 +0000 +++ /tmp/wklog.40.new.5055 2009-11-08 08:40:02.000000000 +0000 @@ -3,6 +3,7 @@ 1. Adding --binlog-with-tables-info option ****************************************** +GLOBAL, read-only option. When set, Query events are to be written in the extended binary format which contains tables_info. When not set, Query events -=-=(Alexi - Thu, 05 Nov 2009, 12:37)=-=- Low Level Design modified. --- /tmp/wklog.40.old.11441 2009-11-05 12:37:16.000000000 +0200 +++ /tmp/wklog.40.new.11441 2009-11-05 12:37:16.000000000 +0200 @@ -1,9 +1,18 @@ OPTION: 2.5 Extend Query Events With Tables Info ================================================ -1. Query_log_event Binary Format -******************************** -Changes to be done: +1. Adding --binlog-with-tables-info option +****************************************** + +When set, Query events are to be written in the extended binary +format which contains tables_info. When not set, Query events +are to be written in usual format (without any changes). + +2. Query event extended binary format +************************************* + +When --binlog-with-tables-info is set, Query events are writen +to binary log in the following (extended) format. Query_log_event binary format --------------------------------- @@ -24,12 +33,12 @@ error_code 2 status_vars_len 2 + query_len 2 (see Note 1) -+ tables_info_len 2 (see Note 2) ++ tables_info_len 2 --------------------------------- BODY: status_vars status_vars_len - db db_len + 1 -+ db db_len (see Note 3) ++ db db_len (see Note 2) query query_len + tables_info @@ -37,7 +46,7 @@ --------------------------------- Name Size (bytes) --------------------------------- - db_len 1 (see Note 4) + db_len 1 (see Note 3) db db_len table_name_len 1 table_name table_name_len @@ -48,19 +57,99 @@ table_name table_name_len NOTES -1. Currently Query_log_event format doesn't include 'query_len' because +1. In usual format, Query_log_event doesn't include 'query_len' because it considers the query to extent to the end of the event. -2. If tables_info is not included in the event (--binlog-with-tables-info - option), tables_info_len = 0. -3. The trailing zero is redundant since the length is already known. -4. In case of db = current db, db_len = 0 and db = empty, because - current db is already included in the current event format. +2. For 'db' (current db) the trailing zero is redundant since the length + is already known. +3. db_len = 0 means that this is the current db. + +When reading Query events from binary log, we can recognize its format +by its post-header length: in extended case the post-header includes 4 +additional bytes. + + #define QUERY_HEADER_LEN (QUERY_HEADER_MINIMAL_LEN + 4) ++ #define QUERY_HEADER_LEN_EXT (QUERY_HEADER_LEN + 4) + ... + #define Q_STATUS_VARS_LEN_OFFSET 11 ++ #define Q_QUERY_LEN_OFFSET Q_STATUS_VARS_LEN_OFFSET + 2 ++ #define Q_QUERY_TABLES_INFO_LEN_OFFSET Q_QUERY_LEN_OFFSET + 2 + +3. Changes in log events +************************ + +3.1. Format description event +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +Changes needed here concern setting post-header length for Query events. +This setting is done in Format description event constructor which creates +the event for writing to binary log: + + if (binlog_with_tables_info) + post_header_len[QUERY_EVENT - 1] = QUERY_HEADER_LEN_EXT; + else + post_header_len[QUERY_EVENT - 1] = QUERY_HEADER_LEN; + +This change is to be done only for case binlog_ver = 4. + +NOTE. The refered above constructor is allowed to be invoked in a client +context for creating "artificial" Format description events in case of +MySQL < 5.0 (e.g. see mysqlbinlog code). To avoid compilation problems +(because of 'binlog_with_tables_info') and taking into account the +"MySQL < 5.0" restriction, we have to #ifdef out the above code in +following manner: + + switch (binlog_ver) { + case 4: /* MySQL 5.0 and higher */ ++ #ifndef MYSQL_CLIENT + ... ++ #else ++ <error> ++ #endif + case 1: + case 3: + ... + } + +3.2. Query event +~~~~~~~~~~~~~~~~ +Changes needed here include adding tables_info and tables_info_len +members (member for query length already exists) and modifying the +following function-members: + +Query_log_event(buf) constructor +-------------------------------- +[Parses binary format written to the 'buf'] + +Getting post-header length from the Format description event (passed +to the constructor as an argument), define whether buf contains an +extended or usual Query event and parse the buf contents accordingly. + +NOTE. Defining Query event format here should be done with taking into +account that this constructor can be called within a Query-derived +event with the event_type argument != QUERY_EVENT. + +Query_log_event(thd) constructor +-------------------------------- +[Creates the event for binlogging] + +In case of binlog_with_tables_info = TRUE, set additionally query_len, +tables_info_len, and tables_info members (the constructor is to have +an additional 'tables_info' argument). + +write() function +---------------- +[Writes the event to binlog] + +In case of binlog_with_tables_info = TRUE, write additional members +(query_len, tables_info_len, and tables_info) to binary log. Also +write corresponding whole event length to the common-header. + +<To be continued> -2. Where to get tables info from? +4. Where to get tables info from? ********************************* -2.1. Case study: CREATE TABLE -****************************** +4.1. Case study: CREATE TABLE +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ *** CREATE TABLE table [SELECT ...] @@ -129,4 +218,4 @@ } } -To be continued +<To be continued> -=-=(Alexi - Wed, 04 Nov 2009, 10:21)=-=- Low Level Design modified. --- /tmp/wklog.40.old.6734 2009-11-04 10:21:20.000000000 +0200 +++ /tmp/wklog.40.new.6734 2009-11-04 10:21:20.000000000 +0200 @@ -21,9 +21,9 @@ slave_proxy_id 4 exec_time 4 db_len 1 -+ query_len 2 (see Note 1) error_code 2 status_vars_len 2 ++ query_len 2 (see Note 1) + tables_info_len 2 (see Note 2) --------------------------------- BODY: -=-=(Alexi - Tue, 03 Nov 2009, 11:19)=-=- Low Level Design modified. --- /tmp/wklog.40.old.7187 2009-11-03 11:19:22.000000000 +0200 +++ /tmp/wklog.40.new.7187 2009-11-03 11:19:22.000000000 +0200 @@ -1 +1,132 @@ +OPTION: 2.5 Extend Query Events With Tables Info +================================================ +1. Query_log_event Binary Format +******************************** +Changes to be done: + + Query_log_event binary format + --------------------------------- + Name Size (bytes) + --------------------------------- + COMMON HEADER: + timestamp 4 + type 1 + server_id 4 + total_size 4 + master_position 4 + flags 2 + --------------------------------- + POST HEADER: + slave_proxy_id 4 + exec_time 4 + db_len 1 ++ query_len 2 (see Note 1) + error_code 2 + status_vars_len 2 ++ tables_info_len 2 (see Note 2) + --------------------------------- + BODY: + status_vars status_vars_len +- db db_len + 1 ++ db db_len (see Note 3) + query query_len ++ tables_info + + tables_info binary format + --------------------------------- + Name Size (bytes) + --------------------------------- + db_len 1 (see Note 4) + db db_len + table_name_len 1 + table_name table_name_len + ... + db_len 1 + db db_len + table_name_len 1 + table_name table_name_len + +NOTES +1. Currently Query_log_event format doesn't include 'query_len' because + it considers the query to extent to the end of the event. +2. If tables_info is not included in the event (--binlog-with-tables-info + option), tables_info_len = 0. +3. The trailing zero is redundant since the length is already known. +4. In case of db = current db, db_len = 0 and db = empty, because + current db is already included in the current event format. + +2. Where to get tables info from? +********************************* + +2.1. Case study: CREATE TABLE +****************************** + +*** CREATE TABLE table [SELECT ...] + + bool mysql_create_table_no_lock( + THD *thd, + const char *db, + const char *table_name, ...) + { + ... + // ------------------------------------- + // WL40: To be included in tables_info: + // * db, table_name + // * thd->lex->query_tables (tables refered to in + // the select-part; empty if no select-part) + // ------------------------------------- + write_bin_log(thd, TRUE, thd->query, thd->query_length); + } + +*** CREATE TABLE table LIKE src-table + + bool mysql_create_like_table( + ... + TABLE_LIST *table, + TABLE_LIST *src_table, + ...) + { + ... + if (thd->current_stmt_binlog_row_based) + { // RBR: In this case we don't replicate temp tables + if (!(create_info->options & HA_LEX_CREATE_TMP_TABLE)) + { + if (src_table->table->s->tmp_table) + { // CREATE normal-table LIKE temp-table: + + // Generate new query without LIKE-part + store_create_info(thd, table, &query, create_info, FALSE); + + // ------------------------------------- + // WL40: To include to tables_info: + // * table (src_table is not included) + // ------------------------------------- + write_bin_log(thd, TRUE, query.ptr(), query.length()); + } + else + { // CREATE normal-table LIKE normal-table + + // ------------------------------------- + // WL40: To include to log_tables_info: + // * table + // * src_table + // ------------------------------------- + write_bin_log(thd, TRUE, thd->query, thd->query_length); + } + } + // CREATE temp-table LIKE ... + // This case is not replicated + } + else + { // SBR: + // ------------------------------------- + // WL40: To include to tables_info: + // * table + // * src_table + // ------------------------------------- + write_bin_log(thd, TRUE, thd->query, thd->query_length); + } + } + +To be continued -=-=(Alexi - Mon, 02 Nov 2009, 11:34)=-=- Worked 2 hours on option 2.5 Worked 2 hours and estimate 48 hours remain (original estimate increased by 50 hours). ------------------------------------------------------------ -=-=(View All Progress Notes, 16 total)=-=- http://askmonty.org/worklog/index.pl?tid=40&nolimit=1 DESCRIPTION: Replication slave can be set to filter updates to certain tables with --replicate-[wild-]{do,ignore}-table options. This task is about adding similar functionality to mysqlbinlog. HIGH-LEVEL SPECIFICATION: 1. Context ---------- (See http://askmonty.org/wiki/index.php/Scratch/ReplicationOptions for global overview) At the moment, the server has these replication slave options: --replicate-do-table=db.tbl --replicate-ignore-table=db.tbl --replicate-wild-do-table=pattern.pattern --replicate-wild-ignore-table=pattern.pattern They affect both RBR and SBR events. SBR events are checked after the statement has been parsed, the server iterates over list of used tables and checks them againist --replicate instructions. What is interesting is that this scheme still allows to update the ignored table through a VIEW. 2. Table filtering in mysqlbinlog --------------------------------- Per-table filtering of RBR events is easy (as it is relatively easy to extract the name of the table that the event applies to). Per-table filtering of SBR events is hard, as generally it is not apparent which tables the statement refers to. This opens possible options: 2.1 Put the parser into mysqlbinlog ----------------------------------- Once we have a full parser in mysqlbinlog, we'll be able to check which tables are used by a statement, and will allow to show behaviour identical to those that one obtains when using --replicate-* slave options. (It is not clear how much effort is needed to put the parser into mysqlbinlog. Any guesses?) 2.2 Use dumb regexp match ------------------------- Use a really dumb approach. A query is considered to be modifying table X if it matches an expression CREATE TABLE $tablename DROP $tablename UPDATE ...$tablename ... SET // here '...' can't contain the word 'SET' DELETE ...$tablename ... WHERE // same as above ALTER TABLE $tablename .. etc (go get from the grammar) .. The advantage over doing the same in awk is that mysqlbinlog will also process RBR statements, and together with that will provide a working solution for those who are careful with their table names not mixing with string constants and such. (TODO: string constants are of particular concern as they come from [potentially hostile] users, unlike e.g. table aliases which come from [not hostile] developers. Remove also all string constants before attempting to do match?) 2.3 Have the master put annotations ----------------------------------- We could add a master option so that it injects into query a mark that tells which tables the query will affect, e.g. for the query UPDATE t1 LEFT JOIN db3.t2 ON ... WHERE ... the binlog will have /* !mysqlbinlog: updates t1,db3.t2 */ UPDATE t1 LEFT JOIN ... and further processing in mysqlbinlog will be trivial. 2.4 Implement server functionality to ignore certain tables ----------------------------------------------------------- We could add a general facility in the server to ignore certain tables: SET SESSION ignored_tables = "db1.t1,db2.t2"; This would work similar to --replicate-ignore-table, but in a general way not restricted to the slave SQL thread. It would then be trivial for mysqlbinlog to add such statements at the start of the output, or probably the user could just do it manually with no need for additional options for mysqlbinlog. It might be useful to integrate this with the code that already handles --replicate-ignore-db and similar slave options. 2.5 Extend Query Events With Tables Info ---------------------------------------- We could extend query events structure with a tables info - a list of tables which the query refers to: <current query event structure> tables_info_len dbase_len dbase table_len table ... dbase_len dbase table_len table Note. In case of <dbase> = current data base, we can set dbase_len = 0 and dbase = empty because current query event structure already includes current data base name. Note. Possibly it is reasonable also to add a --binlog-with-tables-info option which defines whether tables info must be included to the query events. LOW-LEVEL DESIGN: OPTION: 2.5 Extend Query Events With Tables Info ================================================ 1. Adding --binlog-with-tables-info option ****************************************** GLOBAL, read-only option. When set, Query events are to be written in the extended binary format which contains tables_info. When not set, Query events are to be written in usual format (without any changes). 2. Query event extended binary format ************************************* When --binlog-with-tables-info is set, Query events are writen to binary log in the following (extended) format. Query_log_event binary format --------------------------------- Name Size (bytes) --------------------------------- COMMON HEADER: timestamp 4 type 1 server_id 4 total_size 4 master_position 4 flags 2 --------------------------------- POST HEADER: slave_proxy_id 4 exec_time 4 db_len 1 error_code 2 status_vars_len 2 + query_len 2 (see Note 1) + tables_info_len 2 --------------------------------- BODY: status_vars status_vars_len - db db_len + 1 + db db_len (see Note 2) query query_len + tables_info tables_info binary format --------------------------------- Name Size (bytes) --------------------------------- db_len 1 (see Note 3) db db_len table_name_len 1 table_name table_name_len ... db_len 1 db db_len table_name_len 1 table_name table_name_len NOTES 1. In usual format, Query_log_event doesn't include 'query_len' because it considers the query to extent to the end of the event. 2. For 'db' (current db) the trailing zero is redundant since the length is already known. 3. In tables_info, db_len = 0 means that this is the current db. When reading Query events from binary log, we can recognize its format by its post-header length: in extended case the post-header includes 4 additional bytes. #define QUERY_HEADER_LEN (QUERY_HEADER_MINIMAL_LEN + 4) + #define QUERY_HEADER_LEN_EXT (QUERY_HEADER_LEN + 4) ... #define Q_STATUS_VARS_LEN_OFFSET 11 + #define Q_QUERY_LEN_OFFSET Q_STATUS_VARS_LEN_OFFSET + 2 + #define Q_QUERY_TABLES_INFO_LEN_OFFSET Q_QUERY_LEN_OFFSET + 2 *********************************************************************** HELP NEEDED *********************************************************************** The QUERY_HEADER_LEN is used in the definition of MAX_LOG_EVENT_HEADER: log_event.h ~~~~~~~~~~~ #define MAX_LOG_EVENT_HEADER ( /* in order of Query_log_event::write */ \ LOG_EVENT_HEADER_LEN + /* write_header */ \ QUERY_HEADER_LEN + /* write_data */ \ EXECUTE_LOAD_QUERY_EXTRA_HEADER_LEN + /*write_post_header_for_derived */ \ MAX_SIZE_LOG_EVENT_STATUS + /* status */ \ NAME_LEN + 1) which is used only for setting thd->variables.max_allowed_packet mysql->net.max_packet_size Looks like (but I am not quite sure) that QUERY_HEADER_LEN can simply (without making any other changes) be substituted in this definition by QUERY_HEADER_LEN_EXT. Below I list all places where MAX_LOG_EVENT_HEADER is used: slave.cc ~~~~~~~~ static int init_slave_thread(...) { ... /* Adding MAX_LOG_EVENT_HEADER_LEN to the max_allowed_packet on all slave threads, since a replication event can become this much larger than the corresponding packet (query) sent from client to master. */ thd->variables.max_allowed_packet= global_system_variables.max_allowed_packet + MAX_LOG_EVENT_HEADER; /* note, incr over the global not session var */ ... } pthread_handler_t handle_slave_io(...) { ... /* Adding MAX_LOG_EVENT_HEADER_LEN to the max_packet_size on the I/O thread, since a replication event can become this much larger than the corresponding packet (query) sent from client to master. */ mysql->net.max_packet_size= thd->net.max_packet_size+= MAX_LOG_EVENT_HEADER; ... } sql_repl.cc ~~~~~~~~~~~ void mysql_binlog_send(...) { ... /* Adding MAX_LOG_EVENT_HEADER_LEN, since a binlog event can become this larger than the corresponding packet (query) sent from client to master. */ thd->variables.max_allowed_packet+= MAX_LOG_EVENT_HEADER; ... } bool mysql_show_binlog_events(...) { ... /* to account binlog event header size */ thd->variables.max_allowed_packet+= MAX_LOG_EVENT_HEADER; ... } 3. Changes in log events ************************ 3.1. Format description event ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Changes needed here concern setting post-header length for Query events. This setting is done in Format description event constructor which creates the event for writing to binary log: if (opt_binlog_with_tables_info) post_header_len[QUERY_EVENT - 1] = QUERY_HEADER_LEN_EXT; else post_header_len[QUERY_EVENT - 1] = QUERY_HEADER_LEN; This change is to be done only for case binlog_ver = 4. NOTE. The refered above constructor is allowed to be invoked in a client context for creating "artificial" Format description events in case of MySQL < 5.0 (e.g. see mysqlbinlog code). To avoid compilation problems (because of 'binlog_with_tables_info') and taking into account the "MySQL < 5.0" restriction, we have to #ifdef out the above code in following manner: switch (binlog_ver) { + #ifndef MYSQL_CLIENT case 4: /* MySQL 5.0 and higher */ ... break; + #endif case 1: case 3: ... } 3.2. Query event ~~~~~~~~~~~~~~~~ Changes needed here include adding tables_info and tables_info_len members (member for query length already exists) and modifying the following function-members: Query_log_event(buf) constructor -------------------------------- [Parses binary format written to the 'buf'] Getting post-header length from the Format description event (passed to the constructor as an argument), define whether buf contains an extended or usual Query event and parse the buf contents accordingly. NOTE. Defining Query event format here should be done with taking into account that this constructor can be called within a Query-derived event with the event_type argument != QUERY_EVENT. Query_log_event(thd) constructor -------------------------------- [Creates the event for binlogging] In case of opt_binlog_with_tables_info = TRUE, set additionally query_len, tables_info_len, and tables_info members (the constructor is to have an additional 'tables_info' argument). write() function ---------------- [Writes the event to binlog] In case of opt_binlog_with_tables_info = TRUE, write additional members (query_len, tables_info_len, and tables_info) to binary log. Also write corresponding whole event length to the common-header. <To be continued> 4. Where to get tables info from? ********************************* 4.1. Case study: CREATE TABLE ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ *** CREATE TABLE table [SELECT ...] bool mysql_create_table_no_lock( THD *thd, const char *db, const char *table_name, ...) { ... // ------------------------------------- // WL40: To be included in tables_info: // * db, table_name // * thd->lex->query_tables (tables refered to in // the select-part; empty if no select-part) // ------------------------------------- write_bin_log(thd, TRUE, thd->query, thd->query_length); } *** CREATE TABLE table LIKE src-table bool mysql_create_like_table( ... TABLE_LIST *table, TABLE_LIST *src_table, ...) { ... if (thd->current_stmt_binlog_row_based) { // RBR: In this case we don't replicate temp tables if (!(create_info->options & HA_LEX_CREATE_TMP_TABLE)) { if (src_table->table->s->tmp_table) { // CREATE normal-table LIKE temp-table: // Generate new query without LIKE-part store_create_info(thd, table, &query, create_info, FALSE); // ------------------------------------- // WL40: To include to tables_info: // * table (src_table is not included) // ------------------------------------- write_bin_log(thd, TRUE, query.ptr(), query.length()); } else { // CREATE normal-table LIKE normal-table // ------------------------------------- // WL40: To include to log_tables_info: // * table // * src_table // ------------------------------------- write_bin_log(thd, TRUE, thd->query, thd->query_length); } } // CREATE temp-table LIKE ... // This case is not replicated } else { // SBR: // ------------------------------------- // WL40: To include to tables_info: // * table // * src_table // ------------------------------------- write_bin_log(thd, TRUE, thd->query, thd->query_length); } } <To be continued> ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Guest): Add a mysqlbinlog option to filter updates to certain tables (40)
by worklog-noreply＠askmonty.org 25 Nov '09

25 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Add a mysqlbinlog option to filter updates to certain tables CREATION DATE..: Mon, 10 Aug 2009, 13:25 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: Psergey CATEGORY.......: Server-Sprint TASK ID........: 40 (http://askmonty.org/worklog/?tid=40) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 32 ESTIMATE.......: 32 (hours remain) ORIG. ESTIMATE.: 48 PROGRESS NOTES: -=-=(Guest - Wed, 25 Nov 2009, 11:41)=-=- Category updated. --- /tmp/wklog.40.old.5737 2009-11-25 11:41:03.000000000 +0200 +++ /tmp/wklog.40.new.5737 2009-11-25 11:41:03.000000000 +0200 @@ -1 +1 @@ -Server-RawIdeaBin +Server-Sprint -=-=(Bothorsen - Tue, 17 Nov 2009, 17:20)=-=- Alex is closer to a working patch now. Worked 14 hours and estimate 32 hours remain (original estimate unchanged). -=-=(Bothorsen - Thu, 12 Nov 2009, 13:13)=-=- Work hours by Alexi and Bo + estimated time for the task. Worked 16 hours and estimate 46 hours remain (original estimate increased by 14 hours). -=-=(Alexi - Sun, 08 Nov 2009, 15:18)=-=- Low Level Design modified. --- /tmp/wklog.40.old.15787 2009-11-08 15:18:11.000000000 +0200 +++ /tmp/wklog.40.new.15787 2009-11-08 15:18:11.000000000 +0200 @@ -62,7 +62,7 @@ it considers the query to extent to the end of the event. 2. For 'db' (current db) the trailing zero is redundant since the length is already known. -3. db_len = 0 means that this is the current db. +3. In tables_info, db_len = 0 means that this is the current db. When reading Query events from binary log, we can recognize its format by its post-header length: in extended case the post-header includes 4 @@ -75,6 +75,77 @@ + #define Q_QUERY_LEN_OFFSET Q_STATUS_VARS_LEN_OFFSET + 2 + #define Q_QUERY_TABLES_INFO_LEN_OFFSET Q_QUERY_LEN_OFFSET + 2 + +*********************************************************************** +HELP NEEDED +*********************************************************************** +The QUERY_HEADER_LEN is used in the definition of MAX_LOG_EVENT_HEADER: + +log_event.h +~~~~~~~~~~~ +#define MAX_LOG_EVENT_HEADER ( /* in order of Query_log_event::write */ \ + LOG_EVENT_HEADER_LEN + /* write_header */ \ + QUERY_HEADER_LEN + /* write_data */ \ + EXECUTE_LOAD_QUERY_EXTRA_HEADER_LEN + /*write_post_header_for_derived */ \ + MAX_SIZE_LOG_EVENT_STATUS + /* status */ \ + NAME_LEN + 1) + +which is used only for setting + + thd->variables.max_allowed_packet + mysql->net.max_packet_size + +Looks like (but I am not quite sure) that QUERY_HEADER_LEN can simply +(without making any other changes) be substituted in this definition by +QUERY_HEADER_LEN_EXT. + +Below I list all places where MAX_LOG_EVENT_HEADER is used: + +slave.cc +~~~~~~~~ +static int init_slave_thread(...) +{ ... + /* + Adding MAX_LOG_EVENT_HEADER_LEN to the max_allowed_packet on all + slave threads, since a replication event can become this much larger + than the corresponding packet (query) sent from client to master. + */ + thd->variables.max_allowed_packet= global_system_variables.max_allowed_packet + + MAX_LOG_EVENT_HEADER; /* note, incr over the global not session var */ + ... +} +pthread_handler_t handle_slave_io(...) +{ ... + /* + Adding MAX_LOG_EVENT_HEADER_LEN to the max_packet_size on the I/O + thread, since a replication event can become this much larger than + the corresponding packet (query) sent from client to master. + */ + mysql->net.max_packet_size= thd->net.max_packet_size+= MAX_LOG_EVENT_HEADER; + ... +} + +sql_repl.cc +~~~~~~~~~~~ +void mysql_binlog_send(...) +{ ... + /* + Adding MAX_LOG_EVENT_HEADER_LEN, since a binlog event can become + this larger than the corresponding packet (query) sent + from client to master. + */ + thd->variables.max_allowed_packet+= MAX_LOG_EVENT_HEADER; + ... +} +bool mysql_show_binlog_events(...) +{ ... + /* + to account binlog event header size + */ + thd->variables.max_allowed_packet+= MAX_LOG_EVENT_HEADER; + ... +} + 3. Changes in log events ************************ @@ -84,7 +155,7 @@ This setting is done in Format description event constructor which creates the event for writing to binary log: - if (binlog_with_tables_info) + if (opt_binlog_with_tables_info) post_header_len[QUERY_EVENT - 1] = QUERY_HEADER_LEN_EXT; else post_header_len[QUERY_EVENT - 1] = QUERY_HEADER_LEN; @@ -99,12 +170,12 @@ following manner: switch (binlog_ver) { - case 4: /* MySQL 5.0 and higher */ + #ifndef MYSQL_CLIENT + case 4: /* MySQL 5.0 and higher */ ... -+ #else -+ <error> + break; + #endif + case 1: case 3: ... @@ -132,7 +203,7 @@ -------------------------------- [Creates the event for binlogging] -In case of binlog_with_tables_info = TRUE, set additionally query_len, +In case of opt_binlog_with_tables_info = TRUE, set additionally query_len, tables_info_len, and tables_info members (the constructor is to have an additional 'tables_info' argument). @@ -140,7 +211,7 @@ ---------------- [Writes the event to binlog] -In case of binlog_with_tables_info = TRUE, write additional members +In case of opt_binlog_with_tables_info = TRUE, write additional members (query_len, tables_info_len, and tables_info) to binary log. Also write corresponding whole event length to the common-header. -=-=(Alexi - Sun, 08 Nov 2009, 10:40)=-=- Low Level Design modified. --- /tmp/wklog.40.old.5055 2009-11-08 08:40:02.000000000 +0000 +++ /tmp/wklog.40.new.5055 2009-11-08 08:40:02.000000000 +0000 @@ -3,6 +3,7 @@ 1. Adding --binlog-with-tables-info option ****************************************** +GLOBAL, read-only option. When set, Query events are to be written in the extended binary format which contains tables_info. When not set, Query events -=-=(Alexi - Thu, 05 Nov 2009, 12:37)=-=- Low Level Design modified. --- /tmp/wklog.40.old.11441 2009-11-05 12:37:16.000000000 +0200 +++ /tmp/wklog.40.new.11441 2009-11-05 12:37:16.000000000 +0200 @@ -1,9 +1,18 @@ OPTION: 2.5 Extend Query Events With Tables Info ================================================ -1. Query_log_event Binary Format -******************************** -Changes to be done: +1. Adding --binlog-with-tables-info option +****************************************** + +When set, Query events are to be written in the extended binary +format which contains tables_info. When not set, Query events +are to be written in usual format (without any changes). + +2. Query event extended binary format +************************************* + +When --binlog-with-tables-info is set, Query events are writen +to binary log in the following (extended) format. Query_log_event binary format --------------------------------- @@ -24,12 +33,12 @@ error_code 2 status_vars_len 2 + query_len 2 (see Note 1) -+ tables_info_len 2 (see Note 2) ++ tables_info_len 2 --------------------------------- BODY: status_vars status_vars_len - db db_len + 1 -+ db db_len (see Note 3) ++ db db_len (see Note 2) query query_len + tables_info @@ -37,7 +46,7 @@ --------------------------------- Name Size (bytes) --------------------------------- - db_len 1 (see Note 4) + db_len 1 (see Note 3) db db_len table_name_len 1 table_name table_name_len @@ -48,19 +57,99 @@ table_name table_name_len NOTES -1. Currently Query_log_event format doesn't include 'query_len' because +1. In usual format, Query_log_event doesn't include 'query_len' because it considers the query to extent to the end of the event. -2. If tables_info is not included in the event (--binlog-with-tables-info - option), tables_info_len = 0. -3. The trailing zero is redundant since the length is already known. -4. In case of db = current db, db_len = 0 and db = empty, because - current db is already included in the current event format. +2. For 'db' (current db) the trailing zero is redundant since the length + is already known. +3. db_len = 0 means that this is the current db. + +When reading Query events from binary log, we can recognize its format +by its post-header length: in extended case the post-header includes 4 +additional bytes. + + #define QUERY_HEADER_LEN (QUERY_HEADER_MINIMAL_LEN + 4) ++ #define QUERY_HEADER_LEN_EXT (QUERY_HEADER_LEN + 4) + ... + #define Q_STATUS_VARS_LEN_OFFSET 11 ++ #define Q_QUERY_LEN_OFFSET Q_STATUS_VARS_LEN_OFFSET + 2 ++ #define Q_QUERY_TABLES_INFO_LEN_OFFSET Q_QUERY_LEN_OFFSET + 2 + +3. Changes in log events +************************ + +3.1. Format description event +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +Changes needed here concern setting post-header length for Query events. +This setting is done in Format description event constructor which creates +the event for writing to binary log: + + if (binlog_with_tables_info) + post_header_len[QUERY_EVENT - 1] = QUERY_HEADER_LEN_EXT; + else + post_header_len[QUERY_EVENT - 1] = QUERY_HEADER_LEN; + +This change is to be done only for case binlog_ver = 4. + +NOTE. The refered above constructor is allowed to be invoked in a client +context for creating "artificial" Format description events in case of +MySQL < 5.0 (e.g. see mysqlbinlog code). To avoid compilation problems +(because of 'binlog_with_tables_info') and taking into account the +"MySQL < 5.0" restriction, we have to #ifdef out the above code in +following manner: + + switch (binlog_ver) { + case 4: /* MySQL 5.0 and higher */ ++ #ifndef MYSQL_CLIENT + ... ++ #else ++ <error> ++ #endif + case 1: + case 3: + ... + } + +3.2. Query event +~~~~~~~~~~~~~~~~ +Changes needed here include adding tables_info and tables_info_len +members (member for query length already exists) and modifying the +following function-members: + +Query_log_event(buf) constructor +-------------------------------- +[Parses binary format written to the 'buf'] + +Getting post-header length from the Format description event (passed +to the constructor as an argument), define whether buf contains an +extended or usual Query event and parse the buf contents accordingly. + +NOTE. Defining Query event format here should be done with taking into +account that this constructor can be called within a Query-derived +event with the event_type argument != QUERY_EVENT. + +Query_log_event(thd) constructor +-------------------------------- +[Creates the event for binlogging] + +In case of binlog_with_tables_info = TRUE, set additionally query_len, +tables_info_len, and tables_info members (the constructor is to have +an additional 'tables_info' argument). + +write() function +---------------- +[Writes the event to binlog] + +In case of binlog_with_tables_info = TRUE, write additional members +(query_len, tables_info_len, and tables_info) to binary log. Also +write corresponding whole event length to the common-header. + +<To be continued> -2. Where to get tables info from? +4. Where to get tables info from? ********************************* -2.1. Case study: CREATE TABLE -****************************** +4.1. Case study: CREATE TABLE +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ *** CREATE TABLE table [SELECT ...] @@ -129,4 +218,4 @@ } } -To be continued +<To be continued> -=-=(Alexi - Wed, 04 Nov 2009, 10:21)=-=- Low Level Design modified. --- /tmp/wklog.40.old.6734 2009-11-04 10:21:20.000000000 +0200 +++ /tmp/wklog.40.new.6734 2009-11-04 10:21:20.000000000 +0200 @@ -21,9 +21,9 @@ slave_proxy_id 4 exec_time 4 db_len 1 -+ query_len 2 (see Note 1) error_code 2 status_vars_len 2 ++ query_len 2 (see Note 1) + tables_info_len 2 (see Note 2) --------------------------------- BODY: -=-=(Alexi - Tue, 03 Nov 2009, 11:19)=-=- Low Level Design modified. --- /tmp/wklog.40.old.7187 2009-11-03 11:19:22.000000000 +0200 +++ /tmp/wklog.40.new.7187 2009-11-03 11:19:22.000000000 +0200 @@ -1 +1,132 @@ +OPTION: 2.5 Extend Query Events With Tables Info +================================================ +1. Query_log_event Binary Format +******************************** +Changes to be done: + + Query_log_event binary format + --------------------------------- + Name Size (bytes) + --------------------------------- + COMMON HEADER: + timestamp 4 + type 1 + server_id 4 + total_size 4 + master_position 4 + flags 2 + --------------------------------- + POST HEADER: + slave_proxy_id 4 + exec_time 4 + db_len 1 ++ query_len 2 (see Note 1) + error_code 2 + status_vars_len 2 ++ tables_info_len 2 (see Note 2) + --------------------------------- + BODY: + status_vars status_vars_len +- db db_len + 1 ++ db db_len (see Note 3) + query query_len ++ tables_info + + tables_info binary format + --------------------------------- + Name Size (bytes) + --------------------------------- + db_len 1 (see Note 4) + db db_len + table_name_len 1 + table_name table_name_len + ... + db_len 1 + db db_len + table_name_len 1 + table_name table_name_len + +NOTES +1. Currently Query_log_event format doesn't include 'query_len' because + it considers the query to extent to the end of the event. +2. If tables_info is not included in the event (--binlog-with-tables-info + option), tables_info_len = 0. +3. The trailing zero is redundant since the length is already known. +4. In case of db = current db, db_len = 0 and db = empty, because + current db is already included in the current event format. + +2. Where to get tables info from? +********************************* + +2.1. Case study: CREATE TABLE +****************************** + +*** CREATE TABLE table [SELECT ...] + + bool mysql_create_table_no_lock( + THD *thd, + const char *db, + const char *table_name, ...) + { + ... + // ------------------------------------- + // WL40: To be included in tables_info: + // * db, table_name + // * thd->lex->query_tables (tables refered to in + // the select-part; empty if no select-part) + // ------------------------------------- + write_bin_log(thd, TRUE, thd->query, thd->query_length); + } + +*** CREATE TABLE table LIKE src-table + + bool mysql_create_like_table( + ... + TABLE_LIST *table, + TABLE_LIST *src_table, + ...) + { + ... + if (thd->current_stmt_binlog_row_based) + { // RBR: In this case we don't replicate temp tables + if (!(create_info->options & HA_LEX_CREATE_TMP_TABLE)) + { + if (src_table->table->s->tmp_table) + { // CREATE normal-table LIKE temp-table: + + // Generate new query without LIKE-part + store_create_info(thd, table, &query, create_info, FALSE); + + // ------------------------------------- + // WL40: To include to tables_info: + // * table (src_table is not included) + // ------------------------------------- + write_bin_log(thd, TRUE, query.ptr(), query.length()); + } + else + { // CREATE normal-table LIKE normal-table + + // ------------------------------------- + // WL40: To include to log_tables_info: + // * table + // * src_table + // ------------------------------------- + write_bin_log(thd, TRUE, thd->query, thd->query_length); + } + } + // CREATE temp-table LIKE ... + // This case is not replicated + } + else + { // SBR: + // ------------------------------------- + // WL40: To include to tables_info: + // * table + // * src_table + // ------------------------------------- + write_bin_log(thd, TRUE, thd->query, thd->query_length); + } + } + +To be continued -=-=(Alexi - Mon, 02 Nov 2009, 11:34)=-=- Worked 2 hours on option 2.5 Worked 2 hours and estimate 48 hours remain (original estimate increased by 50 hours). -=-=(Alexi - Mon, 02 Nov 2009, 11:20)=-=- High-Level Specification modified. --- /tmp/wklog.40.old.4848 2009-11-02 11:20:24.000000000 +0200 +++ /tmp/wklog.40.new.4848 2009-11-02 11:20:24.000000000 +0200 @@ -90,3 +90,25 @@ It might be useful to integrate this with the code that already handles --replicate-ignore-db and similar slave options. +2.5 Extend Query Events With Tables Info +---------------------------------------- + +We could extend query events structure with a tables info - a list of tables +which the query refers to: + + <current query event structure> + tables_info_len + dbase_len dbase + table_len table + ... + dbase_len dbase + table_len table + +Note. In case of <dbase> = current data base, we can set dbase_len = 0 + and dbase = empty because current query event structure already + includes current data base name. + +Note. Possibly it is reasonable also to add a --binlog-with-tables-info + option which defines whether tables info must be included to the + query events. + ------------------------------------------------------------ -=-=(View All Progress Notes, 15 total)=-=- http://askmonty.org/worklog/index.pl?tid=40&nolimit=1 DESCRIPTION: Replication slave can be set to filter updates to certain tables with --replicate-[wild-]{do,ignore}-table options. This task is about adding similar functionality to mysqlbinlog. HIGH-LEVEL SPECIFICATION: 1. Context ---------- (See http://askmonty.org/wiki/index.php/Scratch/ReplicationOptions for global overview) At the moment, the server has these replication slave options: --replicate-do-table=db.tbl --replicate-ignore-table=db.tbl --replicate-wild-do-table=pattern.pattern --replicate-wild-ignore-table=pattern.pattern They affect both RBR and SBR events. SBR events are checked after the statement has been parsed, the server iterates over list of used tables and checks them againist --replicate instructions. What is interesting is that this scheme still allows to update the ignored table through a VIEW. 2. Table filtering in mysqlbinlog --------------------------------- Per-table filtering of RBR events is easy (as it is relatively easy to extract the name of the table that the event applies to). Per-table filtering of SBR events is hard, as generally it is not apparent which tables the statement refers to. This opens possible options: 2.1 Put the parser into mysqlbinlog ----------------------------------- Once we have a full parser in mysqlbinlog, we'll be able to check which tables are used by a statement, and will allow to show behaviour identical to those that one obtains when using --replicate-* slave options. (It is not clear how much effort is needed to put the parser into mysqlbinlog. Any guesses?) 2.2 Use dumb regexp match ------------------------- Use a really dumb approach. A query is considered to be modifying table X if it matches an expression CREATE TABLE $tablename DROP $tablename UPDATE ...$tablename ... SET // here '...' can't contain the word 'SET' DELETE ...$tablename ... WHERE // same as above ALTER TABLE $tablename .. etc (go get from the grammar) .. The advantage over doing the same in awk is that mysqlbinlog will also process RBR statements, and together with that will provide a working solution for those who are careful with their table names not mixing with string constants and such. (TODO: string constants are of particular concern as they come from [potentially hostile] users, unlike e.g. table aliases which come from [not hostile] developers. Remove also all string constants before attempting to do match?) 2.3 Have the master put annotations ----------------------------------- We could add a master option so that it injects into query a mark that tells which tables the query will affect, e.g. for the query UPDATE t1 LEFT JOIN db3.t2 ON ... WHERE ... the binlog will have /* !mysqlbinlog: updates t1,db3.t2 */ UPDATE t1 LEFT JOIN ... and further processing in mysqlbinlog will be trivial. 2.4 Implement server functionality to ignore certain tables ----------------------------------------------------------- We could add a general facility in the server to ignore certain tables: SET SESSION ignored_tables = "db1.t1,db2.t2"; This would work similar to --replicate-ignore-table, but in a general way not restricted to the slave SQL thread. It would then be trivial for mysqlbinlog to add such statements at the start of the output, or probably the user could just do it manually with no need for additional options for mysqlbinlog. It might be useful to integrate this with the code that already handles --replicate-ignore-db and similar slave options. 2.5 Extend Query Events With Tables Info ---------------------------------------- We could extend query events structure with a tables info - a list of tables which the query refers to: <current query event structure> tables_info_len dbase_len dbase table_len table ... dbase_len dbase table_len table Note. In case of <dbase> = current data base, we can set dbase_len = 0 and dbase = empty because current query event structure already includes current data base name. Note. Possibly it is reasonable also to add a --binlog-with-tables-info option which defines whether tables info must be included to the query events. LOW-LEVEL DESIGN: OPTION: 2.5 Extend Query Events With Tables Info ================================================ 1. Adding --binlog-with-tables-info option ****************************************** GLOBAL, read-only option. When set, Query events are to be written in the extended binary format which contains tables_info. When not set, Query events are to be written in usual format (without any changes). 2. Query event extended binary format ************************************* When --binlog-with-tables-info is set, Query events are writen to binary log in the following (extended) format. Query_log_event binary format --------------------------------- Name Size (bytes) --------------------------------- COMMON HEADER: timestamp 4 type 1 server_id 4 total_size 4 master_position 4 flags 2 --------------------------------- POST HEADER: slave_proxy_id 4 exec_time 4 db_len 1 error_code 2 status_vars_len 2 + query_len 2 (see Note 1) + tables_info_len 2 --------------------------------- BODY: status_vars status_vars_len - db db_len + 1 + db db_len (see Note 2) query query_len + tables_info tables_info binary format --------------------------------- Name Size (bytes) --------------------------------- db_len 1 (see Note 3) db db_len table_name_len 1 table_name table_name_len ... db_len 1 db db_len table_name_len 1 table_name table_name_len NOTES 1. In usual format, Query_log_event doesn't include 'query_len' because it considers the query to extent to the end of the event. 2. For 'db' (current db) the trailing zero is redundant since the length is already known. 3. In tables_info, db_len = 0 means that this is the current db. When reading Query events from binary log, we can recognize its format by its post-header length: in extended case the post-header includes 4 additional bytes. #define QUERY_HEADER_LEN (QUERY_HEADER_MINIMAL_LEN + 4) + #define QUERY_HEADER_LEN_EXT (QUERY_HEADER_LEN + 4) ... #define Q_STATUS_VARS_LEN_OFFSET 11 + #define Q_QUERY_LEN_OFFSET Q_STATUS_VARS_LEN_OFFSET + 2 + #define Q_QUERY_TABLES_INFO_LEN_OFFSET Q_QUERY_LEN_OFFSET + 2 *********************************************************************** HELP NEEDED *********************************************************************** The QUERY_HEADER_LEN is used in the definition of MAX_LOG_EVENT_HEADER: log_event.h ~~~~~~~~~~~ #define MAX_LOG_EVENT_HEADER ( /* in order of Query_log_event::write */ \ LOG_EVENT_HEADER_LEN + /* write_header */ \ QUERY_HEADER_LEN + /* write_data */ \ EXECUTE_LOAD_QUERY_EXTRA_HEADER_LEN + /*write_post_header_for_derived */ \ MAX_SIZE_LOG_EVENT_STATUS + /* status */ \ NAME_LEN + 1) which is used only for setting thd->variables.max_allowed_packet mysql->net.max_packet_size Looks like (but I am not quite sure) that QUERY_HEADER_LEN can simply (without making any other changes) be substituted in this definition by QUERY_HEADER_LEN_EXT. Below I list all places where MAX_LOG_EVENT_HEADER is used: slave.cc ~~~~~~~~ static int init_slave_thread(...) { ... /* Adding MAX_LOG_EVENT_HEADER_LEN to the max_allowed_packet on all slave threads, since a replication event can become this much larger than the corresponding packet (query) sent from client to master. */ thd->variables.max_allowed_packet= global_system_variables.max_allowed_packet + MAX_LOG_EVENT_HEADER; /* note, incr over the global not session var */ ... } pthread_handler_t handle_slave_io(...) { ... /* Adding MAX_LOG_EVENT_HEADER_LEN to the max_packet_size on the I/O thread, since a replication event can become this much larger than the corresponding packet (query) sent from client to master. */ mysql->net.max_packet_size= thd->net.max_packet_size+= MAX_LOG_EVENT_HEADER; ... } sql_repl.cc ~~~~~~~~~~~ void mysql_binlog_send(...) { ... /* Adding MAX_LOG_EVENT_HEADER_LEN, since a binlog event can become this larger than the corresponding packet (query) sent from client to master. */ thd->variables.max_allowed_packet+= MAX_LOG_EVENT_HEADER; ... } bool mysql_show_binlog_events(...) { ... /* to account binlog event header size */ thd->variables.max_allowed_packet+= MAX_LOG_EVENT_HEADER; ... } 3. Changes in log events ************************ 3.1. Format description event ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Changes needed here concern setting post-header length for Query events. This setting is done in Format description event constructor which creates the event for writing to binary log: if (opt_binlog_with_tables_info) post_header_len[QUERY_EVENT - 1] = QUERY_HEADER_LEN_EXT; else post_header_len[QUERY_EVENT - 1] = QUERY_HEADER_LEN; This change is to be done only for case binlog_ver = 4. NOTE. The refered above constructor is allowed to be invoked in a client context for creating "artificial" Format description events in case of MySQL < 5.0 (e.g. see mysqlbinlog code). To avoid compilation problems (because of 'binlog_with_tables_info') and taking into account the "MySQL < 5.0" restriction, we have to #ifdef out the above code in following manner: switch (binlog_ver) { + #ifndef MYSQL_CLIENT case 4: /* MySQL 5.0 and higher */ ... break; + #endif case 1: case 3: ... } 3.2. Query event ~~~~~~~~~~~~~~~~ Changes needed here include adding tables_info and tables_info_len members (member for query length already exists) and modifying the following function-members: Query_log_event(buf) constructor -------------------------------- [Parses binary format written to the 'buf'] Getting post-header length from the Format description event (passed to the constructor as an argument), define whether buf contains an extended or usual Query event and parse the buf contents accordingly. NOTE. Defining Query event format here should be done with taking into account that this constructor can be called within a Query-derived event with the event_type argument != QUERY_EVENT. Query_log_event(thd) constructor -------------------------------- [Creates the event for binlogging] In case of opt_binlog_with_tables_info = TRUE, set additionally query_len, tables_info_len, and tables_info members (the constructor is to have an additional 'tables_info' argument). write() function ---------------- [Writes the event to binlog] In case of opt_binlog_with_tables_info = TRUE, write additional members (query_len, tables_info_len, and tables_info) to binary log. Also write corresponding whole event length to the common-header. <To be continued> 4. Where to get tables info from? ********************************* 4.1. Case study: CREATE TABLE ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ *** CREATE TABLE table [SELECT ...] bool mysql_create_table_no_lock( THD *thd, const char *db, const char *table_name, ...) { ... // ------------------------------------- // WL40: To be included in tables_info: // * db, table_name // * thd->lex->query_tables (tables refered to in // the select-part; empty if no select-part) // ------------------------------------- write_bin_log(thd, TRUE, thd->query, thd->query_length); } *** CREATE TABLE table LIKE src-table bool mysql_create_like_table( ... TABLE_LIST *table, TABLE_LIST *src_table, ...) { ... if (thd->current_stmt_binlog_row_based) { // RBR: In this case we don't replicate temp tables if (!(create_info->options & HA_LEX_CREATE_TMP_TABLE)) { if (src_table->table->s->tmp_table) { // CREATE normal-table LIKE temp-table: // Generate new query without LIKE-part store_create_info(thd, table, &query, create_info, FALSE); // ------------------------------------- // WL40: To include to tables_info: // * table (src_table is not included) // ------------------------------------- write_bin_log(thd, TRUE, query.ptr(), query.length()); } else { // CREATE normal-table LIKE normal-table // ------------------------------------- // WL40: To include to log_tables_info: // * table // * src_table // ------------------------------------- write_bin_log(thd, TRUE, thd->query, thd->query_length); } } // CREATE temp-table LIKE ... // This case is not replicated } else { // SBR: // ------------------------------------- // WL40: To include to tables_info: // * table // * src_table // ------------------------------------- write_bin_log(thd, TRUE, thd->query, thd->query_length); } } <To be continued> ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Guest): Add a mysqlbinlog option to filter updates to certain tables (40)
by worklog-noreply＠askmonty.org 25 Nov '09

25 Nov '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Add a mysqlbinlog option to filter updates to certain tables CREATION DATE..: Mon, 10 Aug 2009, 13:25 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: Psergey CATEGORY.......: Server-Sprint TASK ID........: 40 (http://askmonty.org/worklog/?tid=40) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 32 ESTIMATE.......: 32 (hours remain) ORIG. ESTIMATE.: 48 PROGRESS NOTES: -=-=(Guest - Wed, 25 Nov 2009, 11:41)=-=- Category updated. --- /tmp/wklog.40.old.5737 2009-11-25 11:41:03.000000000 +0200 +++ /tmp/wklog.40.new.5737 2009-11-25 11:41:03.000000000 +0200 @@ -1 +1 @@ -Server-RawIdeaBin +Server-Sprint -=-=(Bothorsen - Tue, 17 Nov 2009, 17:20)=-=- Alex is closer to a working patch now. Worked 14 hours and estimate 32 hours remain (original estimate unchanged). -=-=(Bothorsen - Thu, 12 Nov 2009, 13:13)=-=- Work hours by Alexi and Bo + estimated time for the task. Worked 16 hours and estimate 46 hours remain (original estimate increased by 14 hours). -=-=(Alexi - Sun, 08 Nov 2009, 15:18)=-=- Low Level Design modified. --- /tmp/wklog.40.old.15787 2009-11-08 15:18:11.000000000 +0200 +++ /tmp/wklog.40.new.15787 2009-11-08 15:18:11.000000000 +0200 @@ -62,7 +62,7 @@ it considers the query to extent to the end of the event. 2. For 'db' (current db) the trailing zero is redundant since the length is already known. -3. db_len = 0 means that this is the current db. +3. In tables_info, db_len = 0 means that this is the current db. When reading Query events from binary log, we can recognize its format by its post-header length: in extended case the post-header includes 4 @@ -75,6 +75,77 @@ + #define Q_QUERY_LEN_OFFSET Q_STATUS_VARS_LEN_OFFSET + 2 + #define Q_QUERY_TABLES_INFO_LEN_OFFSET Q_QUERY_LEN_OFFSET + 2 + +*********************************************************************** +HELP NEEDED +*********************************************************************** +The QUERY_HEADER_LEN is used in the definition of MAX_LOG_EVENT_HEADER: + +log_event.h +~~~~~~~~~~~ +#define MAX_LOG_EVENT_HEADER ( /* in order of Query_log_event::write */ \ + LOG_EVENT_HEADER_LEN + /* write_header */ \ + QUERY_HEADER_LEN + /* write_data */ \ + EXECUTE_LOAD_QUERY_EXTRA_HEADER_LEN + /*write_post_header_for_derived */ \ + MAX_SIZE_LOG_EVENT_STATUS + /* status */ \ + NAME_LEN + 1) + +which is used only for setting + + thd->variables.max_allowed_packet + mysql->net.max_packet_size + +Looks like (but I am not quite sure) that QUERY_HEADER_LEN can simply +(without making any other changes) be substituted in this definition by +QUERY_HEADER_LEN_EXT. + +Below I list all places where MAX_LOG_EVENT_HEADER is used: + +slave.cc +~~~~~~~~ +static int init_slave_thread(...) +{ ... + /* + Adding MAX_LOG_EVENT_HEADER_LEN to the max_allowed_packet on all + slave threads, since a replication event can become this much larger + than the corresponding packet (query) sent from client to master. + */ + thd->variables.max_allowed_packet= global_system_variables.max_allowed_packet + + MAX_LOG_EVENT_HEADER; /* note, incr over the global not session var */ + ... +} +pthread_handler_t handle_slave_io(...) +{ ... + /* + Adding MAX_LOG_EVENT_HEADER_LEN to the max_packet_size on the I/O + thread, since a replication event can become this much larger than + the corresponding packet (query) sent from client to master. + */ + mysql->net.max_packet_size= thd->net.max_packet_size+= MAX_LOG_EVENT_HEADER; + ... +} + +sql_repl.cc +~~~~~~~~~~~ +void mysql_binlog_send(...) +{ ... + /* + Adding MAX_LOG_EVENT_HEADER_LEN, since a binlog event can become + this larger than the corresponding packet (query) sent + from client to master. + */ + thd->variables.max_allowed_packet+= MAX_LOG_EVENT_HEADER; + ... +} +bool mysql_show_binlog_events(...) +{ ... + /* + to account binlog event header size + */ + thd->variables.max_allowed_packet+= MAX_LOG_EVENT_HEADER; + ... +} + 3. Changes in log events ************************ @@ -84,7 +155,7 @@ This setting is done in Format description event constructor which creates the event for writing to binary log: - if (binlog_with_tables_info) + if (opt_binlog_with_tables_info) post_header_len[QUERY_EVENT - 1] = QUERY_HEADER_LEN_EXT; else post_header_len[QUERY_EVENT - 1] = QUERY_HEADER_LEN; @@ -99,12 +170,12 @@ following manner: switch (binlog_ver) { - case 4: /* MySQL 5.0 and higher */ + #ifndef MYSQL_CLIENT + case 4: /* MySQL 5.0 and higher */ ... -+ #else -+ <error> + break; + #endif + case 1: case 3: ... @@ -132,7 +203,7 @@ -------------------------------- [Creates the event for binlogging] -In case of binlog_with_tables_info = TRUE, set additionally query_len, +In case of opt_binlog_with_tables_info = TRUE, set additionally query_len, tables_info_len, and tables_info members (the constructor is to have an additional 'tables_info' argument). @@ -140,7 +211,7 @@ ---------------- [Writes the event to binlog] -In case of binlog_with_tables_info = TRUE, write additional members +In case of opt_binlog_with_tables_info = TRUE, write additional members (query_len, tables_info_len, and tables_info) to binary log. Also write corresponding whole event length to the common-header. -=-=(Alexi - Sun, 08 Nov 2009, 10:40)=-=- Low Level Design modified. --- /tmp/wklog.40.old.5055 2009-11-08 08:40:02.000000000 +0000 +++ /tmp/wklog.40.new.5055 2009-11-08 08:40:02.000000000 +0000 @@ -3,6 +3,7 @@ 1. Adding --binlog-with-tables-info option ****************************************** +GLOBAL, read-only option. When set, Query events are to be written in the extended binary format which contains tables_info. When not set, Query events -=-=(Alexi - Thu, 05 Nov 2009, 12:37)=-=- Low Level Design modified. --- /tmp/wklog.40.old.11441 2009-11-05 12:37:16.000000000 +0200 +++ /tmp/wklog.40.new.11441 2009-11-05 12:37:16.000000000 +0200 @@ -1,9 +1,18 @@ OPTION: 2.5 Extend Query Events With Tables Info ================================================ -1. Query_log_event Binary Format -******************************** -Changes to be done: +1. Adding --binlog-with-tables-info option +****************************************** + +When set, Query events are to be written in the extended binary +format which contains tables_info. When not set, Query events +are to be written in usual format (without any changes). + +2. Query event extended binary format +************************************* + +When --binlog-with-tables-info is set, Query events are writen +to binary log in the following (extended) format. Query_log_event binary format --------------------------------- @@ -24,12 +33,12 @@ error_code 2 status_vars_len 2 + query_len 2 (see Note 1) -+ tables_info_len 2 (see Note 2) ++ tables_info_len 2 --------------------------------- BODY: status_vars status_vars_len - db db_len + 1 -+ db db_len (see Note 3) ++ db db_len (see Note 2) query query_len + tables_info @@ -37,7 +46,7 @@ --------------------------------- Name Size (bytes) --------------------------------- - db_len 1 (see Note 4) + db_len 1 (see Note 3) db db_len table_name_len 1 table_name table_name_len @@ -48,19 +57,99 @@ table_name table_name_len NOTES -1. Currently Query_log_event format doesn't include 'query_len' because +1. In usual format, Query_log_event doesn't include 'query_len' because it considers the query to extent to the end of the event. -2. If tables_info is not included in the event (--binlog-with-tables-info - option), tables_info_len = 0. -3. The trailing zero is redundant since the length is already known. -4. In case of db = current db, db_len = 0 and db = empty, because - current db is already included in the current event format. +2. For 'db' (current db) the trailing zero is redundant since the length + is already known. +3. db_len = 0 means that this is the current db. + +When reading Query events from binary log, we can recognize its format +by its post-header length: in extended case the post-header includes 4 +additional bytes. + + #define QUERY_HEADER_LEN (QUERY_HEADER_MINIMAL_LEN + 4) ++ #define QUERY_HEADER_LEN_EXT (QUERY_HEADER_LEN + 4) + ... + #define Q_STATUS_VARS_LEN_OFFSET 11 ++ #define Q_QUERY_LEN_OFFSET Q_STATUS_VARS_LEN_OFFSET + 2 ++ #define Q_QUERY_TABLES_INFO_LEN_OFFSET Q_QUERY_LEN_OFFSET + 2 + +3. Changes in log events +************************ + +3.1. Format description event +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ +Changes needed here concern setting post-header length for Query events. +This setting is done in Format description event constructor which creates +the event for writing to binary log: + + if (binlog_with_tables_info) + post_header_len[QUERY_EVENT - 1] = QUERY_HEADER_LEN_EXT; + else + post_header_len[QUERY_EVENT - 1] = QUERY_HEADER_LEN; + +This change is to be done only for case binlog_ver = 4. + +NOTE. The refered above constructor is allowed to be invoked in a client +context for creating "artificial" Format description events in case of +MySQL < 5.0 (e.g. see mysqlbinlog code). To avoid compilation problems +(because of 'binlog_with_tables_info') and taking into account the +"MySQL < 5.0" restriction, we have to #ifdef out the above code in +following manner: + + switch (binlog_ver) { + case 4: /* MySQL 5.0 and higher */ ++ #ifndef MYSQL_CLIENT + ... ++ #else ++ <error> ++ #endif + case 1: + case 3: + ... + } + +3.2. Query event +~~~~~~~~~~~~~~~~ +Changes needed here include adding tables_info and tables_info_len +members (member for query length already exists) and modifying the +following function-members: + +Query_log_event(buf) constructor +-------------------------------- +[Parses binary format written to the 'buf'] + +Getting post-header length from the Format description event (passed +to the constructor as an argument), define whether buf contains an +extended or usual Query event and parse the buf contents accordingly. + +NOTE. Defining Query event format here should be done with taking into +account that this constructor can be called within a Query-derived +event with the event_type argument != QUERY_EVENT. + +Query_log_event(thd) constructor +-------------------------------- +[Creates the event for binlogging] + +In case of binlog_with_tables_info = TRUE, set additionally query_len, +tables_info_len, and tables_info members (the constructor is to have +an additional 'tables_info' argument). + +write() function +---------------- +[Writes the event to binlog] + +In case of binlog_with_tables_info = TRUE, write additional members +(query_len, tables_info_len, and tables_info) to binary log. Also +write corresponding whole event length to the common-header. + +<To be continued> -2. Where to get tables info from? +4. Where to get tables info from? ********************************* -2.1. Case study: CREATE TABLE -****************************** +4.1. Case study: CREATE TABLE +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ *** CREATE TABLE table [SELECT ...] @@ -129,4 +218,4 @@ } } -To be continued +<To be continued> -=-=(Alexi - Wed, 04 Nov 2009, 10:21)=-=- Low Level Design modified. --- /tmp/wklog.40.old.6734 2009-11-04 10:21:20.000000000 +0200 +++ /tmp/wklog.40.new.6734 2009-11-04 10:21:20.000000000 +0200 @@ -21,9 +21,9 @@ slave_proxy_id 4 exec_time 4 db_len 1 -+ query_len 2 (see Note 1) error_code 2 status_vars_len 2 ++ query_len 2 (see Note 1) + tables_info_len 2 (see Note 2) --------------------------------- BODY: -=-=(Alexi - Tue, 03 Nov 2009, 11:19)=-=- Low Level Design modified. --- /tmp/wklog.40.old.7187 2009-11-03 11:19:22.000000000 +0200 +++ /tmp/wklog.40.new.7187 2009-11-03 11:19:22.000000000 +0200 @@ -1 +1,132 @@ +OPTION: 2.5 Extend Query Events With Tables Info +================================================ +1. Query_log_event Binary Format +******************************** +Changes to be done: + + Query_log_event binary format + --------------------------------- + Name Size (bytes) + --------------------------------- + COMMON HEADER: + timestamp 4 + type 1 + server_id 4 + total_size 4 + master_position 4 + flags 2 + --------------------------------- + POST HEADER: + slave_proxy_id 4 + exec_time 4 + db_len 1 ++ query_len 2 (see Note 1) + error_code 2 + status_vars_len 2 ++ tables_info_len 2 (see Note 2) + --------------------------------- + BODY: + status_vars status_vars_len +- db db_len + 1 ++ db db_len (see Note 3) + query query_len ++ tables_info + + tables_info binary format + --------------------------------- + Name Size (bytes) + --------------------------------- + db_len 1 (see Note 4) + db db_len + table_name_len 1 + table_name table_name_len + ... + db_len 1 + db db_len + table_name_len 1 + table_name table_name_len + +NOTES +1. Currently Query_log_event format doesn't include 'query_len' because + it considers the query to extent to the end of the event. +2. If tables_info is not included in the event (--binlog-with-tables-info + option), tables_info_len = 0. +3. The trailing zero is redundant since the length is already known. +4. In case of db = current db, db_len = 0 and db = empty, because + current db is already included in the current event format. + +2. Where to get tables info from? +********************************* + +2.1. Case study: CREATE TABLE +****************************** + +*** CREATE TABLE table [SELECT ...] + + bool mysql_create_table_no_lock( + THD *thd, + const char *db, + const char *table_name, ...) + { + ... + // ------------------------------------- + // WL40: To be included in tables_info: + // * db, table_name + // * thd->lex->query_tables (tables refered to in + // the select-part; empty if no select-part) + // ------------------------------------- + write_bin_log(thd, TRUE, thd->query, thd->query_length); + } + +*** CREATE TABLE table LIKE src-table + + bool mysql_create_like_table( + ... + TABLE_LIST *table, + TABLE_LIST *src_table, + ...) + { + ... + if (thd->current_stmt_binlog_row_based) + { // RBR: In this case we don't replicate temp tables + if (!(create_info->options & HA_LEX_CREATE_TMP_TABLE)) + { + if (src_table->table->s->tmp_table) + { // CREATE normal-table LIKE temp-table: + + // Generate new query without LIKE-part + store_create_info(thd, table, &query, create_info, FALSE); + + // ------------------------------------- + // WL40: To include to tables_info: + // * table (src_table is not included) + // ------------------------------------- + write_bin_log(thd, TRUE, query.ptr(), query.length()); + } + else + { // CREATE normal-table LIKE normal-table + + // ------------------------------------- + // WL40: To include to log_tables_info: + // * table + // * src_table + // ------------------------------------- + write_bin_log(thd, TRUE, thd->query, thd->query_length); + } + } + // CREATE temp-table LIKE ... + // This case is not replicated + } + else + { // SBR: + // ------------------------------------- + // WL40: To include to tables_info: + // * table + // * src_table + // ------------------------------------- + write_bin_log(thd, TRUE, thd->query, thd->query_length); + } + } + +To be continued -=-=(Alexi - Mon, 02 Nov 2009, 11:34)=-=- Worked 2 hours on option 2.5 Worked 2 hours and estimate 48 hours remain (original estimate increased by 50 hours). -=-=(Alexi - Mon, 02 Nov 2009, 11:20)=-=- High-Level Specification modified. --- /tmp/wklog.40.old.4848 2009-11-02 11:20:24.000000000 +0200 +++ /tmp/wklog.40.new.4848 2009-11-02 11:20:24.000000000 +0200 @@ -90,3 +90,25 @@ It might be useful to integrate this with the code that already handles --replicate-ignore-db and similar slave options. +2.5 Extend Query Events With Tables Info +---------------------------------------- + +We could extend query events structure with a tables info - a list of tables +which the query refers to: + + <current query event structure> + tables_info_len + dbase_len dbase + table_len table + ... + dbase_len dbase + table_len table + +Note. In case of <dbase> = current data base, we can set dbase_len = 0 + and dbase = empty because current query event structure already + includes current data base name. + +Note. Possibly it is reasonable also to add a --binlog-with-tables-info + option which defines whether tables info must be included to the + query events. + ------------------------------------------------------------ -=-=(View All Progress Notes, 15 total)=-=- http://askmonty.org/worklog/index.pl?tid=40&nolimit=1 DESCRIPTION: Replication slave can be set to filter updates to certain tables with --replicate-[wild-]{do,ignore}-table options. This task is about adding similar functionality to mysqlbinlog. HIGH-LEVEL SPECIFICATION: 1. Context ---------- (See http://askmonty.org/wiki/index.php/Scratch/ReplicationOptions for global overview) At the moment, the server has these replication slave options: --replicate-do-table=db.tbl --replicate-ignore-table=db.tbl --replicate-wild-do-table=pattern.pattern --replicate-wild-ignore-table=pattern.pattern They affect both RBR and SBR events. SBR events are checked after the statement has been parsed, the server iterates over list of used tables and checks them againist --replicate instructions. What is interesting is that this scheme still allows to update the ignored table through a VIEW. 2. Table filtering in mysqlbinlog --------------------------------- Per-table filtering of RBR events is easy (as it is relatively easy to extract the name of the table that the event applies to). Per-table filtering of SBR events is hard, as generally it is not apparent which tables the statement refers to. This opens possible options: 2.1 Put the parser into mysqlbinlog ----------------------------------- Once we have a full parser in mysqlbinlog, we'll be able to check which tables are used by a statement, and will allow to show behaviour identical to those that one obtains when using --replicate-* slave options. (It is not clear how much effort is needed to put the parser into mysqlbinlog. Any guesses?) 2.2 Use dumb regexp match ------------------------- Use a really dumb approach. A query is considered to be modifying table X if it matches an expression CREATE TABLE $tablename DROP $tablename UPDATE ...$tablename ... SET // here '...' can't contain the word 'SET' DELETE ...$tablename ... WHERE // same as above ALTER TABLE $tablename .. etc (go get from the grammar) .. The advantage over doing the same in awk is that mysqlbinlog will also process RBR statements, and together with that will provide a working solution for those who are careful with their table names not mixing with string constants and such. (TODO: string constants are of particular concern as they come from [potentially hostile] users, unlike e.g. table aliases which come from [not hostile] developers. Remove also all string constants before attempting to do match?) 2.3 Have the master put annotations ----------------------------------- We could add a master option so that it injects into query a mark that tells which tables the query will affect, e.g. for the query UPDATE t1 LEFT JOIN db3.t2 ON ... WHERE ... the binlog will have /* !mysqlbinlog: updates t1,db3.t2 */ UPDATE t1 LEFT JOIN ... and further processing in mysqlbinlog will be trivial. 2.4 Implement server functionality to ignore certain tables ----------------------------------------------------------- We could add a general facility in the server to ignore certain tables: SET SESSION ignored_tables = "db1.t1,db2.t2"; This would work similar to --replicate-ignore-table, but in a general way not restricted to the slave SQL thread. It would then be trivial for mysqlbinlog to add such statements at the start of the output, or probably the user could just do it manually with no need for additional options for mysqlbinlog. It might be useful to integrate this with the code that already handles --replicate-ignore-db and similar slave options. 2.5 Extend Query Events With Tables Info ---------------------------------------- We could extend query events structure with a tables info - a list of tables which the query refers to: <current query event structure> tables_info_len dbase_len dbase table_len table ... dbase_len dbase table_len table Note. In case of <dbase> = current data base, we can set dbase_len = 0 and dbase = empty because current query event structure already includes current data base name. Note. Possibly it is reasonable also to add a --binlog-with-tables-info option which defines whether tables info must be included to the query events. LOW-LEVEL DESIGN: OPTION: 2.5 Extend Query Events With Tables Info ================================================ 1. Adding --binlog-with-tables-info option ****************************************** GLOBAL, read-only option. When set, Query events are to be written in the extended binary format which contains tables_info. When not set, Query events are to be written in usual format (without any changes). 2. Query event extended binary format ************************************* When --binlog-with-tables-info is set, Query events are writen to binary log in the following (extended) format. Query_log_event binary format --------------------------------- Name Size (bytes) --------------------------------- COMMON HEADER: timestamp 4 type 1 server_id 4 total_size 4 master_position 4 flags 2 --------------------------------- POST HEADER: slave_proxy_id 4 exec_time 4 db_len 1 error_code 2 status_vars_len 2 + query_len 2 (see Note 1) + tables_info_len 2 --------------------------------- BODY: status_vars status_vars_len - db db_len + 1 + db db_len (see Note 2) query query_len + tables_info tables_info binary format --------------------------------- Name Size (bytes) --------------------------------- db_len 1 (see Note 3) db db_len table_name_len 1 table_name table_name_len ... db_len 1 db db_len table_name_len 1 table_name table_name_len NOTES 1. In usual format, Query_log_event doesn't include 'query_len' because it considers the query to extent to the end of the event. 2. For 'db' (current db) the trailing zero is redundant since the length is already known. 3. In tables_info, db_len = 0 means that this is the current db. When reading Query events from binary log, we can recognize its format by its post-header length: in extended case the post-header includes 4 additional bytes. #define QUERY_HEADER_LEN (QUERY_HEADER_MINIMAL_LEN + 4) + #define QUERY_HEADER_LEN_EXT (QUERY_HEADER_LEN + 4) ... #define Q_STATUS_VARS_LEN_OFFSET 11 + #define Q_QUERY_LEN_OFFSET Q_STATUS_VARS_LEN_OFFSET + 2 + #define Q_QUERY_TABLES_INFO_LEN_OFFSET Q_QUERY_LEN_OFFSET + 2 *********************************************************************** HELP NEEDED *********************************************************************** The QUERY_HEADER_LEN is used in the definition of MAX_LOG_EVENT_HEADER: log_event.h ~~~~~~~~~~~ #define MAX_LOG_EVENT_HEADER ( /* in order of Query_log_event::write */ \ LOG_EVENT_HEADER_LEN + /* write_header */ \ QUERY_HEADER_LEN + /* write_data */ \ EXECUTE_LOAD_QUERY_EXTRA_HEADER_LEN + /*write_post_header_for_derived */ \ MAX_SIZE_LOG_EVENT_STATUS + /* status */ \ NAME_LEN + 1) which is used only for setting thd->variables.max_allowed_packet mysql->net.max_packet_size Looks like (but I am not quite sure) that QUERY_HEADER_LEN can simply (without making any other changes) be substituted in this definition by QUERY_HEADER_LEN_EXT. Below I list all places where MAX_LOG_EVENT_HEADER is used: slave.cc ~~~~~~~~ static int init_slave_thread(...) { ... /* Adding MAX_LOG_EVENT_HEADER_LEN to the max_allowed_packet on all slave threads, since a replication event can become this much larger than the corresponding packet (query) sent from client to master. */ thd->variables.max_allowed_packet= global_system_variables.max_allowed_packet + MAX_LOG_EVENT_HEADER; /* note, incr over the global not session var */ ... } pthread_handler_t handle_slave_io(...) { ... /* Adding MAX_LOG_EVENT_HEADER_LEN to the max_packet_size on the I/O thread, since a replication event can become this much larger than the corresponding packet (query) sent from client to master. */ mysql->net.max_packet_size= thd->net.max_packet_size+= MAX_LOG_EVENT_HEADER; ... } sql_repl.cc ~~~~~~~~~~~ void mysql_binlog_send(...) { ... /* Adding MAX_LOG_EVENT_HEADER_LEN, since a binlog event can become this larger than the corresponding packet (query) sent from client to master. */ thd->variables.max_allowed_packet+= MAX_LOG_EVENT_HEADER; ... } bool mysql_show_binlog_events(...) { ... /* to account binlog event header size */ thd->variables.max_allowed_packet+= MAX_LOG_EVENT_HEADER; ... } 3. Changes in log events ************************ 3.1. Format description event ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Changes needed here concern setting post-header length for Query events. This setting is done in Format description event constructor which creates the event for writing to binary log: if (opt_binlog_with_tables_info) post_header_len[QUERY_EVENT - 1] = QUERY_HEADER_LEN_EXT; else post_header_len[QUERY_EVENT - 1] = QUERY_HEADER_LEN; This change is to be done only for case binlog_ver = 4. NOTE. The refered above constructor is allowed to be invoked in a client context for creating "artificial" Format description events in case of MySQL < 5.0 (e.g. see mysqlbinlog code). To avoid compilation problems (because of 'binlog_with_tables_info') and taking into account the "MySQL < 5.0" restriction, we have to #ifdef out the above code in following manner: switch (binlog_ver) { + #ifndef MYSQL_CLIENT case 4: /* MySQL 5.0 and higher */ ... break; + #endif case 1: case 3: ... } 3.2. Query event ~~~~~~~~~~~~~~~~ Changes needed here include adding tables_info and tables_info_len members (member for query length already exists) and modifying the following function-members: Query_log_event(buf) constructor -------------------------------- [Parses binary format written to the 'buf'] Getting post-header length from the Format description event (passed to the constructor as an argument), define whether buf contains an extended or usual Query event and parse the buf contents accordingly. NOTE. Defining Query event format here should be done with taking into account that this constructor can be called within a Query-derived event with the event_type argument != QUERY_EVENT. Query_log_event(thd) constructor -------------------------------- [Creates the event for binlogging] In case of opt_binlog_with_tables_info = TRUE, set additionally query_len, tables_info_len, and tables_info members (the constructor is to have an additional 'tables_info' argument). write() function ---------------- [Writes the event to binlog] In case of opt_binlog_with_tables_info = TRUE, write additional members (query_len, tables_info_len, and tables_info) to binary log. Also write corresponding whole event length to the common-header. <To be continued> 4. Where to get tables info from? ********************************* 4.1. Case study: CREATE TABLE ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ *** CREATE TABLE table [SELECT ...] bool mysql_create_table_no_lock( THD *thd, const char *db, const char *table_name, ...) { ... // ------------------------------------- // WL40: To be included in tables_info: // * db, table_name // * thd->lex->query_tables (tables refered to in // the select-part; empty if no select-part) // ------------------------------------- write_bin_log(thd, TRUE, thd->query, thd->query_length); } *** CREATE TABLE table LIKE src-table bool mysql_create_like_table( ... TABLE_LIST *table, TABLE_LIST *src_table, ...) { ... if (thd->current_stmt_binlog_row_based) { // RBR: In this case we don't replicate temp tables if (!(create_info->options & HA_LEX_CREATE_TMP_TABLE)) { if (src_table->table->s->tmp_table) { // CREATE normal-table LIKE temp-table: // Generate new query without LIKE-part store_create_info(thd, table, &query, create_info, FALSE); // ------------------------------------- // WL40: To include to tables_info: // * table (src_table is not included) // ------------------------------------- write_bin_log(thd, TRUE, query.ptr(), query.length()); } else { // CREATE normal-table LIKE normal-table // ------------------------------------- // WL40: To include to log_tables_info: // * table // * src_table // ------------------------------------- write_bin_log(thd, TRUE, thd->query, thd->query_length); } } // CREATE temp-table LIKE ... // This case is not replicated } else { // SBR: // ------------------------------------- // WL40: To include to tables_info: // * table // * src_table // ------------------------------------- write_bin_log(thd, TRUE, thd->query, thd->query_length); } } <To be continued> ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0