- developers - lists.mariadb.org

[Maria-developers] Updated (by Shinguz): MySQL/Maria Wait Interface (102)
by worklog-noreply＠askmonty.org 10 Mar '10

10 Mar '10

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: MySQL/Maria Wait Interface CREATION DATE..: Wed, 10 Mar 2010, 10:53 SUPERVISOR.....: Bothorsen IMPLEMENTOR....: COPIES TO......: CATEGORY.......: Client-BackLog TASK ID........: 102 (http://askmonty.org/worklog/?tid=102) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Shinguz - Wed, 10 Mar 2010, 12:41)=-=- High Level Description modified. --- /tmp/wklog.102.old.2095 2010-03-10 12:41:15.000000000 +0000 +++ /tmp/wklog.102.new.2095 2010-03-10 12:41:15.000000000 +0000 @@ -10,6 +10,9 @@ http://www.xaprb.com/blog/2009/11/07/a-review-of-optimizing-oracle-performa… http://www.facebook.com/note.php?note_id=355839540932 +Possibly the MySQL performance schema can cover part/all of it yet? +http://dev.mysql.com/doc/performance-schema/en/index.html + Good sources to start with: http://www.amazon.com/Oracle-Wait-Interface-Performance-Diagnostics/dp/0072… http://books.google.ch/books?id=14OmJzfCfXMC&pg=PA6&lpg=PA6&dq=owi+is+a+per… -=-=(Shinguz - Wed, 10 Mar 2010, 10:55)=-=- Title modified. --- /tmp/wklog.102.old.29019 2010-03-10 10:55:18.000000000 +0000 +++ /tmp/wklog.102.new.29019 2010-03-10 10:55:18.000000000 +0000 @@ -1 +1 @@ -MySQL Wait Interface +MySQL/Maria Wait Interface -=-=(Shinguz - Wed, 10 Mar 2010, 10:55)=-=- Version updated. --- /tmp/wklog.102.old.29019 2010-03-10 10:55:18.000000000 +0000 +++ /tmp/wklog.102.new.29019 2010-03-10 10:55:18.000000000 +0000 @@ -1 +1 @@ -Benchmarks-3.0 +Server-9.x DESCRIPTION: What lacks in MySQL is something similar like the Oracle Wait Interface. The basic idea is, that every connection/thread (and the whole system) shows (on request) where it spent how much time doing what. They implemented the wait interface (which is possibly a sub-set of Sun's dtrace) where you can see where you loose your time. Marc Callaghan and Baron Schwartz blogged in this direction recently: http://www.xaprb.com/blog/2009/11/07/a-review-of-optimizing-oracle-performa… http://www.facebook.com/note.php?note_id=355839540932 Possibly the MySQL performance schema can cover part/all of it yet? http://dev.mysql.com/doc/performance-schema/en/index.html Good sources to start with: http://www.amazon.com/Oracle-Wait-Interface-Performance-Diagnostics/dp/0072… http://books.google.ch/books?id=14OmJzfCfXMC&pg=PA6&lpg=PA6&dq=owi+is+a+per… http://www.oracle.com/technology/books/pdfs/sample%20chapter%202729x.pdf http://www.amazon.com/Optimizing-Oracle-Performance-Cary-Millsap/dp/0596005… ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Shinguz): MySQL/Maria Wait Interface (102)
by worklog-noreply＠askmonty.org 10 Mar '10

10 Mar '10

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: MySQL/Maria Wait Interface CREATION DATE..: Wed, 10 Mar 2010, 10:53 SUPERVISOR.....: Bothorsen IMPLEMENTOR....: COPIES TO......: CATEGORY.......: Client-BackLog TASK ID........: 102 (http://askmonty.org/worklog/?tid=102) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Shinguz - Wed, 10 Mar 2010, 12:41)=-=- High Level Description modified. --- /tmp/wklog.102.old.2095 2010-03-10 12:41:15.000000000 +0000 +++ /tmp/wklog.102.new.2095 2010-03-10 12:41:15.000000000 +0000 @@ -10,6 +10,9 @@ http://www.xaprb.com/blog/2009/11/07/a-review-of-optimizing-oracle-performa… http://www.facebook.com/note.php?note_id=355839540932 +Possibly the MySQL performance schema can cover part/all of it yet? +http://dev.mysql.com/doc/performance-schema/en/index.html + Good sources to start with: http://www.amazon.com/Oracle-Wait-Interface-Performance-Diagnostics/dp/0072… http://books.google.ch/books?id=14OmJzfCfXMC&pg=PA6&lpg=PA6&dq=owi+is+a+per… -=-=(Shinguz - Wed, 10 Mar 2010, 10:55)=-=- Title modified. --- /tmp/wklog.102.old.29019 2010-03-10 10:55:18.000000000 +0000 +++ /tmp/wklog.102.new.29019 2010-03-10 10:55:18.000000000 +0000 @@ -1 +1 @@ -MySQL Wait Interface +MySQL/Maria Wait Interface -=-=(Shinguz - Wed, 10 Mar 2010, 10:55)=-=- Version updated. --- /tmp/wklog.102.old.29019 2010-03-10 10:55:18.000000000 +0000 +++ /tmp/wklog.102.new.29019 2010-03-10 10:55:18.000000000 +0000 @@ -1 +1 @@ -Benchmarks-3.0 +Server-9.x DESCRIPTION: What lacks in MySQL is something similar like the Oracle Wait Interface. The basic idea is, that every connection/thread (and the whole system) shows (on request) where it spent how much time doing what. They implemented the wait interface (which is possibly a sub-set of Sun's dtrace) where you can see where you loose your time. Marc Callaghan and Baron Schwartz blogged in this direction recently: http://www.xaprb.com/blog/2009/11/07/a-review-of-optimizing-oracle-performa… http://www.facebook.com/note.php?note_id=355839540932 Possibly the MySQL performance schema can cover part/all of it yet? http://dev.mysql.com/doc/performance-schema/en/index.html Good sources to start with: http://www.amazon.com/Oracle-Wait-Interface-Performance-Diagnostics/dp/0072… http://books.google.ch/books?id=14OmJzfCfXMC&pg=PA6&lpg=PA6&dq=owi+is+a+per… http://www.oracle.com/technology/books/pdfs/sample%20chapter%202729x.pdf http://www.amazon.com/Optimizing-Oracle-Performance-Cary-Millsap/dp/0596005… ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Shinguz): Index usage tracker (103)
by worklog-noreply＠askmonty.org 10 Mar '10

10 Mar '10

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Index usage tracker CREATION DATE..: Wed, 10 Mar 2010, 11:29 SUPERVISOR.....: Bothorsen IMPLEMENTOR....: COPIES TO......: CATEGORY.......: Client-BackLog TASK ID........: 103 (http://askmonty.org/worklog/?tid=103) VERSION........: Benchmarks-3.0 STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Shinguz - Wed, 10 Mar 2010, 11:30)=-=- High Level Description modified. --- /tmp/wklog.103.old.30875 2010-03-10 11:30:45.000000000 +0000 +++ /tmp/wklog.103.new.30875 2010-03-10 11:30:45.000000000 +0000 @@ -35,3 +35,6 @@ Possibly this could/should be implemented on the handler interface level because there we know what we are touching? But I am not familiar with the code. + +If some more specification is need for implementation let me know and I will +collect the stuff from the sources mentioned above. DESCRIPTION: What indexes are needed is often easy to find. What is more difficult is to find which indexes are not used at all. Thus some statistics about the index usage would be nice. I think Percona has already something like this done some time ago. And big O has similar functionality. I could imagine something like this: +------------+------------+--------------+---------------------+---------------------+---------------------+------------+----------+-----------+-----------+------------+ | table_name | index_name | index_length | create_time | update_time | use_time | read_first | read_key | read_next | full_scan | range_scan | +------------+------------+--------------+---------------------+---------------------+---------------------+------------+----------+-----------+-----------+------------+ | test_table | idx1 | 1234567890 | 2010-01-01 00:00:00 | 2010-01-01 00:00:00 | 2010-03-10 11:34:56 | 1234 | 42560 | 2468 | 234 | 321 | | test_table | idx2 | 123456 | 2010-01-01 00:00:00 | NULL | NULL | 0 | 0 | 0 | 0 | 0 | | test_table | idx3 | 234561 | 2010-01-01 00:00:00 | 2010-03-10 11:12:34 | 2010-03-10 11:34:56 | 7890 | 89890 | 15780 | 678 | 321 | | test_table | idx4 | 345612 | 2010-01-01 00:00:00 | 2010-03-10 11:34:56 | NULL | 0 | 0 | 0 | 0 | 0 | | test_table | idx5 | 456123 | 2010-01-01 00:00:00 | 2010-03-10 06:56:12 | NULL | 0 | 0 | 0 | 0 | 0 | | test_table | idx6 | 561234 | 2010-01-01 00:00:00 | 2010-03-10 01:12:34 | 2010-03-10 11:34:42 | 3456 | 12356 | 6912 | 123 | 12 | +------------+------------+--------------+---------------------+---------------------+---------------------+------------+----------+-----------+-----------+------------+ Possibly this could/should be implemented on the handler interface level because there we know what we are touching? But I am not familiar with the code. If some more specification is need for implementation let me know and I will collect the stuff from the sources mentioned above. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Shinguz): Index usage tracker (103)
by worklog-noreply＠askmonty.org 10 Mar '10

10 Mar '10

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Index usage tracker CREATION DATE..: Wed, 10 Mar 2010, 11:29 SUPERVISOR.....: Bothorsen IMPLEMENTOR....: COPIES TO......: CATEGORY.......: Client-BackLog TASK ID........: 103 (http://askmonty.org/worklog/?tid=103) VERSION........: Benchmarks-3.0 STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Shinguz - Wed, 10 Mar 2010, 11:30)=-=- High Level Description modified. --- /tmp/wklog.103.old.30875 2010-03-10 11:30:45.000000000 +0000 +++ /tmp/wklog.103.new.30875 2010-03-10 11:30:45.000000000 +0000 @@ -35,3 +35,6 @@ Possibly this could/should be implemented on the handler interface level because there we know what we are touching? But I am not familiar with the code. + +If some more specification is need for implementation let me know and I will +collect the stuff from the sources mentioned above. DESCRIPTION: What indexes are needed is often easy to find. What is more difficult is to find which indexes are not used at all. Thus some statistics about the index usage would be nice. I think Percona has already something like this done some time ago. And big O has similar functionality. I could imagine something like this: +------------+------------+--------------+---------------------+---------------------+---------------------+------------+----------+-----------+-----------+------------+ | table_name | index_name | index_length | create_time | update_time | use_time | read_first | read_key | read_next | full_scan | range_scan | +------------+------------+--------------+---------------------+---------------------+---------------------+------------+----------+-----------+-----------+------------+ | test_table | idx1 | 1234567890 | 2010-01-01 00:00:00 | 2010-01-01 00:00:00 | 2010-03-10 11:34:56 | 1234 | 42560 | 2468 | 234 | 321 | | test_table | idx2 | 123456 | 2010-01-01 00:00:00 | NULL | NULL | 0 | 0 | 0 | 0 | 0 | | test_table | idx3 | 234561 | 2010-01-01 00:00:00 | 2010-03-10 11:12:34 | 2010-03-10 11:34:56 | 7890 | 89890 | 15780 | 678 | 321 | | test_table | idx4 | 345612 | 2010-01-01 00:00:00 | 2010-03-10 11:34:56 | NULL | 0 | 0 | 0 | 0 | 0 | | test_table | idx5 | 456123 | 2010-01-01 00:00:00 | 2010-03-10 06:56:12 | NULL | 0 | 0 | 0 | 0 | 0 | | test_table | idx6 | 561234 | 2010-01-01 00:00:00 | 2010-03-10 01:12:34 | 2010-03-10 11:34:42 | 3456 | 12356 | 6912 | 123 | 12 | +------------+------------+--------------+---------------------+---------------------+---------------------+------------+----------+-----------+-----------+------------+ Possibly this could/should be implemented on the handler interface level because there we know what we are touching? But I am not familiar with the code. If some more specification is need for implementation let me know and I will collect the stuff from the sources mentioned above. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] New (by Shinguz): Index usage tracker (103)
by worklog-noreply＠askmonty.org 10 Mar '10

10 Mar '10

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Index usage tracker CREATION DATE..: Wed, 10 Mar 2010, 11:29 SUPERVISOR.....: Bothorsen IMPLEMENTOR....: COPIES TO......: CATEGORY.......: Client-BackLog TASK ID........: 103 (http://askmonty.org/worklog/?tid=103) VERSION........: Benchmarks-3.0 STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: DESCRIPTION: What indexes are needed is often easy to find. What is more difficult is to find which indexes are not used at all. Thus some statistics about the index usage would be nice. I think Percona has already something like this done some time ago. And big O has similar functionality. I could imagine something like this: +------------+------------+--------------+---------------------+---------------------+---------------------+------------+----------+-----------+-----------+------------+ | table_name | index_name | index_length | create_time | update_time | use_time | read_first | read_key | read_next | full_scan | range_scan | +------------+------------+--------------+---------------------+---------------------+---------------------+------------+----------+-----------+-----------+------------+ | test_table | idx1 | 1234567890 | 2010-01-01 00:00:00 | 2010-01-01 00:00:00 | 2010-03-10 11:34:56 | 1234 | 42560 | 2468 | 234 | 321 | | test_table | idx2 | 123456 | 2010-01-01 00:00:00 | NULL | NULL | 0 | 0 | 0 | 0 | 0 | | test_table | idx3 | 234561 | 2010-01-01 00:00:00 | 2010-03-10 11:12:34 | 2010-03-10 11:34:56 | 7890 | 89890 | 15780 | 678 | 321 | | test_table | idx4 | 345612 | 2010-01-01 00:00:00 | 2010-03-10 11:34:56 | NULL | 0 | 0 | 0 | 0 | 0 | | test_table | idx5 | 456123 | 2010-01-01 00:00:00 | 2010-03-10 06:56:12 | NULL | 0 | 0 | 0 | 0 | 0 | | test_table | idx6 | 561234 | 2010-01-01 00:00:00 | 2010-03-10 01:12:34 | 2010-03-10 11:34:42 | 3456 | 12356 | 6912 | 123 | 12 | +------------+------------+--------------+---------------------+---------------------+---------------------+------------+----------+-----------+-----------+------------+ Possibly this could/should be implemented on the handler interface level because there we know what we are touching? But I am not familiar with the code. ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] [Branch ~maria-captains/maria/5.1] Rev 2828: Fix some compiler warnings seen in Buildbot.
by noreply＠launchpad.net 10 Mar '10

10 Mar '10

------------------------------------------------------------ revno: 2828 committer: knielsen(a)knielsen-hq.org branch nick: mariadb-5.1 timestamp: Wed 2010-03-10 11:32:14 +0100 message: Fix some compiler warnings seen in Buildbot. Add some extra error output and code cleanup in an attempt to fix/debug a rare random testsuite problem in check_warnings, where the exit code from mysqltest is somehow corrupted inside mysql-test-run.pl. modified: include/my_global.h mysql-test/lib/My/SafeProcess.pm mysql-test/mysql-test-run.pl sql/sql_lex.cc sql/table.cc storage/federatedx/ha_federatedx.cc storage/maria/ma_delete.c storage/maria/ma_loghandler.c storage/myisam/ft_stopwords.c storage/myisam/mi_write.c storage/xtradb/btr/btr0cur.c support-files/compiler_warnings.supp vio/viossl.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] Updated (by Shinguz): MySQL/Maria Wait Interface (102)
by worklog-noreply＠askmonty.org 10 Mar '10

10 Mar '10

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: MySQL/Maria Wait Interface CREATION DATE..: Wed, 10 Mar 2010, 10:53 SUPERVISOR.....: Bothorsen IMPLEMENTOR....: COPIES TO......: CATEGORY.......: Client-BackLog TASK ID........: 102 (http://askmonty.org/worklog/?tid=102) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Shinguz - Wed, 10 Mar 2010, 10:55)=-=- Title modified. --- /tmp/wklog.102.old.29019 2010-03-10 10:55:18.000000000 +0000 +++ /tmp/wklog.102.new.29019 2010-03-10 10:55:18.000000000 +0000 @@ -1 +1 @@ -MySQL Wait Interface +MySQL/Maria Wait Interface -=-=(Shinguz - Wed, 10 Mar 2010, 10:55)=-=- Version updated. --- /tmp/wklog.102.old.29019 2010-03-10 10:55:18.000000000 +0000 +++ /tmp/wklog.102.new.29019 2010-03-10 10:55:18.000000000 +0000 @@ -1 +1 @@ -Benchmarks-3.0 +Server-9.x DESCRIPTION: What lacks in MySQL is something similar like the Oracle Wait Interface. The basic idea is, that every connection/thread (and the whole system) shows (on request) where it spent how much time doing what. They implemented the wait interface (which is possibly a sub-set of Sun's dtrace) where you can see where you loose your time. Marc Callaghan and Baron Schwartz blogged in this direction recently: http://www.xaprb.com/blog/2009/11/07/a-review-of-optimizing-oracle-performa… http://www.facebook.com/note.php?note_id=355839540932 Good sources to start with: http://www.amazon.com/Oracle-Wait-Interface-Performance-Diagnostics/dp/0072… http://books.google.ch/books?id=14OmJzfCfXMC&pg=PA6&lpg=PA6&dq=owi+is+a+per… http://www.oracle.com/technology/books/pdfs/sample%20chapter%202729x.pdf http://www.amazon.com/Optimizing-Oracle-Performance-Cary-Millsap/dp/0596005… ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] Updated (by Shinguz): MySQL/Maria Wait Interface (102)
by worklog-noreply＠askmonty.org 10 Mar '10

10 Mar '10

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: MySQL/Maria Wait Interface CREATION DATE..: Wed, 10 Mar 2010, 10:53 SUPERVISOR.....: Bothorsen IMPLEMENTOR....: COPIES TO......: CATEGORY.......: Client-BackLog TASK ID........: 102 (http://askmonty.org/worklog/?tid=102) VERSION........: Server-9.x STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: -=-=(Shinguz - Wed, 10 Mar 2010, 10:55)=-=- Title modified. --- /tmp/wklog.102.old.29019 2010-03-10 10:55:18.000000000 +0000 +++ /tmp/wklog.102.new.29019 2010-03-10 10:55:18.000000000 +0000 @@ -1 +1 @@ -MySQL Wait Interface +MySQL/Maria Wait Interface -=-=(Shinguz - Wed, 10 Mar 2010, 10:55)=-=- Version updated. --- /tmp/wklog.102.old.29019 2010-03-10 10:55:18.000000000 +0000 +++ /tmp/wklog.102.new.29019 2010-03-10 10:55:18.000000000 +0000 @@ -1 +1 @@ -Benchmarks-3.0 +Server-9.x DESCRIPTION: What lacks in MySQL is something similar like the Oracle Wait Interface. The basic idea is, that every connection/thread (and the whole system) shows (on request) where it spent how much time doing what. They implemented the wait interface (which is possibly a sub-set of Sun's dtrace) where you can see where you loose your time. Marc Callaghan and Baron Schwartz blogged in this direction recently: http://www.xaprb.com/blog/2009/11/07/a-review-of-optimizing-oracle-performa… http://www.facebook.com/note.php?note_id=355839540932 Good sources to start with: http://www.amazon.com/Oracle-Wait-Interface-Performance-Diagnostics/dp/0072… http://books.google.ch/books?id=14OmJzfCfXMC&pg=PA6&lpg=PA6&dq=owi+is+a+per… http://www.oracle.com/technology/books/pdfs/sample%20chapter%202729x.pdf http://www.amazon.com/Optimizing-Oracle-Performance-Cary-Millsap/dp/0596005… ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] New (by Shinguz): MySQL Wait Interface (102)
by worklog-noreply＠askmonty.org 10 Mar '10

10 Mar '10

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: MySQL Wait Interface CREATION DATE..: Wed, 10 Mar 2010, 10:53 SUPERVISOR.....: Bothorsen IMPLEMENTOR....: COPIES TO......: CATEGORY.......: Client-BackLog TASK ID........: 102 (http://askmonty.org/worklog/?tid=102) VERSION........: Benchmarks-3.0 STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: DESCRIPTION: What lacks in MySQL is something similar like the Oracle Wait Interface. The basic idea is, that every connection/thread (and the whole system) shows (on request) where it spent how much time doing what. They implemented the wait interface (which is possibly a sub-set of Sun's dtrace) where you can see where you loose your time. Marc Callaghan and Baron Schwartz blogged in this direction recently: http://www.xaprb.com/blog/2009/11/07/a-review-of-optimizing-oracle-performa… http://www.facebook.com/note.php?note_id=355839540932 Good sources to start with: http://www.amazon.com/Oracle-Wait-Interface-Performance-Diagnostics/dp/0072… http://books.google.ch/books?id=14OmJzfCfXMC&pg=PA6&lpg=PA6&dq=owi+is+a+per… http://www.oracle.com/technology/books/pdfs/sample%20chapter%202729x.pdf http://www.amazon.com/Optimizing-Oracle-Performance-Cary-Millsap/dp/0596005… ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

[Maria-developers] New (by Shinguz): MySQL Wait Interface (102)
by worklog-noreply＠askmonty.org 10 Mar '10

10 Mar '10

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: MySQL Wait Interface CREATION DATE..: Wed, 10 Mar 2010, 10:53 SUPERVISOR.....: Bothorsen IMPLEMENTOR....: COPIES TO......: CATEGORY.......: Client-BackLog TASK ID........: 102 (http://askmonty.org/worklog/?tid=102) VERSION........: Benchmarks-3.0 STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 0 (hours remain) ORIG. ESTIMATE.: 0 PROGRESS NOTES: DESCRIPTION: What lacks in MySQL is something similar like the Oracle Wait Interface. The basic idea is, that every connection/thread (and the whole system) shows (on request) where it spent how much time doing what. They implemented the wait interface (which is possibly a sub-set of Sun's dtrace) where you can see where you loose your time. Marc Callaghan and Baron Schwartz blogged in this direction recently: http://www.xaprb.com/blog/2009/11/07/a-review-of-optimizing-oracle-performa… http://www.facebook.com/note.php?note_id=355839540932 Good sources to start with: http://www.amazon.com/Oracle-Wait-Interface-Performance-Diagnostics/dp/0072… http://books.google.ch/books?id=14OmJzfCfXMC&pg=PA6&lpg=PA6&dq=owi+is+a+per… http://www.oracle.com/technology/books/pdfs/sample%20chapter%202729x.pdf http://www.amazon.com/Optimizing-Oracle-Performance-Cary-Millsap/dp/0596005… ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0