- developers - lists.mariadb.org

[Maria-developers] New (by Monty): Add OpenVos extensions to MariaDB (32)
by worklog-noreply＠askmonty.org 08 Jun '09

08 Jun '09

----------------------------------------------------------------------- WORKLOG TASK -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- TASK...........: Add OpenVos extensions to MariaDB CREATION DATE..: Tue, 09 Jun 2009, 00:48 SUPERVISOR.....: Monty IMPLEMENTOR....: COPIES TO......: CATEGORY.......: Server-BackLog TASK ID........: 32 (http://askmonty.org/worklog/?tid=32) VERSION........: Connector/J-5.1 STATUS.........: Un-Assigned PRIORITY.......: 60 WORKED HOURS...: 0 ESTIMATE.......: 20 (hours remain) ORIG. ESTIMATE.: 20 PROGRESS NOTES: DESCRIPTION: The patches can be found at: ftp://ftp.stratus.com/pub/vos/posix/source/mysql-5.0.75/mysql.html and ftp://ftp.stratus.com/pub/vos/posix/source/mysql-5.1.33/mysql.html We should add those that are stable to 5.1 and others that makes sense to 5.4 ESTIMATED WORK TIME ESTIMATED COMPLETION DATE ----------------------------------------------------------------------- WorkLog (v3.5.9)

1 0

Re: [Maria-developers] Stab at C
by Michael Widenius 08 Jun '09

08 Jun '09

Hi! >>>>> "Paul" == Paul Charles Leddy <pcleddy(a)gmail.com> writes: Paul> Monty, Paul> Is there anyone available to run some questions by about code review, Paul> submission? Shouldn't take long to get all the answers, but finding it Paul> a bit difficult using the website and launchpad. The easiest way to get help with reviews or help with getting tasks done is using the maria-developers list or join us on the #maria channel on Freenode. Regards, Monty

3 3

[Maria-developers] JFYI: InnoDB plugin bug report
by Kristian Nielsen 08 Jun '09

08 Jun '09

Just to follow up on some IRC discussion on this, I posted the following to the InnpDB plugin forum (which the webpage suggests for bug reports). I think it is bad that the create table will work or not depending on zlib version. However, I also think it is best to let InnoDB fix it. Having a different behaviour in InnoDB plugin and XtraDB/MariaDB (as would result if we just fix it ourselves) will not really help to reduce the possible confusion from this. - Kristian. ----------------------------------------------------------------------- I stumbled upon an issue with the compressed row format in the InnoDB plugin. The issue is that the maximum allowed key size depends on the version of libz that InnoDB plugin is linked against. It means that this statement: CREATE TABLE t1(c TEXT, PRIMARY KEY (c(439))) ENGINE=InnoDB ROW_FORMAT=COMPRESSED KEY_BLOCK_SIZE=1 CHARSET=ASCII; will succeed with one version of libz, but fail with another. Whether it is a bug is debatable. It is definitely a corner case but it has some nasty side effects. Eg. a table dump made on one server can refuse to load on another server (not sure of possible effects of a libz upgrade). The reason for the dependency on libz version is in function page_zip_empty_size(): lint size = zip_size /* subtract the page header and the longest uncompressed data needed for one record */ - (PAGE_DATA + PAGE_ZIP_DIR_SLOT_SIZE + DATA_TRX_ID_LEN + DATA_ROLL_PTR_LEN + 1/* encoded heap_no==2 in page_zip_write_rec() */ + 1/* end of modification log */ - REC_N_NEW_EXTRA_BYTES/* omitted bytes */) /* subtract the space for page_zip_fields_encode() */ - compressBound(2 * (n_fields + 1)); On my machine, compressBound() for n_fields=4 returns 21 for the bundled zlib (mysql 5.1.34) but 23 for my system zlib (Ubuntu Hardy x86_64). I would suggest to change the code so that the estimate on max key size does not depend on whatever zlib returns. The zlib manual [www.zlib.net] says this: "destLen is the total size of the destination buffer, which must be at least 0.1% larger than sourceLen plus 12 bytes" Eg, rounding up, this would give the 23 bytes returned by my system libz compressBound(). This is slightly more pessimistic than some versions of compressBound() apparently, but much better to have the same limit on all systems rather than getting slightly larger key sizes only on some. Hope this helps (and that this is the correct place to post it), - Kristian.

1 0

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (psergey:2709)
by Sergey Petrunia 07 Jun '09

07 Jun '09

#At lp:maria based on revid:psergey@askmonty.org-20090603182330-ll3gc91iowhtgb23 2709 Sergey Petrunia 2009-06-07 [merge] Merge MWL#17: Table elimination modified: sql/sql_select.cc === modified file 'sql/sql_select.cc' --- a/sql/sql_select.cc 2009-06-03 13:10:45 +0000 +++ b/sql/sql_select.cc 2009-06-07 18:24:03 +0000 @@ -2586,7 +2586,7 @@ eliminate_tables_for_join_list(JOIN *joi table_map tables_used_outside= used_tables_on_left | used_tables_on_right[i] | used_tables_elsewhere; - table_map cur_tables; + table_map cur_tables= 0; if (tbl->nested_join) {

1 0

[Maria-developers] bzr commit into MariaDB 5.1, with Maria 1.5:maria branch (psergey:2707)
by Sergey Petrunia 07 Jun '09

07 Jun '09

#At lp:maria based on revid:psergey@askmonty.org-20090603131045-c8jqhwlanli7eimv 2707 Sergey Petrunia 2009-06-07 MWL#17: Table Elimination - Fix trivial valgrind warning modified: sql/sql_select.cc per-file messages: sql/sql_select.cc MWL#17: Table Elimination - Fix trivial valgrind warning === modified file 'sql/sql_select.cc' --- a/sql/sql_select.cc 2009-06-03 13:10:45 +0000 +++ b/sql/sql_select.cc 2009-06-07 18:24:03 +0000 @@ -2586,7 +2586,7 @@ eliminate_tables_for_join_list(JOIN *joi table_map tables_used_outside= used_tables_on_left | used_tables_on_right[i] | used_tables_elsewhere; - table_map cur_tables; + table_map cur_tables= 0; if (tbl->nested_join) {

1 0

[Maria-developers] [Branch ~maria-captains/maria/5.1] Rev 2703: Fix accessing ulong enum option as uint, failing on 64-bit big-endian.
by Michael Widenius 07 Jun '09

07 Jun '09

Hi! >>>>> "noreply" == noreply <noreply(a)launchpad.net> writes: noreply> === modified file 'mysys/my_getopt.c' noreply> --- mysys/my_getopt.c 2009-04-25 10:05:32 +0000 noreply> +++ mysys/my_getopt.c 2009-05-20 15:34:34 +0000 noreply> @@ -647,7 +647,7 @@ noreply> return EXIT_OUT_OF_MEMORY; noreply> break; noreply> case GET_ENUM: noreply> - if (((*(int*)result_pos)= find_type(argument, opts->typelib, 2) - 1) < 0) noreply> + if (((*(ulong *)result_pos)= find_type(argument, opts->typelib, 2) - 1) < 0) noreply> return EXIT_ARGUMENT_INVALID; noreply> break; noreply> case GET_SET: noreply> @@ -983,7 +983,7 @@ noreply> *((int*) variable)= (int) getopt_ll_limit_value((int) value, option, NULL); noreply> break; noreply> case GET_ENUM: noreply> - *((uint*) variable)= (uint) value; noreply> + *((ulong*) variable)= (uint) value; noreply> break; noreply> case GET_UINT: noreply> *((uint*) variable)= (uint) getopt_ull_limit_value((uint) value, option, NULL); noreply> @@ -1221,7 +1221,7 @@ noreply> } noreply> break; noreply> case GET_ENUM: noreply> - printf("%s\n", get_type(optp->typelib, *(uint*) value)); noreply> + printf("%s\n", get_type(optp->typelib, *(ulong*) value)); noreply> break; noreply> case GET_STR: noreply> case GET_STR_ALLOC: /* fall through */ Where is the enum used that caused this to break ? (I did a grep of GET_ENUM in the current tree, but couldn't find anything. As an ENUM can only contain a small set of values, it would be more logical that it should be an uint ant not an ulong. (With some 64 bit compilers, uint -> 32 bit while ulong is 64 bit, so there is a clear advantage in using uint) Becasue of this, I am more included to belive the the one that is using GET_ENUM as ulong is doing things wrong instead of the above code. Regards, Monty

2 1

[Maria-developers] how to speed up Maria record loading?
by Giuseppe Maxia 07 Jun '09

07 Jun '09

Hi, I am doing some tests with Maria. Out of the box, without any tuning, the load of records (with multiple record INSERT statements) is 5 times slower than MyISAM, 4 times slower than InnoDB. What can I do to make the loading faster? Thanks Giuseppe -- The Data Charmer http://datacharmer.org/

4 4

[Maria-developers] Query cache test failure in Buildbot for mariadb 5.1
by Kristian Nielsen 07 Jun '09

07 Jun '09

Hi, There is a query cache related failure seen in Buildbot: https://askmonty.org/buildbot/builders/hardy-amd64-dbg/builds/35/steps/test… main.query_cache_28249 [ fail ] Test ended at 2009-05-21 09:13:10 CURRENT_TEST: main.query_cache_28249 --- /home/archivist/archivist-amd64/archivist-amd64/build/mysql-test/r/query_cache_28249.result 2009-05-21 09:30:15.000000000 +0300 +++ /home/archivist/archivist-amd64/archivist-amd64/build/mysql-test/r/query_cache_28249.reject 2009-05-21 11:13:10.000000000 +0300 @@ -46,7 +46,6 @@ 1 0 2 0 3 0 -4 0 RESET QUERY CACHE; SELECT *, (SELECT COUNT(*) FROM t2) FROM t1; a (SELECT COUNT(*) FROM t2) mysqltest: Result length mismatch - saving '/home/archivist/archivist-amd64/archivist-amd64/build/mysql-test/var/1/log/main.query_cache_28249/' to '/home/archivist/archivist-amd64/archivist-amd64/build/mysql-test/var/log/main.query_cache_28249/' Retrying test, attempt(2/3)... main.query_cache_28249 [ retry-pass ] 509 Retrying test, attempt(3/3)... main.query_cache_28249 [ retry-pass ] 404 I checked the bug: http://bugs.mysql.com/bug.php?id=28249 It is about a race condition in query cache and MyISAM (and re-test in Buildbot shows that it is not 100% reproducible). I am thinking that perhaps the bug was re-introduced with Maria, or in any case that someone more familiar with the MyISAM or Maria code would be better at looking into it. Any takers? - Kristian.

2 2

Re: [Maria-developers] Query cache test failure in Buildbot
by Michael Widenius 07 Jun '09

07 Jun '09

In-Reply-To: <87hbzd5mfb.fsf(a)knielsen-hq.org> References: <87hbzd5mfb.fsf(a)knielsen-hq.org> X-Mailer: VM 7.17 under 21.5 (beta28) "fuki" (+CVS-20070806) XEmacs Lucid Reply-To: monty(a)askmonty.org FCC: ~/Mail/outbox --text follows this line-- Hi! >>>>> "Kristian" == Kristian Nielsen <knielsen(a)knielsen-hq.org> writes: Kristian> Hi, Kristian> There is a query cache related failure seen in Buildbot: Kristian> https://askmonty.org/buildbot/builders/hardy-amd64-dbg/builds/35/steps/test… Kristian> main.query_cache_28249 [ fail ] Kristian> Test ended at 2009-05-21 09:13:10 Kristian> CURRENT_TEST: main.query_cache_28249 Kristian> --- /home/archivist/archivist-amd64/archivist-amd64/build/mysql-test/r/query_cache_28249.result 2009-05-21 09:30:15.000000000 +0300 Kristian> +++ /home/archivist/archivist-amd64/archivist-amd64/build/mysql-test/r/query_cache_28249.reject 2009-05-21 11:13:10.000000000 +0300 Kristian> @@ -46,7 +46,6 @@ Kristian> 1 0 Kristian> 2 0 Kristian> 3 0 Kristian> -4 0 Kristian> RESET QUERY CACHE; Kristian> SELECT *, (SELECT COUNT(*) FROM t2) FROM t1; Kristian> a (SELECT COUNT(*) FROM t2) Kristian> mysqltest: Result length mismatch Kristian> - saving '/home/archivist/archivist-amd64/archivist-amd64/build/mysql-test/var/1/log/main.query_cache_28249/' to '/home/archivist/archivist-amd64/archivist-amd64/build/mysql-test/var/log/main.query_cache_28249/' Kristian> Retrying test, attempt(2/3)... Kristian> main.query_cache_28249 [ retry-pass ] 509 Kristian> Retrying test, attempt(3/3)... Kristian> main.query_cache_28249 [ retry-pass ] 404 Kristian> I checked the bug: Kristian> http://bugs.mysql.com/bug.php?id=28249 Kristian> It is about a race condition in query cache and MyISAM (and re-test in Kristian> Buildbot shows that it is not 100% reproducible). Kristian> I am thinking that perhaps the bug was re-introduced with Maria, or in any Kristian> case that someone more familiar with the MyISAM or Maria code would be better Kristian> at looking into it. Sanja, can you create a bug report in launchpad or worklog item to get this fixed? Regards, Monty

1 0

[Maria-developers] [Branch ~maria-captains/maria/5.1] Rev 2711: Added tests to cover more server code
by noreply＠launchpad.net 07 Jun '09

07 Jun '09

------------------------------------------------------------ revno: 2711 committer: Michael Widenius <monty(a)askmonty.org> branch nick: mysql-maria timestamp: Sun 2009-06-07 13:05:19 +0300 message: Added tests to cover more server code Author: Stewart Smith modified: mysql-test/r/alter_table.result mysql-test/r/limit.result mysql-test/t/alter_table.test mysql-test/t/limit.test === modified file 'mysql-test/r/alter_table.result' --- mysql-test/r/alter_table.result 2009-02-19 09:01:25 +0000 +++ mysql-test/r/alter_table.result 2009-06-07 10:05:19 +0000 @@ -1268,4 +1268,21 @@ 4 b 5 a DROP TABLE t1; +SET @save_sql_mode=@@sql_mode; +SET sql_mode=strict_all_tables; +CREATE TABLE t1 (a int NOT NULL default 42); +INSERT INTO t1 values (); +SELECT * FROM t1; +a +42 +ALTER TABLE t1 ALTER COLUMN a DROP DEFAULT; +INSERT INTO t1 values (); +ERROR HY000: Field 'a' doesn't have a default value +INSERT INTO t1 (a) VALUES (11); +SELECT * FROM t1 ORDER BY a; +a +11 +42 +DROP TABLE t1; +SET @@sql_mode=@save_sql_mode; End of 5.1 tests === modified file 'mysql-test/r/limit.result' --- mysql-test/r/limit.result 2008-10-15 21:34:51 +0000 +++ mysql-test/r/limit.result 2009-06-07 10:05:19 +0000 @@ -113,4 +113,36 @@ End of 5.0 tests select 1 as a limit 4294967296,10; a +CREATE TABLE t1 (a int PRIMARY KEY auto_increment); +INSERT INTO t1 VALUES (),(),(),(),(),(),(),(),(),(); +INSERT INTO t1 VALUES (),(),(),(),(),(),(),(),(),(); +SELECT a FROM t1 ORDER BY a LIMIT 10 OFFSET 1; +a +2 +3 +4 +5 +6 +7 +8 +9 +10 +11 +SELECT a FROM t1 ORDER BY a LIMIT 10 OFFSET 10; +a +11 +12 +13 +14 +15 +16 +17 +18 +19 +20 +SELECT a FROM t1 ORDER BY a LIMIT 2 OFFSET 14; +a +15 +16 +DROP TABLE t1; End of 5.1 tests === modified file 'mysql-test/t/alter_table.test' --- mysql-test/t/alter_table.test 2009-02-19 09:01:25 +0000 +++ mysql-test/t/alter_table.test 2009-06-07 10:05:19 +0000 @@ -1000,4 +1000,22 @@ SELECT * FROM t1; DROP TABLE t1; +# +# Test for ALTER column DROP DEFAULT +# + +SET @save_sql_mode=@@sql_mode; +SET sql_mode=strict_all_tables; + +CREATE TABLE t1 (a int NOT NULL default 42); +INSERT INTO t1 values (); +SELECT * FROM t1; +ALTER TABLE t1 ALTER COLUMN a DROP DEFAULT; +--error 1364 +INSERT INTO t1 values (); +INSERT INTO t1 (a) VALUES (11); +SELECT * FROM t1 ORDER BY a; +DROP TABLE t1; +SET @@sql_mode=@save_sql_mode; + --echo End of 5.1 tests === modified file 'mysql-test/t/limit.test' --- mysql-test/t/limit.test 2008-10-16 01:50:56 +0000 +++ mysql-test/t/limit.test 2009-06-07 10:05:19 +0000 @@ -102,4 +102,16 @@ select 1 as a limit 4294967296,10; +# +# Test for LIMIT X OFFSET Y +# + +CREATE TABLE t1 (a int PRIMARY KEY auto_increment); +INSERT INTO t1 VALUES (),(),(),(),(),(),(),(),(),(); +INSERT INTO t1 VALUES (),(),(),(),(),(),(),(),(),(); +SELECT a FROM t1 ORDER BY a LIMIT 10 OFFSET 1; +SELECT a FROM t1 ORDER BY a LIMIT 10 OFFSET 10; +SELECT a FROM t1 ORDER BY a LIMIT 2 OFFSET 14; +DROP TABLE t1; + --echo End of 5.1 tests -- 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