Percona Server moved to https://jira.percona.com/projects/PS

Intermittent percona_changed_pages failures in Jenkins

Bug #1071877 reported by Alexey Kopytov
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Percona Server moved to https://jira.percona.com/projects/PS
Fix Released
High
Laurynas Biveinis
Percona Server moved to https://jira.percona.com/projects/PS 5.5.28-29.2
5.1
Fix Released
High
Laurynas Biveinis
Percona Server moved to https://jira.percona.com/projects/PS 5.1.66-14.2
5.5
Fix Released
High
Laurynas Biveinis
Percona Server moved to https://jira.percona.com/projects/PS 5.5.28-29.2

Related branches

lp:~laurynas-biveinis/percona-server/xtradb-multiple-bmp-files-51
Stewart Smith (community): Approve
Alexey Kopytov (community): Approve
Vlad Lesin: Pending requested
Diff: 2630 lines (+1500/-497)
17 files modified
Percona-Server/mysql-test/include/delete_innodb_bitmaps.inc (+11/-0)
Percona-Server/mysql-test/r/percona_server_variables_debug.result (+1/-0)
Percona-Server/mysql-test/r/percona_server_variables_release.result (+1/-0)
Percona-Server/mysql-test/suite/innodb_plugin/r/percona_changed_page_bmp.result (+32/-2)
Percona-Server/mysql-test/suite/innodb_plugin/r/percona_changed_page_bmp_no_restart.result (+1/-0)
Percona-Server/mysql-test/suite/innodb_plugin/r/percona_changed_pages.result (+142/-28)
Percona-Server/mysql-test/suite/innodb_plugin/t/innodb_bug34300-master.opt (+0/-1)
Percona-Server/mysql-test/suite/innodb_plugin/t/percona_changed_page_bmp.test (+150/-31)
Percona-Server/mysql-test/suite/innodb_plugin/t/percona_changed_page_bmp_no_restart.test (+10/-2)
Percona-Server/mysql-test/suite/innodb_plugin/t/percona_changed_pages.test (+224/-64)
Percona-Server/storage/innodb_plugin/handler/ha_innodb.cc (+6/-0)
Percona-Server/storage/innodb_plugin/handler/i_s.cc (+184/-122)
Percona-Server/storage/innodb_plugin/include/log0online.h (+77/-27)
Percona-Server/storage/innodb_plugin/include/srv0srv.h (+2/-1)
Percona-Server/storage/innodb_plugin/log/log0online.c (+635/-206)
Percona-Server/storage/innodb_plugin/srv/srv0srv.c (+2/-0)
Percona-Server/storage/innodb_plugin/srv/srv0start.c (+22/-13)
lp:~laurynas-biveinis/percona-server/xtradb-multiple-bmp-files-5.5
Alexey Kopytov (community): Approve
Diff: 2798 lines (+1656/-506)
18 files modified
Percona-Server/mysql-test/include/delete_innodb_bitmaps.inc (+11/-0)
Percona-Server/mysql-test/r/percona_server_variables_debug.result (+1/-0)
Percona-Server/mysql-test/r/percona_server_variables_release.result (+1/-0)
Percona-Server/mysql-test/suite/innodb/r/percona_changed_page_bmp.result (+35/-4)
Percona-Server/mysql-test/suite/innodb/r/percona_changed_page_bmp_no_restart.result (+1/-0)
Percona-Server/mysql-test/suite/innodb/r/percona_changed_pages.result (+142/-28)
Percona-Server/mysql-test/suite/innodb/t/percona_changed_page_bmp.test (+155/-39)
Percona-Server/mysql-test/suite/innodb/t/percona_changed_page_bmp_no_restart.test (+10/-2)
Percona-Server/mysql-test/suite/innodb/t/percona_changed_pages.test (+224/-64)
Percona-Server/mysql-test/suite/sys_vars/r/innodb_max_bitmap_file_size_basic.result (+84/-0)
Percona-Server/mysql-test/suite/sys_vars/t/innodb_max_bitmap_file_size_basic.test (+62/-0)
Percona-Server/storage/innobase/handler/ha_innodb.cc (+6/-0)
Percona-Server/storage/innobase/handler/i_s.cc (+184/-122)
Percona-Server/storage/innobase/include/log0online.h (+77/-27)
Percona-Server/storage/innobase/include/srv0srv.h (+2/-1)
Percona-Server/storage/innobase/log/log0online.c (+637/-206)
Percona-Server/storage/innobase/srv/srv0srv.c (+2/-0)
Percona-Server/storage/innobase/srv/srv0start.c (+22/-13)
Revision history for this message
Laurynas Biveinis (laurynas-biveinis) wrote :

The failing test is somewhat nondeterministic: it puts lower and upper bounds on the amount of log bitmap data generated by running a workload that exceeds log capacity twice. These bounds are relatively far removed from each other. The failing test seems like a misprediction of the upper bound.

The multiple bitmap file writer has a different workload that has a short workload and uses server restarts to make InnoDB checkpoints. The number of tablespace pages created is deterministic there.

For now will assume that the multiple file writer branch fixes this and of course will investigate if this comes up again.

Revision history for this message
Laurynas Biveinis (laurynas-biveinis) wrote :

Another source of test case instability: http://jenkins.percona.com/job/percona-server-5.5-param/567/BUILD_TYPE=debug,Host=centos5-32/testReport/junit/%28root%29/innodb/percona_changed_page_bmp_2/

This suggests bitmap files present from the previous (bitmap) test cases in the same MTR run.

Revision history for this message
Laurynas Biveinis (laurynas-biveinis) wrote :

And another source of nondeterminism is bug 1076892.

Revision history for this message
Shahriyar Rzayev (rzayev-sehriyar) wrote :

Percona now uses JIRA for bug reports so this bug report is migrated to: https://jira.percona.com/browse/PS-601

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.
Subscribing...

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.