mysqld crashes with SIG 11

Bug #1410291 reported by Simao
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Percona Server moved to https://jira.percona.com/projects/PS
Expired
Undecided
Unassigned
5.6
Expired
Undecided
Unassigned

Bug Description

Hello,

Our master database crashed after receiving a signal 11. The dump is attached to this bug report.

Previous to this crash, both innodb_table_stats and innodb_index_stats were missing from the filesystem and from the output of a 'SHOW TABLES' command. This went buy unoticed for a very long time until mysqld crashed with this dump.

The messages in the log before the crash were:

2014-10-11T04:19:01+00:00 db04 daemon.err mysqld: 2014-10-11 04:19:01 7f8b9a7ce700 InnoDB: Recalculation of persistent statistics requested for table "<our db>"."geo_" but the required persistent statistics storage is not present or is corrupted. Using transient stats instead.
2014-10-11T04:30:20+00:00 wimdu-db04 daemon.err mysqld: 2014-10-11 04:30:20 7f8b9a7ce700 InnoDB: Error: Table "mysql"."innodb_table_stats" not found.

Messages like these were left unoticed in the logfile for months, but they were there.

We are using the following percona version:

Server version: 5.6.14-rel62.0-log Percona Server with XtraDB (GPL), Release rel62.0, Revision 483

Thank you

Revision history for this message
Simao (simaomm) wrote :
Revision history for this message
Valerii Kravchuk (valerii-kravchuk) wrote :

Please, try to install clean Percona Server 5.6.22-71.0 (latest), load the same data (after making sure persisten statistcis tables are there) and check if the same crashing query:

SELECT DISTINCT COUNT(DISTINCT `messaging_threads`.`id`) FROM `messaging_threads`

(or whatever it was) still leads to crash.

Please, upload also your my.cnf file content. For now this looks to me like a result of somewhat incorrect upgrade from pre-5.6 version.

Revision history for this message
Simao (simaomm) wrote :

Hello,

I am loading the dump into a clean install, it will take some time, and then I will test this query.

I looked further into the logs and it seems that indeed the missing tables error might be there since the upgrade.

But would this cause the crash of the database? It was running almost a year without the tables and it never crashed.

I am attaching our my.cnf.

Thank you

Revision history for this message
Simao (simaomm) wrote :

Hello,

I just tested the exact same query with a fresh dump.

It took 21.82 sec but it did not crash.

Thank you

Revision history for this message
Launchpad Janitor (janitor) wrote :

[Expired for Percona Server 5.6 because there has been no activity for 60 days.]

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-3247

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

Other bug subscribers

Remote bug watches

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