Bug #66402 | Thread locks itself out with semaphore (crash) | ||
---|---|---|---|
Submitted: | 15 Aug 2012 16:14 | Modified: | 28 Mar 2014 20:25 |
Reporter: | Trey Raymond | Email Updates: | |
Status: | No Feedback | Impact on me: | |
Category: | MySQL Server: InnoDB storage engine | Severity: | S2 (Serious) |
Version: | 5.5.21-55-log Percona | OS: | Linux |
Assigned to: | CPU Architecture: | Any | |
Tags: | crash, Semaphore |
[15 Aug 2012 16:14]
Trey Raymond
[15 Aug 2012 16:15]
Trey Raymond
error log
Attachment: bruno.log (application/octet-stream, text), 56.39 KiB.
[15 Aug 2012 16:15]
Trey Raymond
global status (lots of spin waits)
Attachment: bruno.sts (application/octet-stream, text), 9.66 KiB.
[15 Aug 2012 16:15]
Trey Raymond
global variables
Attachment: bruno.var (application/octet-stream, text), 10.56 KiB.
[19 Aug 2012 10:11]
Valeriy Kravchuk
Please, check if this problem ever happens with a recent version from Oracle, 5.5.27.
[20 Aug 2012 19:35]
Trey Raymond
I will try upgrading and watching, but there's no easy way to trigger the fault where I could test it. However, I've gone over the change logs thoroughly, and seen nothing that might address this in any way.
[30 Aug 2012 17:59]
MySQL Verification Team
Are you tried 5.5.27 and repeated the crash?. Thanks.
[3 Sep 2012 18:52]
Trey Raymond
As mentioned earlier, it's impossible to trigger it, it has happened sporadically. What bug fix in the change log do you suspect fixed it? Nothing stood out to me.
[7 Sep 2012 18:27]
Sveta Smirnova
Thank you for the feedback. Well, any InnoDB lock bug fix can fix this issue. Also we need you to test with Oracle's MySQL, because Percona introduces own changes into InnoDB code and we should be sure problem happened in Oracle's part.
[12 Sep 2012 22:04]
Trey Raymond
bug hasn't recurred yet on either the prod install or an identical test host with an up to date oracle-provided package, there's no way I've found to trigger it as mentioned earlier, but clearly something is going wrong with the locks.
[12 Sep 2012 22:50]
Sveta Smirnova
Thank you for the update. Please reopen bug when you meet same issue with latest Oracle MySQL.
[13 Oct 2012 1:00]
Bugs System
No feedback was provided for this bug for over a month, so it is being suspended automatically. If you are able to provide the information that was originally requested, please do so and change the status of the bug back to "Open".
[25 Jan 2013 7:55]
Laurynas Biveinis
Trey - Should you ever again experience this crash, please report it on bugs.percona.com if you are using Percona Server, or here if you are using Oracle MySQL. Thanks.
[26 Feb 2013 1:00]
Bugs System
No feedback was provided for this bug for over a month, so it is being suspended automatically. If you are able to provide the information that was originally requested, please do so and change the status of the bug back to "Open".
[26 Feb 2014 9:29]
Serge Rivest
Experiencing the same problem here with 5.5.27. See innodb dump, I'll attach the file.
[26 Feb 2014 9:31]
Serge Rivest
Massive pile up of threads that self-lock
Attachment: innodb-status (1).log.gz (application/x-gzip, text), 2.30 MiB.
[28 Feb 2014 20:25]
Sveta Smirnova
Thank you for the feedback. Please provide output of SHOW CREATE TABLE for tables teacher_class_activity, r.result, r.activity, a_session, a_data, ce_fulltext_search_log, student_flag, teacher_badges, l_certificate_download, a_reset, student_last_results, student_trial_activities, m_tracking, feedback, accounts_session, accounts_data, m_amazonses_bounced, student_access, parent_link_request, a_autoauth
[1 Apr 2014 1:00]
Bugs System
No feedback was provided for this bug for over a month, so it is being suspended automatically. If you are able to provide the information that was originally requested, please do so and change the status of the bug back to "Open".