[Home]

Summary:ASTERISK-17922: Stuck waiting to execute the sip scheduler queue forever
Reporter:Sébastien Couture (sysreq)Labels:
Date Opened:2011-05-25 10:06:20Date Closed:2011-12-15 10:19:49.000-0600
Priority:MajorRegression?No
Status:Closed/CompleteComponents:Channels/chan_sip/General
Versions:1.8.4 Frequency of
Occurrence
Related
Issues:
Environment:Attachments:( 0) asterisk_backtrace-25052011_095833.txt
( 1) asterisk_backtrace-26052011_111633.txt
( 2) asterisk_backtrace-30052011_081832.txt
( 3) asterisk_backtrace-30052011_090632.txt
( 4) asterisk_backtrace-30052011_091633.txt
( 5) asterisk_backtrace-30052011_101733.txt
( 6) asterisk_backtrace-30052011_135533.txt
( 7) asterisk_showlocks-25052011_095833.txt
( 8) asterisk_showlocks-26052011_111633.txt
( 9) asterisk_showlocks-30052011_081832.txt
(10) asterisk_showlocks-30052011_090632.txt
(11) asterisk_showlocks-30052011_091633.txt
(12) asterisk_showlocks-30052011_101733.txt
(13) asterisk_showlocks-30052011_135533.txt
(14) cli_error_lock.txt
Description:I'm currently having deadlock issues with 1.8.4.1.



****** ADDITIONAL INFORMATION ******

I've attached a backtrace of the running process (before killing it) as well as the output of a 'core show locks'.
Comments:By: Sébastien Couture (sysreq) 2011-05-27 08:26:06

I've had another deadlock issue where Asterisk would again stop responding to any SIP requests. I've attached another backtrace of the then running process as well as a 'core show locks'.
By: Sébastien Couture (sysreq) 2011-05-30 08:43:09

I've had the issue arise three times just this morning. I've attached yet another backtrace and 'core show locks'.
By: Sébastien Couture (sysreq) 2011-05-30 08:58:43

I also get errors directly in the CLI about failed mutex unlocks in chan_sip.c.
By: Sébastien Couture (sysreq) 2011-05-30 09:11:07

Could asterisk_showlocks-30052011_081832.txt be related to issue ASTERISK-16692?
By: Sébastien Couture (sysreq) 2011-05-30 09:26:16

Another deadlock..
By: Sébastien Couture (sysreq) 2011-05-30 13:00:59

And another one.. I've attached the latest backtrace and 'core show locks' output.
By: Sébastien Couture (sysreq) 2011-06-01 10:10:56

Any chance this issue could be related to issue ASTERISK-17297? The patch seem to have been applied to the 1.8 code branch on April 18th, but there's no mention of it in the latest 1.8 ChangeLog. I've tried applying the patch to 1.8.4.1 but it doesn't apply cleanly.
By: Sébastien Couture (sysreq) 2011-06-01 10:41:34

I've ended up modifying the patch for issue ASTERISK-17297 so that it applies cleanly to the 1.8.4.1 code.
By: David Vossel (dvossel) 2011-06-02 12:14:09

Did the patch resolve your issue?
By: Sébastien Couture (sysreq) 2011-06-02 12:32:12

Too early to tell, I've had it running with the patch with no issues only for about 12 hours now. Prior to that, I had had no issues for a full day even without the patch. I'm also watching issue ASTERISK-17950 closely since on that particular server we have a lot of queue activity as well as hints.
By: Sébastien Couture (sysreq) 2011-06-06 10:25:43.743-0500

I've had Asterisk run without any problems for 3½ days now, using the patch from issue #18690. That being said, I'd wait a little more before I consider it fixed completely.
By: Sébastien Couture (sysreq) 2011-06-17 15:14:57.134-0500

It's been two weeks now, and I've had no issues whatsoever. I would consider that Mantis issue #18960's patch resolved this issue.
By: Matt Jordan (mjordan) 2011-12-15 10:19:49.399-0600

I'm closing this as fixed since you noted that the mantis issue patch appeared to have resolved the problem.  If that's not the case, feel free to open the issue again.