|Summary:||ASTERISK-14321: deadlock during call bridge, mutex lock|
|Reporter:||Alan Graham (zerohalo)||Labels:|
|Date Opened:||2009-06-16 15:24:09||Date Closed:||2009-06-23 10:22:09|
|Environment:||Attachments:||( 0) bt_scrubbed.txt|
|Description:||Randomly, but regularly (twice in an hour at worst), with 1.4.26rc2 we deadlock during a call transfer/bridge to our gateway machine (running 184.108.40.206).|
|Comments:||By: Alan Graham (zerohalo) 2009-06-16 15:26:05|
correction - gateway machine is running 220.127.116.11
By: David Brillert (aragon) 2009-06-17 23:55:31
I have been experiencing deadlocks with 1.4.25=>1.4.26rc2 but not 18.104.22.168
I am curious if your deadlock problem goes away if you downgrade to revision r197562 ?
My bug report does not look to have exactly the same core show locks output as yours.
By: Alan Graham (zerohalo) 2009-06-18 02:14:09
aragon: We still seem to get the deadlock in 22.214.171.124, but it is much rarer - it's stable for days, but then will randomly give similar deadlock. We can't keep 1.4.26rc2 up for more than a couple of hours with production load without a crash.
I could lab up r197562, but have a feeling I can't get it to reproduce in the lab.
The core show locks is similar to https://issues.asterisk.org/view.php?id=15213#105717 when I can get it.
By: David Brillert (aragon) 2009-06-19 08:31:51
zerohalo your core show locks looks like mine from 15345.
15213 also looks like my core show locks.
There is a patch on my ticket from mmichelson that looks to have cured my deadlocks.
The more testers the better ;)
The patch is here
By: Russell Bryant (russell) 2009-06-23 10:22:08
I'm going to mark this as fixed by the patch that mmichelson merged yesterday. We'll do a 1.4.26-rc4 with that fix. Please re-open if you still have the problem.