[asterisk-bugs] [JIRA] (ASTERISK-26956) Exceptionally long voice > queue length queuing
Vittorio Giordano (JIRA)
noreply at issues.asterisk.org
Thu Oct 12 10:59:21 CDT 2017
[ https://issues.asterisk.org/jira/browse/ASTERISK-26956?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=239299#comment-239299 ]
Vittorio Giordano commented on ASTERISK-26956:
----------------------------------------------
We have been able to re-produce the issue, so finding the cause: SIP channel hangup is not correctly performed by the phone (hardware) when the digits (DTMF) sent over the AMI for the "hold" function are not correctly read by the system. Due to this the channel of the caller is not correctly closed and it is "frozen". Looking at the Asterisk console we see that the number of active channels starts increasing and the console starts getting full of Exceptionally long queueing message. Once forcing the closure of the caller channel by using the hangup button, we see on the console the "Autodestruct" messages. Then the whole system hang, The asterisk process is still available and the only was to stop and start it is to kill the process with the "-9" option. This has been reproduced with kernel 3.0 and Asterisk 11
> Exceptionally long voice > queue length queuing
> ------------------------------------------------
>
> Key: ASTERISK-26956
> URL: https://issues.asterisk.org/jira/browse/ASTERISK-26956
> Project: Asterisk
> Issue Type: Bug
> Security Level: None
> Affects Versions: 14.4.0
> Environment: Debian GNU/Linux 8.7 (jessie)
> Reporter: Ted G
> Assignee: Unassigned
> Attachments: backtrace-threads.20170423.txt, core-show-locks.20170423.txt
>
>
> Hello,
> I am getting tons of these messages in my logs a few times a day which causes my system to be completely dead:
> [Apr 20 21:39:15] WARNING[24491][C-00000197]: channel.c:1162 _astqueue_frame: Exceptionally long voice queue length queuing to Local/s at announce-0000015b;1
> [Apr 23 15:32:06] WARNING[20645] channel.c: Exceptionally long voice queue length queuing to CBAnn/12-000001f0;1
> I am having a very difficult time figuring out what is causing it other than 'its a channel deadlock' which doesn't really tell me how to fix the issue. This just started happening around ~6 days ago.
> This problem is extremely catastrophic as the only fix is to completely restart the asterisk system.
> Included are backtrace and core-show-locks logs. Please assist this desperate fool. :)
--
This message was sent by Atlassian JIRA
(v6.2#6252)
More information about the asterisk-bugs
mailing list