[asterisk-bugs] [JIRA] (ASTERISK-28395) Asterisk occasionally fails to hangup channels

[Stefan Repke (JIRA)]

Stefan Repke created ASTERISK-28395:
---------------------------------------

             Summary: Asterisk occasionally fails to hangup channels
                 Key: ASTERISK-28395
                 URL: https://issues.asterisk.org/jira/browse/ASTERISK-28395
             Project: Asterisk
          Issue Type: Bug
      Security Level: None
          Components: Channels/chan_iax2, Channels/chan_sip/General, PBX/General
    Affects Versions: 13.8.2
         Environment: Debian Linux
            Reporter: Stefan Repke


Our Asterisk servers occasionally fail to hangup channels (IAX2 or SIP).
We are running version: 13.8.2.

The sympthoms are a follows:
* We observer a HangupRequest event via AMI but the Hangup event does not occur.
* One Asterisk thread starts consuming 100% CPU, i.e. burning one core.
* Strange warning messages occur in CLI and log repeatedly every few seconds:
** IAX case:
{noformat}
WARNING[15982] chan_iax2.c: Max retries exceeded to host 10.101.1.5 on IAX2/125-5832 (type = 6, subclass = 11, ts=119996, seqno=24)
{noformat}
** SIP case:
{noformat}
WARNING[21518]: chan_sip.c:4333 __sip_autodestruct: Autodestruct on dialog 'e8954fa42ffee2d1ace1ac4c6323d990 at 10.249.4.97' with owner SIP/999222-0000149c in place (Method: BYE). Rescheduling destruction for 10000 ms
{noformat}
* The channel stays alive (i.e. still visible via CLI, AMI status, ARI, etc).
* The channel cannot be hungup (via CLI, AMI, etc).
* The only fix is to restart Asterisk.

Up to now, we where facing this problem three times:
# IAX2 channel, thread {{bridge_channel_depart_thread}}
# SIP channel, thread {{pbx_thread}}
# SIP channel, thread {{pbx_thread}}, with thread debugging enabled (includes {{strace}} output).

I've attached the debug info for each of the above cases.

Maybe, this bug is related to:
* ASTERISK-25911
* ASTERISK-27538

---

If I can image, that you need more detailed information to track down this bug.
If so, please let me know how I can provide it and I will try to reproduce this problem.




--
This message was sent by Atlassian JIRA
(v6.2#6252)