[asterisk-bugs] [JIRA] (ASTERISK-28877) Duplicate RINGING DeviceStateChange AMI events from chan_sip.c

[Joshua C. Colp (JIRA)]

     [ https://issues.asterisk.org/jira/browse/ASTERISK-28877?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Joshua C. Colp updated ASTERISK-28877:
--------------------------------------

    Severity: Minor  (was: Major)

> Duplicate RINGING DeviceStateChange AMI events from chan_sip.c
> --------------------------------------------------------------
>
>                 Key: ASTERISK-28877
>                 URL: https://issues.asterisk.org/jira/browse/ASTERISK-28877
>             Project: Asterisk
>          Issue Type: Bug
>      Security Level: None
>          Components: Channels/chan_sip/General
>    Affects Versions: 13.33.0, 16.10.0
>         Environment: Vanilla Asterisk (no FreePBX et al) on Gentoo on QEMU KVM
>            Reporter: Troy Bowman
>            Severity: Minor
>
> I have an AMI event listener that logs additional events to my queue_log to help me tie in agent device availability into my queue reports.
> Whenever I have tried to upgrade Asterisk to a later version, My AMI event listener would log DeviceStateChange events twice whenever the device would ring.  I have been stuck on Asterisk 13.13 for years because I could not understand why RINGING would be duplicated, but other DeviceStateChange events were fine.
> Today, I finally tracked it down to Gerrit change id I547a458fc59721c918cb48be060cbfc3c88bcf9c and ASTERISK-24488.  It seems to me that this was a hack of sorts to get the RINGING Device State Change event through to the newer (at the time) Stasis Message Bus code because of some kind of race condition.
> I realize that I should migrate to chan_pjsip, but I am still not ready to do that in production yet.  I fear that since chan_sip is EoL, no one will care about this, but I am reporting it anyway.  For now, I am simply commenting out those lines of code that duplicate the event.
> If it is intended that the same AMI events should be sent multiple times, then I guess I'll have to write code that somehow debounces events on my side.  It is my opinion that the AMI events should only be sent once per event.



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