[asterisk-bugs] [JIRA] (ASTERISK-30221) Channel masquerades screw up groups

[N A (JIRA)]

N A created ASTERISK-30221:
------------------------------

             Summary: Channel masquerades screw up groups
                 Key: ASTERISK-30221
                 URL: https://issues.asterisk.org/jira/browse/ASTERISK-30221
             Project: Asterisk
          Issue Type: Bug
      Security Level: None
          Components: Core/Bridging, Functions/func_groupcount
    Affects Versions: 18.14.0
            Reporter: N A
            Severity: Major


RELATED to: ASTERISK-30210

So circling back to the weird thing with masquerades:

I don't think it has anything to do with optimizations, per se, I think it's the masquerade itself. For some reason, it seems to consistently screw up groups.

Here is an example of a system caught red handed where, after a masquerade occurs, suddenly all of the groups on all channels on the system suddenly become assigned to a *SINGLE* channel. This includes the bizarre scenario of also having the same group assignment twice on the same channel, such that GROUP_COUNT returns 2 even those there is only one channel in the group. This has the effect afterwards of putting the channel into an infinite loop due to dialplan logic that expects 2 channels.

None of this happened prior to upgrading from 18.9.0 to 18.14.0, so a regression is possible but I'm not yet positive either. It also seems weird that masquerades would be broken in such a a fundamental way.

The group screwing up seems to vary: the example in ASTERISK-30210 simply made all the groups disappear, in this case they all hop onto a single channel even from other channels that still exist (as well as groups that no longer do so).

Note also that this system includes the group variables patch on Gerrit and testing without that is difficult/impossible. The bug could very well be there too, though that doesn't explain why it's only triggered by masquerades.



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