[asterisk-bugs] [JIRA] (ASTERISK-24195) bridge_native_rtp: Removing mixmonitor from a native RTP capable smart bridge doesn't cause the bridge to resume being a native rtp bridge

[Matt Jordan (JIRA)]

    [ https://issues.asterisk.org/jira/browse/ASTERISK-24195?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=221504#comment-221504 ] 

Matt Jordan commented on ASTERISK-24195:
----------------------------------------

Using UNBRIDGE when removing the MixMonitor is the correct solution. We've done that elsewhere with other framehooks as well (namely, the PJSIP transfer framehook).

> bridge_native_rtp: Removing mixmonitor from a native RTP capable smart bridge doesn't cause the bridge to resume being a native rtp bridge
> ------------------------------------------------------------------------------------------------------------------------------------------
>
>                 Key: ASTERISK-24195
>                 URL: https://issues.asterisk.org/jira/browse/ASTERISK-24195
>             Project: Asterisk
>          Issue Type: Bug
>      Security Level: None
>          Components: Applications/app_mixmonitor, Bridges/bridge_native_rtp
>    Affects Versions: SVN, 12.4.0
>            Reporter: Jonathan Rose
>            Severity: Minor
>
> So basically it's like this...
> 1. Start a PJSIP call from PJSIP/A to PJSIP/B without features or other hooks so that it can start a native RTP bridge
> 2. Start a mixmonitor on one of the two channels above.  The bridge should change to a simple bridge.
> 3. Stop mixmonitor.  Currently the bridge remains a simple bridge until the next time something occurs that would make the bridge re-evaluate.
> I've already determined a simple fix (which I'm not 100% certain of the correctness of) for this problem where I simply apply an UNBRIDGE soft hangup flag to the channel when closing the MixMonitor. On the other hand, I'm in the middle of writing a patch right now that pulls the unbridge flag out of the soft hangup flags and into its own member of the structure, so the exact implementation of this would change a little anyway.



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