Hello All,
I have run into a problem with rtpproxy's timeout notification feature.
The problem appears when the call establishes two separate RTP sessions
but only uses one of them the other is dead.
Such a scenario can appear when an endpoint requests a session for audio
and the other one for video, but during call setup the endpoints decide
not to use the video stream (called end does not support it).
In this case the call is disconnected after -T seconds because of the video
stream being idle, but the call is proceeding properly with audio only.
This can be considered a bug or a feature, depending on network policy.
I created a patch to change this behaviour, by simply checking all sessions
of the same call_id and accumulating their TTL's. If it's at 0 then we have
a timeout, but if any one of them is active timeout notification is canceled.
The new behaviour can be activated using the -X parameter, the default
being the old behaviour.
Please review, and if you believe it's worthy commit it to the codebase.
The patch was created against latest git version from
git://sippy.git.sourceforge.net/gitroot/sippy/rtpproxy
Thanks,
Kiss Karoly
_______________________________________________
Devel mailing list
[email protected]
http://lists.rtpproxy.org/mailman/listinfo/devel
