[asterisk-bugs] [JIRA] (ASTERISK-30119) res_pjsip_sdp_rtp: Timeout doesn't handle bundled streams

[Torbjörn Abrahamsson (JIRA)]

    [ https://issues.asterisk.org/jira/browse/ASTERISK-30119?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=259927#comment-259927 ] 

Torbjörn Abrahamsson commented on ASTERISK-30119:
-------------------------------------------------

We encountered the same problem using 18.12.1, while not having the issue with 18.8.0. Is someone looking at this? We will probably revert the patch referred to above to solve our needs, as we feel unsure if we can construct a correct solution at the moment.

> res_pjsip_sdp_rtp: Timeout doesn't handle bundled streams
> ---------------------------------------------------------
>
>                 Key: ASTERISK-30119
>                 URL: https://issues.asterisk.org/jira/browse/ASTERISK-30119
>             Project: Asterisk
>          Issue Type: Bug
>      Security Level: None
>          Components: Resources/res_pjsip_sdp_rtp
>    Affects Versions: 16.24.0, 16.25.0, 16.26.0, 16.27.0
>            Reporter: Joe Searle
>            Severity: Major
>              Labels: webrtc
>         Attachments: endpoint.txt, video_recvonly.pcap, video_sendrecv.pcap
>
>
> Since upgrading from 16.22.0 to 16.27.0 when making outbound calls from my sip client (webrtc) I have been noticing Asterisk has been dropping the calls slightly after the 60 second mark with the notice
> {quote}[Jun 30 13:05:34] NOTICE[20479]: res_pjsip_sdp_rtp.c:149 rtp_check_timeout: Disconnecting channel 'PJSIP/softphone-web-6-00000002' for lack of video RTP activity in 65 seconds{quote}
> I used git bisect to track down the issue to this commit
> {quote}14156f9827eb0caa4e3b64bafce318e475a12b68{quote}
> The related jira issue is ASTERISK-28890
> Prior to this change I wouldn't have any issue with Asterisk dropping the calls.
> I thought that the problem was that the INVITE sent from the endpoint to Asterisk was set to *recvonly* on the video stream and that Asterisk wasn't checking whether it should be receiving the video RTP before it adds the applies the rtp_timeout to it. This does not seem to be the case however because I tested enabling video in the softclient before making another attempt so it was set to *sendrecv* but I still have the problem with video being sent to Asterisk.
> Reading the linked jira issue the change relates to *rtp_keepalive* which I don't have enabled on the endpoints.
> I will attach the endpoints settings and a pcap once I figure out how.
> The only work around I have found so far is to disable rtp_timeout on the endpoints which isn't ideal.



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