[Live-devel] BasicTaskScheduler doEventLoop gets stuck for >10s after some time period while receiving an RTP AVC stream via RTSP

Nathan Commons ncommons at t-21.biz
Thu Feb 28 15:14:53 PST 2019


My setup: my MediaSink subclass for my application (calling it RTPSink) has
a H264VideoRTPSource (gotten from the MediaSubession, determined by SDP
parsing). After calling startPlaying, it runs fine for some period of time.
However, after some random amount of time, the BasicTaskScheduler that's
used will enter SingleStep and not exit for over 10s (and this is possibly
only because stopPlaying or some other shutdown command is called by the
application after not seeing data). I've confirmed this by subclassing
BasicTaskScheduler to MyTaskScheduler, which just overwrites doEventLoop
with an identical function with some debug print statements. I've also
confirmed that the data is still being sent by the source (no RTSP timeout
causing a BYE) by capturing traffic with tcpdump during the period where
event loop is stuck in SingleStep. As the event loop gets stuck in
SingleStep, no more data is made available to my application via my
afterGettingFrame function on my RTPSink, despite the tcpdump of that
interface still showing packets being delivered as normal.

I'm trying to troubleshoot further now, but I was wondering if this sounds
like a familiar problem (user error or something already reported) or if
anyone had any advice on troubleshooting or could point me in the right
direction.

This was tested with the 2019.02.03 release.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.live555.com/pipermail/live-devel/attachments/20190228/2a414b8d/attachment.html>


More information about the live-devel mailing list