Should RTP timeout detection work when generate-RTCP=on?

[Andrew Pogrebennyk]

Hi,

it seems we can not get the RTP timeout detection to work ( having set the b2b-url = http://%%:8046/RPC, xmlrpc-format = 2 in daemon config), I see no logs about callback child being forked after timeout=60 but just logs about rtcp-generation constantly popping.

The daemon version is mr13.4.1.

Is it expected that when rtcp generation is activated, it is preventing the timeout detection mechanism to work? Not sure if I've missed something, but i see no incoming packets to the RTP and RTCP ports any more both ways, just egress RTCP packets generated locally and no notice about timeout detection which does not kick in.

Regards,

[Richard Fuchs]

On 01/10/2025 10.05, Andrew Pogrebennyk wrote:
> Is it expected that when rtcp generation is activated, it is
> preventing the timeout detection mechanism to work? Not sure if I've
> missed something, but i see no incoming packets to the RTP and RTCP
> ports any more both ways, just egress RTCP packets generated locally
> and no notice about timeout detection which does not kick in.

It is supposed to work, and I've just tested it on both current master
and 13.4, and it does work. So something else must be afoot.

Cheers