Yes, you can’t change these things as these are purely related to webrtc internal mechanism, not related to mediasoup directly.
1 Like
The client will do its best to restart the connection, but if they fail several times it’ll give up. You would wait several seconds to thirty seconds to re-call ICE. Now if they truly lost connection a 30 second ping check and at 60 seconds if no ping is noticed, re-fresh them. Problems all diverted.
Why do we need to restart ICE? I mean if we are restarting connections then everything will be created fresh from start like transport, producer, consumer etc, so the ice candidates will be auto generated with them. Why restart ICE is necessary? Can you please explain the scenario?
You just do, if you need more of an answer you best get to testing and doing traffic limits to fake states.