We currently have enough to get started with testing projectrtp - especially on a reliable network. But we need to add timeouts and reconnects. Both the node and proxy have keep-alive enabled on their connections. But this will tear down a whole connection if the keep-alive is interrupted. Ideally, requests should be queued for a small period of time and/or redistributed to other RTP servers on longer failure.
We currently have enough to get started with testing projectrtp - especially on a reliable network. But we need to add timeouts and reconnects.
Both the node and proxy have keep-alive enabled on their connections. But this will tear down a whole connection if the keep-alive is interrupted.
Ideally, requests should be queued for a small period of time and/or redistributed to other RTP servers on longer failure.