We are trying to stream video from one server to another (one is a local LAN server, the other is public).
Unfortunately, a server may be behind a NAT which does port replacing. Therefore, we cannot predict the actual remote IP or PORT tuple for the incoming UDP RTP packages on the public server. I was wondering why this information is required in the connect call of the pipe transport?
Would it be a security issue if one would make the check in /worker/src/RTC/PipeTransport.cpp#L609 optional?
What other solution could help us to have the local LAN server (which serves local webrtc clients) also pipe the stream to a “cloud” server which serves public clients?