You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
With our Janus FTL setup, we figured out that we could use janus servers to serve other servers, instead of viewers, effectively creating a mesh-like network of relays that deliver video to users. With Janus FTL, a streamer would stream to an ingest, and an edge close to the viewer would connect to that ingest to proxy the video to the viewer. Any additional viewers to that edge would utilize the same video stream as to not duplicate the connections.
With Waveguide the distributed setup can be very similar, just instead of utilizing FTL, utilizing our already existing WHEP output. A Waveguide server close to a user could negotiate with WHEP to have the stream sent to the Waveguide server, and then create another peer connection to the user to deliver the video to them. The benefits of doing it this way are a more reliable network connection DC to DC, and reducing the number of raw consumers on the Waveguide servers that are accepting inputs.
There's still more to brainstorm on how this would work.
The text was updated successfully, but these errors were encountered:
With our Janus FTL setup, we figured out that we could use janus servers to serve other servers, instead of viewers, effectively creating a mesh-like network of relays that deliver video to users. With Janus FTL, a streamer would stream to an ingest, and an edge close to the viewer would connect to that ingest to proxy the video to the viewer. Any additional viewers to that edge would utilize the same video stream as to not duplicate the connections.
With Waveguide the distributed setup can be very similar, just instead of utilizing FTL, utilizing our already existing WHEP output. A Waveguide server close to a user could negotiate with WHEP to have the stream sent to the Waveguide server, and then create another peer connection to the user to deliver the video to them. The benefits of doing it this way are a more reliable network connection DC to DC, and reducing the number of raw consumers on the Waveguide servers that are accepting inputs.
There's still more to brainstorm on how this would work.
The text was updated successfully, but these errors were encountered: