JVB RTT aggregate

Most of the times JVB shows very high RTT and jitter aggregate statistics, we have setup 6 video-bridges and according to charts:






Many times we have extreme jitter aggregate and RTT aggregate while few of conferences being failed but many users being dropped from conferences continuously. Also sometimes Jicofo does not receive health check responses from video bridges and destroys all of the conferences although all video-bridges are connected and Prosody does not shows any bridge disconnection log.
Most of the users join the conferences between 8:00 and 16:00.
As it shows, We have very high RTT aggregates (25 seconds, 15 seconds, 5 seconds)
see


for more info.
Anyone has any idea to fix that?
Thanks

Those are indeed some high RTTs, but these should just be a function of the network itself. It may be that the connections have issues, or that deploying JVBs in more geographical regions can help with this–but those are very extreme.

Thanks for the reply.
Most of out participants reside in the capital, so they are close to each other.
“geographical locations” means different data centers in a city or something else?

If you’ve got all the participants and the JVB in the same general area, then it’ll likely be something network related: either the JVB’s network or on the client side.

1 Like