Intermittent error with remote participants

I am running jitsi stable, with p2p mode disabled.

After a restart of the jvb service, everything (seems to) works correctly.

However, after an unknown period of time, I find that remote participants (e.g. someone using the Jitsi Meet iOS app, over the cellular network) are able to connect, but without audio or video in either direction. Participants joining from the LAN (and on a VPN into the LAN) are able to join and transmit correctly.

Looking in jvb.log, I see this:

JVB 2019-07-22 06:32:14.769 INFO: [16] org.jitsi.videobridge.health.Health.log() Performed a successful health check in 22ms. Sticky failure: false
JVB 2019-07-22 06:32:16.069 WARNING: [5042138] org.jitsi.videobridge.EndpointMessageTransport.log() SCTP connection with ce6d10ed not ready yet.
JVB 2019-07-22 06:32:16.069 WARNING: [5042138] org.jitsi.videobridge.EndpointMessageTransport.log() No available transport channel, can’t send a message
JVB 2019-07-22 06:32:18.073 WARNING: [5042139] org.jitsi.videobridge.EndpointMessageTransport.log() SCTP connection with ce6d10ed not ready yet.

I have a brute-force workaround, of using a cronjob to restart the server each morning before it is likely to be used, but I’d welcome any help in tracking down the actual problem.

I tested the cronjob approach overnight, and it worked. But it is a hack, and only works because our system does not get used very heavily, so I can be confident that it will not be in use when the automatic restart happens.

u can deploy multiple bridge so when u restarting a bridge the other can be used till comes back online and since multiple bridge deployment jicofo will automatically assign less loaded bridge for new conference…here is the tutorial lasttime i checked about a month ago it still works as the video described

cheers

Much appreciated — thank you. That’s definitely overkill for my limited use case, but really good to know.