What could cause a "schedule Endpoint Message Transport Timeout"

We’ve had some stability issues lately. Restarting the cluster resolves them for now.
But it would great to understand what is going on.

And suggestion what this log entry on the jvb could mean?

2021-02-16 12:00:56.506 SEVERE: [20] [confId=d7485eee230f481a epId=ba0082ad gid=6395 stats_id=Miracle-eEP conf_name=ksh-elefanten@conference.foo.com] Endpoint.lambda$scheduleEndpointMessageTransportTimeout$9#994: EndpointMessageTransport still not connected.
Got sctp association state update: 1
sctp is now up.  was ready? false
Got sctp association state update: 1
sctp is now up.  was ready? false

Also jicofo is acting up

Jicofo 2021-02-16 10:29:17.278 SEVERE: [28] org.jitsi.jicofo.JitsiMeetConferenceImpl.log() Error adding SSRCs from: ihs_e4@conference.foocom/c919bea7: MSID mismatch detected in group SourceGroup[FID, ssrc=2458450837, ssrc=1972344875, ]@6be76c20

Any input much appreciated.

We’ve also see trouble in the jicofo log

Caused by: org.jivesoftware.smack.XMPPException$XMPPErrorException: XMPP error reply received from conference.jitsi.foo.com.jitsi.foo.com: XMPPError: remote-server-not-found - cancel. Generated by auth.jitsi.foo.com

The identifier conference.jitsi.foo.com.jitsi.foo.com looks very wrong. But I am not sure why that appeared all of a sudden. Where would that be coming from?

org.jitsi.videobridge.xmpp.user.shard.HOSTNAME=jitsi.foo.com
org.jitsi.videobridge.xmpp.user.shard.DOMAIN=auth.jitsi.foo.com