In a goBGP instance (version 2.10.0 ) running in our network, a few iBGP sessions went down and stayed down. This was noticed after a week of the sessions being down. This goBGP instance has 11 iBGP sessions with a peer, and 8 of those 11 sessions went down and stayed down. The goBGP daemon had to be restarted to bring all iBGP sessions back up.
I saw the following log message being logged periodically by goBGP for the 8 down sessions – while they were down:
Nov 5 18:05:01 gc-rm-res-01-alpha.nae02.spprod.viasat.io bgpd[17195]: {"Key":"10.95.236.96","Topic":"Peer","level":"warning","msg":"accepted conn is closed to avoid be blocked","time":"2020-11-05T18:05:01Z"}
A restart of the goBGP instance fixed this issue and allowed all 11 iBGP sessions to come back up again. Questions:
What would have caused 8 of 11 iBGP sessions at one goBGP instance to go down and stay down?
What are some good practices to not have this problem happen again?
In a goBGP instance (version 2.10.0 ) running in our network, a few iBGP sessions went down and stayed down. This was noticed after a week of the sessions being down. This goBGP instance has 11 iBGP sessions with a peer, and 8 of those 11 sessions went down and stayed down. The goBGP daemon had to be restarted to bring all iBGP sessions back up.
I saw the following log message being logged periodically by goBGP for the 8 down sessions – while they were down: Nov 5 18:05:01 gc-rm-res-01-alpha.nae02.spprod.viasat.io bgpd[17195]: {"Key":"10.95.236.96","Topic":"Peer","level":"warning","msg":"accepted conn is closed to avoid be blocked","time":"2020-11-05T18:05:01Z"}
A restart of the goBGP instance fixed this issue and allowed all 11 iBGP sessions to come back up again. Questions:
Thanks.