Dear people,
I am new on this list but I believe it is the right audience for my question. If not,
throw your paperballs at me.
I have a setup of Routinator 0.10.2 talking RTR to BIRD v2.0.8 (yes - both not recent but
upgrading is sometimes complicated). I have run into a problem where BIRD sometimes has
far too litle ROAs in its RPKI table but the connection is up and alive. Another BIRD
talking to the same Routinator has the full ROA dump so I think the database on Routinator
is fine.
If I restart the rpki protocol on BIRD, it comes right back up but keeps running into
"protocol-error"s, reestablishing, protocol-error and so on. I created a PCAP from this
back and forth which my Wireshark 3.6.6 has trouble dissecting parts of it. I can share
the PCAP if needed.
So after some time and lots of reconnects, the table was suddenly up and I was back to the
expected number of ROAs.
Has anyone run into that issue before? I know it's good to run recent software but before
doing any upgrades out of hope, I would like to understand the cause of this. The TCP
connection seems clean (beside one end sending FIN at some point) and there does not seem
to be any packetloss or similar between BIRD and Routinator.
Again, happy to share the PCAP if anyone is willing to take a peek.
Thank you!
Sascha
--
RPKI mailing list
[email protected]
https://lists.nlnetlabs.nl/mailman/listinfo/rpki
