Lucas Servén Marín

Hmm the funny thing here is that your private IPs are being used as the public endpoints for WireGuard. This creates some tricky situations. The problem here (i think) is...

Also, can you share the logs from the Kilo pod on master? Ideally with debug log level :))

Ack thanks a lot for reporting this. You provided tons of helpful details. Could you share some additional pieces of info: * What tag of the squat/kilo image are you...

@3rmack thanks for the quick reply. It's comforting that restarting doesn't resolve the issue, otherwise we might not have a convincing solution. The kubelet seems to complain that it can't...

Hmm in that case we'll need a bit more inspection. Can you please share the output of `cat /etc/cni/net.d/10-kilo.conflist` from the broken node? I need to double check nothing changed...

Took a look and CNI v0.3.1 should still work. We need to ensure that the CNI configuration file exists for the broken node and check if there are recent kubelet...

Ack, yes i somehow missed them earlier. Thanks. I spun up a two node kubeadm cluster last night and haven't been able to replicate this issue yet.

I haven't been able to replicate this :/ this seems like it may be an issue with the specific environment, perhaps a container runtime issue

@hhstu this seems like the kilo0 device is not available / doesn't exist. Are there any logs from the boringtun container?

Hmmm can you please show a list of the devices available in the erroring Kilo Pod? `ip l`