Lucas Servén Marín

Let me know if you’d like to PR it :) I’d be happy to iterate in it afterwards

> Maybe just setting a route for the service CIDR towards an in-cluster node/kube-proxy would be sufficient? That's very true, this will be sufficient. We can absolutely point the entire...

I think we could also add a new guide/tutorial about balancing scheduling across different clouds/locations. This would help people see the value of multi-cloud and get started. We could use...

hi @idcmp can you share the Kilo logs from the laptop VM? This is the most important piece of information here for why the node is unready.

Hi @idcmp, you mentioned that the Kilo pod on the laptop VM was crashlooping, but the logs you shared don't mention anything about a crash. Are there more logs that...

Thanks @idcmp those logs are really helpful :) It's quite funny that Kilo is getting terminated. It seems that Kubernetes is shutting it down for some reason. Did you by...

Ah, it sounds like maybe it's related to this: > Pod sandbox changed, it will be killed and re-created.

I'll try testing on k8s v1.24 (and updating our ewe tests). So far our tests have been on v1.23. I wonder if this has caused an issue.

I wanted to add that this issue sounds to me like it is a bit more fundamental than WireGuard/Kilo/networking. It seems like something is breaking with the container runtime interrupting...

> I really think the problem relates to the EC2 instance having an Endpoint= in its kilo0 config for the laptop VM Peer, so when rekeying happens it tries to...