I’d considered doing something similar at some point but couldn’t quite figure out what the likely behaviour was if the workers lost connection back to the control plane. I guess containers keep running, but does kubelet restart failed containers without a controller to tell it to do so? Obviously connections to pods on other machines will fail if there is no connectivity between machines, but I’m also guessing connections between pods on the same machine will be an issue if the machine can’t reach coredns?
Comment on K3s+Wireguard(?)
johntash@eviltoast.org 8 months ago
Make sure you read up on latency requirements for k8s too. It’s definitely doable but etcd for example needs really low latency between nodes to not cause issues by default.
If you only have one master like you described, it’d probably be fine.
RegalPotoo@lemmy.world 8 months ago
peter@feddit.uk 8 months ago
I’ve done it with docker swarm and it was awful, the connection latency would break the cluster constantly
OrthoStice@feddit.it 8 months ago
Interesting. I didn’t have time to test it yet, anyway