3 ms·
> together, dealing with their weird failure modes, etc. Basically, RAFT sucks. It really isn't very resilient AT ALL. You can literally take down an entire RAF
by klysm 4y ago
> together, dealing with their weird failure modes, etc. Basically, RAFT sucks. It really isn't very resilient AT ALL. You can literally take down an entire RAFT cluster with a single packet. One freakin packet.
etcd (the core of k8s) uses raft as its consensus algorithm. I’m not even sure what you are conflating, or what this magical single packet you are referring to is. Kubernetes will very much not continue to work if etcd goes down which is why people tend to host it outside of k8s itself for production use cases.
- withinboredom 4y agoConsul, Nomad, and Vault all use RAFT to function /at all/ so it’s really easy to break. Kubernetes will continue to run in a degraded state if etcd loses quorum. Your services will still be reachable.
- klysm 4y agoIf nomad loses quorum do your services become unreachable? I don’t think so.
- withinboredom 4y agoYou have to turn off the Nomad node to restore quorum, so yeah, the do become unreachable.