Garvin Pang

@InsomniaCoder and @flomsk can you post both the node and the ec2nodeclass + nodepool hash for when this fails

@ibalat Have you attempted cutting a ticket with AWS to investigate how the node got partitioned from the control plane?

Is this issue only showing up with Karpenter? What is weird to me is that this is the normal way with how autoscaler handles partitioned nodes as mentioned by @dcherniv...

@steveannett This is common issue with Kubernetes where if the node has issue or in general kubelet stop responding to control plane nodes get "stuck" in terminating. https://kubernetes.io/docs/concepts/architecture/nodes/#node-controller

Thanks for trying to tackle this! I wonder if there is a cleaner way to represent the overlays to reduce sprawl. Thinking more generally, each dimension can be additive, adding...

Hey @msvticket, I think its important to mention the role Karpenter plays within a cluster. Karpenter handles the creation of nodes based on what the cluster needs and it doesn't...

@jmdeal Updated the doc lmk if we want more info

Chatting during working hours. - We think fully configurable is likely fine - We wondered if moving this to Node overlay make sense. Currently, we are not sure however it...

Note i swapped the options number (2 1)

If you look https://github.com/kubernetes-sigs/karpenter/issues/357 there are a few cases there. For myself it is GPU bootstrapping due to having to install drivers