Which component are you using?:

/area cluster-autoscaler

What version of the component are you using?:

Component version: 1.32.0

What k8s version are you using (kubectl version)?:

$ kubectl version
Client Version: v1.31.2
Kustomize Version: v5.4.2
Server Version: v1.32.5-eks-5d4a308

What environment is this in?:

We are running cluster-autoscaler on AWS EKS cluster.

What did you expect to happen?:


cluster-autoscaler should be able to delete placeholder nodes from ASG that contains / in its name.

What happened instead?:


cluster-autoscaler failed to delete placeholder nodes from ASG that contains / in its name, because it was not recognized as a known ASG.

W0709 11:37:32.508503   30155 static_autoscaler.go:867] Error while trying to delete nodes from k8s-node/edge-prod-aeus2a/rika-test: aws:///eu-south-2a/i-placeholder-k8s-node/edge-prod-aeus2a/rika-test-1 doesn't belong to a known asg

Due to above error, cluster-autoscaler stopped all scaling operations after failing to fix the node group size when the MaxNodeStartupTime window elapsed.

E0709 11:42:02.380948   30155 static_autoscaler.go:441] Failed to fix node group sizes: failed to decrease k8s-node/edge-prod-aeus2a/rika-test: attempt to delete existing nodes targetSize:3 delta:-2 existingNodes: 3

How to reproduce it (as minimally and precisely as possible):

1. Create an ASG that includes / in its name.
2. Trigger a cluster-autoscaler scale-up event to the ASG that lacks capacity (in our case, we used a machine type that does not have sufficient resources available).
3. At this point, cluster-autoscaler will try to delete the placeholder node, but but it will fail because the node is not recognized as part of a known ASG.
4. If this situation persists beyond the MaxNodeStartupTime duration, cluster-autoscaler will halt all scaling operations for all ASGs.

Anything else we need to know?: