You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Okay I have found the reason there was still some long running pod with a do not evict annotation which prevented it from beeing removed. A corresponding log entry / metric would help to discover this problem quickly.
Can you try upgrading to the latest version of Karpenter? There should now be events that are fired over the node that indicate the reason why drift deprovisioning would be blocked. kubernetes-sigs/karpenter#224
Version
Karpenter Version: v0.24.0
Kubernetes Version: v1.24.10
Expected Behavior
A drifted node should be removed timely.
Actual Behavior
One old node (using k8s version 1.23 is still running). It should be removed due too beeing running for more than 24 hours and for beeing drifted...
Steps to Reproduce the Problem
The cron provisioner looks like this:
Resource Specs and Logs
there are no specific logs for this instance
Community Note
The text was updated successfully, but these errors were encountered: