Parallel pod deletion makes kubectl stuck forever about kubernetes HOT 3 OPEN

This issue is currently awaiting triage.

If a SIG or subproject determines this is a relevant issue, they will accept it by applying the triage/accepted label and provide further guidance.

The triage/accepted label can be added by org members by writing /triage accepted in a comment.

from kubernetes.

Try to delete some static pod kubectl delete pod -n kube-system kube-controller-manager-k8s-control-plane-1

note that these are mirror pods, which serve more of an informal representative purpose. deleting them will tell the kubelet to re-create them at API level, but this delete action doesn't have much value.

It seems that the problem is caused by parallel deletion of the pod by kubelet.

your test could be triggering some corner case around static pod sync.
as long as regular pods (non-mirror) can be deleted fine in parallel, without exhibiting this problem, then maybe this could be claimed as a problem around static pods and a "won't fix".

leaving this to sig node for triage, though. maybe there is a race that should be avoided.
/sig node

if you really want to delete these pods and stop the related containers, you should move/remove the YAML files in /etc/kubernetes/manifests.

from kubernetes.

This was a first attempt to do not really delete the pods and containers, but restart the containers and make sure they are restarted and ready. Without pod deletion, immediate attempt to check the pods might show an outdated status. Finally, other solution was developed avoiding the pod deletion.

from kubernetes.