I deploy a v1.11 kube-cluster by kubeadm with one master and three node.

My kube-controller pods frequently log:

E1012 08:17:42.897796 node_lifecycle_controller.go:889] Error updating node izwz95wx4ufpnolro1iekmz: Operation cannot be fulfilled on nodes "izwz95wx4ufpnolro1iekmz": the object has been modified; please apply your changes to the latest version and try again

"izwz95wx4ufpnolro1iekmz" node will be NotReady periodically and kubectl describe node provide nothing useful information but just NotReady status.

How to debug my cluster and find out the root cause?