CoreDNS Restarts Repeatedly

Symptom

When Kubernetes is started, Pod CoreDNS restarts repeatedly and is in the crashloopbackoff state.

Key Process and Cause Analysis

For details, visit https://github.com/coredns/coredns/issues/2087.

Conclusion and Solution

1. Check whether /etc/resolve.conf is correctly configured.
2. Modify the Kubernetes configuration file.

   kubectl -n kube-system edit configmap coredns

   Delete the loop field, save the configuration, and delete the Pod. The Pod will automatically restart.