Issues connecting to API

[nvollmar]

Some pods like CoreDNS or Promtail occasionally fail to connect to API endpoint after a rolling restart:

W1014 08:33:30.375710       1 reflector.go:539] github.com/prometheus/prometheus/discovery/kubernetes/kubernetes.go:586: failed to list *v1.Pod: Get "https://10.43.0.1:443/api/v1/pods?fieldSelector=spec.nodeName%3Dk3s-master-1&resourceVersion=229439985": dial tcp 10.43.0.1:443: connect: connection refused

This usually persists until I delete the pod manually, the new incarnation in most cases can then connect.

Network policies to access the API server are in place.