How to reproduce:

1. Install microshift
2. Wait for all pods to be ready
3. Reboot
4. Check pods

Symptoms:

Router:

openshift-ingress          router-default-65fb8995d5-6rkzh       0/1     ContainerCreating   1          4m42s

If you describe the pod  router-default-65fb8995d5-6rkzh:

  Warning  FailedMount       26s (x7 over 58s)  kubelet            *MountVolume.SetUp failed for volume "service-ca-bundle" : configmap references non-existent config key:* service-ca.crt

 
For the ServiceCA, the issues are not visible unless you look at the pod logs; if you look at those, you will find what seems regular activity, but it will only start functioning when this happens:

I1102 14:03:58.778069       1 leaderelection.go:258] successfully acquired lease openshift-service-ca/service-ca-controller-lock

So it seems that during the shutdown, the service-ca controller lock is not released correctly (shutdown issues somewhere), and also the lock time is probably very high for a single-host environment.

Also, service CA starts way too many workers for everything (5)