Issue

• kube-service-catalog is only able to run on first master node in 3 master OCP cluster.
  Reboot of the first master node results in oc commands hang and the GUI shows timeouts.
  The control-manager pod from kube-service-catalog project seems not to run on any other master nodes.

# oc adm drain master1.example.com --ignore-daemonsets --force  --delete-local-data
node "master1.example.com" already cordoned
WARNING: Ignoring DaemonSet-managed pods: apiserver-xxx, controller-manager-xxx, logging-fluentd-xxx, apiserver-xxx; Deleting pods with local storage: apiserver-xxx
node "master1.example.com" drained

Even deleting the controller-manager pod after master1 was drained does not start this pod on another master.

# oc get pod -o wide
NAME                       READY     STATUS    RESTARTS   AGE       IP            NODE
apiserver-xxx            1/1       Running   2          14d       10.xx.xx.xx   master1.example.com
controller-manager-xxx   1/1       Running   0          4m        10.xx.xx.xx   master1.example.com