6

I've deployed a Kubernetes cluster using the Azure CLI (az acs create command). The nodes in the cluster are running Windows.

I want to shutdown and restart a worker node.

I tried kubectl drain to remove the node from the cluster. This works and the node status changes to 'Ready, SchedulingDisabled'

I then shutdown the node using the Azure portal. At this point the node status changes to 'NotReady, SchedulingDisabled'

I then restart the node using the Azure portal. However, the node status remains at 'NotReady, SchedulingDisabled'. I was expecting it to become 'Ready, SchedulingDisabled' and be then able to run kubectl uncordon to make it available again.

What is the recommended process for shutting down and restarting nodes in a Kubernetes cluster?

Saca
  • 10,355
  • 1
  • 34
  • 47
HarshaP
  • 91
  • 6
  • 1
    The process described above is correct. When the node restarts it should automatically reach 'Ready, SchedulingDisabled' state. check if both Docker and Kubelet process inside the node is running. if not restart those process to make it to "Ready, SchedulingDisabled" – codenio Apr 26 '18 at 06:17
  • Thanks. I will look into manually restarting the kube processes. – HarshaP May 03 '18 at 16:45

0 Answers0