OpenShift Container Platform 4.14 upgrade is stuck when dhcp option domain-name is used in AWS

Issue

• When upgrading OpenShift 4.13 to 4.14 on AWS, where domain-name dhcp option is set, we notice that the upgrade is stuck, once the nodes are restarting. When checking we found that several nodes, including master are failing to become ready again because their hostname has changed.

  $ oc get nodes
  NAME                                         STATUS                        ROLES                  AGE     VERSION
  ip-10-0-132-230.us-east-2.compute.internal   Ready                         worker                 6h30m   v1.27.10+28ed2d7
  ip-10-0-135-56.us-east-2.compute.internal    Ready                         control-plane,master   6h36m   v1.27.10+28ed2d7
  ip-10-0-152-125.foo.bar                      NotReady                      worker                 73s     v1.27.10+28ed2d7
  ip-10-0-152-125.us-east-2.compute.internal   NotReady,SchedulingDisabled   worker                 3h31m   v1.26.13+8f85140
  ip-10-0-153-28.us-east-2.compute.internal    Ready                         worker                 118m    v1.26.13+8f85140
  ip-10-0-171-219.us-east-2.compute.internal   Ready                         control-plane,master   6h36m   v1.27.10+28ed2d7
  ip-10-0-187-99.us-east-2.compute.internal    Ready                         worker                 6h26m   v1.26.13+8f85140
  ip-10-0-216-47.us-east-2.compute.internal    Ready                         control-plane,master   6h36m   v1.27.10+28ed2d7
  ip-10-0-223-250.us-east-2.compute.internal   Ready                         worker                 6h30m   v1.26.13+8f85140
  

• When updating OCP to 4.14 on AWS, we noticed that hostname of OCP nodes are changing and csr need to be approved to allow the nodes to join the cluster again, enforcing manual, unexpected effort during OCP updates.

• The OpenShift 4.14 update on AWS is stuck when domain-name dhcp option is customized.