Sounds like maybe the some type of timing issuing in stopping (relinquishing) services, that failed before the node was actually shutdown. If that specific node still shows being offline in hx connect, but you can still ssh into the storage controllers mgmt ip then try confirming a couple of the services are running:
priv service storfs status
priv service exhibitor status
If either of these are stopped, you can try restarting the services and see if the cluster will return to healthy, allowing you to retry the upgrade:
priv service storfs start (only if stopped)
priv service exhibitor start (only if stopped)
Eventually the cluster should go back to being healthy, allowing the upgrade to be retried.
stopping/starting services should normally include TAC guidance.
Kirk...
