vCenter Server becomes unresponsive/inaccessible due to a host/hardware/resource(storage/network)/VM failure and is not restarted by vSphere HA

Products

• VMware vCenter Server

Issue/Introduction

Symptoms:

• If your vCenter VM is hosted on an HA enabled cluster and becomes unresponsive/inaccessible due to a host/hardware/resource(storage/network)/VM failure, vSphere HA might not be able to restart the virtual machine under a corner case race condition.
• If you run into this race/corner case you will see the NoCompatibleHost event. The summary of the event will say "No host is compatible with the virtual machine."
• Note: Any powered on virtual machine could potentially run into this corner case race.

Cause

This occurs due to a potential corner case race between vCenter server and HA primary on the HA enabled cluster which could be triggered by frequent cluster membership changes.

Resolution

To workaround the issue manually re-register the VCenter Server virtual machine on another healthy host and power it on via the ESX level apis/command line interface.