Users may observe that the default [vSphere HA virtual machine monitoring action] alarm does not trigger when an ESXi host is rebooted or fails, even though virtual machines are successfully failed over to other hosts in the cluster. This is because that specific alarm is designed for VM Monitoring events (such as I/O timeouts or guest OS unresponsiveness) rather than host-level failovers.
The [vSphere HA virtual machine monitoring action] alarm is specifically tied to the VM Monitoring feature. To receive notifications when a virtual machine is restarted due to a host failure such as reboot, a custom alarm using a different trigger is required.
You can create a custom alarm to notify you when a virtual machine is restarted by vSphere HA by using the "vSphere HA restarted a virtual machine" trigger.
Procedure to create the alarm: