While attending the HA: Internals and Best Practices session (BC3197) on Tuesday afternoon, I learned something new about HA’s VM health monitoring. With a cluster’s HA, you can set it to monitor the health of your VMs and have them automatically restart if it detects an OS hang-up. To do this it uses a heartbeat from the Guest’s VMware tools.
The cool thing I never knew before is a screenshot of the VM’s console is taken before it is reset, with up to 10 being saved (in the same directory as the vmx file). So when you are trying to troubleshoot the issue that caused the VM to become unresponsive, you should have a screenshot of the BSOD or any onscreen messages. Cool.