There are very few times that you will ever need to do this but when you do, you will have the information that you need. I frequently use this when testing HA during a greenfield implementation.  I do this prior to removing a blade so I can test software failure prior to physical failure. Another instance of you having to do this is if you are having to work with VMware support and they need to get some full cores to see if there is a driver / software issue. Either way it’s pretty fun so let’s do it!

SSH to your host

SSH

Warning: If you type the following command and hit enter on a host it will cause it to panic. Use caution when doing this.

Type: vsish -e set /reliability/crashMe/Panic 1

Crash Me

 

After doing so you will see the following screen on the host

PSOD

 

After this you can gather a VM support or wait and watch HA fail the VMs over if it is enabled for this cluster. This will allow you to test a failure without pulling any hardware.