VM Stopping with HA

[zemotard]

Hi, How does HA to turn off vm in case of failure ? It uses power on / power script ?

Regards

[esiebert7625]

It powers the VM off to release the lock on the disk file, this is like pressing the power button on the server. This is necessary for the VM to quickly start on another ESX host. The default isolation response is power off the VM and can be changed to leave powered on if you do not have the VM set to use HA. Using a clean OS shutdown is not an option for this.

Fyi…if you find this post helpful, please award points using the Helpful/Correct buttons.

-=-=-=-=-=-=-=-=-=-=-==-=-=-=-=-=-=-=-=-=-=-=-

Thanks, Eric

Visit my website: http://vmware-land.com

-=-=-=-=-=-=-=-=-=-=-==-=-=-=-=-=-=-=-=-=-=-=-

[VirtualNoitall]

Hello,

HA starts up virtual machines that are on a host that is no longer responding. Typically this means the virtual machines were not properly or gracefully stopped. HA simply restarts the virtual machines on another host to get you back up and running with minimal intervention.

[zemotard]

"Typically this means the virtual machines were not properly or gracefully stopped" Not necessary, if just your console connection is down, that doesn't mean you vm are stopped .... That's why I wish to know if it's possible to define a script to shutdown vm when the host doesn't respond.

[esiebert7625 · Accepted solution]

Yeah alot of times it's simple network connectivity that is lost and the host is OK. The power down of the VM is needed though to ensure the other ESX host can start it. Running a script or trying to do a shutdown would take too long and there is no guarantee that it will complete. Power off is a definitive method to quickly and definitely make sure the VM is shutdown. Remember HA is not clustering, it's simply high availability. Perhaps in a future release they will let us change the HA trigger from 12 seconds to something higher to allow for brief network outages.

[VirtualNoitall]

Hopefully you have setup redundancy to cover that. If not though then that is a valid concern. Service Console pnic redundancy is a typical config.