We have a problem which has not been solved yet.. I have searched for a lot of forums without success. So basically I would like to make some VM policy or triggered job which will power off an individual VM on event of any host failure. the reason is: We have a VmWare cluster with two hosts, protected by HA. One of our VM has a big resource consumption, but the system itself is not a business critical one. Unfortunately all of our VMs not fit to one host, therefore we are running everything on host1 and the mentioned big resource consumer on host2.
So we would like to prevent our business critical VMs in case of host failure - if host2 fails no problem, the big resource consumer will not fail over to host1 because of VM override policy... But If host 1 fails, all of our business critical VMs have to be moved to host2 - and here is the problem because they cannot be moved all to host2 because of limited resource on host2.
We would like to poweroff the big resource consumer if any of hosts fails in cluster.
Is there any possibility to achieve this kind of behavior?