We have a 2-node ESXi 4.1 cluster and at this point in time we've not got enough RAM for one of the hosts to support all VMs should one of the hosts become unavailable, but it really is quite important to have HA running so we can provide some redundancy.
This obviously means that we have to try and tweak settings for HA for individual VMs so that only the important VMs are started on the other host. This seems quite simple to setup and configure but it has just struck me that because the VMs are distributed between hosts (we also have DRS enabled), this may not be as effective as I first thought - although the HA behaviour can be set for Vms individually, not all VMs will be moved (as the healthy host will contain a bunch of VMs already) - only the VMs from the 'dead' host.
What I want to know is if I set VM Restarting Priorty to 'Disabled' for a number of non-critical VMs, in the case where HA has kicked-in (and 2 hosts have been redcued to one), is there any way to tell the cluster that I want al non-critical VMs to be switched-off on the healthy host (even the ones that haven't been moved)?
Cheers,
Tim