After upgrading VirtualCenter from 2.0.2 to 2.5 and before upgrading my ESX 3.0.1 hosts to 3.5, my DRS/HA enabled cluster initially fails to configure HA with status "An error occured during configuration of the HA Agent on the host" in Tasks & Events for the cluster. Disabling and then re-enabling HA on the cluster successfully enables HA but the process must be must be repeated on the cluster anytime a host is rebooted because HA fails to configure when that host exits maintenance mode following a reboot. My lab environment is as follows:
VirtualCenter 2.5
2 ESX 3.0.1, 60875 hosts
Since disabling/re-enabling HA on the cluster resolves the issue and once the hosts are upgraded to ESX 3.5, this isn't a critical issue. In the production environment I'm testing this for however, we have 30 clusters and approximately 90 ESX 3.0.1 hosts to upgrade so I'd like to find a proactive solution to avoid issues in the timeframe between the VirtualCenter upgrade and the completion of the host upgrades.