I've been running into complications with upgrading ESX servers in a clustered scenario.
Typically I simply flip the server into maintenance mode which auto-migrates the VMs off then run esxupdate and reboot the box if necessary. When I'm done I simply pull it out of maintenance mode and move on to the next ESX Host.
For some reason, however, on this latest tarball to upgrade from 3.x to 3.0.2 following this same basic process doesn't work. Instead I get an error that it cannot upgrade while VMs are running on this machine (they are not, it's in maintenance mode).
Has anyone run into this? Can you not upgrade 3.0.2 while a member of the cluster? If not, has anyone tried removing a box form the cluster, then readding it after the upgrade and running updates systematically that way?
I basically simply want to avoid breaking down my cluster and resource pool and remake it if I can avoid it.
