DO-ESX4 (cluster host) lost High Availability

2 days ago, I noticed that do-esx4 had an error on it which, shortly after, disappeared, with no trace of it in the logs.

Unfortunately, the error going away,  did not eliminate the problem with the ability to vMotion virtual machines from within the cluster. That server had on it, at the time, cvusd-citrix6, cvusd-foodsvr and do-trackit.

The server seemed to be pretty flaky, and I really needed to remove the VMs from it. In order to do that, I went ahead and powered off the VMs on it, and moved them over to another host, then powered them back on.

Now I was able to apply all patches and update ESX4 to ESX4 Update 1, and reboot the server.

After the reboot, I re-enabled HA, which threw a cryptic error at first, but after going through with the process, HA finally enabled. Now, in conjunction with the working DRS, we are able to vMotion VMs within the CVUSD cluster again.

Over the next few days, I will be applying the same updates to the remaining servers in the VMware cluster.

• This update was performed on 01/26/2010 @ 20:40