After an ESX host migration this weekend I started receiving an error when trying to power on machines that were in a DRS cluster:-
“The request refers to an object that does no longer exist or has never existed”
At first of course I filled my pants thinking I’d just lost the entire VM I was moving accross hosts, but a quick check told me it was all still there and that it fired up fine on a host which wasn’t a cluster member.
Turning DRS off cured the problem, but that’s not very useful when you’ve got 4 spanky new BL460 Blades with more processors and memory than you can shake a stick at and you want DRS to magic all your VMs around to keep things as sweet as can be.
Turns out the fix is simply to restart your VC service. Seeing as I needed to bounce my VC server to apply Windows Hotfixes that was actually convenient to bounce the whole thing and watch as the error bothered me no more. Winner 🙂