Greetings, fellow vm'rs.
I'm hoping you can help me. I do have an SR open for this.
In fact, the reason I'm posting my problem here is because maybe some of you have seen this, even though a search didn't really find anything relevant in the KB.
Scenario:
7-nodes (four Dell M600 blades and three Dell 2950s, fed by Brocade switches backed by fc san with two tiers.
Symptoms:
After storage issue with one of the tiers, tehre are a few datastores that display data, but when one goes to do anything with that data -- rename an nvram file, copy a vmdk, copy a vmx, ANYTHING, we get a "INVALID ARGUMENT" from the VC and from the OS itself.
Then there's one datasstore that appears invisible, yet size numbers hint at something being in it. When one tries to go to that datastore and do an ls on it, one gets "device is busy or in use"
How can we beat this? We've restarted each host in turn.. migrate people out, restart, repeat on other systems. That's failed to achieve the goal of freeing up the "stuck" storage.
Is our only option a "Master Reset?" Stop all guests, stop all hosts, stop storage, then bring storage up, ,then hosts? We're thinking this may be the only way to fix this so we can migrate the most critical systems out of the jetstor.. does anyone know of any other way to clear this up?
Any help at all will be appreciated.
