Best way to manage MSCS virtuals

[NuggetGTR]

Ok, Basically I am just after what people recommend is the best way to handle a large amount of MSCS virtuals in a large environment.

For eg I am suggesting that the MSCS virtuals have their own ESX cluster. my resoning is the administration of them.

Because they run physical mode RDM's they can not be migrated or moved due to scsi bus sharing. So now I find I can not put hosts into maintainance mode to preform preemtive maintainance on them, DPM is affect and DRS to a point also. For me to put a host in maintainance mode I require to turn off the MSCS virtuals on the host which would be an outage which would require a butt load of paper work and then might not even be approved. sort of defeats the purpose of having all this cool tech.

So by having their own ESX cluster there would be only 10 hosts which will be a pain instead of the 50 they care currently spread over.

what have other people done?

Cheers

[bulletprooffool]

Do you require masses of paperwork even for shutting down a single node of the cluster? (I guess as it is a cluster, it is probably mission critical, so yes . . )

It does seem that limiting the number of hosts is a good idea - or perhaps creating an ESX cluster that specifically hosts MCSC clusters? This way maintenance for that ESX cluster (as in general when we do maintenance on a node, we do the same for all nodes in the cluster) impacts all MSCS nodes in the same Change request?

Ideally, you want to be creating the paperwork just once, to cover the islocation of all nodes of all MCSC clusters, so basing the paperwork on a VM cluster maintenance schedule that covers all MSCS nodes will limit the number of paperwork requests (but of course make getting the change authorised significantly more difficult)