upgrade drives in pv220s attached to SQL cluster

The upgrade is very disruptive.

Note: Replacing each drive with a larger one (one at a time) doesn't give you more space. The raid controller (in a non-clustered mode) would allow creating another virtual disk in the extra free space, but in cluster mode that option has been disabled, so you wouldn't be able to use 'extra' the space at all.

Assumptions:
- you're keeping your Quorum drives

Steps:
- upgrade to the latest bioses, firmwares, drivers, etc (particularly the PERC drivers and firmware)
- back up your data
- disable cluster services on one node and shut this system down
- remove the disk(s) that will be upgraded from the cluster using the cluster administrator (so that after the change the cluster isn't looking for the disks with the old disk signature)
- stop and disable cluster services on this server
- power everything down
- remove the 'old' drives (keep track which one went where)
- put in the 'new' drives
- power up 1 server and go into the perc bios to create the new raid setup
- boot this server and partition and format the disk
- now you add it to the cluster as a cluster resource and bring this disk online (otherwise you can't read/write to and from it as the cluster disk driver protects the disk)
- restore your data
- verify your SQL can start

At this point you power up the other node.
- check in the perc bios for this other node that the raid config is recognized
- let the server boot up
- re-enable cluster services and let it join the cluster

It's not 'pretty', but I know of any other way to do this due to the way the percs handle virtual disks.

From my experience with PERCs, they don't specify just which drive, they specify where the start and end point on each disk is, but then won't allow extending the section on the physical disk that the virtual disk can use. You can expand by adding more drives to the raid container, but not by swapping to larger drives