We have several servers with 5i raid controllers. Each system has 3 partitions, sysvol, syspage and data. The sysvol of all needs increasing in size. We are considering either imaging or the use of server magic to resize the sysvol partition. We will also be adding 2 additional disks into each array - 1 hot spare - so we have a number of unused disks to hand.
One plan is to image the existing disk (partitions) to a network share. Remove the old disks from the array (careully noting their physical postition in the array) insert the new drives, establish the new array, partition the logical drive to the new sizes and image back again, allowing the partions to expand as they come back. If successful the old drives could then be wiped and used in the next machine for the same process.
In this way we hope that if anything goes wrong, we could just unplug the new disks and replace them with the old, since they would be unchanged.
However this supposes that all array configuration data is kept on the disk and none in any raid controller flash memory (or the like).
Has anyone done this, is it feasible, does the raid controller 'remember' raid configuration data?
All array configuration located at drives itself in RIS table, so each Drive know his address and presence of other Drives and where it assigned to ARRAY X, Logical Drive, or Online Spare. No Information stored in Array Controller.
