Our lab is using a CRD-5440 raid controller on our file server (RH 7.0 AXP),
with three disks (Quantum
Atlas V, 36.7 GB SCSI), configured as RAID 5. One of these three disks died on
us, and we finally received our replacement (a Seagate Cheetah 36ES,
36.704 GB SCSI). However, what was supposed to be a simple rebuild operation
is becoming a nightmare, because our replacement drive is seen as having
less capacity than the original two Atlas drives! As seen from the raid
controller, the Atlas drives each have 35020 MB capacity, and the new
Cheetah has only 35003 MB.
The two options open to us now, as I see it, are to either re-build the
whole RAID and restore from back-up, or purchase a larger drive --
neither of which I want to do because we've already purchased a drive,
and it sort of defeats the whole purpose of having the RAID in the first
place.
Is it possible to work around the 17 MB difference and rebuild the lost
drive on the Cheetah, or have I been foiled by Murphy's Law once again?
Thanks in advance,
Trevor Hamm
Of course it is possible. Simply back everything up in file form (no
partition images), repartition all the drives with equal-sized partitions,
and restore.
BTW this is not how RAID is supposed to work. Ideally, the drives should be
identical. You can probably make this work, but dissimilar drives often
cause problems that are difficult to diagnose and even harder to fix.
--
Paul Lutus
www.arachnoid.com
> Is it possible to work around the 17 MB difference and rebuild the lost
> drive on the Cheetah,
No.
What you can do is ring the supplier who sent you the new drive and
explain the situation an see if they'll agree to RMA it for a drive of
slightly larger capacity.
--
Trevor Hemsley, Brighton, UK.
Trevor-...@dial.pipex.com