We have a Dell 1600SC server. Currently, we have the PERC4/SC with 3 x
73GB 10,000 RPM U320 SCSI Drives, configured as RAID5.

In a routine check of the system today I observed that Disk "0" of the
array was reported by Dell OpenMange as having a status of "Prrdictive

A quickie on-lin chat with Dell Tech support confirmed that the drive
should be replaced ASAP.

Fair enough. However, the server is 3 years old. Since all 3 drives
are identical make and model, and were put into service at the same
time, I don't see much point in replacing the one 73 GB drive that is
about to fail.

What I am going to do is add a second RAID5 array at the hardware
level - comprised of 3 x 146 GB U320 SCSI drives - likely Seagate (the
73 GB drives are Fujitsu). I will then migrate the entire O/S and data
to the new array, and decommission the old array.

The $64 question is....

When the system was commisioned, the cost differential between 10K
drives and 15K drives was substanital. NOW, however, the difference is
a mere $15 per drive! (for the 146 GB models)

So, do 15K drives give of significantly more heat than 10K drives? If
so, will it be of any concern in the Dell 1600SC Server Case? I note
that the drive mounting is vertical, and there is all kinds of
internal cowling to driect the air flow.