Bill Pitz dell at billpitz.com
Wed Mar 13 19:02:01 CST 2002

On Wed, Mar 13, 2002 at 02:47:07PM -0600, Michael E Brown <michael_e_brown at dell.com> wrote:
> Bad disk? If this is still under warranty, you may want to call it in and 
> get a replacement.

On that note, I've got a related (dumb) question that doesn't seem to be
clearly documented anyplace.

I'm working on a PE2550 with two 73gb drives that are running in mirrored
(raid 1) mode with the on-board PERC3/Di.  Works great, but I'm in the
process of writing disaster recovery documentation for the company and
one of the things that we need to address is the procedure to follow if
a drive fails.  I know the obvious, and I know what needs to be done, but
I'm not 100% sure how to accomplish it using the CLI.

Here's how my steps are set up right now:

1. Determine which disk is bad.
2. Power down the disk using the enclosure prepare slot command.
3. Remove bad disk, insert good disk.
4. Activate the slot.
5. Initialize the new (empty) disk.

This is where things are a little fuzzy.  Obviously, I need to get the
mirror to update from the existing disk to the new disk.  What's the best
way to achieve this?  The scrub command? (That worried me, because there
doesn't seem to be any way to specify which disk has the "correct" data
and which disk doesn't.)  Is there another command to use?  Does the
controller figure it out by itself?



More information about the Linux-PowerEdge mailing list