On Mon, 14 May 2001 08:35:20 -0500, Wu, Jie wrote:

>....Disk2 is mounted as /u01 and contains the TSM db and log. disk3 is
>mounted as /u02 and contains dbcopy and logcopy of what are on disk1. Disk3
>is mounted as /u03 and acts as a storage pool. The drive having problem is
>disk2. What is the procedure of replacing the drive? Can I assume that by
>shutting down TSM, umounting /u01, replacing the drive, creating logic
>volume and fite system, mounting /u01, and starting TSM, the TSM server will
>automatically regenerate what were on /u01 using dbcopy and logcopy on /u02?
>Jie

I am just getting back to ADSM so take my comments with a grain
of salt.

disk1 is root volume. TSM db and log
Disk2 is mounted as /u01 dbcopy and logcopy
disk3 is mounted as /u02 storage pool
Disk3 is mounted as /u03

I am a little confused given that you seem to have listed disk3
twice.
On any case you could
Down TSM
Copy the contents of Disk2 to tape or another disk.
Get a new drive and prepare it at the OS level
Restore the dbcopy and logcopy
Bring back TSM

Alternatively
Delete dbcopy and logcopy (from within TSM, not from the OS
level)
Get a new drive and prepare it at the OS level
Create new dbcopy and logcopy

The first approach is the safest.
The second approach allows you to do the operation without ever
having to down the TSM server, but creates the risk of doing
operations with only one DB and one log.

If you have space you could do even a different/safer approach
Create new dbcopy and logcopy on different disk from disk1 or
disk2
Delete dbcopy and logcopy from disk2 (from within TSM, not
from the OS level)
Get a new drive and prepare it at the OS level
Create new dbcopy and logcopy in disk2
Delete the other dbcopy and logcopy you had created.

Good luck!