If the data "was only on the disk", I would say your disaster recovery
preparation needs serious fix!
The idea is that you backup the whole storage pool *hierarchy* to a copy
pool. For example hierarchy DISKPOOL -> TAPEPOOL must be backed up with
1. ba stg diskpool copypool
2. ba stg tapepool copypool
3. ba db t=dbs

Without backing up the diskpool to a onsite copy pool you are exposed to
data loss in case of disk storage failure (even if there is no site-wide
disaster).

Zlatko Krastev
IT Consultant






Marc Levitan <marc.levitan AT PFPC DOT COM>
Sent by: "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>
24.10.2003 16:12
Please respond to "ADSM: Dist Stor Manager"


        To:     ADSM-L AT VM.MARIST DOT EDU
        cc:
        Subject:        Re: Move nodedata - what is moved first


What would happen if there was a site disaster and the data was only on
the
disk which is no longer available to perform restores?
I guess what I am asking is, without sending DIRMC off-site, can you
recover from a site disaster?





|---------+--------------------------->
|         |           Deon George     |
|         |           <dgeorge AT AU1 DOT IBM|
|         |           .COM>           |
|         |           Sent by: "ADSM: |
|         |           Dist Stor       |
|         |           Manager"        |
|         |           <[email protected]|
|         |           T.EDU>          |
|         |                           |
|         |                           |
|         |           10/23/2003 08:07|
|         |           PM              |
|         |           Please respond  |
|         |           to "ADSM: Dist  |
|         |           Stor Manager"   |
|         |                           |
|---------+--------------------------->

>-------------------------------------------------------------------------------------------------------------------------------|
  |                                                        |
  |        To:      ADSM-L AT VM.MARIST DOT EDU                       |
  |        cc:                                                        |
  |        Subject: Re: Move nodedata - what is moved first      |

>-------------------------------------------------------------------------------------------------------------------------------|



Peter,

> servers . Currently, our main file server has data on over 200 3590
> tapes therefore a directory restore can potentially have hours added to
> the process directly related to tape mounts.

Is the directory information you referring about related to Windows
systems? You should use the DIRMC client option to store all your
directory information in a DISK based storage pool (DISK or FILE), so that
it remains on faster quicker access media for restore purposes. (Dont let
that stuff go to tape for the reasons you have outlined below.)

The DIRMC client option is not really required for Unix based systems, as
the database has enough space to store that information.

...deon
---
Have you looked at the A/NZ Tivoli User Group website?
http://www.tuganz.org

Deon George, IBM Tivoli Software Engineer, IBM Australia
Office: +61 3 9626 6058, Fax: +61 3 9626 6622, Mobile: +61 412 366 816,
IVPN +70 66058
mailto:dgeorge AT au1.ibm DOT com, http://www.ibm.com/tivoli