Hi Steve,

If using the GUI, you can use the "/TSMOPTFILE=xxx.opt"
and the "/CONFIGFILE=xxx.cfg" GUI parameters to solve this.
This would enable you to have a set of DSM.OPT files
and DP/SQL configuration files that do exactly what you want.
You would have a set of these two files for each SQL server instance.
Then you would launch the GUI like this:

  TDPSQL -TSMOPTFILE=srv1inst1.opt -CONFIGFILE=srv1inst1.cfg
  TDPSQL -TSMOPTFILE=srv1inst2.opt -CONFIGFILE=srv1inst2.cfg
  .
  TDPSQL -TSMOPTFILE=srv2inst1.opt -CONFIGFILE=srv2inst1.cfg
  TDPSQL -TSMOPTFILE=srv2inst2.opt -CONFIGFILE=srv2inst2.cfg
  .
  TDPSQL -TSMOPTFILE=srv3inst1.opt -CONFIGFILE=srv3inst1.cfg
  TDPSQL -TSMOPTFILE=srv3inst2.opt -CONFIGFILE=srv3inst2.cfg
  .
  .

- Each of the DSM.OPT files would have the correct
       NODENAME srvXinstY
  for the SQL Server that the data is stored under.

- Each of the config files would have the correct
       SQLSERVER srvX\instY
  setting to specify which SQL Server to connect to.


As for your single NODENAME idea.... you could store the data
all under one TSM NODENAME, but then it becomes difficult for
scheduling and management of data.
I would keep them under different NODENAMEs.

Thanks,

Del

----------------------------------------------------

"ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU> wrote on 12/19/2005
04:48:54 PM:

> TSM serv 5.2.2.0 & 5.2.4.3
> TSM client 5.3.0.5
> TDP client 5.2.1.0
>
> 12 SQL Server machines, each of which has 2 or 3 SQL Server instances -
> some in the 600GB+ range.
>
> Currently, each machine is defined to TSM as a node named <machine>_SQL
>
> Problem: TSM is used for "cloning" prod databases to test machines
> (cross-machine restores).  This is done by our SQL admin via the TDP
> gui.  Frequently, he forgets to "undo" the changes this process makes to
> the .opt and .cfg files, and backups get sent to the wrong node name, or
> fail.
>
> Question:
> 1. I want to script a front-end that will allow him to plug in the
> machine & instance and then execute the gui using overrides, but I'm not
> sure exactly which ones are needed and which ones are not.
> 2. Assume I want to restore a database from instance "ADAM\EVE" into
> instance "CAIN\ABEL".
> 3. If I'm on server CAIN and I specify SQLSERVER=CAIN\ABEL and
> FROMSQLSERVER=ADAM\EVE , do I still have to somehow override the node
> name as well?
> 4. If I used a generic node name like SQLSERVER on all the backups,
> would that buy me anything on the restore side as far as any machine
> being able to see all the instance backups?
>
> thanks.
>
> Steve Schaub
> Systems Engineer, WNI
> BlueCross BlueShield of Tennessee
> 423-752-6574 (desk)
> 423-785-7347 (cell)
>
> Please see the following link for the BlueCross BlueShield of Tennessee
E-mail
> disclaimer:  http://www.bcbst.com/email_disclaimer.shtm