ADSM-L

Re: Backup Question

1999-06-30 02:58:04
Subject: Re: Backup Question
From: Clarence Beukes <clarence AT ZA.IBM DOT COM>
Date: Wed, 30 Jun 1999 08:58:04 +0200
Hi Henry...

Keep your implementation simple and easy to manage.
Rather define one PD; which means one PS. Looking at your information, I
would assume that you want to treat all client data in the same way(6
copies of each file on each node). If you do have different requirements
for certain nodes or files on a node; define seperate MGMT classes. Seeing
that you only have one tape drive in your library; you would want to keep
the mgmt classes' copy groups that are linked to storage pools, to a
minimal; reducing tape mount requests; which might end up in a bottleneck
within ADSM.

Hope this helps.

   Have a nice day
   Clarence Beukes
   IT Specialist (Storage Solutions)
   IBM Product Support Services
   Tel : +2711 302 5333   Fax : +2711 302 8209
   Mobile : +2782 573 5665

"The success of men never lies in falling, but in falling and rising up
again."



Henry Chui <Henry_Chui AT VANDAGROUP DOT COM> on 06/30/99 08:11:12 AM

Please respond to "ADSM: Dist Stor Manager" <ADSM-L AT VM.MARIST DOT EDU>

To:   ADSM-L AT VM.MARIST DOT EDU
cc:    (bcc: Clarence Beukes/South Africa/IBM)
Subject:  Backup Question





Hello all,

     I have a server with ADSM server for NT installed with only one tape
drive, if I want to keep six copies of backup of the server  (i.e.
increment backup of each copy of six in each day).  Does any one have good
idea to do it? The current solution I can think of is just define six
nodes, belong to six different PD, PS, MG..etc , six different stgpools.
Finally, use backup client to backup it.  Any better solutions?


Thanx a lots!
Henry Chui
<Prev in Thread] Current Thread [Next in Thread>