John,

   I have heard a request for what you require made to IBM before, but
there is no mechanism in ADSM to somehow equate files across one node to
the other.

However, all is not lost.  Consider these options:
1) Do your application installs all in a common directory and exclude them
in the     DSM.OPT.

2) Use a more standard management class and let ADSM do its magic.  Due to
the fact that these files (executables, dll's etc) do not change often,
there will usually only be one copy stored for each file on each server.

I have done some thinking about running two backups on one server.  The
first excludes the application execution directories and the second backs
up only these directories, but with a common NODE name that will be used to
backup this common directory across all servers.  The caveat is that there
may be files stored in what is assumed to be a static subdirectory that the
application changes and needs.

I saw an interesting idea float across the newsgroup that outlined defining
a management class that pointed to storage groups that did not actually
save any data.

Does anyone else have any ideas on how to implement this type of solution.


IBM, can you tell us if there is any upcoming support?

Dan T.

----------
> From: John Dawson <jdawson AT TKG DOT COM>
> From: John Dawson <jdawson AT TKG DOT COM>
> To: ADSM-L AT VM.MARIST DOT EDU
> Subject: ADSM backing up redundant files
> Date: Wednesday, March 26, 1997 3:28 PM
>
> How does ADSM handle the situation where you have copies of a given set
> of files on many machines in the network?  For example, say you've
> deployed MS Office across 2000 desktop machines.  What's the best way
> to make sure you don't waste time doing 1999 backup operations, chewing
> up your network and server storage area for no good reason?
>
> --
> John Dawson // Internet: jdawson AT tkg DOT com // The Kernel Group, Inc.