Hi,

Does anyone use the generate backupset command - whether it is for DRP test
or for ad-hoc archiving for user data?

I basically have two questions:
1)     If the backupset is supposed to speed recovery time - has anyone
found this to be true?
2)    Has anyone really tested the process and had some files not included
in the backupset - files that should have been there?

Some background info first:
Hardware:
Client
RS/6000 H40 - I think
Server(TSM)
Dell optiplex GX1
Software:
O/S - AIX4.3
Tsm client V3.7.2.0
Tsm Server 3.7.0.0

The process we followed was as follows:

The idea was to take data from our off_site pool and restore the Tsm server
database backup onto a server with identical hardware/software. After the
Tsm server was restored the client would be restored on a test box.
This process works - but restoring the Tsm server first to get to the data
on the client was a bit of a pain for the guys doing the DR tests for the
client.
I thought of backupsets - this is what the Tivoli website has to say about
backupsets:
=======================================================
Network-Free Rapid Recovery

Rapid Recovery provides the ability to create a backup set which
consolidates a client's files onto a set of media that is portable and may
be directly readable by the clients system for fast, LAN-free
(non-networked) restore operations. The portable backup set, which is
synthesized from existing backups, is tracked and policy managed by the
Tivoli Storage Manager server.
======================================================
Now - here comes my first problem. I generated a backup set from the NT
server onto 8mm mammoth casettes. I took the backup set to the client and
started the restore. It takes about 1/2 an hour before any data is
transferred and then stops after a few files had been restored and then
sends data after a pause of about 15 minutes or so. What happened to rapid
recovery?
Before we start analyzing the problem etc. the same files we tried restoring
took 2 minutes from the Adsm server(via the network), and 3 hours from the
backupset directly from a mammoth tape streamer connected to the client.

My second question is regarding the completeness of the backupset that was
generated:
Some files were not on the backupset - while they have been backed up which
was verified by a select * from backups for this node. The file is marked in
the Tsm database as an active copy of the file. In theory this file should
have been included in the backupset but was not.

Now - has anyone done what I'm attempting to do - and had success with it?
It seems like the restore of data from the client is very inefficient!

Can somebody from developement maybe shed some light on why this should be
happening?

Thanks
Christo Heuer