Hi,
In December 2005 we migrated our Netware cluster as follows:

Previous system:
* Netware 6.0 SP 4
* NCS 1.6
(all files are stored on an EMC-CX600 SAN and on NSS-Volumes)
* Tivoli Storage Manager Backup/Archive Client
DSMC.NLM Version 5.1.5
* SMDR 6.50.03 28 January 2003
* TSA600 6.00.05 6 February 2003
(with /cluster off)
* TSANDS 10110.95 26 September 2002

Actual system:
* Netware 6.5 SP 3
* NCS 1.7
(all files are stored on an EMC-CX600 SAN and on NSS-Volumes)
* Tivoli Storage Manager Backup/Archive Client
DSMC.NLM Version 5.3 17 February 2005
* SMDR 6.54.02 17 May 2005
* TSAFS 6.51 8 June 2005
(with /nocluster)
* TSANDS 10551.44.05 3 February 2004

>From this time, we observed that our incremental Tivoli backup

works substantially (about factor 10) slowlier.

For instance, processing about 720,000 files lasts about 7 h
(in comparison with about 45 min on our previous system).
It turned out
that this time is consumed only for inspecting the objects,
not for backupping the files itself.
Excluding files from backup does not reduce the number of objects
inspected
and the amount of time needed.
We use the same dsm.opt files as before.

Has anybody observed a similar behavior?
Can anybody guess the reason?
Is anybody able to provide a hint
which modifications of parameters/options may be successful?

B. Kuhlmann
University of Hamburg