Our ADSM storage pools have been configured with MAXSIZE=50m so
that huge files don't overflow our limited tape capacity.

This is working as expected.  The client gives some error messages:
   ANS1228E Sending of object '/u/raid/ashildur/Rufe/2x2/N/wfn-0.5fm2.gz' failed
   ANS1310E Object too large for server limits
and the server reports:
   ANR0521W Transaction failed for session 47 for node ASRV (DEC UNIX) - object
   excluded by size in storage pool BACKUPPOOL and all successor pools.

The PROBLEM with the ADSM implementation is that the client first
sends (almost) 50 MB of the file to the server, before the server
discovers that it doesn't really want to backup the file.
This is really hurting our backups, because the files excluded
by size are in excess of 100GB, and the ADSM backup process spends
99% of its time needlessly sending files across the network.

QUESTION:  Can't the ADSM server/client protocol exchange the
MAXSIZE=50m parameter once and for all, so that the client will
never send the large files across the network in the first place ?
Is there a way to configure such a behavior with ADSM ?

We're running the ADSM Server Version 3, Release 1, Level 2.90
on AIX-RS/6000.  The clients run Version 3, Release 1, Level 0.8.
Do later versions of the Tivoli ADSM solve this problem ?

Thanks for sharing your insights.

Ole Holm Nielsen
Department of Physics, Technical University of Denmark,
Building 307, DK-2800 Lyngby, Denmark