It seems that many of us have our "rapid DB growth" war stories. Mine
was a HPUX node whose sysadmin decided to NFS-mount a very large
filesystem (tens of millions of files, which were already backed up to
TSM on their native node) but neglected to add it to his node's TSM
client exclude list.

He was mystified when his node's regular backups did not finish for a
week running, until I saw the TSM database filling up rapidly, the
scratch tape pool rapidly disappearing, and I found this happening. It
took a CPU-week to delete the unwanted filespace and reclaim all those
tapes.

Roger Deschner      University of Illinois at Chicago     rogerd AT uic DOT edu
=============== You can observe a lot just by watching. ================