On Mar 13, 2010, at 8:47 PM, yoda woya wrote:

if I were to be looking  to assess the health of TSM, what would be
the  top
things to look for

With solid planning and serious hardware allocation, followed by a
regimen of good administration, you should have a solid system.  The
Performance Tuning Guide and Problem Determination Guide will give you
a good sense of things to watch for.  Most problems result from
hardware failures and client administrators not being as diligent with
their end of things (e.g., runaway backups, arbitrarily renaming file
systems, deficient client environment planning and administration).
You must regularly watch the server, to at least have a sense of how
things seem when all is well, to know right away when things are not
okay ("knowing the pulse of the system").  Running out of scratches is
a cardinal sin of failed server administration, as is not noticing
that TSM db backups are not happening.  It's not hard to develop a
basic TSM health monitor based upon dsmadmc, and there are commercial
packages to do a thorough job of it.

     Richard Sims