If you look at the results of "q proc f=d" or "q se ### f=d", you might come
close to knowing how much has transpired -- but you need something (like
ServerGraph?) to track progress so you can see what the instantaneous
performance level of a given task... kinda like vmtune/vmstat every 5 sec's,
then compare the delta and run a continuing, smooth curve graph connecting
the dots across the intervals.

I agree with Mark's post;  while this is "nice to have", and there IS more
instrumentation being incorporated in latest versions, I would put the
recent rash of relentless, recurring, regression bugs at the very HIGHEST
priority -- new features aren't worth the effort when they come at the
expense of serious breakage (eg, the recent/ongoing saga with expiration &
conflict-lock!).

Don France
Technical Architect -- Tivoli Certified Consultant
San Jose, Ca
(408) 257-3037
mailto:don_france AT att DOT net

Professional Association of Contract Employees
(P.A.C.E. -- www.pacepros.com)