Today we have a tool called TBLMON which can monitor the usage
of several system tables. It does not do "all tables", but it
does do many tables that (over the years) have at one time or
another been an issue. This tool is stable (*), easy to setup
and run, and relatively useful. (* This tool has been run on
many production systems and has never caused a problem. The
worst thing that has ever happened is that the an out of date
version of the tool reported a wrong number - no impact on
the system.)

TBLMON is not officially distributed in any manner, though
lots of people have copies of it (especially various members
of the HP3000 support community). We are considering making this
tool available on Jazz (with some supporting documentation).

The longer term plan is to have an officially supported tool
that monitors system resource utilization, which is distributed
with the system (FOS). The tenative plan is to have this tool
be submitted to the next release of the OS (no official name/
number for that release has been announced). This tool will
probably be called RMT (for Resource Monitoring Tool) --
clever eh? :-)

The main improvement of RMT (?) over TBLMON will be official
integration into the OS (implying supportable). In addition
we are hoping to expand the list of "resources" monitored by
RMT over the current list. The user interface and output from
TBLMON will not be different in RMT - but the concepts will
be similar.

We are also (tenatively) planning on making a copy of RMT
available which can execute on 5.0, 5.5, and 6.0 systems
(hopefully without a patch). We will be trying to recruit
"customer partners" to run this tool on their production
systems to gather data that helps us plan and prioritize
future "capacity" improvements. Ideally these customers
would be "large customers" (in the sense that they have
a lot of users on their systems and are closer to pushing
existing system limits).

If any of you have experience with TBLMON and would like
to suggest some enhancements which would be particularly
useful, this is a good opportunity. In addition if you
know of "resources" that it does not monitor that you would
like to see monitore, let me know. We will take your
suggestion into account in implmenting RMT (no promises
that we will do all suggestions).

Also if any of you think your environment is a good
example of a large, growing, HP3000 shop and you would
be willing to run either TBLMON or RMT to collect data
please let me know.

One final note, if any of you presently use the tool
called "threshold manager" please let me know. I am still
trying to decide how important the threshold manager
functionality would be in RMT (namely the ability to
set "thresholds" and be alerted if any resource hits
a particular "threshold" -- like 'alert me when I
am using over 75% of the maximum number of processes).

Thanks
Scott McClellan