--------------------------<snip>----------------------------
I'm in agreement with what Kees said - and TMM is *not* $0. After much
effort (years) one of my customers finally jumped from a large TMM
implementation to (duplexed) VTSs. The DASD cost of TMM was about 30% of
the installation - maybe more. The HSM (CPU) cost was also very high -
in real chargeable mainframe M-thingies. And TMM pollutes the ML1
processing big time. And space management runs far more than should be
necessary.
TMM worked when it was introduced at this shop, but because it was
"free" it grew too big, and was then too hard to cost-justify removing.
Big motivator in the end was the dictate from above to reduce the cost
of IT by lowering the cap on the machine. They finally went and worked
out what all that HSM CPU time really cost.
--------------------------<unsnip>-------------------------
Here, as in many other areas, YMMV.
We had a very time-dependant daytime computing schedule and tape mounts
were causing frequent delays. Due to the nature of our business, delays
translated to fines and penalties that could get really large really
fast. TMM allowed us to eliminate tape-mount delays completely from most
of our entire production schedule. When I got "riffed", our only tape
mounts were full-volume and incremental backups, both of which left our
time-dependant processing to roll merrily forward without delay.
When fines and penalties can add up to as much as $50,000 per minute,
which would you rather spend money on? For us, time was TRULY money!
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
