We have chosen to use Netview on MVS to automate the ADSM maintenance
cycle. It kicks off at a certain time of day (depending on
day-of-week), invokes the cycle steps as prerequisite steps finish,
takes incremental DB dumps at certain intervals depending on
time-of-day, and issues commands to DEL VOLH, scratch offsite empty
tapes, etc. at certain times of the day. Some cycle steps are
trickier than others, such as migration. We actually count the
processes as they start and stop so we know our "migration" cycle step
completes.
This mechanism allows us to put conditional logic to work for us. For
example, if a migration process starts (and we haven't initiated it as
a set step of the cycle), we can E-Mail appropriate parties so that
they can monitor the need for more space in the pools. We can also
automatically restart a DB-Backup operation that fails due to IO
errors. We set reclamation to 90% after completing the cycle on
weekdays only if it is early enough in the day and then cancel the
process(es) prior to the next day's cycle.
Another nice reason to use a console automation product: we can issue
operator alerts based on the MVS descriptor class with a small ANR*
message id exception list. In fact, our exception list can specify
what messages warrant an alert, which ones generate e-mail, and which
ones we ignore.
Etc.
Jim Kleberg
State Street
______________________________ Forward Header __________________________________
Subject: Automated Steps to Create Tapes for Disastere Recovery
Author: jlawson (jlawson AT THEHARTFORD DOT COM)
Date: 6/30/97 9:30 AM
|