maybe my english is not sufficient, I will try toclearify:
- the database monitor doesn’t have any measurment for RLA, that did not change! (BTW: I’m not very interested in RLA workload, I didn’t use RLA for > 10 years).
some other aspects:
its not of interest to me how many query engines a database has (silly idea to have more than one and even more silly talking about this) and how they are named. Using SQL I describe a set of records and the query engine does it work, hopefully with good performance.
In an environment with heavy transaction activity, thosands of SQL accesses, done by hundreds of programms I’ve never seen before, tools like Visual Explain don’t help anything.
My firts look would be to WRKSYSSTS, WRKDSKSTS and WRKACTJOB, the number of Serverjobs QZDASOINIT and QPWFxxx has some information of the number of connections, the total database workload is not easy to find, because the local connects and RLA could not be identified and seperated from the other workload.
nexr step would be to just start monitoring all jobs for an hour or more and then having a look to the results and there is only one tool, IBM renames it from time to time (they seem to have fun with this activity), some call it database Navigator, I called it “oops nerv” because its buggy.