On 5/3/2016 4:30 AM, Peter Hunkeler wrote:
I'm analyzing a job using CA's MA-Tune (similar to Strobe). Question is if
there is that could be optimized. Looking at the MA-Tune reports which are
based on 100 samples per second for 1 minute, I see that the job is seen in
IEAVEPS1 for some 20%. IEAVEPS1 is POST, IIRC.
I thought that POST would be a quick process and therefore I'm wondering it
appears at the top of the list. I'd like to understand a bit better what I'm
seeing and I'm *not* saying this is a problem.
The program does a lot with DB2, and when a DB2 call is made, I'd expect the
TCB to go into a WAIT until the SRB in DB2 completes and POSTs the TCB.
Watching from 9000 ft, is this about right?
Why do I not see any entry reporting the WAIT (IEAVEWT1 or similar)? Is this
just bad luck with sampling?
I would have thought that POSTing is quick, so I wonder why I often see a
couple of IEAVEPS1 entries in a row? Of yourse 1/100s is a long time so other
things may have happened inbetween, not being caught ba MA-Tune's sampling. I
wonder, however, if and what could case this to take longer that what I would
expect.
Any clue?
The tool is improperly counting samples in Pause processing as using CPU
rather than counting them as delays. Strobe had the same issue years
ago, when Pause/Release was first introduced.
Greg
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
