SMF30 CPU statistics and the JESLOG TOTAL CPU TIME.
Posted: Thu Jul 31, 2014 1:47 pm
HI,
As a performance analyst, I've a question that should the CPU seconds that get recorded in SMF type 30 subtype 5 record for a job, be the same when you look at the TOTAL CPU TIME reported for the same job in the JESLOG? I'm noticing that the number of CPU seconds is always a little more (like 3-4%) in the SMF 30 record, then what shows up in JESLOG. Is it shop specific or there is some general rule for this?
As a performance analyst, I've a question that should the CPU seconds that get recorded in SMF type 30 subtype 5 record for a job, be the same when you look at the TOTAL CPU TIME reported for the same job in the JESLOG? I'm noticing that the number of CPU seconds is always a little more (like 3-4%) in the SMF 30 record, then what shows up in JESLOG. Is it shop specific or there is some general rule for this?