i can

Dawn May

Dawn May

Bookmark and Share

October 2014

Sun Mon Tue Wed Thu Fri Sat
      1 2 3 4
5 6 7 8 9 10 11
12 13 14 15 16 17 18
19 20 21 22 23 24 25
26 27 28 29 30 31  

« Viewing Job Level SQL Metrics with the Performance Data Investigator | Main | POWER and PowerVM Reference Information »

July 24, 2013


We've had an issue for over a year involving QINACTITV and Java. There was a change made by IBM in 2012 that forced the JVM to become active at certain intervals. This invalidates the use of QINACTITV because the job never times out as inactive. According to many discussions with support this is working as designed and there are no plans to change it.

Sounds like some good thinking done here but that bit about break handling programs will require some thinking as this PTF is applied......

This change should help with the problem caused by Java consuming CPU time. Java is part of the reason that the change is enabled by default.

Oops, details. Obviously, if the break message reaches the screen, there is I/O and therefore activity. I keep forgetting that the message delivery mechanism does a suspend/restore of the screen around the message handling. That is also I/O so even when the message gets handled by the program it looks active. Better example would have been a signal processing procedure or a job interrupt event. Tried to use a more common example and messed up on accuracy. Sorry about that. Still requires thought about what might have previously been considered active but will now be considered inactive.

The comments to this entry are closed.