RE: CPU monitor >90% for 5 minutes - attach to the alert the jobs and/or tasks causing the issue.

Rob,

1) PRTSYSACT does not exist on my V7R1 LPAR.

2) One problem with WRKSYSACT to an outfile is that it automatically submits the job instead of creating QPFRDATA/QAITMON outfile in the same job.
This, I will not be able to utilize the results within my monitor.

However, I did find this.

When you run WRKSYSACT to an outfile, it submits a job, calling this program.
CALL PGM(QPYMON) PARM(*FILE 005 *CPU *ALL 00060 QAITMON QPFRDATA)
taking a snapshot every 5 seconds for 60 intervals (5 minutes).

So instead of calling WRKSYSACT to an outfile, I could call this instead.
CALL PGM(QPYMON) PARM(*FILE 005 *CPU *ALL 00001 QAITMON QPFRDATA)
Changing to 1 interval.

The next issue or problem is that when WRKSYSACT runs interactively, the field it sorts on, CPU util, does not exist in the outfile.
Average CPU Utilization
The Average CPU utilization field shows the average CPU utilization
for all the active processors.

I'm still reviewing the fields available from QPFRDATA/QAITMON
An SQL by Cpu Time For Job Or Task may give me the result needed.

Any thoughts from the group.

Paul

-----Original Message-----
From: MIDRANGE-L [mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of Rob Berendt
Sent: Monday, January 09, 2017 1:55 PM
To: Midrange Systems Technical Discussion
Subject: Re: CPU monitor >90% for 5 minutes - attach to the alert the jobs and/or tasks causing the issue.

If using a file is too straightforward you can try the command PRTSYSACT to take the file you created with WRKSYSACT OUTPUT(*FILE) and generate a spool file.


Rob Berendt
--
IBM Certified System Administrator - IBM i 6.1 Group Dekko Dept 1600 Mail to: 2505 Dekko Drive
Garrett, IN 46738
Ship to: Dock 108
6928N 400E
Kendallville, IN 46755
http://www.dekko.com





From: Rob Berendt <rob@xxxxxxxxx>
To: Midrange Systems Technical Discussion <midrange-l@xxxxxxxxxxxx>
Date: 01/09/2017 01:53 PM
Subject: Re: CPU monitor >90% for 5 minutes - attach to the alert
the jobs and/or tasks causing the issue.
Sent by: "MIDRANGE-L" <midrange-l-bounces@xxxxxxxxxxxx>



Oddly enough I failed to see a Work Management Service which fits that
bill. Sure Active_Job_Info comes close but it fails to cover system tasks

like SysTmpStg does.
https://www.ibm.com/support/knowledgecenter/ssw_ibm_i_73/rzajq/rzajqudfactivejobinfo.htm


I suggest you submit an RFE for this.

Instead of being fixated on analyzing printer output why don't you just
use WRKSYSACT OUTPUT(*FILE) and read that instead?


Rob Berendt