Brainstorming, I thought of another method.
Check QHST for CPF1164, then if the job that ended matches one of the jobs from the susbsystem, send the alert.
What would really be nice if IBM included the subsystem name as an additional variable on the CPF1164.
Then you would only have to check CPF1164 ending from subsystem X.
Any chance IBM could add the subsystem name to the CPF1164?
From: MIDRANGE-L [mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of Musselman, Paul
Sent: Thursday, June 07, 2018 11:30 AM
To: Midrange Systems Technical Discussion
Subject: RE: V7R3 - Halcyon Subsystem job monitor - determine which job is not running
We wrote our own monitor, but we just check min/ideal/max jobs in the subsystem. The subsystem only runs jobs from 1 application, so any deviation is an issue to be dealt with. If the number of jobs isn't perfect, we want to know.
Our process uses 2 programs-- first is our 'RTVSBSINF' command, which is a wrapper for the QWDRSBSD API. This is a 'nice' API, and doesn't require a User Space. The second program we call SBSRUNCHK. It uses the RTVSBSINF command and sends a lot of messages if the number of jobs running isn't correct. It also checks to make sure the SBS is running in the first place!
Paul E Musselman
From: MIDRANGE-L [mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of Steinmetz, Paul
Sent: Thursday, June 07, 2018 10:46 AM
To: 'Midrange Systems Technical Discussion' <midrange-l@xxxxxxxxxxxx>
Subject: V7R3 - Halcyon Subystem job monitor - determine which job is not running
I have a Halcyon subsystem job monitor, which triggers if all of the jobs in the subsystem are not running.
The trigger is determined if the number of jobs falls below X. (X is currently 50)
I would like to add the alert sent the job name of the job which job is NOT running,
Is there way to determine which job is not running?
One offered solution was to create a separate monitor for every job, doable, but overkill.
IBM i Systems Administrator
Pencor Services, Inc.
462 Delaware Ave
Palmerton Pa 18071