IBM Mainframe Forum Index
 
Log In
 
IBM Mainframe Forum Index Mainframe: Search IBM Mainframe Forum: FAQ Register
 

Reports to keep track if JOB takes more time to complete


IBM Mainframe Forums -> All Other Mainframe Topics
Post new topic   Reply to topic
View previous topic :: View next topic  
Author Message
nartcr

New User


Joined: 06 Jun 2007
Posts: 83
Location: anada

PostPosted: Mon Jul 16, 2012 7:27 pm
Reply with quote

Hello,

I have a requirement to perform below action. At this point i do not know how to achieve this, but if you can point me in some direction that would be great. We use ZEKE for job scheduling, CICS for online processing,MQ as a communication channel between online and batch.

Say JOB1 is defined in ZEKE, and JOB1 takes more than 1 hour to complete execution ( when on an avg, it completes execution in 30 minutes ).

I want to develop something which would track any jobs taking more than a specified time in system. The idea is that, we should be knowing if there are any issues around JOB1 which would help us take corrective measures.

May be schedule a batch job, and run it every 15 minutes - this would check SDSF active jobs, gather the start time and check if it exceeds normal time limit? Is this doable?

My understanding is that CICS online region wouldnt be able to track the batch job details, however, i wasnt able to find anything which says ZEKE can communicate with CICS.

Any help is appreciated
Back to top
View user's profile Send private message
dick scherrer

Moderator Emeritus


Joined: 23 Nov 2006
Posts: 19244
Location: Inside the Matrix

PostPosted: Mon Jul 16, 2012 9:30 pm
Reply with quote

Hello,

For all of the jobs you want to track, create a file that has the jobname and the "bogey" time (how long is acceptable for a run).

You can run your tracker in real time for batch jobs by looking in the system and comparing current time running against the bogey.

You can also run the tracker against "yesterday's" smf data (you would need help frlom your system support people) for the jobnames in the file that has the jobnames and acceptable times.

Keep in mind that often when a job runs longer than expected it is because of something else in the system. Wall time is not a very reliable source for measuring. . .
Back to top
View user's profile Send private message
nartcr

New User


Joined: 06 Jun 2007
Posts: 83
Location: anada

PostPosted: Mon Jul 16, 2012 10:52 pm
Reply with quote

Thanks Dick,

Are you suggesting something like this -

Schedule batch job every 15 minutes - use some tool like EZACFM1 to capture only ACTIVE jobs, and record the start time, and compare it with bogey?

Is this my only option?

If a job runs in infinite loop, at present, we are manually monitoring it - I was looking for automating it. As far as SMF data goes, i would expect only after job/step completes execution the data goes in SMF.
Back to top
View user's profile Send private message
expat

Global Moderator


Joined: 14 Mar 2007
Posts: 8797
Location: Welsh Wales

PostPosted: Mon Jul 16, 2012 11:00 pm
Reply with quote

Seems to me like another munch the resource project.

A job takes as long as a job takes, and with sufficient and applied testing should NEVER go into an infinite loop.

Get the basics right and you won't need to spend time and resource monitoring what should not happen.
Back to top
View user's profile Send private message
dbzTHEdinosauer

Global Moderator


Joined: 20 Oct 2006
Posts: 6966
Location: porcelain throne

PostPosted: Mon Jul 16, 2012 11:16 pm
Reply with quote

wooden wheels and stone knives.

ZEKE is a rather sophisticated piece of software. It is not some
piece of homegrown BS that attempts to automate job submission.

From the ZEKE Website a few quotes:
Quote:
Presents all schedule information for every supported platform on a single screen and highlights all exceptions (abends, late jobs, etc.)
Quote:
Tracks jobs enterprise-wide, regardless of platform
it appears obvious that the TS has limited understanding of the system as a whole.
ZEKE has all the facts and figures at its command.
Suggest that the TS start by learning ZEKE's capabilities.

I would venture to guess that ZEKE is underutilized

another example of Mcillian's famous quote:
Quote:
could not fix your brakes, so I made your horn louder
Back to top
View user's profile Send private message
dick scherrer

Moderator Emeritus


Joined: 23 Nov 2006
Posts: 19244
Location: Inside the Matrix

PostPosted: Mon Jul 16, 2012 11:40 pm
Reply with quote

Hello,

Quote:
Is this my only option?

No, but choosing what to run/when to run it depends on just what you want to accomplish. . .

If all you want is a "snapshot" periodically, you can look at jobs currently running in the system.

If you want to track a job's performance over time, the current snapshot won't be of much use.

Suggest you work with your scheduling people and learn what might be available thru ZEKE.
Back to top
View user's profile Send private message
nartcr

New User


Joined: 06 Jun 2007
Posts: 83
Location: anada

PostPosted: Wed Jul 18, 2012 12:15 am
Reply with quote

Thanks.
Back to top
View user's profile Send private message
View previous topic :: :: View next topic  
Post new topic   Reply to topic View Bookmarks
All times are GMT + 6 Hours
Forum Index -> All Other Mainframe Topics

 


Similar Topics
Topic Forum Replies
No new posts To get the the current time DFSORT/ICETOOL 13
No new posts RC query -Time column CA Products 3
No new posts C Compile time time stamps Java & MQSeries 10
No new posts Parallelization in CICS to reduce res... CICS 4
No new posts Insert system time/date (timestamp) u... DFSORT/ICETOOL 5
Search our Forums:

Back to Top