ESP Workload Automation

When a job is submitted with a JCL error there is no indication of the error.

The job in CSF displays a "pnode" = input and "job status" = submitted but nothing to indicate an error.

Reading through the manual I found a "sysmsgs" command which would appear to fix the "no indication" problem.

I set up a test using sysmsgs 'IEFC452I' ccfail desc(2) assuming that this would show up as a failed job in ESP and send a WTO message but I can't get it to trigger.

Does anyone have any suggestions or ideas on how to handle the JCL error notification.

Running ESP 11.3 on a z/os 1.12 system

Thanks

David.Mendoza:

When a job is submitted with a JCL error there is no indication of the error.

The job in CSF displays a "pnode" = input and "job status" = submitted but nothing to indicate an error.

Reading through the manual I found a "sysmsgs" command which would appear to fix the "no indication" problem.

I set up a test using sysmsgs 'IEFC452I' ccfail desc(2) assuming that this would show up as a failed job in ESP and send a WTO message but I can't get it to trigger.

Does anyone have any suggestions or ideas on how to handle the JCL error notification.

Running ESP 11.3 on a z/os 1.12 system

 

Thanks

 

Depending on the type of the type of JCL error you may not get an indication if you use the CCFAIL mechanism. The rationale is that CCFAIL inserts a step named ESPCCFCK as the last step of the job to catch the condition code errors. If you get a JCL error that causes all the remaining steps in the job to be flushed then there is not going to be any CCFAIL indication as the ESPCCFCK step is going to be flushed.  One way to get around this problem is to use CCCHK rather than CCFAIL. CCCHK causes a PGM=CYBES080 step to be inserted right before the first step in the job and as a consequence can catch errors that CCFAIL cannot catch like the scenario I described above. In addition to using CCCHK I would also suggest you use a DUEOUT statement and an NOTIFY OVERDUE statement.  This will catch JCL errors wherein you submit a job from one JES node that is transmitted to another JES node. That is to say the overdue notification will signify that you probably got a JCL error at the destination/executing node rather than the orgin/submitting node.

Hope this helps.

Michael E. Ellis, Systems Programmer, Deere & Company

 

 

Thanks Michael ,

I'll look into using your suggested method as it appears to be the best way to go. Just playing devil's advocate, I'm not sure the Production group will go back and update the thousands of jobs in the scheduler.

The "SYSMSGS" parameter appears to be able to handle the error globally (read as: no added work to their group) rather than modifying each individual application, I just can't get it to fire.

Thank You

In order for SYSMSGS to work I believe the message must appear in the JESYSMSG output of the job not the JESMSGLG and I don't see a situation where that message appeared in JESYSMSG (although I did not do exhaustive testing).

 

Good luck,

Greg

 

David.Mendoza:

Thanks Michael ,

I'll look into using your suggested method as it appears to be the best way to go. Just playing devil's advocate, I'm not sure the Production group will go back and update the thousands of jobs in the scheduler.

The "SYSMSGS" parameter appears to be able to handle the error globally (read as: no added work to their group) rather than modifying each individual application, I just can't get it to fire.

Thank You

Hi Gregg,

This is the output of the LSYSMSGS command (I was hoping to catch either msg):

lsysmsgs                                                      
'IEFC452I' ID(0001), DESC(2), CCFAIL                          
'IEFC605I' ID(0002), DESC(2), NAME(P-), WTO, JOBNAME, JCLERROR 

This is what appears in the JESYSMSG ddname:

********************************* TOP OF DATA **********************************
 STMT NO. MESSAGE                                                              
        2 IEFC605I UNIDENTIFIED OPERATION FIELD                                
        3 IEFC001I PROCEDURE ENCORED WAS EXPANDED USING SYSTEM LIBRARY SYS1.SHAW
******************************** BOTTOM OF DATA ********************************

From the JESMSGLG ddname:

11.50.50 JOB27613 ---- TUESDAY,   18 FEB 2014 ----                  
11.50.50 JOB27613  IRR010I  USERID STCESP   IS ASSIGNED TO THIS JOB.
11.50.50 JOB27613  IEFC452I POPS999R - JOB NOT RUN - JCL ERROR  289  

This is all my test job produces except for the JCL but I would think something should be captured or reflected in the CSF/Director.

The test appears similar to the examples in the book.  

Thanks

We don't use this facility the way you appear to be trying to.  Are these jobs running from ESP, if so you should be able to get what you want with the following NOTIFY:

Notify   Failure -
Route(02) Desc(02)

 

You can even code the notify to send an email to a mailbox if that's of any value to you (and I think you can have mutilpe notifies).

 

With the above notify (and actually I'm not sure the notify is required at all for this) you should see the job in CSF like:

Job Name Account  ApplName Gen# P Node   Job Qual Job Status                    Jobno
___ PJCLERR  ?        TESTGFS    53 FAIL              FAILED, JCL ERROR AT 10.20 1  12457

 

For sure the notify will generate the following at the operator console:

*ESP_1195W Job PJCLERR(JOB12461) ENDED, CMPC=JCLERROR AT 10.35.31 ON 19 FEB 2014, APPL TESTGFS

 

The above job failed with a JCL error before it began execution.  If you have one that has something like a DSN not found and you have Encore (you didn't say if you do) you would get something like:

Job Name Account  ApplName Gen# P Node   Job Qual Job Status                    Jobno
___ PJCLERR  ?        TESTGFS    53 FAIL              ENCORE PREDICTS ERROR, CC146  12458

 

and a message like so at the console:

ESP_1195W Job PJCLERR(JOB12460) ENDED, CMPC=146 AT 10.35.03 ON 19 FEB 2014, APPL TESTGFS.54

 

If you still need to get the SYSMSGS working then the only idea I woud have for you (other than opening a ticket with CA) would be for you to check your TRACKOPT setting, per the manual (are any SYSMSGS working, I don't recall if you have said previously?):

If TRACKOPT SYSMSGS is not specified, system messages are not intercepted.

 

Again, good luck,

Greg

 

David.Mendoza:

Hi Gregg,

This is the output of the LSYSMSGS command (I was hoping to catch either msg):

lsysmsgs                                                      
'IEFC452I' ID(0001), DESC(2), CCFAIL                          
'IEFC605I' ID(0002), DESC(2), NAME(P-), WTO, JOBNAME, JCLERROR 

This is what appears in the JESYSMSG ddname:

********************************* TOP OF DATA **********************************
 STMT NO. MESSAGE                                                              
        2 IEFC605I UNIDENTIFIED OPERATION FIELD                                
        3 IEFC001I PROCEDURE ENCORED WAS EXPANDED USING SYSTEM LIBRARY SYS1.SHAW
******************************** BOTTOM OF DATA ********************************

From the JESMSGLG ddname:

11.50.50 JOB27613 ---- TUESDAY,   18 FEB 2014 ----                  
11.50.50 JOB27613  IRR010I  USERID STCESP   IS ASSIGNED TO THIS JOB.
11.50.50 JOB27613  IEFC452I POPS999R - JOB NOT RUN - JCL ERROR  289  

This is all my test job produces except for the JCL but I would think something should be captured or reflected in the CSF/Director.

The test appears similar to the examples in the book.  

Thanks

Hi Greg,

Thanks for the information on using the "notify". Since I saw the sysmsgs command in the documentation I thought it would be a good solution to our issue,

I was just wondering if anyone else was using it for the same purpose and could point me in the right direction.

Thanks for your testing and explanation. This is a great community, I'm glad to be a part of it.

Thanks 

David,

Out of curiosity did you ever check your TRACKOPT setting or are other SYSMSGS working?

Cheers,

Greg

Hi Greg,

This is the first time that we've tried to use this command so no other messages are set-up for capture.

This is the output of the "trackopt sysmsgs" command:

 trackopt sysmsgs                                                                 
ESP2558I TRACKING OPTIONS: NOSTC NOTSU SYSMSGS MASTER NOTRACK_PURGE NOPOST_OLDEST 

I've also tried to enable the specific sysmsgs with the "sysmsgs enable id(0002)" command eventhough the manual indicates that they are enabled at definition time.

David,

After some testing in my not so spare time (I guess I'm a Squirrel, shiny objects distract me), it seems that messages in the section of JESYSMSG as follows

do not trigger the SYSMSGS:

STMT NO. MESSAGE
          3 IEFC621I EXPECTED CONTINUATION NOT RECEIVED

Where messages like the following do:

IEF253I SGSBR14A STEP TEST - DUPLICATE NAME ON DIRECT ACCESS VOLUME 

Go figure. I do not know which you were looking for but I expect it was the first and I could not trap that message.  If thats what you were

doing and you would like this to work for those messages I suggest you confirm my results with CA and if confirmed, request an enhancememt.

Good luck,

Greg

Thank You very much Gregg, I'll contact CA.