ESP Workload Automation

  • 1.  ESP - Simulation consternation

    Posted Oct 31, 2018 10:46 AM
      |   view attached

    Greetings Everyone!

     

    I had a request over the weekend to change the time dependency on a job from its regular time to a different time, but just for that day.  The job normally runs at midnight, but they wanted it pushed to 18:30 the same day.  I checked to ensure that the event loads at midnight, and it does, so it would certainly support the change. Normally, we accomplish this by adding an statement to the job definition like this:

     

    IF TODAY('OCT 28 2018') THEN DELAYSUB 18:30

     

    Once I added this line to the job definition and simulated, the results of the simulation indicate that the job would indeed get kicked off as I intend:

     

    SIMULATION OF EVENT ESPPROD.HIPA5010 AT 00.00.00 ON SUNDAY OCTOBER 28TH, 2018

     

    TIME RELATED DEPENDENCIES

    HE800X03  EARLYSUB(08.45.01 SUN 28 OCT 2018)

    HE800X53  EARLYSUB(13.30.01 SUN 28 OCT 2018)

    HE706W01  EARLYSUB(07.00.01 SUN 28 OCT 2018)

    HE805X51  EARLYSUB(18.30.01 SUN 28 OCT 2018)

     

    However, the event loaded at midnight and the job ran at it's normal time:

     

    AJ HE805X51 APPL(HIPA5010.2572)

    Job=HE805X51(45941028), Wob=AX, Completed->CC 0

       Submit Overdue ........... 2018/10/28 00:20

       Started .................. 2018/10/28 00:29

       Completed ................ 2018/10/28 00:34

       AverageRunTime=7

       Agent=PESP027, EspProc=SS.CAI.ESP.PROD.APPLLIB(HIPA5010), Doclib=SS.CAI.ESP.PROD.DOCLIB

     

    My colleges tell me that the reason this happened is becomes of the order of the statements (i.e. I added the IF statement BELOW the DELAYSUB statement):

     

    AIX_JOB HE805X51

    RUN DAILY

    DELAYSUB 00:00

    IF TODAY('OCT 28 2018') THEN DELAYSUB 18:30

    LATESUB 00:20

    AGENT PESP027

    SCRIPTNAME /translator/ediprod/scripts/HE805X51.ksh

    USER ecoper

    ENDJOB

     

    I am not complaining about this behavior.  It is my opinion, however, that the statements should have been read in the correct order that they would be during submission at the time of simulation.  If this had been the case, the simulation would have shown the time dependency at midnight as opposed to 18:30, and I would know to move the statement.

     

    I am not sure if this is a problem or if I need to submit an idea for it.

     

    Jonathan Calloway

    Batch Scheduling Specialist

    Operations Support

    Office: 423.535.7342

    Cell: 423.309.2547

     

     

     

    -


    Please see the following link for the BlueCross BlueShield of Tennessee E-mail disclaimer:  http://www.bcbst.com/email_disclaimer.shtm

     



  • 2.  Re: ESP - Simulation consternation

    Posted Oct 31, 2018 11:12 AM

    The SIMULATION should be how it runs, maybe you should open a Support Case with CA.

     

    The job ran 00:29 not at 00:00, could something else be in play?

     

    AJ HE805X51 APPL(HIPA5010.2572)

    Job=HE805X51(45941028), Wob=AX, Completed->CC 0

       Submit Overdue ........... 2018/10/28 00:20

       Started .................. 2018/10/28 00:29

       Completed ................ 2018/10/28 00:34


    Did you check the ESP Job Output for what happened around this time?

     

    The LATESUB statement indicates a late submission time for a job. This is the latest acceptable submission time before the job is flagged as overdue.

     

    Was the job flagged as overdue at 00:20 on 10/28 and someone manually intervened and started it?
     

    I have not tested this but I would think it would be better to group the DELAYSUB/LATESUB commands and then inlcude the appropriate times.

     

    AIX_JOB HE805X51
    RUN DAILY
    IF TODAY('OCT 28 2018') THEN DO
      DELAYSUB 18:30
      LATESUB  18:50         /* 20 minutes after DELAYSUB */
    ENDDO
    ELSE DO
      DELAYSUB 00:00
      LATESUB  00:20         /* 20 minutes after DELAYSUB */
    ENDDO


    This way you do one set of commands or the other and eliminates depending on the order ESP reads/processes commands.



  • 3.  Re: ESP - Simulation consternation

    Posted Oct 31, 2018 01:45 PM

    Rick,

     

    We IPL'd on Sunday night due to Z13 to Z14 upgrade.  So I do not have any logs to help answer your questions.

     

    <JC>



  • 4.  Re: ESP - Simulation consternation

    Posted Oct 31, 2018 11:24 AM

    Hi, 

    I agree with Rick. I think someone changed the start time. 

     

    Coding the job this way works as you expected. 



  • 5.  Re: ESP - Simulation consternation

    Posted Nov 01, 2018 08:32 AM

    Well, I have not been able to duplicate this.  I cannot find any requests or other evidence that an operator changed anything.  I've got one more scheduled test set for tomorrow morning.  I'll mark this helpful if it is successful. 

     

    <JC>