CA Jobtrac Job management r11: How to detect loops in a successor chain

Products

• Jobtrac

Issue/Introduction

This Knowledge Document explains how to detect loops within a schedule issuing 'Loop Detected' or 'Limit Reached. Loop?'.

Environment

Release:
Component: JOBTRC

Resolution

 

CA Jobtrac JM currently offers two methods to detect successor loops:

The 'L' line command and the (sample REXX) Imod LOOPDET as explained below.

1. The 'L' line command:
   
   The primary purpose of this Scheduled Event Network panel is to list all the event's predecessors and successors. If an event belonging to a loop is selected, then message "Loop Detected" is issued. If there's a loop, but the selected event doesn't directly belong to it, then the message "Limit Reached. Loop?" is issued.
   
   
2. Imod LOOPDET:
   
   Because of the recursive call of an internal subroutine, the online execution (SRVEDIT or JOBTRAC) may result in error message SRV277/GRX0011I (with "Control stack full"), so it is best to run in batch. It either issues message "LOOPDET00 CHECKPOINT VERIFIED: NO LOOPS" or "LOOPDET99 LOOP HAS BEEN DETECTED: SUCCESSOR CHAIN FOLLOWS". The JCL for the batch job is as follows:

   //jobname   JOB . . . your accounting data . . .        
   //STEP1    EXEC PGM=SRVBATCH,PARM='LOOPDET'  
   //STEPLIB    DD DISP=SHR,DSN=your.Jobtrac.CAILIB 
   //           DD DISP=SHR,DSN=your.GSS.CAILIB 
   //           DD DISP=SHR,DSN=your.DATACOM.CUSLIB 
   //           DD DISP=SHR,DSN=your.DATACOM.CAILIB 
   //IMOD       DD DISP=SHR,DSN=your.sample.REXX.ISET.file 
   //JOBTRACx   DD DUMMY with 'x' as your Jobtrac Token 
   //PARMLIB    DD *   ADDRESS JOBTRAC GJTRGCUU 15 MAXTASK 5 TYPE 3 
   //ISRVLOG    DD SYSOUT=* 
   //OUTPUT     DD SYSOUT=* 
   //