The Automation of SPARCS!
      I have built members in PMHH.REXX.JCL/LIB to automatically build JCL to create the SPARCS files and download them to the server.

      Late each month, Medical Records (or whatever they are calling themselves at the moment) will open an Infopoint ticket requesting the SPARCS runs for the previous month. On the requested date, run the following jobs one at a time from PMHH.REXX.JCL:

  SPARCSOP - creates/submits job 'OPmmmyy'
  SPARCSFK - creates/submits job 'FKmmmyy'
  SPARCSCT - creates/submits job 'CTmmmyy'
  SPARCSIP - creates/submits job 'IPmmmyy'

Each job MUST complete successfully before the next job is run!
(Jobs are created with 'TYPRUN=HOLD', so they can be released one at a time).

      The followiing jobs will create the JCL in datasets instead of submitting them. If the jobs need to be created then modified before submission, this technique should be used.

  ZPARCSOP - creates job 'OPmmmyy' in PMHH.OPMMMYYD.JCL
  ZPARCSFK - creates job 'FKmmmyy' in PMHH.FKMMMYYD.JCL
  ZPARCSCT - creates job 'CTmmmyy' in PMHH.CTMMMYYD.JCL
  ZPARCSIP - creates job 'IPmmmyy' in PMHH.IPMMMYYD.JCL

Example: Request is to create September SPARCS data on Friday, October 27th, 2006.
     Submitting the SPARCSxx jobs in October would create 'xxSEP06' jobs and download to server as 'xxsep06.txt'

      If for some reason, the September data was not ready until after October, we would run the ZPARCSnn jobs, which would create PMHH.xxMMMYYD.JCL with:
     //xxOCT06 JOB ...
     DATE=11/01/06,
     xxoct06.txt

These would be changed to:
     //xxSEP06 JOB ...
     DATE=10/01/06,
     xxsep06.txt
before being submitted and run one at a time.


Always run IPmmmyy job last. It creates a file needed for Thomson.