Submitted job processing (z/OS®)

On z/OS®, to improve ODF performance, you can use a submitted job and the persistence feature.

When you use a job name on distributions, ODF uses a feature of z/OS that allows jobs to run in created address spaces. The started job persistence feature allows a started job to remain active for a period of time waiting for additional distributions to be run. You can control the period of time by using the timing parameters in the ARS.CFG file.

The started job feature is designed for distributions that have JOB NAME controls that are established by report administration. Address spaces that are dynamically created for Job Name controlled distribution processes remain persistent for as long as there is distribution work to be handled. Persistent distribution report bundle processing improves performance by reusing resources.

If the distribution is defined with a jobname, then the report bundle processor is started by a submitted job. There might be multiple started jobs running at one time. The maximum number of jobs that can run at one time is specified in the ARS.CFG file parameter named ARSODF_MAXJOBS.

If the report bundle processor is started as a submitted job then the jobname is used for the JES spool file data set name as well as the stepname. If the jobname within a distribution changes by having different jobnames defined in the report bundle definitions, then a new JES spool file data set will be created with the new jobname. The naming convention for JES spool file data sets with a jobname is P99nnnnn, where nnnnn is an incremental number that is generated by ODF to identify a unique JES spool file data set name. If the jobname or output definitions are different within the report bundles, then a new JES spool file data set is allocated for each new jobname or output parameters. The Report Break Indicator is not used when the report bundle processor is running as a submitted job (jobname specified).