Goal: Assume that you irregularly have to schedule a chain of jobs, that is, a set of jobs linked by predecessor-successor relationships.

You’d like to consolidate the scheduling of this chain of jobs into a single job that sees to the scheduling of all jobs in the chain. The advantage of this consolidation: you need only to schedule a single job to launch the chain of jobs.

Implementation: You’d implement this job in three small ABAP reports. These show both:

• a loose, informal predecessor-successor relationship (between Jobs 1 and 2, no guarantee that Job 1 completes before Job 2); and
• a tight relationship using the "start after predecessor" mechanism of the background processing system (between jobs 2 and 3). Here, there is a guarantee that Job 2 finishes before Job 3. In this case, there’s even a guarantee that Job 2 has completed successfully before Job 3 can start.

The first report schedules Job 1 to run immediately. Job 1 does some work and schedules both Job 2 and Job 3. There is no formal predecessor/successor relationship between Jobs 1 and 2. Usually, Job 2 won’t start until Job 1 processing has been completed. But it’s possible for steps in Jobs 1 and 2 to overlap. For example, if "Program_X" in Job 1 is an external program and Job 1 does not wait for Program_X to complete, then "Program_Y" in Job 2 and Program_X in Job 1 could overlap.

Job 1 must schedule both Jobs 2 and 3 because a job cannot schedule its own successor. This is because a predecessor job must have the status Scheduled or Released when a successor job is scheduled. If you want to set up chains of jobs, do so from a single initial report or job, as in this example.