This section explains basic job flow operations that you should be familiar with before creating a job flow (job execution order).
Basic job flow operations
The job flow you create is essentially operated as explained below.
The jobs are started in the order in which they are joined by a connection line, starting with the first job.
If the preceding job terminates normally, the succeeding job is started. If the preceding job terminates abnormally, processing is canceled.
However, if the succeeding job is a recovery job, it is started even if the preceding job terminates abnormally.
Normal and abnormal termination of jobs
The completion code of a job is used to determine whether a job is "Completed" or "Abended".
The following table lists the default completion code settings.
"Completed" | Completion code is 0 |
" Abended " | Completion code is 1 or higher |
The range for which the job will be regarded as "Completed" can be changed.
Refer to "Changing the "Completed" range" for information on how to change the range.
Changing the "Completed" range
You may want to, for example, use the completion code to create a branching flow. By default, only jobs with a completion code of 0 are regarded as "Completed", so jobs with any other completion code are regarded as "Abended" and processing is canceled. For a branching flow, you must therefore configure the settings so that a job is not regarded as "Abended" and processing is not canceled even when the completion code is not 0.
In this case, change the range for which the job will be regarded as "Completed". You can use either of the following methods to change the "Completed" range.
Change "Upper limit of completion code" in the job information.
After the change, jobs with completion codes from 0 to the specified upper limit will be regarded as "Completed".
Set "Ignore completion code" in the job information.
After you set this value, jobs will be regarded as "Completed" regardless of the completion code, and processing will not be canceled.
Information
"Pseudo-normal"
If you increase the upper limit of the completion codes that will be regarded as "Completed", you can treat the completion codes from 1 to the upper limit as "Pseudo-normal" to differentiate these completion codes from the original "Completed" (completion code 0). Setting "Validate pseudo-normal" causes the job status to be differentiated as indicated below.
"Completed" | 0 |
"Pseudo-normal" | 1 to the set upper limit |
"Pseudo-normal" is differentiated from the "Completed" status, but processing is regarded as having terminated normally.
