Enterprise Scheduler Service and Jobs - Oracle Fusion Applications Administration Essentials - page 60

To perform these kinds of scheduled jobs, a scheduling engine is required. Historic-
ally, the Operating System ( OS ) would provide this kind of utility service. In UNIX-
based systems, this was called cron . This approach worked for simplistic non-rule
based scheduling. The UI for interacting with the cron was usually text-based, and
required a highly structured programming construct.

The OS-based scheduler would not work for complex ERP processing. The ERP ap-
plications required a scheduler that would have the following features:

• Complex Schedules : This provides the ability to process complex schedul-

ing needs, for example, every third week or every alternate day, except on
the last week of the quarter.

• Dependencies : This provides the ability to understand dependencies

between scheduled jobs. This ensures that the processing follows business
rules. It ensures that the logic dependencies are maintained.

• Grouping : This provides the ability to group tasks and execute them togeth-

er. This provides for parallelism as well as the ability to enforce data integrity.

• Reprocessing : This provides the mechanism to reprocess jobs when tasks

fail to complete successfully. A clean restart is required.

• Monitoring : This allows us to monitor jobs that are running. It also helps us

to monitor the upcoming schedules and status of jobs that have been com-
pleted.

• Logging and Alerts : This provides detailed logging of the jobs. It also in-

cludes alerts for job completion and failures.