The event-driven process

Event-driven workload automation is an ongoing process and can be reduced to the following steps:

1. The administrator defines the event rule, that is the criteria to correlate event conditions and actions.
2. The administrator builds and activates triggering selection data based on the previous criteria, starting the event-driven process: on each system where the scheduler runs, specific components act as event or action provider. The action provider access the data correlating event conditions and actions, according to the following schema:

   	Event Provider	Action Provider	Correlation data
   Data set triggering	SMF	Tracker, requesting to change special resource availability	EQQJCLIB member in the system where the tracker runs
   HFS or ZFS file triggering	Tracker, handling jobs that run file watching	Controller, releasing dependencies on operations that run file watching	Dependencies defined on operations that run file watching file watching utility parameters
   Event-triggered tracking (ETT)	JES user request to change special resource availability	Controller, adding an occurrence to the current plan	Database ETT record in the system where the controller runs

The following sections describe how you can set up this process. In particular:

• To trigger an action on a special resource, see Data set triggering.
• To trigger an action based on HFS or ZFS file changes, see HFS or ZFS file triggering.
• To trigger workload submission (adding an occurrence), define the correlation as described in Adding occurrences by event-triggered tracking.