You create a workflow in the Developer tool. You add objects to the workflow and connect the objects with sequence flows. The Data Integration Service uses the instructions that you configure in the workflow to run the objects.

A workflow object is an event, task, or gateway. An event starts or ends the workflow. A task is an activity that runs a single unit of work in the workflow, such as running a mapping, sending an email, or running a shell command. A gateway defines a decision to split or merge the data paths in the workflow.

A sequence flow connects one workflow object to another. Sequence flows specify the order in which the Data Integration Service runs the objects. You can define a condition on a sequence flow to determine whether the Data Integration Service runs the object that the sequence flow specifies.

You can define and use workflow variables and parameters to make workflows more flexible. A workflow variable represents a value that records run-time information and that can change during a workflow run. A workflow parameter represents a constant value that you define before you run a workflow. You use workflow variables and parameters in conditional sequence flows and object fields. You also use workflow variables and parameters to pass data between a task and the workflow.

You can configure a workflow for recovery so that you can complete an interrupted workflow instance. A running workflow instance can be interrupted when an error occurs, when you cancel the workflow instance, or when a service process shuts down unexpectedly. You cannot recover an aborted workflow.

After you deploy a workflow, you run an instance of the workflow from the deployed application using the infacmd wfs command line program. You monitor the workflow instance run in the Monitoring tool.