Create an Integration Task to Configure and Run a Data Flow
In Data Integration, an integration task lets you prepare a data flow for running.
You can add parameters to a data flow, so that it can be used with different values each time you run the task. Default parameter values can be updated within an integration task. You can then validate the data flow with the updated parameters, before you publish the task to an Application.
In this tutorial, you:
- Create an integration task to configure runtime context for a data flow.
- Create an Application to contain and run published tasks.
- Publish a task to an Application.
- Create and monitor a task run.
Before You Begin
To create an integration task, you must have the following:
- Access to a Data Integration workspace. See Connect to Data Integration.
- A project and data flow created.
1. Creating an Integration Task
2. Creating an Application
An Application is a container for published tasks, data flows, and their dependencies. You can run published tasks in an Application for testing, or roll them out into production.
3. Publishing a Task to an Application
4. Creating and Monitoring a Task Run
Additional Resources
To learn more, see:
What's Next
Now that you've created an integration task to configure and run a data flow, you can Load Data into Autonomous Data Warehouse Using a Data Loader Task.