Use a Notebook activity in a pipeline
After you've created a linked service in Azure Data Factory for your Azure Databricks workspace, you can use it to define the connection for a Notebook activity in a pipeline.
To use a Notebook activity, create a pipeline and from the Databricks category, add a Notebook activity to the pipeline designer surface.
Use the following properties of the Notebook activity to configure it:
Category | Setting | Descriptions |
---|---|---|
General | Name | A unique name for the activity. |
Description | A meaningful description. | |
Timeout | How long the activity should run before automatically canceling. | |
Retries | How many times should Azure Data Factory try before failing. | |
Retry interval | How long to wait before retrying. | |
Secure input and output | Determines if input and output values are logged. | |
Azure Databricks | Azure Databricks linked service | The linked service for the Azure Databricks workspace containing the notebook. |
Settings | Notebook path | The path to the notebook file in the Workspace. |
Base parameters | Used to pass parameters to the notebook. | |
Append libraries | Required code libraries that aren't installed by default. | |
User properties | Custom user-defined properties. |
Running a pipeline
When the pipeline containing the Notebook activity is published, you can run it by defining a trigger. You can then monitor pipeline runs in the Monitor section of Azure Data Factory Studio.