DatabricksTaskOperator¶

Use the DatabricksTaskOperator to launch and monitor task runs on Databricks as Airflow tasks. This can be used as a standalone operator in a DAG and as well as part of a Databricks Workflow by using it as an operator(task) within the DatabricksWorkflowTaskGroup.

Examples¶

Running a notebook in Databricks using DatabricksTaskOperator¶

tests/system/databricks/example_databricks.py[source]

    task_operator_nb_1 = DatabricksTaskOperator(
        task_id="nb_1",
        databricks_conn_id="databricks_conn",
        job_cluster_key="Shared_job_cluster",
        task_config={
            "notebook_task": {
                "notebook_path": "/Shared/Notebook_1",
                "source": "WORKSPACE",
            },
            "libraries": [
                {"pypi": {"package": "Faker"}},
                {"pypi": {"package": "simplejson"}},
            ],
        },
    )

Running a SQL query in Databricks using DatabricksTaskOperator¶

tests/system/databricks/example_databricks.py[source]

    task_operator_sql_query = DatabricksTaskOperator(
        task_id="sql_query",
        databricks_conn_id="databricks_conn",
        task_config={
            "sql_task": {
                "query": {
                    "query_id": QUERY_ID,
                },
                "warehouse_id": WAREHOUSE_ID,
            }
        },
    )