DatabricksSubmitRunOperator
DatabricksSubmits a Spark job run to Databricks using the api/2.1/jobs/runs/submit API endpoint.
Access Instructions
Install the Databricks provider package into your Airflow environment.
Import the module into your DAG file and instantiate it with your desired params.
Parameters
tasksArray of Objects(RunSubmitTaskSettings) <= 100 items. See also https://docs.databricks.com/dev-tools/api/latest/jobs.html#operation/JobsRunsSubmit
jsonA JSON object containing API parameters which will be passed directly to the api/2.1/jobs/runs/submit endpoint. The other named parameters (i.e. spark_jar_task, notebook_task..) to this operator will be merged with this json dictionary if they are provided. If there are conflicts during the merge, the named parameters will take precedence and override the top level json keys. (templated) See also For more information about templating see Jinja Templating. https://docs.databricks.com/dev-tools/api/latest/jobs.html#operation/JobsRunsSubmit
spark_jar_taskThe main class and parameters for the JAR task. Note that the actual JAR is specified in the libraries. EITHER spark_jar_task OR notebook_task OR spark_python_task OR spark_submit_task OR pipeline_task OR dbt_task should be specified. This field will be templated. See also https://docs.databricks.com/dev-tools/api/2.0/jobs.html#jobssparkjartask
notebook_taskThe notebook path and parameters for the notebook task. EITHER spark_jar_task OR notebook_task OR spark_python_task OR spark_submit_task OR pipeline_task OR dbt_task should be specified. This field will be templated. See also https://docs.databricks.com/dev-tools/api/2.0/jobs.html#jobsnotebooktask
spark_python_taskThe python file path and parameters to run the python file with. EITHER spark_jar_task OR notebook_task OR spark_python_task OR spark_submit_task OR pipeline_task OR dbt_task should be specified. This field will be templated. See also https://docs.databricks.com/dev-tools/api/2.0/jobs.html#jobssparkpythontask
spark_submit_taskParameters needed to run a spark-submit command. EITHER spark_jar_task OR notebook_task OR spark_python_task OR spark_submit_task OR pipeline_task OR dbt_task should be specified. This field will be templated. See also https://docs.databricks.com/dev-tools/api/2.0/jobs.html#jobssparksubmittask
pipeline_taskParameters needed to execute a Delta Live Tables pipeline task. The provided dictionary must contain at least pipeline_id field! EITHER spark_jar_task OR notebook_task OR spark_python_task OR spark_submit_task OR pipeline_task OR dbt_task should be specified. This field will be templated. See also https://docs.databricks.com/dev-tools/api/2.0/jobs.html#jobspipelinetask
dbt_taskParameters needed to execute a dbt task. The provided dictionary must contain at least the commands field and the git_source parameter also needs to be set. EITHER spark_jar_task OR notebook_task OR spark_python_task OR spark_submit_task OR pipeline_task OR dbt_task should be specified. This field will be templated.
new_clusterSpecs for a new cluster on which this task will be run. EITHER new_cluster OR existing_cluster_id should be specified (except when pipeline_task is used). This field will be templated. See also https://docs.databricks.com/dev-tools/api/2.0/jobs.html#jobsclusterspecnewcluster
existing_cluster_idID for existing cluster on which to run this task. EITHER new_cluster OR existing_cluster_id should be specified (except when pipeline_task is used). This field will be templated.
librariesLibraries which this run will use. This field will be templated. See also https://docs.databricks.com/dev-tools/api/2.0/jobs.html#managedlibrarieslibrary
run_nameThe run name used for this task. By default this will be set to the Airflow task_id. This task_id is a required parameter of the superclass BaseOperator. This field will be templated.
idempotency_tokenan optional token that can be used to guarantee the idempotency of job run requests. If a run with the provided token already exists, the request does not create a new run but returns the ID of the existing run instead. This token must have at most 64 characters.
access_control_listoptional list of dictionaries representing Access Control List (ACL) for a given job run. Each dictionary consists of following field - specific subject (user_name for users, or group_name for groups), and permission_level for that subject. See Jobs API documentation for more details.
wait_for_terminationif we should wait for termination of the job run. True by default.
timeout_secondsThe timeout for this run. By default a value of 0 is used which means to have no timeout. This field will be templated.
databricks_conn_idReference to the Databricks connection. By default and in the common case this will be databricks_default. To use token based authentication, provide the key token in the extra field for the connection and create the key host and leave the host field empty. (templated)
polling_period_secondsControls the rate which we poll for the result of this run. By default the operator will poll every 30 seconds.
databricks_retry_limitAmount of times retry if the Databricks backend is unreachable. Its value must be greater than or equal to 1.
databricks_retry_delayNumber of seconds to wait between retries (it might be a floating point number).
databricks_retry_argsAn optional dictionary with arguments passed to tenacity.Retrying class.
do_xcom_pushWhether we should push run_id and run_page_url to xcom.
git_sourceOptional specification of a remote git repository from which supported task types are retrieved. See also https://docs.databricks.com/dev-tools/api/latest/jobs.html#operation/JobsRunsSubmit
Documentation
Submits a Spark job run to Databricks using the api/2.1/jobs/runs/submit API endpoint.
There are three ways to instantiate this operator.
See also
For more information on how to use this operator, take a look at the guide: DatabricksSubmitRunOperator
Example DAGs
Execute ML Pipelines in Databricks
Run an end-to-end pipeline from BigQuery data ingest, feature engineering, model training, to publishing all within a Databricks cluster.
AI + Machine Learning
Retrain ML Model in Databricks
Demonstrates orchestrating ML retrain pipelines executed on Databricks with Airflow.
AI + Machine Learning
Execute ML Pipeline in Databricks with AutoML
Demonstrates orchestrating ML pipelines executed on Databricks with Airflow using Databricks' AutoML.
AI + Machine Learning