Airflow Summit 2025 is coming October 07-09. Register now for early bird ticket!

airflow.providers.databricks.hooks.databricks_sql

Attributes

LIST_SQL_ENDPOINTS_ENDPOINT

T

Classes

DatabricksSqlHook

Hook to interact with Databricks SQL.

Functions

create_timeout_thread(cur, execution_timeout)

Module Contents

airflow.providers.databricks.hooks.databricks_sql.LIST_SQL_ENDPOINTS_ENDPOINT = ('GET', 'api/2.0/sql/endpoints')[source]
airflow.providers.databricks.hooks.databricks_sql.T[source]
airflow.providers.databricks.hooks.databricks_sql.create_timeout_thread(cur, execution_timeout)[source]
class airflow.providers.databricks.hooks.databricks_sql.DatabricksSqlHook(databricks_conn_id=BaseDatabricksHook.default_conn_name, http_path=None, sql_endpoint_name=None, session_configuration=None, http_headers=None, catalog=None, schema=None, caller='DatabricksSqlHook', **kwargs)[source]

Bases: airflow.providers.databricks.hooks.databricks_base.BaseDatabricksHook, airflow.providers.common.sql.hooks.sql.DbApiHook

Hook to interact with Databricks SQL.

Parameters:
  • databricks_conn_id (str) – Reference to the Databricks connection.

  • http_path (str | None) – Optional string specifying HTTP path of Databricks SQL Endpoint or cluster. If not specified, it should be either specified in the Databricks connection’s extra parameters, or sql_endpoint_name must be specified.

  • sql_endpoint_name (str | None) – Optional name of Databricks SQL Endpoint. If not specified, http_path must be provided as described above.

  • session_configuration (dict[str, str] | None) – An optional dictionary of Spark session parameters. Defaults to None. If not specified, it could be specified in the Databricks connection’s extra parameters.

  • http_headers (list[tuple[str, str]] | None) – An optional list of (k, v) pairs that will be set as HTTP headers on every request

  • catalog (str | None) – An optional initial catalog to use. Requires DBR version 9.0+

  • schema (str | None) – An optional initial schema to use. Requires DBR version 9.0+

  • kwargs – Additional parameters internal to Databricks SQL Connector parameters

hook_name = 'Databricks SQL'[source]
supports_autocommit = True[source]
session_config = None[source]
http_headers = None[source]
catalog = None[source]
schema = None[source]
additional_params[source]
query_ids: list[str] = [][source]
get_conn()[source]

Return a Databricks SQL connection object.

run(sql: str | collections.abc.Iterable[str], autocommit: bool = ..., parameters: collections.abc.Iterable | collections.abc.Mapping[str, Any] | None = ..., handler: None = ..., split_statements: bool = ..., return_last: bool = ..., execution_timeout: datetime.timedelta | None = None) None[source]
run(sql: str | collections.abc.Iterable[str], autocommit: bool = ..., parameters: collections.abc.Iterable | collections.abc.Mapping[str, Any] | None = ..., handler: Callable[[Any], T] = ..., split_statements: bool = ..., return_last: bool = ..., execution_timeout: datetime.timedelta | None = None) tuple | list[tuple] | list[list[tuple] | tuple] | None

Run a command or a list of commands.

Pass a list of SQL statements to the SQL parameter to get them to execute sequentially.

Parameters:
  • sql – the sql statement to be executed (str) or a list of sql statements to execute

  • autocommit – What to set the connection’s autocommit setting to before executing the query. Note that currently there is no commit functionality in Databricks SQL so this flag has no effect.

  • parameters – The parameters to render the SQL query with.

  • handler – The result handler which is called with the result of each statement.

  • split_statements – Whether to split a single SQL string into statements and run separately

  • return_last – Whether to return result for only last statement or for all after split

  • execution_timeout – max time allowed for the execution of this task instance, if it goes beyond it will raise and fail.

Returns:

return only result of the LAST SQL expression if handler was provided unless return_last is set to False.

abstract bulk_dump(table, tmp_file)[source]

Dump a database table into a tab-delimited file.

Parameters:
  • table – The name of the source table

  • tmp_file – The path of the target file

abstract bulk_load(table, tmp_file)[source]

Load a tab-delimited file into a database table.

Parameters:
  • table – The name of the target table

  • tmp_file – The path of the file to load into the table

get_openlineage_database_info(connection)[source]

Return database specific information needed to generate and parse lineage metadata.

This includes information helpful for constructing information schema query and creating correct namespace.

Parameters:

connection – Airflow connection to reduce calls of get_connection method

get_openlineage_database_dialect(_)[source]

Return database dialect used for SQL parsing.

For a list of supported dialects check: https://openlineage.io/docs/development/sql#sql-dialects

get_openlineage_default_schema()[source]

Return default schema specific to database.

get_openlineage_database_specific_lineage(task_instance)[source]

Generate OpenLineage metadata for a Databricks task instance based on executed query IDs.

If a single query ID is present, attach an ExternalQueryRunFacet to the lineage metadata. If multiple query IDs are present, emits separate OpenLineage events for each query instead.

Note that get_openlineage_database_specific_lineage is usually called after task’s execution, so if multiple query IDs are present, both START and COMPLETE event for each query will be emitted after task’s execution. If we are able to query Databricks for query execution metadata, query event times will correspond to actual query’s start and finish times.

Args:

task_instance: The Airflow TaskInstance object for which lineage is being collected.

Returns:

An OperatorLineage object if a single query ID is found; otherwise None.

Was this entry helpful?