SQLExecuteQueryOperator for ClickHouse¶

Use the SQLExecuteQueryOperator to execute SQL commands in a ClickHouse database.

Because ClickHouseHook extends DbApiHook, no dedicated ClickHouse operator is needed — the generic SQLExecuteQueryOperator handles DDL, DML, and analytical queries.

Using the Operator¶

Set the conn_id argument to the ID of a ClickHouse connection.

ClickHouse Airflow Connection Metadata¶
Parameter	Input
Host: string	ClickHouse server hostname
Port: integer	HTTP port (default: `8123`, TLS: `8443`)
Login: string	ClickHouse username (default: `default`)
Password: string	ClickHouse user password
Database (Schema): string	Default database (default: `default`)
Extra: JSON dict	`secure`, `verify`, `connect_timeout`, `send_receive_timeout`, `compress`, `client_name`

An example usage of the SQLExecuteQueryOperator to connect to ClickHouse:

tests/system/clickhouse/example_clickhouse.py[source]

create_table = SQLExecuteQueryOperator(
    task_id="create_table",
    sql=CREATE_TABLE_SQL,
)

insert_rows = SQLExecuteQueryOperator(
    task_id="insert_rows",
    sql=INSERT_ROWS_SQL,
)
read_rows = SQLExecuteQueryOperator(
    task_id="read_rows",
    sql=f"SELECT id, name FROM {CLICKHOUSE_TABLE} ORDER BY id",
)

drop_table = SQLExecuteQueryOperator(
    task_id="drop_table",
    sql=DROP_TABLE_SQL,
)

Querying Data¶

Use a handler to return query results:

tests/system/clickhouse/example_clickhouse.py[source]

read_rows = SQLExecuteQueryOperator(
    task_id="read_rows",
    sql=f"SELECT id, name FROM {CLICKHOUSE_TABLE} ORDER BY id",
)

Note

session_settings passed to ClickHouseHook directly (via hook_params) take precedence over any session_settings defined in the connection’s extra JSON field. Conflicting keys are resolved in favor of the constructor argument.