Procedures + Tasks

Use Case	Benefit
Daily or hourly ETL jobs	Native orchestration, no external scheduler
Conditional logic	Use JavaScript or SQL logic in procedures
Backfill or cleanup routines	Automate data maintenance
Audit, logging, or exception handling	Add structured error management

---

Snowflake stored procedures support JavaScript with embedded SQL, or Snowflake Scripting (SQL-like syntax).

CREATE OR REPLACE PROCEDURE copy_data()
RETURNS STRING
LANGUAGE SQL
AS
$$
BEGIN
  INSERT INTO target_table
  SELECT * FROM source_table
  WHERE load_date = CURRENT_DATE;

  RETURN 'Data copied successfully';
END;
$$;

---

Once your procedure is created, you can schedule it with a Snowflake Task.

CREATE OR REPLACE TASK daily_copy_task
  WAREHOUSE = my_wh
  SCHEDULE = 'USING CRON 0 2 * * * UTC'
AS
  CALL copy_data();

ALTER TASK daily_copy_task RESUME;

---

Check status and logs:

SHOW TASKS;

SELECT *
FROM TABLE(INFORMATION_SCHEMA.TASK_HISTORY())
WHERE TASK_NAME = 'DAILY_COPY_TASK'
ORDER BY SCHEDULED_TIME DESC;

---

You can chain tasks using the AFTER clause:

CREATE TASK parent_task
  WAREHOUSE = my_wh
  SCHEDULE = 'USING CRON 0 * * * * UTC'
AS
  CALL stage_data();

CREATE TASK child_task
  WAREHOUSE = my_wh
  AFTER parent_task
AS
  CALL transform_data();

Now, child_task runs automatically after parent_task completes.

---

Tip	Explanation
Use separate warehouses for task execution	Avoid resource contention
Enable error handling in procedures	Use TRY/CATCH or RETURN codes
Monitor TASK_HISTORY() regularly	Track failures and durations
Chain tasks for multi-step workflows	Mimics DAG-style logic
Use SUSPEND during maintenance	Prevent unintended task runs

---

• Snowflake Stored Procedures
• Snowflake Tasks
• Snowflake Scripting (Procedural SQL)

---