Categories:

System functions (System Information)

SYSTEM$SET_RETURN_VALUE¶

Explicitly sets the return value for a task.

In a task graph, a task can call this function to set a return value. Another task that identifies this task as the predecessor task (using the AFTER keyword in the task definition) can retrieve the return value set by the predecessor task using SYSTEM$GET_PREDECESSOR_RETURN_VALUE.

Syntax¶

SYSTEM$SET_RETURN_VALUE( '<string_expression>' )

The value for the string_expression argument can be a string literal or a variable; for example, SYSTEM$SET_RETURN_VALUE(:VARIABLE).

Arguments¶

string_expression

The string to set as the return value. The string size must be <= 10 kB (when encoded in UTF8).

Examples¶

Create a task that sets a return value. Create a second, child task that runs after the predecessor task has completed. The child task retrieves the return value set by the predecessor task (by calling SYSTEM$GET_PREDECESSOR_RETURN_VALUE) and inserts it into a table row:

Copy codeExpand code block

-- Create a table to store the return values.
CREATE OR REPLACE TABLE return_values_table (str VARCHAR);

-- Create a task that sets the return value for the task.
CREATE TASK set_return_value_task
  WAREHOUSE = return_task_wh
  SCHEDULE = '1 MINUTE'
  AS
    CALL SYSTEM$SET_RETURN_VALUE('The quick brown fox jumps over the lazy dog');

-- Create a task that identifies the first task as the predecessor task and retrieves the return value set for that task.
CREATE TASK get_return_value_task
  WAREHOUSE = return_task_wh
  AFTER set_return_value_task
  AS
    INSERT INTO return_values_table VALUES(SYSTEM$GET_PREDECESSOR_RETURN_VALUE());


-- Note that if there are multiple predecessor tasks that are enabled, you must specify the name of the task to retrieve the return value for that task.
CREATE TASK get_return_value_by_pred_task
  WAREHOUSE = return_task_wh
  AFTER set_return_value_task
  AS
    INSERT INTO return_values_table VALUES(SYSTEM$GET_PREDECESSOR_RETURN_VALUE('get_return_value_task'));

-- Resume task (using ALTER TASK ... RESUME).
-- Wait for task to run on schedule.

SELECT DISTINCT(str) FROM return_values_table;
+-----------------------------------------------+
|                      STR                      |
+-----------------------------------------------+
|  The quick brown fox jumps over the lazy dog  |
+-----------------------------------------------+

SELECT DISTINCT(RETURN_VALUE)
  FROM TABLE(information_schema.task_history())
  WHERE RETURN_VALUE IS NOT NULL;


+-----------------------------------------------+
|                  RETURN_VALUE                 |
+-----------------------------------------------+
|  The quick brown fox jumps over the lazy dog  |
+-----------------------------------------------+

Show lessSee more

Scroll to top

Example 2: Call by using a separate stored procedure¶

Similar to the first example, but set the return value for the task and retrieve it by calling separate stored procedures:

-- Create a table to store the return values.
CREATE OR REPLACE TABLE return_values_sp (str VARCHAR);

-- Create a stored procedure that sets the return value for the task.
CREATE OR REPLACE PROCEDURE set_return_value_sp()
RETURNS STRING
LANGUAGE JAVASCRIPT
EXECUTE AS CALLER
AS $$
var stmt = snowflake.createStatement({sqlText:`CALL SYSTEM$SET_RETURN_VALUE('The quick brown fox jumps over the lazy dog');`});
  var res = stmt.execute();
$$;

-- Create a stored procedure that inserts the return value for the predecessor task into the 'return_values_sp' table.
CREATE OR REPLACE PROCEDURE get_return_value_sp()
RETURNS STRING
LANGUAGE JAVASCRIPT
EXECUTE AS CALLER
AS $$
var stmt = snowflake.createStatement({sqlText:`INSERT INTO return_values_sp VALUES(SYSTEM$GET_PREDECESSOR_RETURN_VALUE());`});
var res = stmt.execute();
$$;

-- Create a task that calls the set_return_value_sp stored procedure.
CREATE TASK set_return_value_t
WAREHOUSE=warehouse1
SCHEDULE='1 MINUTE'
AS
  CALL set_return_value_sp();

-- Create a task that calls the get_return_value stored procedure.
CREATE TASK get_return_value_t
WAREHOUSE=warehouse1
AFTER set_return_value_t
AS
  CALL get_return_value_sp();

-- Resume task.
-- Wait for task to run on schedule.

SELECT DISTINCT(str) FROM return_values_sp;
+-----------------------------------------------+
|                      STR                      |
+-----------------------------------------------+
|  The quick brown fox jumps over the lazy dog  |
+-----------------------------------------------+

SELECT DISTINCT(RETURN_VALUE)
  FROM TABLE(information_schema.task_history())
  WHERE RETURN_VALUE IS NOT NULL;

+-----------------------------------------------+
|                  RETURN_VALUE                 |
+-----------------------------------------------+
|  The quick brown fox jumps over the lazy dog  |
+-----------------------------------------------+

Example 3: Use a variable to set the return value¶

The following example demonstrates how to dynamically generate a return value based on the task’s execution and set the return value by using a variable. In this example, the task loads data from a stream into a landing table and sets the return value to indicate the number of rows loaded:

CREATE OR REPLACE TASK load_raw_data
WAREHOUSE = 'WH'
WHEN
    SYSTEM$STREAM_HAS_DATA('NEW_WEATHER_DATA')
AS
    DECLARE
        rows_loaded NUMBER;
        result_string VARCHAR;
    BEGIN
        INSERT INTO raw_weather_data ( -- our landing table
            row_id)
        SELECT
            row_id
        FROM
            new_weather_data  -- our source stream
        ;

        -- to see the number of rows loaded in the UI
        rows_loaded := (SELECT $1 FROM TABLE(RESULT_SCAN(LAST_QUERY_ID())));
        result_string := :rows_loaded || ' rows loaded into RAW_WEATHER_DATA';
        -- show result string as task return value
        CALL SYSTEM$SET_RETURN_VALUE(:result_string);
    END;