Integration Service Activities

Last updated May 5, 2025

Technical references

Working with Snowflake activities

Tip: This page includes some tips you may find useful when working with Snowflake activities.

When you expect your query to result in a large set of rows, always use LIMIT and OFFSET to paginate through your results. Refer to the Snowflake documentation for query examples for this use case.
Generate Output Schema works successfully only for query statements that do not contain any workflow variables. Once the output schema is generated, you can replace the query with variables. We recommend using a query that you’re successfully able to run in Snowflake directly.
Use Execute Any Query Asynchronously for any queries that take more than 30 seconds to execute.
You can execute non-Select commands (such as insert, update, stored procedures, etc.) only with the Execute Any Query Asynchronously activity.
Use Get Asynchronous Query Execution Status with the query ID obtained from the output of Execute Any Query Asynchronously.
You can use the Asynchronous Query Status Changed trigger to initiate a process when a query was successfully executed or encountered an error.
To get the results of an asynchronous query, use Get Asynchronous Query Execution Status. Use the query ID from the output of Execute Any Query Asynchronously to generate the output schema. Once the schema is generated, you can use a variable query ID. Changing the input to a variable does not affect the output properties, which remain the same. This means the generated output can't be changed unless you manually regenerate it. We recommend setting up different workflow processes if needed. If you’re using Studio Desktop, you can use the Native Search activity from Snowflake package version 5.0.1.
For the Record Created and Record Updated triggers to run successfully, you must execute the following query on the tables you are monitoring for events: ALTER TABLE [TABLE_NAME] SET CHANGE_TRACKING = TRUE;. Additionally, make sure you have the required permissions configured in Snowflake.

Identifying the Primary Key output while inserting a row in Snowflake

If you have a table with an auto-incrementing ID column, a common use case is wanting to insert a row in that table and output the ID of the created row.

The simplest workaround to this problem is to maintain a foreign key relationship in the table where you want to offload your data. In the example below, you may find that in the Snowflake Insert Record activity, even though you have an Output SNOWFLAKE_ID, there is no real output during execution when the row gets added. This is a Snowflake limitation. Therefore, we maintain a Foreign Key Column (SFDC_OPPORTUNITY_ID) for this table.

Once the row is added, we can use the Snowflake Native search activity to query the Primary Key (SNOWFLAKE_ID) in exchange for the Foreign Key (SFDC_OPPORTUNITY_ID) and using a subsequent Get Row Item system activity.

The returned SNOWFLAKE_ID (Primary Key) can also be stored in the upstream application in a custom field (Salesforce in this case), so that each time an update needs to be performed in a row, the ID does not need to be queried with Foreign Key.

On this page