tasks-streams

Tasks and Streams

This Snowflake Native Application sample demonstrates how to execute a task and visualize changes using streams within a native application.

Getting Started

Installation

Execute the following command:

snow app run

First-time setup

You should grant some permissions and bind some references before start with the app.

• warehouse_reference: You must choose the warehouse that your tasks will execute on periodically. This reference grants usage on the selected warehouse to the application. (See setup_script.sql on line 70)

• EXECUTE TASK permission: You must grant your application the permission to execute a task.

These references and permissions are defined in manifest.yml.

Once permissions and bindings are set, execute the create_task procedure to create and start a task that runs every minute. The application cannot create the task directly in the setup script, since it does not have the EXECUTE TASK permission by default.

The created task updates the internal.incremented_values table by incrementing its only value by one calling the increment_by_one procedure.

Test the app

In the main page of the app, you will see a dataframe with the values from the internal.incremented_values_stream, a stream based on the internal.incremented_values table.

The dataframe has two columns:

• VALUE: The incremented number by the task every minute.
• ACTION: The last action applied to the table in each row at a specific time. The action should always be insert.

At the end of the page, there are three buttons:

• Start task: Executes the create_task procedure and creates a new task if not exists.
• Stop Task: Executes the stop_task procedure and stops the already created task if it exists.
• Refresh: Reloads the page to see the updated value by the task.

Additional Resources

• Grant access to a Snowflake Native App
• Create a user interface to request privileges and references
• CREATE TASK
• CREATE STREAM