Skip to main content

How to connect to Snowflake

Prerequisites​

To connect Snowflake with Rhombus AI, you need to:

  1. A valid Snowflake account with access credentials
  2. Username and Password for authentication
  3. Access to at least one Database in your Snowflake account
  4. A Warehouse configured in your Snowflake account (required for query execution)
info

The Snowflake Connector is an enterprise-grade data warehouse native connector. Unlike cloud storage connectors (such as S3), it doesn't just "read files" but integrates directly with the Snowflake database through SQL protocol, enabling table-level data retrieval.

Step 1. Access the Connection Wizard​

  1. In the Workflow interface, navigate to the canvas.

  2. Click + Add Node from the left sidebar or top toolbar.

  3. Select Data Input from the node menu.

  4. In the configuration panel, click Connect Third Party.

  5. In the popup window, select Snowflake.

    Snowflake Connection Wizard

Step 2. Enter Credentials​

The connection wizard uses a 3-step process. The first step is credential verification.

  1. Enter your Snowflake Account (account ID).

    tip

    Your Snowflake account identifier can be found in your Snowflake URL or account settings.

  2. Enter your Username.

  3. Enter your Password.

  4. Click Continue to verify your credentials.

info

The system will validate your credentials in real-time by calling the Snowflake API. Upon successful authentication, it will automatically retrieve all databases visible to your account.

caution

All required fields must be filled before the Continue button becomes active. The form uses strict validation to ensure data integrity.

Step 3. Select Database and Warehouse​

After successful credential verification, you'll proceed to the resource selection step.

  1. Select Database: Choose a target database from the dropdown list.

    info

    The database list is automatically populated based on your account permissions. If you add a new database in Snowflake, it will appear in the list after refreshing.

  2. Select Warehouse: Choose a compute warehouse from the dropdown.

    tip

    Snowflake requires a warehouse to execute queries. Make sure the selected warehouse is running or can be started automatically.

info

After selecting a database, the system automatically retrieves all schemas and tables within that database. This auto-discovery feature ensures you always see the latest database structure.

Step 4. Select Tables​

The final step allows you to choose specific tables to sync.

  1. The interface displays all Tables and Views available in the selected database.

  2. Browse through the list (displayed as a tree or list structure).

  3. Select one or more tables by checking the boxes next to them.

  4. Click Connect to complete the connection.

info

You can select multiple tables in a single connection session. All selected tables will be synchronized to your project's Data Sources.

After Connection​

Once you click Connect successfully:

Data Sources Management​

  • Selected tables will appear in your project's Data Sources list
  • Table metadata is synchronized to Rhombus AI's Data Source management system
  • You can view and manage these data sources from the project settings

Using Snowflake Tables in Workflows​

  1. In the Workflow canvas, add a Data Input node.
  2. In the node configuration, select your connected Snowflake table from the dropdown menu.
  3. The system will use Snowflake's SQL interface to extract data directly into your workflow.

Data Extraction​

  • Full or Incremental Reads: The system supports both full table reads and incremental data extraction based on your configuration
  • Direct SQL Integration: Data is retrieved through Snowflake's native SQL protocol, ensuring efficient and reliable data transfer
  • Real-time Discovery: If new tables are added to your Snowflake database, they will be available after refreshing the connection

Key Features​

Auto-Discovery​

The connector automatically discovers databases, schemas, and tables without requiring manual input. This means:

  • New databases appear automatically after refresh
  • Schema changes are reflected in real-time
  • No need to manually update table names or structures

Type Validation​

The connection wizard uses strict form validation:

  • All required fields must be completed before proceeding
  • Invalid credentials are caught at the frontend level
  • Empty inputs are prevented before API calls

Asynchronous Loading​

The interface handles loading states gracefully:

  • Loading indicators appear when fetching databases and tables
  • Large table structures are loaded asynchronously to prevent UI freezing
  • Clear feedback is provided during each step transition

Troubleshooting​

Connection Fails at Credentials Step​

  • Verify your Snowflake account identifier is correct
  • Ensure your username and password are valid
  • Check that your Snowflake account is active and accessible

No Databases Appear​

  • Confirm your account has access to at least one database
  • Verify your user permissions in Snowflake
  • Try refreshing the connection

Warehouse Selection Issues​

  • Ensure the selected warehouse exists and is accessible
  • Check that the warehouse has sufficient credits (if applicable)
  • Verify warehouse permissions for your user account

Tables Not Loading​

  • Confirm the selected database contains tables or views
  • Check your database and schema permissions
  • Wait for the auto-discovery process to complete (may take time for large databases)
tip

If you encounter persistent issues, try disconnecting and reconnecting with fresh credentials. The system validates all inputs before proceeding to the next step.