C3 AI Documentation Home

Authenticate to Snowflake Using Key Pair or PAT in Data Fusion

A Snowflake connector enables ingestion of structured data from a Snowflake data warehouse into Data Fusion. Snowflake connectors are categorized under SQL/JDBC connectors and are commonly used for batch-oriented ingestion workflows.

Use this connector to access tables and views stored in Snowflake and ingest the data into target types for downstream processing, transformation, and analytics workflows.

When to Use a Snowflake Connector

Use a Snowflake connector when:

  • Your data resides in Snowflake
  • You need to ingest structured, table-based data
  • Your data can be queried using SQL

Typical use cases include:

  • Batch ingestion from warehouse tables
  • Analytics and reporting pipelines
  • Data synchronization workflows
  • Enterprise data warehouse integrations

Connector Configuration Model

Snowflake connectors are configured using the Configure Snowflake Connector wizard, which organizes connector setup into separate stages.

StagePurpose
ConnectorSelect the connector type
DestinationConfigure the connector identity
AuthenticationConfigure connection and authentication settings
Review & TestValidate and save the connector

Snowflake uses separate Destination and Authentication configuration sections in the connector wizard.

Connection endpoint and session configuration fields are configured on the Destination page. These fields include:

  • serverEndpoint
  • port
  • database
  • warehouse

Identity and secret-based authentication fields are configured on the Authentication page.

Authentication Model

Supported authentication methods include:

  • KeyPair — Uses a private key for authentication
  • ProgrammaticAccessToken (PAT) — Uses a token for authentication

The Authentication page always includes the following identity fields:

  • username
  • password

Selecting an authentication method reveals additional credential fields specific to that method.

Authentication Method Comparison

MethodDescriptionBest Suited ForSecurity ProfileNotes
KeyPairUses a private key (PEM format) with a passphraseProduction workloads, automation, and service accountsHighRecommended for secure non-interactive access. Eliminates password-only dependency. Simplifies authentication without requiring private key management. Token lifecycle management is still required.
ProgrammaticAccessToken (PAT)Uses a token for authenticationService integrations and automated workflowsHighToken lifecycle management is still required.

KeyPair Authentication Fields

When KeyPair is selected, the following additional fields appear:

FieldDescriptionRequired
privateKeyPEM content including -----BEGIN PRIVATE KEY----- and -----END PRIVATE KEY-----Yes
passphrasePassphrase for the key if encrypted. Leave blank if the key is not encrypted.No

ProgrammaticAccessToken Authentication Fields

When ProgrammaticAccessToken is selected, the following additional field appears:

FieldDescriptionRequired
programmaticAccessTokenProgrammatic access tokenYes

Connect to Snowflake

Before You Begin

Before configuring the connector, ensure that:

  • You have the Snowflake server hostname (serverEndpoint)
  • You have valid Snowflake credentials
  • You have the Snowflake warehouse and database names
  • Your Snowflake account is accessible from the platform environment

Ensure that:

  • The corresponding public key is registered for the Snowflake user
  • Your user has access to the specified warehouse and database
  • If using an encrypted private key, the passphrase is available

Step 1: Select the Connector Type

  1. On the Data Fusion canvas, click + Add Data Source.
  2. In the connector list, locate Snowflake under the Database & Data Warehouse section, or search for snowflake.
  3. Click Snowflake.

The wizard advances to the Destination step.

Note: Connector names must be unique, alphanumeric, and must not begin with a number. If you enter a duplicate or invalid name, the wizard displays a validation error.

Step 2: Configure Destination Settings

On the Destination page, configure the connector identity.

FieldDescriptionRequired
nameUnique name for the connectorYes
descriptionOptional description for the connectorNo

Also configure the following connection endpoint fields on the Destination page:

FieldDescriptionRequired
serverEndpointSnowflake server hostname (for example, account.snowflakecomputing.com)Yes
portPort number for the Snowflake connectionNo
databaseName of the Snowflake database to connect toYes
warehouseName of the Snowflake virtual warehouse to use for query executionYes

Automatically Populate Fields

To automatically populate Authentication and Advanced fields:

  1. Click the Auto-populate panel on the left side of the wizard.
  2. Select JSON or JDBC URL as the input format.
  3. Paste the configuration.
  4. Click Extract & Suggest Mapping.
  5. Review the suggested mappings.
  6. Click Apply Mapping & Fill Fields.

Click Continue.

Step 3: Configure Authentication

On the Authentication page, configure the connection details and authentication method.

Base Connection Fields

The following fields are always displayed:

FieldDescription
usernameSnowflake username
passwordSnowflake password

Select an Authentication Method

From the Authentication Method dropdown, select one of the following:

  • KeyPair
  • ProgrammaticAccessToken

KeyPair Authentication

If KeyPair is selected, complete the following fields:

FieldDescriptionRequired
privateKeyPaste PEM content including -----BEGIN PRIVATE KEY----- and -----END PRIVATE KEY-----Yes
passphrasePassphrase for the key if encrypted. Leave blank if the key is not encrypted.No

ProgrammaticAccessToken Authentication

If ProgrammaticAccessToken is selected, complete the following field:

FieldDescriptionRequired
programmaticAccessTokenProgrammatic access tokenYes

Click Continue.

Step 4: Review and Test the Configuration

On the Review & Test page:

  1. Review the connector configuration summary.
  2. Verify the Destination and Authentication sections.
  3. Review any validation issues displayed by the wizard.

To save and validate the connection:

  • Click Save & Test

To save without validating the connection:

  • Click Save & Close

If Save & Test is used, the wizard validates:

  • Authentication
  • Network connectivity
  • Access to the Snowflake account

A success or failure status is displayed after validation completes.

If the test fails:

  1. Return to the relevant step.
  2. Update the configuration.
  3. Run the test again.

[!NOTE] If validation fails during Save & Test, the connector is still saved and the wizard remains open so you can correct the configuration.

After saving, the connector appears on the Connectors page.

Next Steps

After creating the connector:

  • Create a Source Collection for Snowflake data

  • Configure transformations and mappings

  • Run and monitor ingestion jobs

Was this page helpful?