🔌Setting up a Snowflake Connector

Configuring the Snowflake connector in Toucan

As a security best practice, we recommend you create a specific user and role to connect to Toucan; check out our guide on granting access if you need to do this.

Fill out the form with the following info:

  • name: for example, my_snowflake_connector

  • account: in the form account_id.region_id.cloud_platform. When you connect to your snowflake UI, it is the first part of the url <account_identifier>.snowflakecomputing.com. You can read more about it here

  • authentication method: between Snowflake (ID+Password) (deprecated by Snowflake in November 2025) "Key pair"

    • Snowflake (ID+Password)

      • User (string) mandatory: the user with rights to access to the Snowflake database

      • Password(string) mandatory: password associated to the user

    • Key pair

      • User (string) mandatory: the user with rights to access to the Snowflake database

      • Private Key(string) mandatory: the key generated to access to your Snowflake database the key in PEM-encoded PKCS#8 format

      • Password (string) optional: if your key is protected by a password you can add it to the password field

  • Token endpoint (string):

  • Token endpoint content type: ('application/json')

  • Role (string) optional: fill with the role associated to your user

  • Default Warehouse (string) optional: fill with the Warehouse you want to access

  • Retry policy (boolean) optional:

Snowflake Connector Form
Snowflake Connector Form

  • Click save. You should see the following check succeeding:

Snowflake Connection Test
Snowflake connection test
  • Once the check has succeeded, you can close the popup window and save the form.

Selecting data from Snowflake

To create a dataset from Snowflake, click on the "create from" icon, you will then be able to:

  • Select the Database

  • Select the Data warehouse

  • Select the Schema

  • Select Table or Views

  • Only keep the columns you need

For more info, see the dedicated section Creating datasets

Last updated

Was this helpful?