What is Databricks Partner Connect?

Partner Connect lets you create trial accounts with select Databricks technology partners and connect your Databricks workspace to partner solutions from the Databricks UI. This allows you to try partner solutions using your data in the Databricks lakehouse, then adopt the solutions that best meet your business needs.

Partner Connect provides a simpler alternative to manual partner connections by provisioning the required Databricks resources on your behalf, then passing resource details to the partner. Required resources might include a Databricks SQL warehouse (formerly Databricks SQL endpoint), a service principal, and a personal access token.

Not all Databricks partner solutions are featured in Partner Connect. For a list of partners that are featured in Partner Connect, with links to their connection guides, see Databricks Partner Connect partners.

Tip

If you have an existing partner account, Databricks recommends that you follow the steps to connect to the partner solution manually in the appropriate connection guide. This is because the connection experience in Partner Connect is optimized for new partner accounts.

Note

Some partner solutions allow you to connect using Databricks SQL warehouses or Databricks clusters, but not both. For details, see the partner’s connection guide.

Requirements

To connect your Databricks workspace to a partner solution using Partner Connect, you must meet the following requirements:

Your Databricks account must be on the Premium plan. This is because Partner Connect uses Databricks service principals, which are available only on this plan. To view your Databricks account details, see Manage your Databricks account.
To create new connections to partner solutions, you must first sign in to your workspace as a Databricks workspace admin. For information about Databricks workspace admins, see Manage users.
For all other Partner Connect tasks, you must first sign in to your workspace as a Databricks workspace admin or a Databricks user who has at least the Workspace access entitlement. For more information, see Manage users.

Quickstart: Connect to a partner solution using Partner Connect

Make sure your Databricks account, workspace, and the signed-in user all meet the requirements for Partner Connect.
In the sidebar, click Partner Connect.
Click the tile for the partner that you want to connect your workspace to. If the tile has a check mark icon, stop here, as your workspace is already connected. Otherwise, follow the on-screen directions to finish creating the connection.

Note

Partner solutions that use a locally-installed application instead of a web-based one (such as Power BI Desktop and Tableau Desktop) do not display a check mark icon, even after you connect your workspace to them.
To work with your new connection, see the concluding or Next steps section of the corresponding Setup guide in the list of Databricks Partner Connect partners.

Common tasks required to create and manage partner connections

This section describes common tasks you might need to complete to create and manage partner connections using Partner Connect.

Allow users to access partner-generated databases and tables

Partner solutions in the Data Ingestion category in Partner Connect can create databases and tables in your workspace. These databases and tables are owned by the partner solution’s associated Databricks service principal in your workspace.

By default, these databases and tables can be accessed only by the service principal and by workspace admins. To allow other users in your workspace to access these databases and tables, use the SQL GRANT statement. To get access details for an existing database or table, use the SQL SHOW GRANTS statement.

Create an access token

Databricks partner solutions require you to provide the partner with a Databricks personal access token. The partner uses this token to authenticate with your Databricks workspace.

For cloud-based partner solutions in Partner Connect (such as Fivetran, Labelbox, Prophecy, and Rivery), Partner Connect automatically creates the token (along with a Databricks service principal that is associated with that token) and then shares the token’s value with the partner. You cannot access the token’s value. If for any reason the token expires or the token’s value is no longer shared with the partner, you must create a replacement token for the service principal; to do this, see Manage service principals and personal access tokens. To share the replacement token with the partner, see the partner’s documentation.

Only Databricks workspace administrators can generate replacement tokens for Databricks service principals. If you cannot generate a replacement token, contact your administrator. See also Manage service principals.

For desktop-based partner solutions in Partner Connect (such as Power BI and Tableau), you must create the token and then share the token’s value with the partner. To create the token, see the Token management API. To set up the partner solution so that it shares the new token with the partner, follow the on-screen instructions in Partner Connect or see the partner’s documentation.

Important

When you create the token and share the token’s value with the partner, the partner can take whatever actions that the related entity (such as your Databricks user or a Databricks service principal) can normally take within your Databricks workspace. Do not share token values with partners whom you do not trust.

Note

As a security best practice, when you authenticate with automated tools, systems, scripts, and apps, Databricks recommends that you use personal access tokens belonging to service principals instead of workspace users. To create tokens for service principals, see Manage tokens for a service principal.

Databricks workspace administrators can disable token generation. If you cannot generate a token, contact your administrator. See also Monitor and revoke personal access tokens.

Allow a SQL warehouse to access external data

To allow a Databricks SQL warehouse to access data outside of Databricks, see Enable data access configuration.