What inspired this wish list request?
I was told that JMP is working on such a connector, and want to advocate for as accelerated development as possible.
Our organization uses Databricks as our primary data platform, and connecting JMP to Databricks currently requires configuring the Simba Spark ODBC driver manually for each user. This process involves downloading and installing the driver, locating workspace-specific server hostname and HTTP path values, configuring DSNs through the Windows ODBC Data Source Administrator, and managing personal access token authentication — all before a user can run their first query. For analysts and scientists who aren't IT-savvy, this is a significant barrier. Any change to the Databricks workspace (cluster changes, token expiration, workspace migration) requires users to repeat or troubleshoot the process. The overhead of supporting this setup across a team is substantial and ongoing.
What is the improvement you would like to see?
A native Databricks connector in JMP that allows users to authenticate directly using their existing Databricks credentials (personal access token or, ideally, Entra). The connector should:
- Present a simple connection dialog where the user enters their Databricks workspace URL and authenticates
- Automatically discover available catalogs, schemas, and tables based on the user's Unity Catalog permissions
- Respect existing Unity Catalog row/column-level security and access controls without any additional configuration
- Store connection profiles so users can reconnect with one click
- Support both interactive SQL queries and table import workflows within JMP
Why is this idea important?
Will increase efficiencies and reduce complexity for end users. Reduce IT support to help users connect, manage access tokens, and document lengthy checklists to guide users. From a data governance perspective, honoring the Unity Catalog permissions would ensure that access controls are respected and still able to be centrally managed.