Connect to cluster from PowerBI

Firstly a connection to a databricks cluster is different to a connection to a databricks sql endpoint.

Here we use databricks cluster connection.

There are two ways to get the connection details:

(1) Go the 'Partner Connect' and choose 'Power BI' to export a data source file (.pbids). This is plain text so open it with notepad to see the 'host' and 'path' for the cluster

(2) Go the 'Compute' tag and click into the computing node. Expand the 'Advanced options' and go to the JDBC/ODBC part. It has connection details too.

Note the 'host' is the same for cluster connection and sql endpoint connection, but the path (API url) is different.

personal access token

To generate a personal access token, go to the 'settings' -> "user settings", the "Access Tokens". Note the location might have changed in the latest databricks.

If using Power BI desktop, go to 'Get data' and search for 'Azure databricks'. On connection, it will ask for the host the path, and then you can use access token / login using active directory.

If connected successfully, a navigator will show up the data on the cluster, and you can go from there for Power BI data transformation, and visualization, etc.

This is not showing the DBFS files, but the database and database tables on databricks. Pretty handy.