Databricks

The Databricks connector provides read-only access to data and metadata from the Databricks data analytics platform.

A connection admin can set up a connection to Databricks in the Zero Copy Connector Hub and grant data stewards access to this connection. Data stewards can then use the established connection to create a data fabric table and map data from Databricks. This allows users to access Databricks data through the table list view or by using GlideRecord scripts. For details on creating data fabric tables and mapping data, see Managing data fabric tables.

When you configure the connection with an access token, you can choose whether the connection authenticates using a shared, system-level OAuth entity profile or personal authentication.

Note: With personal authentication, each user’s access to Databricks data is authenticated and audited against their own credentials, not a shared account. This supports stricter access control and clearer audit trails.

The connector has been enhanced to improve the performance of the following Glide queries and list view operations. These improvements allow the majority of queries to be executed at the data source.

• Sort

• Limit

• Filter

• GroupBy

• avg()

• count()

• max()

• min()

• sum()

• References

Supported data types

The following table lists supported Databricks data types and the default matching data types in a data fabric table.

Important: Databricks data types not included in the table aren’t supported for data mapping in Zero Copy Connector Hub.

Databricks Data fabric table bigint Long Integer String int Long Integer String smallint Long Integer String tinyint Long Integer String timestamp_ntz Basic Date/Time char Char timestamp Basic Date/Time date Date decimal Decimal string String boolean True/False double Floating Point Number float Floating Point Number

Related topics

Create a Databricks connection