Azure Databricks' approach to query federation: foreign catalogs and Unity Catalog connections expose outside databases and catalogs for governed, read-only queries, with work pushed down to the source when possible.
Read more: Microsoft Learn
In the Ultra Transcenders books
Each book explains Lakehouse Federation in context, with comparison tables and the common traps.
Terms in this definition
- Databricks
Analytics platform built on Apache Spark where data is processed in notebooks; Unity Catalog is the governance model it recommends.
- Federation
An arrangement in which separate identity providers trust one another, letting people use the account from their home organisation to reach services in another domain. In a hybrid setup, Entra ID can pass sign-ins over to a federation service like AD FS.
- Unity Catalog
Azure Databricks' governance solution covering both data and AI in one place, with centralised permissions, auditing, data discovery and lineage.
Related terms
- Foreign catalog
Mirrors a database living in another system so it can be queried read-only from Unity Catalog over a Lakehouse Federation connection. Making one takes
CREATE CATALOG, and on the connection you need ownership orCREATE FOREIGN CATALOG. - JDBC
Java Database Connectivity, a widely used standard API through which applications and tools reach databases. SQL warehouses and clusters accept connections via the Databricks JDBC Driver, and Lakehouse Federation relies on JDBC when pushing queries down.
- Unity Catalog connection
Securable, made with
CREATE CONNECTION, holding credentials and host details that Lakehouse Federation uses to query an outside database; foreign catalogs are built from it.