Unity Catalog managed tables let teams automate table management while keeping data in cloud storage they own. You control where new managed tables and volumes are written, while Unity Catalog manages governance, lifecycle, and optimization across Delta Lake and Apache Iceberg.
Key Highlights
- Keep data in storage you control: Managed table data can live in your own S3 bucket, ADLS container, or GCS bucket, at a managed storage location defined for your metastore, catalog, or schema.
- Set storage placement by domain: Configure a managed location at the metastore, catalog, or schema level. The most specific location takes precedence, making it possible to separate data for compliance, residency, administration, or cost attribution.
- Change where new data lands: Use ALTER CATALOG or ALTER SCHEMA ... SET MANAGED LOCATION when requirements change. Existing tables remain in their current locations; new tables and volumes use the updated path.
- Convert external tables deliberately: Before converting an external table to managed, set the target managed location. During conversion, Databricks copies the table data and transaction log into that location.
- Stay open and interoperable: Managed tables remain accessible through open APIs, including the Unity REST API and Iceberg REST Catalog, with credential vending and governance controls for supported external engines.
- Automate optimization and maintenance: Unity Catalog managed tables support capabilities such as predictive optimization, metadata management, and automatic lifecycle operations to help reduce manual administration.
๐ Read the full post here