The Convergence of Open Table Formats and Open Catalogs: Catalog Commits is Generally Available

Catalog Commits unifies Delta Lake and Iceberg by establishing catalogs as the coordination layer for modern lakehouse architectures.

• •Eliminates the "split-brain" problem where external engines writing directly to storage cause catalog metadata to diverge from actual table state
• •Enables consistent governance and data discovery across multiple engines and AI agents through standardized catalog APIs
• •Unlocks multi-table ACID transactions on the lakehouse, allowing organizations to consolidate traditional data warehousing workloads
• •Supported by major engines including Spark, Flink, Trino, DuckDB, and StreamNative, with easy integration via Delta Kernel
• •Generally available on Databricks Runtime 16.4+ with features including external engine writes, enhanced governance, and multi-statement transactions