Unity Catalog 0.7.0 enabled the Iceberg REST catalog by default; Databricks transactions reached GA

Catalog interoperability was the main thread of October. Unity Catalog 0.7.0 shipped on 13 October with native Apache Iceberg REST catalog support enabled by default, so external query engines could read, write and manage Iceberg tables. The release also made dropped managed tables, volumes and models purge their cloud storage files after a default seven-day delay [9]. Databricks said Unity Catalog was the first catalog to support the Iceberg REST catalog spec for REGISTER and UNREGISTER, which lets tables move between catalogs without duplicating data [5].

On 16 October, multi-statement transactions became generally available for Delta Lake and Iceberg tables on Databricks, backed by row-level concurrency and Unity Catalog governance [8]. A Workday Data Connect connector entered Beta and let teams query Workday's shared Iceberg tables through zero-copy federation [10]. At the specification level, the Iceberg project introduced read restrictions and catalog labels. They aim to standardize policy enforcement across engines and make governance metadata portable across federated catalogs [12].

The dbt ecosystem focused on Iceberg setup. A catalogs.yml v2 spec separated dbt's Iceberg configuration from Snowflake-specific catalog integrations. A dbt Mesh demo showed three teams on Snowflake and Databricks sharing one set of Iceberg tables through catalog federation, with no data copied [2]. Another post argued that Iceberg on Snowflake had become simple enough for a small team to run in an afternoon [1]. Obie Insurance reported cutting a month-end close process from over 25 hours to 2 hours after rebuilding on dbt and Iceberg in a lake house architecture [3].

Community writing turned to operations, with pieces on the cost of table maintenance [4] and on running Iceberg in production for real-time workloads [6].