Advancing Apache Iceberg on Databricks: Iceberg v3 GA, Open Sharing, and Unified Governance

2026-05-28 · Source: Databricks · Field: Technology & Digital — Data Science & Analytics, Cloud Computing & IT Infrastructure · Depth: Intermediate, medium

Summary

Databricks has announced a comprehensive suite of Apache Iceberg capabilities within its Unity Catalog, aiming to establish it as the most interoperable Iceberg catalog available. Key general availability (GA) features include Managed Iceberg for direct table operations with Predictive Optimization and Liquid Clustering, and native support for Iceberg v3, introducing deletion vectors, row tracking, and the VARIANT type. Foreign Iceberg and Credential Vending are also GA, enabling governance and secure querying of externally managed Iceberg tables. Additionally, External Sharing to Iceberg clients is GA via the Delta Sharing protocol, with Public Preview for sharing Foreign Iceberg tables. Other advancements include Beta for Cross-engine Attribute-Based Access Control and new Preview catalog federation connectors for Google Cloud Lakehouse and Palantir, expanding existing support for AWS Glue, Snowflake Horizon, Hive Metastore, and Salesforce Data Cloud. These updates position Unity Catalog to deliver open APIs, federated estates, cross-engine governance, secure sharing, and continuous performance innovation.

Key takeaway

For MLOps Engineers or AI Architects managing diverse data ecosystems, Databricks' Unity Catalog enhancements for Apache Iceberg offer a unified solution. You can now centralize governance, securely share data, and automate performance optimization across various engines and external catalogs, including AWS Glue and Google Cloud Lakehouse. This eliminates data duplication and complex authentication, allowing you to leverage Iceberg v3 features and cross-engine ABAC. Evaluate these GA capabilities to streamline your lakehouse operations and ensure consistent data access for AI and agentic applications.

Key insights

Unity Catalog unifies Apache Iceberg governance, sharing, and performance across diverse engines and external catalogs.

Principles

• Open APIs are crucial for engine flexibility.
• Centralized governance must span multiple catalogs.
• Automated optimization maintains table performance.

Method

Unity Catalog enforces Attribute-Based Access Control by evaluating policies during server-side scan planning, returning filtered scan plans to external Iceberg engines for authorized data access.

In practice

• Utilize Managed Iceberg for direct table operations.
• Register foreign Iceberg tables for unified governance.
• Share live data via Delta Sharing to Iceberg clients.

Topics

• Apache Iceberg
• Unity Catalog
• Lakehouse Architecture
• Data Governance
• Catalog Federation
• Delta Sharing

Best for: CTO, VP of Engineering/Data, Director of AI/ML, Data Engineer, MLOps Engineer, AI Architect

Related on AIssential

• The Next Era of the Open Lakehouse: Apache Iceberg™ v3 in Public Preview on Databricks — Iceberg v3 enhances open lakehouse capabilities with features for incremental processing and semi-structured data.
• Iceberg is harder than you think — Iceberg implementation is complex, requiring careful data transport, catalog integration, and engine optimization.
• Google Cloud Introduces Cross-Engine Iceberg Support in BigQuery — Google Cloud enhances BigQuery with cross-engine Apache Iceberg support, simplifying lakehouse operations and enabling multi-cloud, multi-tool data access.
• Backstage with Lakebase, part 2 — Unified governance for operational databases simplifies security, compliance, and cost attribution while enabling rapid development workflows.
• Why Databricks is Winning the Data Wars in 2026 — A Technical Breakdown — Databricks' Lakehouse architecture unifies data warehousing and data lake functionalities on open standards, driven by governance and AI needs.
• Designing a Modern Enterprise Lakehouse with Databricks, DABs, and Medallion Architecture — Platform scalability in enterprise lakehouses relies on reusable engineering patterns and consistent deployment strategies, not just individual technologies.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Databricks.