What problem does it solve?
Databricks provides multiple pathways to work with Apache Iceberg across managed tables, Delta interoperability, and external access, creating fragmented patterns and integration friction. This Skill consolidates guidance, best practices, and interoperability patterns to simplify deploying and operating Iceberg across Unity Catalog, external readers/writers, and cross-cloud integrations.
Core Features & Use Cases
- Native Managed Iceberg: create and manage Iceberg tables inside Databricks with full read/write capability and external engine compatibility.
- UniForm (Delta-to-Iceberg): make Delta tables readable as Iceberg without data migration; metadata generation happens asynchronously to expose Iceberg views externally.
- Compatibility Mode: enable streaming tables and materialized views to be accessible as Iceberg via UniForm-compatible metadata.
- Iceberg REST Catalog (IRC): external engines connect to UC-managed Iceberg data through a standard REST endpoint for reads and selective writes.
- Iceberg v3 and external tooling: supports newer v3 features (deletion vectors, VARIANT, row lineage) for advanced analytics scenarios.
- Snowflake interoperability: catalog integrations and foreign catalogs to enable cross-platform data sharing with Snowflake.
- PyIceberg and OSS Spark interoperability: programmatic access to Iceberg data from external tools and libraries.
Quick Start
Create a managed Iceberg table in Unity Catalog and enable UniForm/IRC to expose it to external engines.
