Yufei Gu

Apache Polaris is designed as a next generation open metadata and governance layer for modern lakehouses. While Polaris started with strong roots in Apache Iceberg, real world data platforms rarely live in a single table format.

Iceberg’s UDF spec introduces a self-contained, versioned metadata format for SQL scalar and table functions that can move across catalogs and engines like Spark and Trino. This talk walks through the core model, definitions, parameter naming rules, return types, and dialect-specific representations, and explains how versioning, rollback, determinism, and null-handling hints make UDFs portable wit

Apache Polaris (Incubating) implements the Apache Iceberg REST catalog specification and allows users to manage Iceberg tables, views, and other related metadata. Role-based access control governs which users have access to what data, and Polaris supports attaching policies to tables in order to categorize data and to configure table maintenance.

As data lakes and lakehouses grow in importance, seamless access and strong governance are critical. Within the Apache Iceberg™ ecosystem, governance solutions are often fragmented and complex.

This talk explores the evolving landscape of Iceberg REST specifications, delving into key advancements such as view APIs, server capabilities, and server-side planning. We will also examine Apache Polaris (Incubating), a leading reference implementation, highlighting its key features and the vibrant community supporting it.