Managed Apache Iceberg at Scale: How Spanner Powers Lakehouse Runtime Catalog
Google, Tuesday, October 6th, 2026
Google explains how its serverless Lakehouse runtime catalog uses Spanner to provide a scalable managed Apache Iceberg REST catalog.
Google Cloud examines the challenges a managed Apache Iceberg catalog faces as AI agents drive lakehouse queries to new scale.
Because the catalog maintains table pointers, handles atomic commits and is the source of truth for table locations, large enterprises need atomicity, consistency, availability and concurrency at scale.
Google's serverless Lakehouse runtime catalog is built on Spanner, implements the open Iceberg REST catalog specification and is designed to handle large volumes of repeated small queries from agents.