Cloudflare Basin, its Iceberg-based analytics platform on R2, is now GA
Serguey Asael Shinder2 min readCloudflare has made its data platform, announced a year ago, generally available and renamed it Cloudflare Basin. It is a serverless analytics platform built on Apache Iceberg, the open table format for data lakes, with data stored in R2 object storage.
Basin has three parts, each renamed from an earlier product. Basin Pipelines, formerly Cloudflare Pipelines, receives events from Workers, HTTP or Cloudflare Logpush, transforms them with SQL and writes them as Iceberg tables or files in R2. Basin Catalog, formerly R2 Data Catalog, manages Iceberg metadata and maintains tables automatically, compacting metadata and data files and generating statistics for query planning. Basin SQL, formerly R2 SQL, is a serverless distributed SQL engine that uses those statistics to split queries into smaller tasks spread across Workers.
Cloudflare's pitch rests on openness and egress. Because the data is stored as Iceberg, it can be read and written with any compatible engine; the post names PyIceberg, DuckDB, Snowflake and Apache Spark. Cloudflare argues that this portability only works in practice because R2 does not charge for egress, so the same data can be used from other tools, regions and clouds.
The company says its own billing and infrastructure teams adopted the platform during the beta, for billing metrics and infrastructure telemetry, and quotes a customer that replaced an S3 and Athena setup with it. It also says it worked over the past year on speed, openness and cost, and that a catalog, a pipeline and a first query can be set up in seconds.
Why it matters
Iceberg on cheap object storage has become the default design for analytics that should outlive any one query engine. A platform that keeps the data in an open format and does not charge to read it out lowers the cost of leaving, which is the usual hidden price of a managed analytics service.