Apache Polaris is now a Top-Level Project at the Apache Software Foundation. For anyone building on Apache Iceberg, this is one of the most important catalog milestones since the REST Catalog spec itself.
Graduation means Polaris has cleared Apache's bar for contributor diversity, governance maturity, and long-term sustainability. It's no longer incubating. It's production-grade open source with a community that can sustain itself independent of any single vendor. That distinction matters more than most people realize.
Dremio and Snowflake originally co-created Polaris and donated the code to the Apache Software Foundation in August 2024. The bet was straightforward: the Apache Iceberg catalog layer should be open and vendor-neutral, governed by the community rather than tied to any one company's product roadmap.
What happened next was fast by Apache standards. In roughly 18 months of incubation, the project shipped six releases (0.9.0 through 1.3.0), closed over 2,800 pull requests, and attracted around 100 contributors. The GitHub repo sits at 1.8k stars and 357 forks, healthy traction for infrastructure software.
More telling than the code volume is who showed up. The project's PMC now includes engineers from Dremio, Snowflake, Google, Microsoft, Confluent, and LanceDB. Committers represent Bloomberg, Starburst, and several independents. Of the 13 PMC members, six were elected during incubation. That means the community grew its own leadership rather than inheriting it from the original donors. Five of the eight committers were also elected during incubation.
The graduation vote on the Apache Incubator mailing list passed with 27 +1 votes and zero objections. JB Onofré, who served as the project's champion and mentor, shepherded the process. Ryan Blue, Holden Karau, Kent Yao, and François Papon served as mentors alongside him.
Explore this interactive demo and see how Dremio's Intelligent Lakehouse enables Agentic AI
Apache Polaris implements the Iceberg REST Catalog spec. You define your Iceberg tables, namespaces, and credentials in one place. Every engine that connects (Dremio, Spark, Flink, Trino) sees the same metadata.
Without a shared catalog, metadata scatters across Hive Metastores, AWS Glue configurations, and engine-specific stores. Each copy drifts independently. Permissions fall out of sync. Schema changes in one engine don't appear in another. These aren't edge cases; they're the default state for most multi-engine Iceberg deployments today.
Polaris solves this with a deliberately narrow scope. It manages Iceberg metadata. Engines handle compute. Data stays in your object storage. There's no lock-in because there's nothing proprietary in the stack.
Open source projects live or die by their communities, not their code. Top-Level status at Apache is the clearest signal the community is healthy.
The numbers back that up. Contributors from at least eight organizations across the data industry. A PMC that elected nearly half its members during incubation. Six production releases in under two years. This isn't a project that depends on one or two companies to keep the lights on.
For data architects, this is the green light to standardize on Polaris as the catalog layer across your Iceberg stack. For executives evaluating options, it removes the biggest adoption risk: project continuity. Your investment in Polaris won't be stranded if one company shifts priorities. The community governs the roadmap now.
If you've been waiting for the right moment to commit to an open catalog standard, this is it.
Dremio co-created Polaris because we believe the catalog should be open. We built Open Catalog on that conviction: Polaris at the core, with the operational and governance layers enterprises need to run it in production.
That means fine-grained access controls, automated Iceberg table maintenance (compaction, cleanup, repartitioning), lineage and governance across your full data estate, and a managed service so your team isn't running catalog infrastructure themselves.
If you're already using Polaris or evaluating it, Open Catalog is the fastest path from open standard to production.
Deploy agentic analytics directly on Apache Iceberg data with no pipelines and no added overhead.
With its capabilities in on-prem to cloud migration, data warehouse offload, data virtualization, upgrading data lakes and lakehouses, and building customer-facing analytics applications, Dremio provides the tools and functionalities to streamline operations and unlock the full potential of data assets.
Dremio Arctic bring new features to Dremio Cloud, including Apache Iceberg table optimization and Data as Code.
Unveiling Dremio Cloud on Azure, a Fast, Scalable and Secure Lakehouse Platform Introduction Today we introduce Dremio Cloud on Azure, newly landing on Azure in November 2023, in public preview. Dremio Cloud is a powerful lakehouse platform providing streamlined self-service, unparalleled SQL performance, centralized data governance and seamless lakehouse management. In this blog post, we […]
