Glossary Entry

MinIO and Apache Iceberg: S3-Compatible Storage After the Community Edition

MinIO is an S3-compatible object storage server often used for local Iceberg lakehouses. Here is its 2026 status and what to run for dev and production.

Written by Alex Merced

What is MinIO?

MinIO is an object storage server that implements the Amazon S3 API. Because it speaks S3, any engine or library that reads and writes S3 (Apache Spark, Trino, Dremio, Flink, PyIceberg, DuckDB) can point at a MinIO endpoint instead of AWS and work unchanged. That made MinIO the default storage container in local and on-premises Apache Iceberg stacks for years: a Docker Compose file with MinIO, a REST catalog, and an engine gave you a working data lakehouse on a laptop.

MinIO stores objects, not tables. It has no idea what an Iceberg snapshot or manifest is. The table format and the catalog provide the table semantics; MinIO provides the bucket that the data files and metadata files live in. See object storage for how that layer fits the rest of the stack.

Project Status in 2026

The open-source MinIO server changed a lot between 2025 and 2026, and anyone reading older tutorials should know the timeline:

  • May 2025: most administration features were removed from the community edition’s web console and moved to the commercial product.
  • October 2025: MinIO stopped publishing community Docker images and prebuilt binaries.
  • December 2025: the community repository entered maintenance mode, with no new features or accepted pull requests.
  • 2026: the minio/minio GitHub repository was marked “no longer maintained” and archived (read-only). The source remains under AGPLv3.

MinIO Inc. now develops AIStor, offered as a free standalone edition and a commercially supported enterprise edition. Existing community deployments keep running, but they no longer receive releases or security fixes from upstream.

MinIO in a Lakehouse Stack

UseCommunity MinIO (archived)AIStorCloud object storage (S3, ADLS, GCS)
Local Iceberg development and CIStill works with existing images; no updatesFree edition is an optionNot local
On-premises production lakehouseNot advisable without a support contractCommercially supportedNot applicable
Cloud production lakehouseNot typicalPossibleThe usual choice
Credential vending from a REST catalogWorks through S3-compatible STS settings, with configurationSameNative in Polaris, Glue, Unity Catalog

For local work, the role MinIO plays is narrow: it exercises the S3 code path (path-style access, endpoints, multipart uploads, credentials) that a plain local filesystem skips. Other S3-compatible servers, such as Ceph Object Gateway, SeaweedFS, and Garage, can fill the same role. Whatever you choose, test that the engine settings that matter for Iceberg (s3.endpoint, s3.path-style-access, region) behave the same way they will in production.

A Concrete Example

A typical local stack runs four containers: an S3-compatible store with a warehouse bucket, an Iceberg REST catalog (the apache/iceberg-rest-fixture image or Apache Polaris) configured with the store’s endpoint, Spark or Dremio as the engine, and a one-shot client container that creates the bucket on startup. The engine creates a table through the catalog, the catalog records the metadata location, and the engine writes Parquet data files and Iceberg metadata files into the bucket. Swapping the store for AWS S3 later changes only the endpoint and credentials, not the table.

For a full walkthrough, see Local Iceberg Development Environments: Docker, MinIO, and In-Memory Catalogs for CI and Block vs. Object Storage.

Frequently Asked Questions

Is MinIO still open source?

The archived community source code is still licensed under AGPLv3, but the project is no longer maintained upstream. MinIO’s active product is AIStor.

Do I need MinIO to run Apache Iceberg?

No. Iceberg needs a file system or object store the engine can reach. Locally that can be the filesystem; in the cloud it is usually S3, ADLS, or GCS. MinIO is one S3-compatible option.

Does MinIO understand Iceberg tables?

No. MinIO stores the files. The Iceberg catalog and the engine handle snapshots, schema, and transactions.

Learn More

To learn more about the Data Lakehouse, read The Lakehouse Built for Everyone by Alex Merced. You can find this and other books by Alex Merced at books.alexmerced.com.

Work with Alex