Open Source is Not Enough: The Post-MinIO Era

The End of the “Default” Choice

For years, MinIO has been the reflex choice for S3-compatible object storage in Kubernetes setups, SaaS platforms, and data lakes. It was the standard we all reached for.

However, recently (December 3, 2025), the landscape shifted. MinIO officially moved into maintenance mode. This means no new features and, critically, community pull requests are no longer being merged. The development focus has shifted entirely to their commercial offerings.

For me, this serves as a stark reminder: Open Source alone isn’t enough — governance matters.

The Governance Gap

We often conflate “source code availability” with “safety.” But as we are seeing now, if a single vendor controls the keys, the community is always at risk of a pivot.

As Alexey Minin recently put it:

“In 2025, ‘Open Source’ isn’t enough. We need Open Governance. Projects hosted by foundations (CNCF, Apache, Linux Foundation) are immune to these rug-pulls. Single-vendor projects are not.”

If your infrastructure relies on single-vendor open source, you aren’t just betting on the technology; you are betting on their business model remaining aligned with your needs forever.

Reassessing the Landscape

If you are reassessing your options for 2026, the focus should be on projects with sustainable governance models or clear licenses. Here are three open-source S3 alternatives worth exploring:

• Garage (AGPLv3, Rust) Small, robust, and designed for distributed storage. It is excellent for self-hosting and multi-site setups where lightweight resilience is key.

• RustFS (Apache 2.0, Rust) A performance-focused contender built for modern data lakes and AI workloads. If raw throughput is your bottleneck, this is the one to watch.

• SeaweedFS (Apache 2.0, Go) A highly versatile veteran. It handles distributed blobs, objects, and massive quantities of small files efficiently.

Moving Forward

The lesson here isn’t just about picking a new storage tool; it’s about looking at the foundation underneath it. When architecting for the long term, we need to ask: who really owns this project?

The shift away from MinIO is inconvenient, but it is an opportunity to move toward true open governance.¹