MinIO repository is no longer maintained

Original Article ↗ Hacker News Discussion ↗

Status change and AIStor pivot

  • Commenters note the shift from “maintenance mode” to “THIS REPOSITORY IS NO LONGER MAINTAINED” as confirmation of a full pivot to the closed-source AIStor product.
  • Several see AIStor Free as essentially an upsell funnel and rebrand of MinIO, with skepticism about “free” claims and “fool me once” sentiment.
  • Some argue that, given MinIO’s commercial goals, the wording nuance doesn’t matter—long-term users should assume the open-source line is dead.

Alternatives and their tradeoffs

  • Frequently mentioned replacements: Garage, RustFS, SeaweedFS, Ceph, LocalStack, S3Proxy, rclone serve s3, S3 Ninja, versitygw, filestash, plus hosted services like Wasabi.
  • Garage is praised for simple distributed setups, good docs, and self-hosting philosophy, but lacks some S3 features (ACLs, CORS) and fine-grained access controls.
  • SeaweedFS is liked for simplicity (weed server -s3, weed mini) and speed; some warn it’s still “personal-project-like” with questionable code structure and potential data-loss risks.
  • RustFS is seen as the closest MinIO-style rewrite and very easy to run; others flag its CLA and licensing scaffolding as a likely future rug-pull.

Local development and “just a filesystem” S3

  • Many only used MinIO as a local S3 emulator and now want a minimal drop-in Docker service.
  • Suggestions: Garage single-node, SeaweedFS weed mini, S3 Ninja, rclone serve s3, versitygw, simple home-grown servers, and LocalStack (though it too is moving toward a more restricted free tier).
  • Several want a trivial “S3 over local filesystem” implementation, but note incompatibilities between S3 object names and POSIX filenames and the complexity of full S3 semantics.

Ceph and heavy-duty object storage

  • Ceph is repeatedly recommended for serious, large-scale or high-integrity use; users report multi‑PB clusters and strong resilience.
  • Downsides: steep learning curve, operational complexity, hardware/network demands; seen as inappropriate for “just toss it into a customer environment” use cases.
  • Some argue Ceph is overkill if you don’t need distributed block storage; others see it as the only truly battle‑tested open alternative at MinIO’s scale.

Licensing, CLAs, and rug-pull risk

  • Strong suspicion toward CLAs with copyright assignment (e.g., RustFS), viewed as enabling relicensing to closed source.
  • Debate over whether CLAs are mostly for legal hygiene or explicitly to allow future commercial relicensing.
  • AGPL is seen by some as a deterrent to hyperscalers, but others note it didn’t prevent MinIO’s pivot.

Business models, ethics, and “social contract”

  • Repeated pattern noted: VC-backed, single-vendor OSS (Elastic, Redis, Terraform, now MinIO) going source-available or closed once adoption is high.
  • One camp: maintainers owe nothing; licenses disclaim any promise of continued work; users relying on free infra without contingency are naive.
  • Other camp: while legally allowed, using “open source” to build trust and adoption and then cutting off the open version is described as a bait‑and‑switch that erodes community trust.
  • Advice offered:
    • Evaluate funding and governance (foundations like CNCF/Linux Foundation vs single vendor).
    • Be wary of CLAs and single‑company control.
    • Always assume you may need to migrate; keep a plan B.

Migrations and operational lessons

  • Several report already moving from MinIO to Garage, Ceph, SeaweedFS, or RustFS, generally with manageable friction but nontrivial data migration.
  • Some share active large‑scale migrations (hundreds of TB) and stress the need for monitoring and infra expertise, especially with Ceph.
  • Observers note how many are now scrambling to test unfamiliar alternatives, highlighting that few had pre-planned migration runbooks despite critical dependency on MinIO.