The open-source MinIO project is no longer maintained. Your backups, artefacts and application data may still sit on it. We pick the replacement that fits your size, copy the data with checks, and switch your applications to the new endpoint.
From the MinIO repository on GitHub:1
minio/minio repository was archived and is read-only. The README says it is “no longer maintained”.All three speak the S3 API. They differ in how they grow and how much running they need.
| Option | Usually fits | Trade-off |
|---|---|---|
| SeaweedFS | Many objects, small to large clusters, teams that want simple operations | Fewer enterprise features than Ceph |
| Garage | Small, lightweight clusters, also across several sites | Implements a subset of the S3 API — we check yours against it |
| Ceph RGW | Large scale, or you already run Ceph for block or file storage | More moving parts; needs real Ceph skills |
| Stay on MinIO / AIStor | You prefer the vendor and its terms work for you | Licence and price set by one company |
Which S3 features your applications really use — versioning, object lock, lifecycle rules, presigned URLs — tested against the chosen option before any data moves.
Installed as code, with monitoring, alerts and capacity planning.
Buckets copied in the background, object counts and checksums compared, a final short sync at cut-over.
Access keys, bucket policies and service accounts re-created; old keys revoked at the end.
Applications, GitLab, Harbor, backup tools pointed to the new endpoint and tested.
Adding capacity, replacing a disk or node, restoring a bucket, upgrading.
The order is always the same. How long each step takes depends on your volume — the measurement in step 1 sets the price and the dates. Our free scope script gives the first numbers.
Buckets, sizes, clients and S3 features in use. Option chosen together; final fixed price.
New cluster built; your applications tested against it with real operations.
Background copy with verification. Clients switched in batches, with a short final sync.
MinIO set to read-only, last clients found and fixed, then switched off.
Fixed price after the measurement in step 1. Paid by milestone.
Yes, and during the migration we do exactly that. Just don't treat it as a long-term home: without maintenance, new vulnerabilities stay open.
Usually only a short write freeze per client at the final sync. Reads can continue.
Most only need a new endpoint and keys. If an app uses an S3 feature the target does not support, we find it in week one, before any data moves.
It depends on size and team. We give a clear recommendation after the inventory, with the reasons.
Rough size, number of buckets and what writes to them — enough for a first estimate.