MinIO

What is it?

MinIO is an open-source object storage server that speaks the Amazon S3 API byte-for-byte. It runs as a single binary on Linux, macOS, and Windows, scales horizontally to exabytes across a distributed cluster, and deploys cleanly to Kubernetes via an operator or Helm chart.

What does it do?

MinIO provides S3-compatible buckets, objects, versioning, lifecycle rules, server-side encryption (SSE-S3, SSE-KMS, SSE-C), identity and access management, bucket notifications, and site-to-site replication. Any tool built for S3 — Spark, Presto, TensorFlow, Backup software, Terraform — works without modification.

Where is it used?

MinIO is deployed inside banks, telcos, and hyperscalers who need S3-style object storage behind their own firewalls. It powers AI/ML training pipelines, CDN origin stores, observability data lakes, Kubernetes-native persistent storage, and air-gapped government clouds where AWS is not an option.

When & why it emerged

MinIO was founded in 2014 to democratise object storage. S3 had won the cloud-storage API war, but only in AWS's own cloud. Enterprises wanted the same programming model on-prem for regulatory, performance, or cost reasons — MinIO gave them a pure-software, open-source implementation with commercial support.

Why we use it at Internative

We ship MinIO for clients who need S3-compatible storage without the 'everything must live in AWS' tax. It powers the object layer of our Koordex platform, backs several client data lakes, and sits behind CDN setups we run for clients with strict data-residency requirements. Our Kubernetes stacks default to the MinIO operator when object storage is on the bill of materials.