Introduction

Alarik is a high-performance, S3-compatible object storage system written in Swift.

Alarik is a high-performance, S3-compatible object storage system written in Swift, licensed under the Apache 2.0 License.

What is Alarik?

Alarik is an object storage server that speaks the S3 protocol. If you've used Amazon S3 or self-hosted alternatives like MinIO, you already know how to use Alarik. Any tool, SDK, or application that works with S3 will work with Alarik out of the box.

Object storage is ideal for storing unstructured data like images, videos, backups, logs, and static assets. Unlike traditional file systems, object storage scales horizontally and provides simple HTTP-based access to your data.

Why Alarik?

Truly Open Source

Alarik is licensed under Apache 2.0 with no asterisks. The object storage landscape has seen its share of licensing changes and restrictions that caught users off guard. We believe infrastructure software should be genuinely open-no licensing traps, no moving goalposts, and no "open core" bait-and-switch.

Built with Swift

Swift isn't just for iOS apps. It compiles to highly optimized native code with performance comparable to Rust or Go, while providing memory safety by design. This eliminates entire classes of vulnerabilities common in C/C++ implementations. Combined with Swift's modern async/await concurrency and SwiftNIO, Alarik is built on a solid foundation for high-performance systems programming.

Performance First

Even in beta, benchmarks show Alarik delivering competitive and often superior throughput compared to established solutions. We're not just building another object store-we're building a fast one.

Simple by Design

Alarik focuses on doing one thing well: storing and serving objects via the S3 API. No bloated feature set, no complex clustering requirements to get started. Deploy a single instance and scale when you need to.

Features

S3-Compatible API

Full object operations, multipart uploads, versioning with delete markers, conditional requests, range reads, presigned URLs, bucket policies, and lifecycle rules - works with any S3 SDK or tool out of the box.

Web Console

A modern browser UI for managing buckets and objects - drag-and-drop upload, recursive search, bucket & folder stats, version history, shared links, and a live admin dashboard.

Authentication & Access Control

Local accounts or OIDC SSO with multiple identity providers, per-user S3 access keys with optional expiration, fine-grained bucket policies, and public access blocks.

Webhooks

AWS-event-shaped HTTP notifications with HMAC-signed deliveries, a persistent outbox, automatic retries, and delivery health you can inspect from the console or API.

Bucket Replication

Continuous, SigV4-signed replication to any S3-compatible target - synchronous or asynchronous per rule, with opt-in delete and existing-object replication, and full task health visibility.

Horizontal Scaling

Opt-in cluster mode - erasure-coded object data, replicated control-plane metadata, quorum writes, automatic rebalancing, and any-node-serves-any-request routing.

Erasure Coding

Cluster mode protects object data with Reed-Solomon erasure coding - the durability of replication at ~1.3-1.5x storage overhead instead of 3x, with automatic read-repair and background bit-rot scrubbing.

Durable & Memory-Bounded

Uploads are fsynced to stable storage before they're acknowledged, and object bodies of any size stream to and from disk with bounded memory - a multi-gigabyte upload or download never has to fit in RAM.

Who is Alarik For?

  • Developers who need local S3-compatible storage for development and testing
  • Self-hosters who want control over their data without vendor lock-in
  • Teams looking for a simple, performant object store they can trust
  • Anyone frustrated with the licensing direction of existing solutions