Introduction

Record Store stores files — objects — and serves them over an S3-compatible HTTP API. If your application can talk to Amazon S3, it can talk to Record Store.

What it is for

Record Store is useful when you want object storage that you run yourself:

• application file storage (uploads, exports, generated documents, media)
• storage that must stay on infrastructure you control
• an S3 endpoint for development that behaves like the one in production
• storage for a small team or a single application, without operating a distributed system from day one

It is a single Rust binary with an embedded metadata database. There is no external database, message queue, or object gateway to run alongside it.

How it is organised

Record Store exposes three listeners. Each has a distinct audience.

Port	Listener	Who talks to it
7600	S3 API	Applications, AWS SDKs, embed links
7601	Management API	The console, the CLI, automation
7602	Web console	Administrators, share-link recipients

Only 7600 and 7602 are normally reachable from outside your network. See Ports for the full picture and Reverse Proxy and TLS for publishing them.

The shape of a deployment

A deployment is one process with one copy of your data, on one machine.

flowchart LR
    App[Application] -->|S3 API :7600| RS[Record Store]
    Admin[Administrator] -->|Console :7602| RS
    RS --> Disk[(Data directory)]

There is nothing else to stand up: no external database to provision, no coordination service, and no gateway in front. What you are responsible for is the data directory underneath it. Put it on durable storage, and read Durability for exactly what a successful write promises.

Redundant storage is not backup

Redundant disks protect against hardware failure. They do not protect against a mistaken delete, which is written through just as reliably as anything else. Read Backup and Restore.

Next

• Installation — the supported ways to obtain and run Record Store
• Quick Start — running and storing an object in about five minutes