Deploy and Host Open Archiver on Railway

Open Archiver copies your organisation's mail out of Google Workspace, Microsoft 365, a generic IMAP mailbox, or a PST, EML or mbox export, and keeps it in storage you control. Everything it archives becomes full-text searchable, attachments included, and stays searchable after the original message is deleted from the mailbox it came from. This is a community-maintained template; it is not affiliated with the Open Archiver project.

About Hosting Open Archiver

Open Archiver is a real platform rather than a single app, and hosting it reflects that. One container runs the web interface, the API, and three background workers that ingest, index and schedule. Around it sit a PostgreSQL database for metadata, a Valkey instance for the job queues, a Meilisearch index for the search, and a volume for the archived mail itself. All four must agree on connection strings and two shared keys.

The part that needs care is the first minute. A fresh instance shows a setup page that creates the first Super Admin. The project closes that page correctly once an account exists, but on a hosting platform the public address is live from the moment the service starts, and what sits behind that page is an organisation's entire mail history. This template completes the setup inside the container, before the public port is opened, so the instance is claimed before anybody can reach it. It also refuses to start on the placeholder secrets shipped in the project's example configuration, which are easy to copy and quiet when wrong.

Be aware of the cost. This is a four-service deployment, the application container idles at several hundred megabytes before any mail is ingested, and the image reinstalls its dependencies on every start, so deploys take minutes rather than seconds.

Why Deploy Open Archiver on Railway?

Railway is a singular platform to deploy your infrastructure stack. Railway will host your infrastructure so you don't have to deal with configuration, while allowing you to vertically and horizontally scale it.

By deploying Open Archiver on Railway, you are one step closer to supporting a complete full-stack application with minimal burden. Host your servers, databases, AI agents, and more on Railway.

Concretely, this template provisions all four services, wires the connection strings and shared keys, attaches the volumes, generates the administrator password and the encryption keys, creates your account, and points the healthcheck at the one route that answers without a session.

Common Use Cases

Meet a mail retention obligation without paying per mailbox for a commercial archiving product.
Keep a permanent, searchable record of correspondence after employees leave and their mailboxes are deleted.
Load a decade of PST or mbox exports into one place and actually find things in them.
Give auditors or legal a read-only account scoped to search rather than access to live mailboxes.

Dependencies for Open Archiver Hosting

A PostgreSQL database for users, roles, ingestion sources and job metadata.
A Valkey or Redis instance for the ingestion and indexing queues.
A Meilisearch instance for the full-text index.
A persistent volume for the archived messages and attachments, or S3-compatible object storage.

Deployment Dependencies

Open Archiver upstream project and documentation: https://github.com/LogicLabs-OU/OpenArchiver
Configuration reference: https://github.com/LogicLabs-OU/OpenArchiver/blob/main/.env.example
Caddy, used as the gate that holds the public port shut until setup is complete: https://caddyserver.com
Template repository, wrapper image and tests: https://github.com/youssefsiam38/openarchiver-railway
Published image: ghcr.io/youssefsiam38/openarchiver-railway
Open Archiver is licensed AGPL-3.0-only. Running it as a network service makes you a distributor under section 13, so keep the corresponding source of the version you run available.

Implementation Details

The wrapper adds no application code. It validates the configuration and refuses the example secrets and malformed keys, starts the application with its frontend on loopback, waits for the API, and completes the setup through the application's own endpoint. Only then does it open the public port. Completing the setup is idempotent, so a redeploy never disturbs an account whose password you have since changed, and a signalled stop exits cleanly rather than looking like a crash.