Deploy and Host OpenOrange with Railway

OpenOrange runs Cloudflare Workers and Durable Objects on Node.js, with each Durable Object as a Rivet Actor on a self-hosted Rivet Engine. This template runs your Worker as a service you can scale with replicas, a private Rivet Engine to place and store your Durable Objects, and a Railway Bucket for their durable copy, with every variable configured for you and a starter Worker to try it.

About Hosting OpenOrange

OpenOrange needs a Rivet Engine, and the Engine needs local disk for recent writes plus S3-compatible object storage for the durable copy of every Durable Object. The template runs the Engine on Railway's private network with a volume at /var/lib/rivet/log and a Railway Bucket, and generates the token that the Worker's nodes use to reach it. The openorange service serves the Worker on a public domain. Each replica is a node of the Engine, which runs every Durable Object on one node at a time and routes calls to it from any replica. Nodes create their namespace and node pool on first start, drain on every deploy, and keep their Durable Objects' data across deploys. Patch releases of both images install automatically during Railway's nightly update window.

Common Use Cases

Self-hosting Cloudflare Workers and Durable Objects
Moving a Worker off Cloudflare, starting from its Wrangler configuration
Stateful applications with one SQLite database per object
Chat rooms, counters, coordinators, and WebSocket services

Dependencies for OpenOrange Hosting

A Rivet Engine, which the template runs as the rivet-engine service
A Railway volume for the Engine's recent writes and identity
A Railway Bucket for the durable copy of your Durable Objects
A Wrangler project and the Railway CLI or a GitHub repo to deploy your Worker

Deployment Dependencies

Implementation Details

Add this Dockerfile next to your wrangler.jsonc, with the .dockerignore from the template's repo so local secrets stay out of the image:

FROM ghcr.io/joeychilson/railway-openorange:1.0
COPY . .
RUN openorange-build

openorange-build installs your packages, runs your build.command, and bundles the Worker. Deploy it to the openorange service, disconnecting the starter image once first so Railway doesn't deploy it again over your Worker:

railway link
railway service source disconnect --service openorange
railway up --service openorange

Or connect your GitHub repo to the service, and every push deploys. Scale with railway scale --service openorange us-east=2 eu-west=2. Keep rivet-engine at one replica and off the public internet.

Why Deploy OpenOrange 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 OpenOrange 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.