Deploying to Node.js

Run your Rails app on Node.js, Bun, or Deno with traditional server databases.

Table of Contents

• Deploying to Node.js
  • Overview
  • Runtime Options
  • Database Options
  • Development
  • Migrations
  • Production Build
  • Deployment Options
    • Docker
    • Fly.io
    • Railway / Render
    • Traditional VPS
  • Environment Variables
  • Configuration
    • config/database.yml
  • npm Ecosystem
  • Performance

Overview

Server deployment runs your application on a JavaScript runtime with access to server-side databases. This is closest to traditional Rails deployment.

Use cases:

• Traditional web applications
• Apps requiring server-side processing
• Integration with npm ecosystem (Puppeteer, LangChain, etc.)
• Self-hosted deployment

Runtime Options

Database Options

Development

bin/juntos up -d sqlite

This builds the app and starts a server on port 3000.

Migrations

Run migrations before starting the server:

bin/juntos migrate -d sqlite
bin/juntos up -d sqlite

For remote databases:

# Set connection string in .env.local
echo "DATABASE_URL=postgres://user:pass@host/db" >> .env.local

bin/juntos migrate -d pg
bin/juntos up -d pg

Production Build

bin/juntos build -t node -d pg

Creates a deployable application in dist/:

cd dist
npm install
node lib/rails.js  # Or use package.json scripts

Deployment Options

Docker

FROM node:22-slim

WORKDIR /app
COPY dist/ .
RUN npm install --production

EXPOSE 3000
CMD ["node", "lib/rails.js"]

docker build -t myapp .
docker run -p 3000:3000 -e DATABASE_URL=... myapp

Fly.io

cd dist
fly launch
fly deploy

Railway / Render

1. Push dist/ to a Git repository
2. Connect the repository to the platform
3. Set DATABASE_URL environment variable
4. Deploy

Traditional VPS

# On server
git clone your-repo
cd your-repo/dist
npm install --production
DATABASE_URL=... node lib/rails.js

Use PM2 or systemd for process management:

npm install -g pm2
pm2 start lib/rails.js --name myapp
pm2 save

Environment Variables

Configuration

config/database.yml

development:
  adapter: sqlite3
  database: db/development.sqlite3

production:
  adapter: pg
  url: <%= ENV['DATABASE_URL'] %>

The adapter is selected at build time. Different environments can use different adapters by building with the appropriate configuration.

npm Ecosystem

Server targets can use the full npm ecosystem:

// In a controller or service
import puppeteer from 'puppeteer';
import { ChatOpenAI } from 'langchain/chat_models/openai';

// Generate PDFs, call LLMs, etc.

Add dependencies to dist/package.json after building, or maintain a separate package.json that gets merged during build.

Performance

• Cold start: ~100-500ms depending on app size
• Request handling: Similar to Express.js
• Memory: Depends on database connection pooling

For high-traffic applications, consider:

• Connection pooling (pg-pool, mysql2 pooling)
• Clustering (PM2 cluster mode)
• Caching (Redis, in-memory)