Deploying the API to Railway

Railway is a platform-as-a-service that deploys Node.js applications from a GitHub repo. It handles servers, scaling, and SSL automatically. (Referral link — signing up through it supports this platform at no cost to you.)

Before you deploy

Make sure your project has:

1. A start script in package.json:

   "scripts": {
     "start": "node dist/index.js",
     "build": "tsc"
   }

2. A Procfile (optional but explicit):

   web: npm start

3. Your dist/ directory is in .gitignore (Railway will build it).

4. The app listens on process.env.PORT:

   const PORT = parseInt(process.env.PORT || '3000', 10)
   app.listen(PORT, ...)

Deploying to Railway

1. Push your code to GitHub.

2. Go to railway.com and create a free account.

3. Click New Project → Deploy from GitHub repo → select your bulletin-api repository.

4. Railway detects Node.js and automatically runs npm install && npm run build && npm start.

5. Click on the deployment → Settings → Variables and add:

   NODE_ENV=production
   JWT_SECRET=<a long random string — never share this>
   DATABASE_PATH=./bulletin.db
   FRONTEND_URL=https://your-username.github.io

6. In the Settings tab, generate a domain (e.g., bulletin-api-production.up.railway.app).

7. Wait for the deployment to complete — check the build logs for errors.

Verifying the deployment

# Health check
curl https://bulletin-api-production.up.railway.app/health
# → { "status": "ok" }

# Register a user
curl -X POST https://bulletin-api-production.up.railway.app/auth/register \
  -H "Content-Type: application/json" \
  -d '{"username": "testuser", "password": "testpassword"}'
# → { "token": "eyJ...", "userId": 1, "username": "testuser" }

Important: SQLite in production

SQLite works on Railway but with one limitation: the database file (bulletin.db) is stored on the container’s filesystem. If Railway restarts the container (which it does occasionally), the database persists IF it’s in the project directory committed to git — but it usually isn’t (databases aren’t committed to git).

For a learning project, this is fine. For production apps, you’d use a persistent database service (Railway PostgreSQL, PlanetScale, Supabase, etc.).

Add your bulletin.db to .gitignore and accept that the database may reset on deployment. For learning purposes, this is perfectly acceptable.

Updating the deployment

Every push to your GitHub main branch triggers a new deployment automatically.

Environment variables summary

Variable	Value
NODE_ENV	production
JWT_SECRET	Long random string (never commit)
DATABASE_PATH	./bulletin.db
FRONTEND_URL	Your GitHub Pages URL
PORT	Set automatically by Railway

Exercise

1. Push your bulletin-api project to GitHub.
2. Deploy to Railway following the steps above.
3. Set all environment variables.
4. Test POST /auth/register against the live URL.
5. Note the Railway API URL — you’ll use it in Part 2 to configure the React frontend.

Recap

• Railway deploys Node.js apps from GitHub repos — auto-detects, auto-builds.
• PORT is set by Railway — use process.env.PORT in your app.
• Set environment variables (JWT_SECRET, FRONTEND_URL) in the Railway dashboard.
• SQLite persists as long as the container doesn’t restart — acceptable for learning, use a managed DB for production.
• Save your Railway URL — the React frontend needs it in Module 08.