Heroku

Heroku Best Practices

Developing on Heroku requires adherence to specific patterns to leverage its Platform-as-a-Service (PaaS) strengths. This guide focuses on the Twelve-Factor App methodology and Heroku's runtime principles to ensure your applications are scalable, resilient, and maintainable.

1. Code Organization and Structure

Maintain a clean, version-controlled codebase and clearly define your application's processes.

1.1. Procfile for Process Types

Explicitly declare your application's process types using a Procfile. This allows Heroku to run different types of dynos (web, worker, etc.) and scale them independently.

❌ BAD: Relying on default web process or running multiple concerns in one dyno.

# No Procfile, or a single 'web' process doing everything
web: npm start

✅ GOOD: Separate web and worker processes.

web: node src/server.js
worker: node src/worker.js
release: node src/db-migrate.js

• Context: Use web for HTTP traffic, worker for background jobs, and release for one-off tasks like database migrations before the app starts.

1.2. app.json for Declarative Setup

Use app.json for Heroku Button deployments, Review Apps, and to declare add-ons and environment variables.

❌ BAD: Manual setup of add-ons and config vars for each environment.

# No app.json, requiring manual configuration

✅ GOOD: Declarative app.json for consistent deployments.

// app.json
{
  "name": "My Awesome App",
  "description": "A description of my awesome app.",
  "repository": "https://github.com/myorg/my-awesome-app",
  "keywords": ["node", "express", "heroku"],
  "addons": [
    "heroku-postgresql",
    "heroku-redis"
  ],
  "env": {
    "NODE_ENV": {
      "description": "The environment the app is running in.",
      "value": "production"
    },
    "LOG_LEVEL": {
      "description": "Logging verbosity.",
      "value": "info"
    }
  },
  "scripts": {
    "postdeploy": "npm run db:seed"
  },
  "formation": {
    "web": {
      "quantity": 1,
      "size": "standard-1x"
    },
    "worker": {
      "quantity": 1,
      "size": "standard-1x"
    }
  }
}

• Context: app.json ensures consistency across environments and simplifies onboarding for new developers.

2. Common Patterns and Anti-patterns

Adhere to the Twelve-Factor App principles for robust cloud-native applications.

2.1. Configuration in the Environment

Store all configuration that varies between deploys (database credentials, API keys) in environment variables, not in code.

❌ BAD: Hardcoding secrets or committing config files.

// config.js
const DB_URL = "postgres://user:pass@host:port/db"; // BAD: hardcoded secret
const API_KEY = "super-secret-key"; // BAD: hardcoded secret

✅ GOOD: Fetch configuration from environment variables.

// config.js
const DB_URL = process.env.DATABASE_URL;
const API_KEY = process.env.EXTERNAL_API_KEY;

if (!DB_URL || !API_KEY) {
  console.error("Missing critical environment variables!");
  process.exit(1);
}

• Context: Heroku automatically injects add-on credentials (e.g., DATABASE_URL, REDIS_URL) as environment variables. Always use process.env.PORT for the server port.

2.2. Stateless Processes

Ensure your application processes are stateless and share-nothing. Any data that needs to persist must be stored in a backing service.

❌ BAD: Storing user sessions or uploaded files on the local filesystem.

// server.js
const express = require('express');
const session = require('express-session');
const fs = require('fs');

const app = express();
app.use(session({ secret: 'my-secret', resave: false, saveUninitialized: false })); // BAD: default session store is in-memory
app.post('/upload', (req, res) => {
  fs.writeFileSync(`/tmp/${req.file.name}`, req.file.buffer); // BAD: writing to ephemeral filesystem
});

✅ GOOD: Use external backing services for state.

// server.js
const express = require('express');
const session = require('express-session');
const RedisStore = require('connect-redis').default;
const { createClient } = require('redis');
const { S3Client, PutObjectCommand } = require('@aws-sdk/client-s3'); // Or any cloud storage

const app = express();

// Use Redis for session storage
const redisClient = createClient({ url: process.env.REDIS_URL });
redisClient.connect().catch(console.error);
app.use(session({
  store: new RedisStore({ client: redisClient }),
  secret: process.env.SESSION_SECRET,
  resave: false,
  saveUninitialized: false
}));

// Upload files to S3
const s3 = new S3Client({ region: process.env.AWS_REGION });
app.post('/upload', async (req, res) => {
  const command = new PutObjectCommand({
    Bucket: process.env.S3_BUCKET_NAME,
    Key: req.file.name,
    Body: req.file.buffer
  });
  await s3.send(command);
  res.send('File uploaded to S3');
});

• Context: Heroku's filesystem is ephemeral and unique to each dyno. Data written to it will be lost on dyno restart or scale events.

3. Performance Considerations

Optimize for Heroku's dyno model to achieve high performance and scalability.

3.1. Background Jobs

Offload long-running tasks to worker dynos to keep web dynos responsive.

❌ BAD: Performing heavy computations or sending emails directly in a web request.

// web_route.js
app.post('/process-data', async (req, res) => {
  const result = await performHeavyComputation(req.body.data); // Blocks web dyno
  await sendEmail(req.user.email, result); // Blocks web dyno
  res.json({ status: 'completed', result });
});

✅ GOOD: Queue background jobs for worker dynos.

// web_route.js
const { Queue } = require('bullmq'); // Example with BullMQ (requires Redis)
const myQueue = new Queue('my-processing-queue', { connection: { host: process.env.REDIS_HOST, port: process.env.REDIS_PORT } });

app.post('/process-data', async (req, res) => {
  await myQueue.add('process-task', { data: req.body.data, userEmail: req.user.email });
  res.json({ status: 'processing', message: 'Task queued successfully.' });
});

// worker.js (running on a worker dyno)
const { Worker } = require('bullmq');
const worker = new Worker('my-processing-queue', async job => {
  const result = await performHeavyComputation(job.data.data);
  await sendEmail(job.data.userEmail, result);
}, { connection: { host: process.env.REDIS_HOST, port: process.env.REDIS_PORT } });

• Context: Web dynos should respond quickly (ideally < 500ms). Use a queueing add-on (e.g., Heroku Redis with BullMQ/Celery) for any task that takes longer.

3.2. Database Optimization

Use Heroku Postgres efficiently. Optimize queries, add indexes, and consider connection pooling.

❌ BAD: Opening a new database connection for every request, N+1 queries.

// In a Node.js Express app without connection pooling
app.get('/users/:id', async (req, res) => {
  const client = new pg.Client({ connectionString: process.env.DATABASE_URL });
  await client.connect();
  const user = await client.query('SELECT * FROM users WHERE id = $1', [req.params.id]);
  // ... potentially more queries for related data without proper joins
  await client.end();
  res.json(user.rows[0]);
});

✅ GOOD: Use a connection pool and optimize queries.

// In a Node.js Express app with pg-pool
const { Pool } = require('pg');
const pool = new Pool({
  connectionString: process.env.DATABASE_URL,
  ssl: { rejectUnauthorized: false } // Required for Heroku Postgres
});

app.get('/users/:id', async (req, res) => {
  const user = await pool.query('SELECT * FROM users WHERE id = $1', [req.params.id]);
  res.json(user.rows[0]);
});

• Context: Heroku Postgres is a managed service. Focus on application-level database performance. Always use SSL when connecting to Heroku Postgres.

4. Common Pitfalls and Gotchas

Avoid these common mistakes to ensure smooth operation on Heroku.

4.1. Binding to PORT Environment Variable

Your web process must listen on the port specified by the PORT environment variable. Heroku dynamically assigns this port.

❌ BAD: Hardcoding a port (e.g., 3000).

// server.js
const PORT = 3000; // BAD: hardcoded port
app.listen(PORT, () => console.log(`Listening on ${PORT}`));

✅ GOOD: Use process.env.PORT.

// server.js
const PORT = process.env.PORT || 5000; // Use Heroku's PORT, fallback for local dev
app.listen(PORT, () => console.log(`Listening on ${PORT}`));

• Context: Heroku routes traffic to your web dyno on a dynamically assigned port. Failing to bind to process.env.PORT will cause your app to crash.

4.2. Handling Graceful Shutdown

Heroku sends a SIGTERM signal to dynos before shutting them down. Your app should catch this signal and perform cleanup.

❌ BAD: Abruptly exiting on SIGTERM.

// server.js
process.on('SIGTERM', () => {
  console.log('Received SIGTERM. Exiting immediately.');
  process.exit(0); // BAD: no time for cleanup
});

✅ GOOD: Implement graceful shutdown.

// server.js
const server = app.listen(PORT, () => console.log(`Listening on ${PORT}`));

process.on('SIGTERM', () => {
  console.log('Received SIGTERM. Shutting down gracefully...');
  server.close(() => {
    console.log('HTTP server closed.');
    // Close database connections, flush logs, etc.
    // Ensure all pending requests are handled or rejected.
    process.exit(0);
  });

  // Force close if server doesn't close within a timeout
  setTimeout(() => {
    console.error('Forcefully shutting down after timeout.');
    process.exit(1);
  }, 10000); // 10 seconds grace period
});

• Context: Heroku gives your dyno 10 seconds to shut down. This is crucial for preventing data loss and ensuring a smooth user experience during deploys or scaling events.

5. Testing Approaches

Integrate testing into your Heroku development workflow, especially with Heroku Flow.

5.1. Automated CI/CD with Heroku Flow

Leverage Heroku Flow for automated testing, linting, and deployment. This ensures code quality and prevents regressions.

❌ BAD: Manual testing and deployment, no automated checks.

# No CI/CD setup, manual deploys to production

✅ GOOD: Configure Heroku CI for automated testing on every push.

// app.json (excerpt for Heroku CI)
{
  "scripts": {
    "test": "npm test"
  },
  "environments": {
    "test": {
      "addons": ["heroku-postgresql:mini"],
      "env": {
        "NODE_ENV": "test"
      }
    }
  }
}

• Context: Heroku CI runs your test script (defined in package.json or app.json) on a fresh dyno for every push to GitHub, providing immediate feedback.

5.2. Review Apps for Feature Testing

Use Review Apps to automatically spin up a disposable Heroku app for every pull request, enabling easy testing and collaboration.

❌ BAD: Developers testing features locally or on a shared staging environment.

# No Review Apps, PRs are reviewed without a dedicated deploy

✅ GOOD: Enable Review Apps in Heroku Dashboard.

// app.json (ensure this file exists and is configured)
{
  "name": "My Awesome App",
  "description": "A description of my awesome app.",
  "repository": "https://github.com/myorg/my-awesome-app",
  "keywords": ["node", "express", "heroku"],
  "addons": [
    "heroku-postgresql:hobby-dev"
  ],
  "env": {
    "NODE_ENV": "development"
  }
}

• Context: Review Apps provide an isolated, production-like environment for each feature, making code reviews more effective and catching environment-specific bugs early.