Stick to the basics. Only include the dependencies you really need for your function. Extra packages just bloat your function and slow things down. Run npm audit and npm prune to find and ditch those unused packages.

Grab a tool like Webpack or Rollup to bundle your dependencies. Bundling cuts down the number of modules, making them quicker to load during execution.

Keep it lean. Break down large functions into smaller, more focused ones. This makes each function's environment simpler and faster to start up.

Load dependencies based on the environment. For instance, use different setups for development and production to keep things light during critical operations.

if (process.env.NODE_ENV === 'production') {
  // Load production-only dependencies
}

Shift non-essential initialization logic outside the main function body. This way, only the necessary setup steps run during cold starts.

const config = require('config');
const db = require('db');

exports.myFunction = (req, res) => {
  // Main function logic here
};

HTTPS functions are your friends. They tend to start up faster because they pre-warm better than background triggers. Use them when you can for quicker cold starts.

const functions = require('firebase-functions');

exports.myHttpFunction = functions.https.onRequest((req, res) => {
  res.send("Hello World");
});

Set the right memory and timeout settings for your functions. Overdoing it with resources can actually slow down initialization.

exports.myFunction = functions
  .runWith({ timeoutSeconds: 30, memory: '256MB' })
  .https.onRequest((req, res) => {
    res.send("Hello World");
  });

Reuse connections to external resources like databases. Creating new connections every time adds a lot of overhead.

let dbConnection;

exports.myFunction = functions.https.onRequest(async (req, res) => {
  if (!dbConnection) {
    dbConnection = await initializeDatabaseConnection();
  }
  // Use dbConnection
});

Take advantage of Firebase Performance Monitoring to profile your cloud functions. This helps you spot bottlenecks and areas that need improvement. Turn on profiling in the Firebase console and check the logs for insights.

Keep an eye on function performance and cold start times using Firebase console metrics. Make small improvements regularly and stay updated with the latest Firebase releases for new optimizations.