การนำไปใช้งานแบบไร้เซิร์ฟเวอร์

Welcome to serverless deployment! This strategy lets you build and run applications without managing servers.

Instead of provisioning VMs, you write code that runs in response to events, like an HTTP request.

The cloud provider (e.g., AWS, Google Cloud) handles all the server management for you.

Deploying NestJS on serverless platforms offers compelling advantages, especially for APIs:

• Automatic Scaling: Your application scales instantly with demand, from zero to thousands of requests.
• Cost-Efficiency: You only pay for the compute time consumed by your code, not for idle servers.
• Reduced Operations: No server maintenance, patching, or scaling to worry about. Focus on your code!

Several major cloud providers offer robust serverless computing services:

• AWS Lambda: The most popular choice, often used with API Gateway for HTTP endpoints.
• Google Cloud Functions: Google's offering, deeply integrated with the Google Cloud ecosystem.
• Azure Functions: Microsoft's solution for event-driven serverless applications.

We'll primarily focus on concepts applicable to AWS Lambda.

NestJS applications are typically long-running processes, constantly listening for requests. Serverless functions, however, are designed to be short-lived, executing only when triggered.

The challenge is to make NestJS's robust architecture fit into this 'function-as-a-service' model. We need a way to 'boot up' our NestJS app within the context of a single serverless invocation.

To run a NestJS app on AWS Lambda, we use an adapter. This adapter acts as a bridge, translating incoming Lambda events into a format NestJS understands (like an Express.js or Fastify request) and then translating NestJS's response back for Lambda.

Popular tools like @vendia/serverless-express (for Express) or @nestjs/platform-fastify with aws-lambda-fastify help achieve this.

The Serverless Framework is a powerful CLI tool that simplifies deploying serverless applications to various providers.

It handles packaging your code, creating cloud resources (like Lambda functions and API Gateway endpoints), and managing deployments.

You can install it globally via npm:

The core of a Serverless Framework project is the serverless.yml file. This YAML configuration defines your service, provider, and functions.

Here's a simplified structure:

This is the actual entry point for your Lambda function. It initializes your NestJS application and wraps it with an adapter.

The goal is to initialize NestJS only once (a 'warm start') to reduce latency on subsequent calls.

Once your serverless.yml and handler are configured, deploying your NestJS application is straightforward:

1. Build: Compile your TypeScript code (e.g., npm run build).
2. Deploy: Run serverless deploy from your terminal. The CLI handles packaging and provisioning.
3. Test: The CLI will output your API Gateway endpoint, which you can use to test your deployed NestJS API.

In this lesson, we explored how to deploy NestJS applications using serverless platforms like AWS Lambda.

You learned about the benefits (cost, scaling, reduced ops), the challenge of adapting NestJS for short-lived functions, and how the Serverless Framework and adapters help bridge this gap. This approach enables highly scalable and cost-efficient backend services.