Serverless Architectures & APIs

What

An architecture where developers do not have to manage the servers themselves.

• The cloud provider automatically provisions and manages the server environment. Developers simply write individual pieces of code (as Functions) which are then executed on an event-driven basis.
• 요청하기 전까지 서버는 죽어있음. api요청이 올떄 그때 서버가 부팅되서 열리고 다시 꺼짐!

• Main services
  • AWS Lambda
  • Microsoft Azure Functions
  • Google Cloud Functions
• How it works
  • User Request → API Gateway → Lambda Function → External DB, Authentication, 3rd Party API
• Java Spring is not fit for serverless
  • 실행되기 위해서 서블릿 컨테이너가 필요함 → 톰캣이 초기화 될때 리소스가 많이 들어감
  • 규모가 클수록 응답이 느려질거임

Characteristics of Serverless

Feature	Description
No Maintenance Burden	The cloud provider is responsible for all server operations, including security patches.
Billing Model	You pay only for what you use, based on execution time and the number of invocations.
Scalability	The infrastructure automatically scales up or down in response to traffic (Auto-scaling).
Deployment Speed	You can deploy quickly at the function level, rather than deploying an entire application.

Disadvantages

• Cold Start: If a function has not been invoked for a while, there can be a noticeable delay on the first execution as the cloud provider has to initialize a new environment for it.
• Execution Time Limits: Functions are designed for short-running tasks. For example, AWS Lambda has a maximum execution time of 15 minutes. Longer processes require orchestration with other services like AWS Step Functions.
• Difficulty with State Management: Functions are stateless by design. To maintain any state or data between executions, you must use an external service like a database or a cache (e.g., Redis).
• Monitoring Complexity: Tracing and debugging logic across a complex, distributed set of functions can be difficult. It requires specialized observability tools like AWS CloudWatch and X-Ray.
  • But CloudWatch can be… expensive

Example (AWS Lambda + API Gateway)

☕Java based Lambda Handler

public class HelloHandler implements RequestHandler<Map<String,Object>, APIGatewayProxyResponseEvent> {
    @Override
    public APIGatewayProxyResponseEvent handleRequest(Map<String,Object> input, Context context) {
        return new APIGatewayProxyResponseEvent()
            .withStatusCode(200)
            .withBody("Hello from Lambda!");
    }
}

• The handleRequest() method is the most fundamental handler interface for building AWS Lambda functions in Java.
• It is designed to receive the incoming event data (as a JSON object) and a context object (containing runtime info), and then returns a response object after processing.
• An API Gateway is used to connect this Lambda function to the web, translating HTTP Requests into the event object that the handler receives.

Serverless Backend Platforms: Supabase & Firebase

Note

The term “Serverless” means more than just executing functions with services like AWS Lambda (Function-as-a-Service).

• Serverless also includes platforms that provide a complete suite of backend features (Database, Authentication, Storage, Real-time updates, etc.) without requiring you to manage servers.

• Firebase
• Supabase

Practical Tips for Using Serverless Services

• Firebase is excellent for mobile app startups that need to build and launch a Minimum Viable Product (MVP) quickly.
• Supabase is well-suited for web services, internal admin panels, and SaaS platforms where a relational (SQL) database is beneficial.
• Remember their core difference: Firebase is NoSQL-centric, while Supabase is SQL-centric.
  • This fundamental design difference will influence your entire application architecture.
  • SQL vs NoSQL
• Security Warning: Since both platforms allow the client to connect directly to the database, you must separate sensitive business logic into serverless functions (Cloud Functions or Edge Functions) to maintain security.

Summary

Feature	Firebase	Supabase
Database	Firestore / RealtimeDB (NoSQL)	PostgreSQL (SQL)
Authentication	Firebase Auth	Supabase Auth
Function Execution	Cloud Functions (Node.js based)	Edge Functions (Deno based)
Real-time Processing	Powerful real-time synchronization	Table change event detection (Slightly slower)
Deployment	Google Cloud	Self-hosted server or Supabase Platform
Target Development	Mobile apps, Real-time apps	Web services, SaaS