Heldinhow/aws-lambda

AWS Lambda Development

Overview

AWS Lambda allows you to run code without provisioning or managing servers, executing only when triggered by events. This skill should be invoked when building microservices, event-driven architectures, background jobs, or API backends that require rapid scaling.

Core Principles

• Statelessness: Lambda functions are ephemeral. Any persistent state must be stored in external services (like DynamoDB, S3, or RDS).
• Event-Driven: Functions should be designed to react to specific triggers (API Gateway requests, S3 bucket uploads, SQS messages).
• Fast Execution: AWS charges based on execution time and memory. Optimize code for fast startup and quick processing.
• Least Privilege: Grant the Lambda execution role only the specific AWS permissions required to perform its task.

Preparation Checklist

• [ ] Determine the trigger source (e.g., API Gateway, EventBridge, SQS).
• [ ] Select the appropriate runtime (Node.js, Python, Go, etc.) based on team expertise and performance needs.
• [ ] Identify dependencies and plan how they will be packaged (e.g., Lambda Layers, bundling with Webpack/esbuild).

Step-by-Step Process

1. Define the Handler: Write the core function signature handler(event, context) ensuring it correctly parses the incoming event payload.
2. Implement Business Logic: Separate the core logic from the handler. The handler should only deal with event parsing and response formatting.
3. Manage Configuration: Use Environment Variables for dynamic configuration (database strings, API keys).
4. Local Testing: Mock the specific AWS event payload locally to test the logic before deployment.
5. Optimize Cold Starts:
   • Move database connections or heavy initializations outside the handler function (in the global scope) so they are reused in warm invocations.
   • Keep the deployment package size small.
6. Deployment & Logging: Deploy via IaC tools (AWS SAM, Serverless Framework, CDK, or Terraform). Ensure robust logging using console.log or a structured logger to CloudWatch.

Do's and Don'ts

• ✅ Do use proper error handling to ensure visibility into failures and to correctly trigger DLQs (Dead Letter Queues) if needed.
• ✅ Do configure concurrency limits and timeouts appropriately to control costs and prevent runaway processes.
• ❌ Don't store sensitive secrets in plain text environment variables; use AWS Secrets Manager or Parameter Store.
• ❌ Don't build massive "monolithic" Lambda functions that route dozens of different endpoints internally.

Anti-Patterns

• The Recursive Loop: A Lambda function that triggers itself (e.g., reading from an S3 bucket and writing back to it, which triggers the function again), leading to massive bills.
• Timeouts: Setting a 3-second timeout for a function that calls a slow third-party API, causing constant failures.
• Heavy Frameworks: Using heavy web frameworks (like full Spring Boot) inside a Lambda, causing multi-second cold starts.