Designing and building resilient backend APIs in the cloud
summary
Shows a serverless stack using S3, CloudFront, API Gateway, AWS Lambda and DynamoDB with EventBridge-driven decoupling, logging, metrics and tracing pushed to CloudWatch. Demonstrates AWS Lambda Power Tools for TypeScript to emit JSON logs and custom metrics, recommends exponential backoff with jitter for retries, and runs a chaos experiment with the fault injection simulator that injects 3–10s latency and reveals an unexpected Lambda timeout set to 3s, producing real failures to fix.
AI-written summary · about
chapters (8)
- 0:00Resilient backend APIs matter
- 5:36Resource exhaustion handling
- 11:05Decoupling long tasks with events
- 15:05Real-time system health monitoring
- 21:49Detecting downstream failures
- 24:38Retry strategies with backoff
- 30:00Chaos engineering validation
- 40:22Key takeaways
related talks
23mSimplifying design complexities using Serverless Architecture on AWS
50mBuild, Deploy, Debug and Tune Serverless Microservices
04mServerless applications, Microservices, AWS
47mBuild your next microservices application with modern AWS services
17mHow do we use Serverless Architectures on AWS Lambda
41mYou Don’t Want to Build a Serverless Monolith
51mServerless architectural patterns
54mServerless beyond Functions