Top REST API Design Pitfalls
summary
Collected from over 150 companies, the talk lists concrete mistakes and fixes: use OpenTelemetry trace IDs to find old clients, run compare OpenAPI in CI and consumer-driven contract tests (pact.io or Spring Cloud Contract) to catch breaking changes, and run bots that open PRs for client-jar upgrades. Avoid sequential numeric IDs, cap batch POSTs (e.g. 1,000 ids) to prevent 1GB requests and 300-thread DoS, separate DTOs and apply CQRS/read-models, and don't return data from PUT; never leak stack traces and return consolidated 4xx validation errors (422/400).
AI-written summary · about
chapters (10)
- 0:00Introduction
- 1:21API backwards compatibility
- 5:13Versioning strategies
- 7:40Contract management
- 12:46Performance pitfalls
- 19:23Domain model decoupling
- 25:15CQRS principles
- 31:04API design and user intent
- 37:35Endpoint semantics and patch
- 43:32Error handling and summary
related talks
51mTop 10 Rest API Design Pitfalls
52mTop 10 Rest API Design Pitfalls
52mTop REST API Design Pitfalls
42mREST API design good practices
52mDesigning and Implementing REST APIs For the Long Haul
18mBuild a terrible API for people you hate
58m15 Best Practices for Designing Enterprise RESTful APIs in the Cloud
18mUsing APIs appropriately