How to Avoid Doing Microservices Completely Wrong
summary
Claims that over 90% of teams who attempt microservices accidentally build a distributed monolith and presents 10 common mistakes with concrete examples: shared database causing schema coupling, synchronous HTTP calls causing cascading latency, and breaking event contracts. Alternatives and fixes include an event bus, circuit breakers, OpenTelemetry for tech-agnostic logging, domain-driven boundaries to size services, strangler-fig gradual migration from a modular monolith, and automated CI/CD to enable safe, independent deploys.
AI-written summary · about
chapters (15)
- 0:00Avoiding distributed monoliths
- 1:01Client story: distributed monolith
- 3:48Monoliths vs microservices vs distributed monoliths
- 13:0010 microservices mistakes
- 13:21Mistake 1: microservices always better
- 17:12Mistake 2: shared data stores
- 21:29Mistake 3: microservices too big
- 22:53Mistake 4: microservices too small
- 25:23Mistake 5: starting from scratch
- 29:58Mistake six coupling through crosscutting concerns
- 32:05Mistake 7: sync communication pitfalls
- 36:21Mistake 8: breaking event contracts
- 38:48Mistake nine missing automated builds and releases
- 39:51Mistake ten mismatched team structures
- 41:35Summary
related talks
53mDon’t Build a Distributed Monolith: How to Avoid Doing Microservices Wrong
01h 04mDon’t Build a Distributed Monolith
43mMicroMonolith - Top anti-patterns of distributed systems
55mMicroservices, Where Did It All Go Wrong
44mMicroservices, Where Did It All Go Wrong
40m10 tips for failing at microservices
01h 03mMicroservices vs The Distributed Monolith
57mEmbracing Microservices