A practical tale of assessing a Microservices Architecture
summary
Case study of an event-driven microservices system uses source-code metrics, message-graph extraction and interviews to find concrete problems: a dozen services (some ~20,000 LOC), 65 message types with many unused or multiply-consumed messages, a shared library coupling logic, per-service databases and no distributed tracing. Data volume hits hundreds of GB and load testing is unfinished, exposing message-broker overload and single-region risk. Assessor builds a DDD-based reference design, maps aggregates to services, and delivers a prioritized action plan: define quality attributes, consolidate owners, add tracing, finish load tests and redesign boundaries where needed.
AI-written summary · about
chapters (6)
- 0:00Introduction
- 1:06Assessment goals
- 9:22Microservices structure
- 25:24Quality attributes
- 31:48Summary
- 32:08Q&A
related talks
41mA practical tale of assessing a Microservices Architecture
52mWhat Does a Good Microservice Architecture Look Like?
40mMeasuring Micro-services
30mMeasuring Microservices
31mHundreds of Microservices without breaking your APIs
51mResponsible Microservices
54mThe Long Sad History of Microservices
23mWhy You Had a Bad Time With Microservices