Why You Had a Bad Time With Microservices
summary
Argues microservices pay off well below “Google scale,” shows an HPC case where an API, message bus, event sourcing and CQRS replace a tightly coupled MVC monolith, yielding materialized read models, easier tests, faster onboarding, and new features (SMS) added without touching core logic. Demonstrates moving file jobs through events, using Elasticsearch for read views, and explains autonomy: services neither query nor depend on others, which is the practical success criterion.
AI-written summary · about
chapters (11)
- 0:00Intro
- 0:21Software is hard to change
- 2:51Whats the model
- 5:40Extract Microservices
- 8:13Autonomy
- 9:23No Kittens
- 10:03Commands and Events
- 11:51Domain Events
- 13:12Case Study
- 17:58System Components
- 21:15Summary
related talks
54mThe Long Sad History of Microservices
43mPanel: Microservices, Where Did It All Go Wrong
38mMicroservices, a bittersweet symphony
45mMicroservices: A retrospective
48mChallenges in Implementing MicroServices
44mMicroservices, Where Did It All Go Wrong
50mChallenges in Implementing Microservices
26mPros/Cons of Microservices
