Microservices and Rules Engines – a blast from the past
summary
Microservices often just redistribute coupling rather than remove it, so use tiny pluggable rule components that a generic engine loads at startup to handle cross-service concerns. Examples: search uses IFilterProductIds components from each domain, pricing composes multiplicative factors via IProvidePriceFactor implementations so order doesn’t matter, and fraud aggregates IProvideFraudMultiple scores. Implementation is simple assembly scanning and runtime composition (Angular composite UI and a GitHub demo exist), which preserves service autonomy while enabling search/pricing/fraud without tight coupling.
AI-written summary · about
chapters (13)
- 0:00Intro
- 2:57and now, everybody's doing it
- 5:21How is the coupling different?
- 6:53Presto - coupling disappear-o!
- 18:50Services communicate via events
- 20:25We need a much wider perspective
- 21:00Looking across the enterprise
- 23:27Composing components from services - systems
- 26:16For example, implementing a product search
- 36:53Enter the rules engine
- 38:19Think design pattern
- 39:03Let's start with search
- 45:32Fraud detection can work the same way
related talks
45mResponsible Microservices
44mMicroservices - How to create problems by solving problems?
16mMicroservices Paradise: is it possible to achieve?
32mMicroservices Orchestration with
03mMicroServices - Let’s Build Some!
38mMicroservices, a bittersweet symphony
35mWhy I Wouldn't Use MicroServices
23mMICROSERVICES ARCHITECTURE AS IF YOU MEANT IT