From Monolith to Microservices With the Strangler Pattern
summary
Explains a concrete Strangler migration: add a proxy, extract UI and services (recommendation, inventory, order, cart), introduce an anti-corruption layer to preserve interfaces and a synchronization agent that uses event sourcing to keep the old DB coherent, switch extracted logic to Lambda/containers, use circuit breakers/service mesh and Saga workflows with AWS Step Functions and GraphQL via AppSync. Shows Amazon’s thousands-of-teams transition and practical tradeoffs that make the approach work.
AI-written summary · about
chapters (24)
- 0:00Modernization with Strangler pattern
- 1:02Applications need modernization now
- 3:13Cloud migration journey
- 4:38Monoliths hinder cloud-native innovation
- 5:21Defining monoliths and limits
- 7:28Microservices solve monolith limits
- 9:38When microservices are right
- 11:23Microservices harder than they seem
- 13:11Amazon’s microservices success
- 15:15Strangler pattern with Storefront
- 18:48Domain-driven design for decomposition
- 21:16Strangling via user interfaces
- 21:59Routing proxies for migration
- 23:04Anti-corruption layers for dependencies
- 24:29Data sync between services
- 26:16Finalizing strangulation
- 26:58Trade-offs and distributed complexity
- 28:01Key microservice patterns
- 35:00Four takeaways for adoption
- 36:28Q&A on vendor lock-in
- 40:40Modular monoliths as step
- 42:04When full rewrite needed
- 43:09Domain challenges in decomposition
- 46:22Closing remarks
related talks
24mModernise your monolith using strangler pattern
24mPatterns of Evolution from Monolith to Microservices
42mQ&A - Monolith to Microservices
30mJourney from monolith to microservices on AWS
39mConsidering Migrating a Monolith to Microservices?
54mFrom Monolith to (µ)Services. What happens with my Data?
30mApplication Modernization Using Cloud Strangler Pattern and DevOps Best Practices
26mDDD + Microservices + DevOps – The Strangler Figs Way