Beyond Microservices: Streams, State and Scalability
summary
Breaks down when HTTP call-chains create a distributed monolith and prescribes concrete fixes: deploy sidecar proxies/service mesh for routing, rate limiting and retries; front external traffic with an API Gateway; publish state changes to an event stream and give each service its own local database populated from that stream; use a schema registry (Confluent) for contracts. Result: independent deploys, fewer cross-team rollouts, retry storms avoided; tradeoffs are eventual consistency and more complex data repair/auditing.
AI-written summary · about
chapters (15)
- 0:00Introduction
- 1:06Microservices evolution and goals
- 3:55Distributed monolith risks
- 10:21Speaker background
- 11:02API Gateway patterns
- 15:18Service mesh and sidecars
- 20:37Central control plane importance
- 21:00Event-driven microservices
- 27:43Database per service
- 30:11Event-driven stateful services
- 33:29Message contracts and schemas
- 37:45Serverless microservices
- 42:02State management in serverless
- 46:00Design patterns toolbox
- 47:04Q&A session
related talks
37mBeyond Microservices: Streams, State & Scalability
24mPatterns of Evolution from Monolith to Microservices
01h 03mMicroservices vs The Distributed Monolith
59mHow to Escape The Distributed Monolith
44mMicroservices - How to create problems by solving problems?
50mAddressing Recent Events
26mBuilding Scalable Microservice APIs From The Beginning
11mLightning Talk: Event Based Communication in Complex Micro-service...