From Synchronous to Asynchronous Distributed Microservices
summary
Starts from why synchronous request-response breaks at scale and shows concrete alternatives: replace blocking calls with fire-and-forget or pub/sub over a durable transport (Proofpoint uses Kafka), implement asynchronous request/response by attaching context or persisting it, and use patterns like CQRS and Event Sourcing to separate write and read models. Covers delivery guarantees (at-least-once, exactly-once caveats), backpressure strategies, retries and dead-letter queue handling, and the consistency versus availability tradeoffs you actually face when migrating.
AI-written summary · about
chapters (13)
- 0:00Intro
- 1:44Synchronous vs Asynchronous Systems
- 6:47Trading consistency for availability
- 10:58CAP Theorem in microservices
- 17:56Asynchronous Message Patterns
- 19:23Fire and Forget (Notification)
- 23:00Request/Response
- 28:27Distributed Log
- 30:31Command Query Responsibility Segregation
- 33:41Event Sourcing
- 40:03Transactions
- 40:52Backpressure
- 42:21Error Honding
related talks
01h 02mAsynchronous Distributed Systems
01h 01mThe Unsung Hero of Modern Software: Asynchronous Messaging
51mPatterns & Practices of Implementing Event-driven MicroSvc...
47mEvent-Driven Microservices = Microservices Done Right
11mLightning Talk: Event Based Communication in Complex Micro-service...
37mBeyond Microservices: Streams, State & Scalability
47mThe Rise of Reactive Microservices
53mYou Keep Using That Word: Asynchronous And Interprocess Comms