Reacting to an Event-Driven World
summary
Claim: dropping an event backbone into a monolith doesn’t make end-to-end non‑blocking—configuration and architecture do. Demonstrations use Apache Kafka (3+ brokers, topics with partitions), reactive libraries like RxJava and flowable.interval, and frameworks Vert.x and MicroProfile Reactive Messaging on Open Liberty with Docker and Zookeeper. Built a coffee‑shop example and a system/inventory microservice lab that publishes average system load every 15s and consumes it; ordering, offsets, a three‑partition/three‑consumer idle case, and manual offset commits show the reactive patterns work and where surprises (idle consumers, duplicate vs lost messages) occur.
AI-written summary · about
chapters (23)
- 0:00Introduction
- 1:18Coffee shop demo: blocking vs async
- 3:03Event-driven architecture with Kafka
- 5:30Kafka limits for reactive systems
- 6:12Reactive manifesto's four behaviors
- 8:36Barista scaling under high load
- 10:00Elasticity in reactive systems
- 11:47Resiliency and circuit breakers
- 13:08Building reactive systems overview
- 13:51Reactive architecture patterns
- 16:17Reactive programming tools
- 17:41Resilient Kafka config
- 21:13Resilient Kafka producers/consumers
- 26:48Kafka producers/consumers scalability
- 31:02Reactive Kafka apps with frameworks
- 32:04Akka Alpaca Kafka connector
- 33:05MicroProfile reactive messaging
- 35:12Vert.x reactive Kafka toolkit
- 36:14Demo: reactive Java Kafka
- 40:45Hands-on lab: MicroProfile reactive
- 52:54Running demo and testing
- 53:57Free ebook and resources
- 54:17Q&A
related talks
01h 00mReacting to an Event-Driven World
39mReacting to an Event-Driven World
50mBuilding reactive microservices with MicroProfile
44mReacting to the future of application architecture
46mGet Reactive! Programming, Systems, and Microservices
48mReactive Microservice in Action
41mThe Promise of Event Driven Programming with Fast Data
26mEvent Driven APIs