REACTIVE A New Hope! Learn From a Performance History
summary
Claims that switching Java services to Spring WebFlux with Project Reactor and R2DBC drivers cuts thread usage dramatically — a 10,000 concurrent request bench shows the reactive service holding ~23 threads versus the blocking Tomcat stack that spikes to hardware limits. Shows code changes (Flux/Mono wrappers, reactive repositories), uses Netty, H2 and Spring Data R2DBC, builds a reactive API gateway handling ~5 million requests/day in small 1 CPU 2 GB containers and it works in production.
AI-written summary · about
chapters (10)
- 0:00Reactive programming in Java
- 1:29Monolithic to reactive apps
- 6:10Why Java devs should embrace reactive
- 7:59What makes reactive apps different
- 12:38Blocking to reactive with Spring Boot
- 22:48Performance gains in reactive apps
- 27:03Tech stack behind Project Reactor
- 29:09When to use reactive programming
- 30:11Securing reactive APIs with OAuth 2.0
- 35:20Key learnings and future of reactive Java
related talks
45mReactive web services
40mReactive programming: lessons learned
54mReactive for CRUD: fantasies and reality
15mReactifying your application
46mGet Reactive! Programming, Systems, and Microservices
01h 27mReactive Oriented Architecture with Grails
42mReacting to the Future of Application Architecture
34mReactive Architecture