5 Reasons to Use Reactive Programming if You Are Not eBay
summary
eBay Classifieds evaluates reactive stacks like Project Reactor and RxJava alongside actor systems such as Akka and finds five practical wins: handle high-latency external calls with Kafka gateways (used at Gumtree UK), scale CPU-bound jobs via ForkJoin executor for faster stream/batch processing (Motor Talk DE), use back-pressure in stream pipelines, implement parallel async conversation + notifications (mobile de), and build an Android BFF; debugging is harder and Tomcat’s default 200-thread model often makes blocking DB paths cheaper.
AI-written summary · about
chapters (16)
- 0:00Introduction
- 3:13Reactive Systems
- 4:29Project Reactor
- 9:18Debugging
- 11:13How expensive is blocking
- 12:25Threads Overhead
- 18:08External call example
- 19:17How fast is Fork Join Executor
- 20:48Stream processing example from Gumtree UK, Motor Talk DE
- 22:52Back-Pressure
- 24:22Conversation Service Example from mobile de
- 26:08Android BFF Example from Gumtree UK
- 27:50Client-side reactivity
- 29:34Use-cases for reactive programming
- 30:46What might be a good candidate for reactivity
- 31:45When I (personally) would NOT use Reactive Programming
related talks
40mReactive programming: lessons learned
41mThe definite guide to reactive programming in Java
49mReactive Programming in Java
52mReactive Programming in Java
01h 03mReactive Enterprise Java
48mReactive for the Impatient
24mBringing Reactive Programming to Java and Microservices
01h 27mReactive Oriented Architecture with Grails