Everything Everywhere All at Once
summary
Comparing concurrency models on the JVM: the talk examines Thread Per Request (Spring MVC), Reactor (Project Reactor), Kotlin Coroutines, and Java Loom. It explains how each works internally (event loop, continuations), why they are considered asynchronous, differences and similarities, and traces what happens when a Spring Boot REST endpoint is handled under those four scenarios to guide choice for existing and new projects.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 1:20Thread and context basics
- 4:10CP vs IO Bound models
- 7:30Spring MVC vs Reactor
- 10:03Rewriting app for Reactor
- 16:30Switching to Kotlin
- 17:13How coroutines work
- 26:46Coroutines with Spring and JVM
- 27:48Virtual threads in Java 19
- 32:05Performance test scenarios
- 36:49Choosing optimal solution
- 39:00Summary and Q&A
related talks
45mEverything Everywhere All at Once
36mHow Do People Handle Concurrency?
38mJava Project Loom: Why Should I Care?
58mConcurrency in Spring Boot Applications: Making the Right Choice
34mConcurrency in Spring Boot Applications: Making the Right Choice
56mFlavors of Concurrency in Java
46mTour d’horizon de la gestion de la concurrence dans différents langages
50mFlavors of Concurrency in Java