Cloud Native Java with OpenJ9: Fast, lean and definitely mean
summary
Shows measurements comparing HotSpot JDK8 vs OpenJ9: typical memory use about half for OpenJ9 with similar throughput, CPU often equal or slightly lower. Demonstrates shareclasses cuts startup/footprint ~20%, quickstart gives very fast cold starts for functions at the cost of throughput, and JITaaS prototypes show up to ~40% CPU reduction by centralizing JIT work. Includes Docker examples, JVM flags (Xmx, container tuning, compact/on-idle GC) and runnable demo code on GitHub.
AI-written summary · about
chapters (21)
- 0:00Introduction
- 0:39Why care about OpenJ9
- 6:00OpenJ9 overview
- 7:11History of OpenJ9
- 13:02Repetitive work
- 14:11R class content
- 15:12Share classes
- 15:37Reduce startup
- 15:48Share optimizations
- 16:05Quick start
- 16:38Xtuned for virtualized
- 16:57Container support
- 17:22Idle tuning
- 18:20Configurations
- 18:42Demos
- 19:02Toon
- 23:57Jetts
- 25:15JIT as Service
- 28:43Additional Reading
- 29:02Wrapup
- 30:22Questions
related talks
38mAre you ready for Cloud Native Java?
55mBoost Developer Productivity with Open Cloud Native Java
38mAdopt Open J9 for Spring Boot performance!
42mSimple tweaks to get the most out of your JVM
06mSimple Tweaks to Get the Most Out of your JVM
46mSimple Tweaks to Get the Most Out of Your JVM
31mFast and Distroless Java in Containers: the Recipe!
48mThe Next Generation of Java Runtimes