A deep-dive conference talk covering advanced optimization techniques for Spring Boot applications running in the cloud. Topics include benchmarking methodology using the Spring Pet Clinic, JVM-level optimizations (JDK version upgrades from 17 to 26, garbage collector tuning, tiered compilation, compact object headers, jlink for custom JVM images, AppCDS and AOT caching via Project Leyden), Spring Boot-specific optimizations (lazy initialization, AOT processing, removing actuators, dependency cleanup), and container image optimizations using Paketo Buildpacks. The talk compares multiple JVM runtimes (OpenJ9/Semeru, GraalVM JIT, Oracle JDK, Azul Zulu) and different base images (Ubuntu, Alpine via BellSoft Liberica) to show their impact on startup time, memory usage, throughput, and image size. Key takeaway: combining multiple simple optimizations yields significant cumulative improvements.
Sort: