Théo, Renaud and Claire, Dubois (2024) JAVA PERFORMANCE TUNING: JVM GARBAGE COLLECTORS, JIT OPTIMIZATIONS, AND PROFILING TOOLS. Journal of Adaptive Learning Technologies, 1 (5). pp. 35-49. ISSN 2997-3902

Text 35-49+JAVA+PERFORMANCE+TUNING+JVM+GARBAGE+COLLECTORS,+JIT+OPTIMIZATIONS,+AND+PROFIL.pdf Download (339kB)

Abstract

In performance-critical Java applications, fine-tuning the Java Virtual Machine (JVM) can mean the difference between scalable success and operational bottlenecks. This article provides a comprehensive exploration of Java performance tuning, with a focus on three pivotal pillars: garbage collection (GC), just-in-time (JIT) compilation optimizations, and modern profiling tools. We begin by demystifying the internal workings of JVM garbage collectors—including G1, ZGC, and Shenandoah—highlighting their trade-offs, suitability for various workloads, and tuning strategies. We then dive into the role of JIT compilers (C1, C2, and Graal), examining how they dynamically optimize bytecode for runtime efficiency through techniques like method inlining, escape analysis, and speculative optimizations. The discussion continues with a practical guide to leveraging profiling and monitoring tools such as Java Flight Recorder (JFR), VisualVM, and async-profiler to diagnose bottlenecks and uncover inefficiencies at scale. Real-world scenarios and actionable tuning methodologies are presented throughout to help developers identify performance hot spots, reduce GC pause times, and fine-tune execution paths. Whether you are optimizing enterprise-grade applications or latency-sensitive systems, this article offers a pragmatic, expert-level roadmap to unlocking the full potential of the JVM.

Item Type:	Article
Subjects:	Q Science > Q Science (General)
Divisions:	Postgraduate > Master's of Islamic Education
Depositing User:	Journal Editor
Date Deposited:	04 Jun 2025 12:06
Last Modified:	04 Jun 2025 12:06
URI:	http://eprints.umsida.ac.id/id/eprint/16178

Actions (login required)

View Item