Scaling Java Virtual Machine on a many-core system

Author

Ganesan, K. ; Yao-Min Chen ; Xiaochen Pan

Author_Institution

Oracle Corp., Austin, TX, USA

fYear

2014

fDate

10-12 Dec. 2014

Firstpage

336

Lastpage

339

Abstract

In order to effectively leverage the abundant parallelism provided by large many-core enterprise servers, Java applications, libraries and the virtual machine need to be carefully architected to avoid single-thread bottlenecks. Among the challenges tackled in scaling a single Java Virtual Machine (JVM), we discuss the most rewarding ones including converting shared data objects to per-thread independent objects, applying scalable memory allocators, utilizing appropriate concurrency frameworks, parallelizing garbage collection phases and enhancing data affinity to CPUs through a NUMA aware garbage collector. In this paper, we use SPECjbb2013 to illustrate the importance of various scaling bottlenecks and demonstrate the performance gain from the discussed solutions.

Keywords

Java; multiprocessing programs; multiprocessing systems; storage management; virtual machines; CPU; Java applications; Java virtual machine scaling; NUMA aware garbage collector; SPECjbb2013; concurrency frameworks; data affinity enhancement; garbage collection phase parallelization; many-core enterprise servers; many-core system; per-thread independent objects; performance gain; scalable memory allocators; shared data object conversion; Benchmark testing; Concurrent computing; Hardware; Java; Libraries; Program processors; Scalability;

fLanguage

English

Publisher

ieee

Conference_Titel

Integrated Circuits (ISIC), 2014 14th International Symposium on

Conference_Location

Singapore

Type

conf

DOI

10.1109/ISICIR.2014.7029556

Filename

7029556