An Efficient Disk I/O Characteristics Collection Method Based on Virtual Machine Technology

Author

Yu-Liang Shen ; Lu Xu

Author_Institution

Inst. of Comput. Technol., Chinese Acad. of Sci., Beijing

fYear

2008

fDate

25-27 Sept. 2008

Firstpage

943

Lastpage

949

Abstract

The disk I/O performance may be the bottleneck in computer system, because of the mechanism characters of hard disk system. In order to tune the system performance effectively, the collection of the disk I/O workload characteristics is the essential step for the performance optimization work. Compared to other I/O characteristics collection method, this paper presents a kind of efficient on-line I/O characteristics analysis method based on Xen 3.0 virtual machine monitor. In virtual machine environment, this method can be transparent to the unmodified operating system. This method can be used to collecting a comprehensive set of essential I/O characteristics metrics. Through the testing, we found that this method has little system overhead and impact on the application system I/O performance. In addition, we demonstrate our I/O characteristics analysis results under the large file copy and the filebench benchmark workload.

Keywords

input-output programs; virtual machines; Xen 3.0 virtual machine monitor; computer system performance tuning; disk I/O performance; disk I/O workload characteristic collection method; filebench benchmark workload; hard disk system; large file copy; operating system; performance optimization; Application software; Computers; Hard disks; High performance computing; Operating systems; Optimization; System performance; System testing; Virtual machine monitors; Virtual machining;

fLanguage

English

Publisher

ieee

Conference_Titel

High Performance Computing and Communications, 2008. HPCC '08. 10th IEEE International Conference on

Conference_Location

Dalian

Print_ISBN

978-0-7695-3352-0

Type

conf

DOI

10.1109/HPCC.2008.92

Filename

4637807