Performance Modeling of Apache Web Server Affected by Aging

Author

Zhao, Jing ; Trivedi, Kishor S.

Author_Institution

Dept. of Comput. Sci. & Technol., Harbin Eng. Univ., Harbin, China

fYear

2011

fDate

Nov. 29 2011-Dec. 2 2011

Firstpage

56

Lastpage

61

Abstract

A number of studies have reported the phenomenon of "software aging", characterized by progressive software performance degradation. Response time (RT) as a customer-affecting metric can be used to detect the onset of software aging. Alberto Avritzer et al. proposed several software aging detection and rejuvenation algorithms based on the measured variation of RT. In this paper, we extend the result by Ablerto Avritzer et al. by using a validated M/E2/1/K queueing model of Apache web server that is used to help detect aging. Simulation results show the effectiveness of these algorithms and to help set various control parameters.

Keywords

file servers; queueing theory; software metrics; software performance evaluation; Apache Web server; M/E2/1/K queueing model; customer-affecting metric; progressive software performance degradation; rejuvenation algorithms; software aging; Aging; Degradation; Software algorithms; Time factors; Time measurement; Web servers; queueing model; response time distribution; software aging detection; software performance degradation; software rejuvenation algorithm;

fLanguage

English

Publisher

ieee

Conference_Titel

Software Aging and Rejuvenation (WoSAR), 2011 IEEE Third International Workshop on

Conference_Location

Hiroshima

Print_ISBN

978-1-4673-0739-0

Type

conf

DOI

10.1109/WoSAR.2011.13

Filename

6141726