Sang-myeon Park, Young-song Mun, Journal of Internet Computing and Services, Vol. 17, No. 1, pp. 7-14, Feb. 2016
10.7472/jksii.2016.17.1.07, Full Text:  HTML
Keywords: Cloud computing, data center, power consumption, power meter

Abstract

Recently, as cloud computing technologies are developed, it enable to work anytime and anywhere by smart phone and computer. Also, cloud computing technologies are suited to reduce costs of maintaining IT infrastructure and initial investment, so cloud computing has been developed. As demand about cloud computing has risen sharply, problems of power consumption are occurred to maintain the environment of data center. To solve the problem, first of all, power consumption has been measured. Although using power meter to measure power consumption obtain accurate power consumption, extra cost is incurred. Thus, we propose prediction method about power consumption without power meter. To proving accuracy about proposed method, we perform CPU and Hard disk test on cloud computing environment. During the tests, we obtain both predictive value by proposed method and actual value by power meter, and we calculate error rate. As a result, error rate of predictive value and actual value shows about 4.22% in CPU test and about 8.51% in Hard disk test.

Statistics
Cite this article