Evaluation of the Improved EPCLB Algorithm for Distributed Applications

Author

Enokido, Tomoya ; Takizawa, Makoto

Author_Institution

Rissho Univ., Tokyo, Japan

fYear

2011

fDate

Nov. 30 2011-Dec. 2 2011

Firstpage

496

Lastpage

501

Abstract

The extended power consumption laxity-based (EPCLB) algorithm is discussed for reducing the total power consumption of servers in general type applications in this paper. However, each time a load balancer receives a new request, the load balancer has to collect the state on computation and transmission laxities of processes and traffic in each server and estimate the power consumption to perform the request. Since the state of each server might be changed during estimation to perform the request, it is difficult to correctly estimate the power consumption. In this paper, we improve the EPCLB algorithm so that the total power consumption of servers and the communication overhead of a load balancer can be reduced. We evaluate the improved EPCLB (IEPCLB) algorithm compared with the EPCLB and round-robin (RR) algorithms.

Keywords

client-server systems; energy consumption; network servers; power aware computing; resource allocation; telecommunication traffic; communication overhead; distributed applications; extended power consumption laxity-based algorithm; green IT technology; improved EPCLB algorithm; load balancer; server power consumption reduction; server traffic; transmission laxity; Biological system modeling; Computational modeling; Degradation; Load modeling; Power demand; Servers; Digital ecosystems; Distributed systems; EPCLB algorithm; Green IT technology; Power consumption models;

fLanguage

English

Publisher

ieee

Conference_Titel

Intelligent Networking and Collaborative Systems (INCoS), 2011 Third International Conference on

Conference_Location

Fukuoka

Print_ISBN

978-1-4577-1908-0

Type

conf

DOI

10.1109/INCoS.2011.88

Filename

6132857