Leveraging Mobile Agents for Log Analysis in
Data Center Networks

Zhifeng Xiao¹⁺ and Yun Wang²

¹Penn State Erie, The Behrend College, Erie, Pennsylvania, USA
zux2@psu.edu

²Bradley University, Peoria, Illinois, USA

ywang2@bradley.edu

Abstract

Recent advances have witnessed the success of Data Center Network (DCN). In this paper, we propose a novel approach for distributed log analysis in DCN. Our method leverages Mobile Agents (MAs) to perform log analysis. MAs are able to travel across the network, fulfill their tasks on the host machines, and report to the system operator. Compared to the traditional centralized method, MA-based log analysis takes advantage of the processing ability of distributed machines, offloading log analysis from the central node. We evaluate both centralized method and MA-based log analysis in Emulab to process log files of a cluster of Apache HTTP servers, the result shows that the MAbased method outperforms the centralized method in terms of efficiency, scalability, and bandwidth consumption.

Keywords: Mobile Agent, Log Analysis, Data Center Network

+: Corresponding author: Zhifeng Xiao
Penn State Erie, The behrend College, School of Engineering, 242 Burke Center, 5101 Jordan

Road, Erie, PA 16563, USA, Tel: +1-814-898-6252, Web: http://cs.bd.psu.edu/~zux2/

Journal of Internet Services and Information Security (JISIS), 4(3): 59-71, August 2014 [pdf]