Performance Evaluation of Clustered Web Servers Using the Round Robin Scheme

Hristoski, Ilija and Dimovski, Tome and Kotevski, Zoran and Markoska, Ramona and Rendevski, Nikola (2018) Performance Evaluation of Clustered Web Servers Using the Round Robin Scheme. Proceedings of The Eighth International Conference on Applied Internet and Information Technologies (ICAIIT 2018). pp. 42-47.

Text AIIT_2018_Proceedings_pp_42-47.pdf Download (383kB)

Abstract

Contemporary e-Commerce systems are constantly facing huge and bursty workloads. Continuous performance evaluation of Web server clusters is a cornerstone of capacity planning methodology, whilst their performance modeling and simulation strive to foresee the system’s behavior under various scenarios. In this paper, we present a simulation model of a generic e-Commerce Web server cluster, consisting of a variable number of Web servers, using the class of Generalized Stochastic Petri Nets (GSPNs). The model also implements one of the basic static load balancing algorithms, the Round Robin algorithm, for equal distribution of incoming HTTP requests across the cluster’s Web servers. We convey a steady-state and transient analysis of the GSPN model by numerical simulations using dedicated software, TimeNET, and present the obtained results. This way, we accomplish a threefold aim, (1) to establish a modeling framework suitable for performance analysis of arbitrary e-Commerce Web server cluster; (2) to get valuable insights into the dynamics and behavior of such systems; (3) to demonstrate the modeling and evaluation power of TimeNET regarding the performance analysis of arbitrary Discrete-Event Dynamic Systems (DEDSs).

Item Type:	Article
Subjects:	Scientific Fields (Frascati) > Natural sciences > Computer and information sciences
Divisions:	Faculty of Economics Faculty of Information and Communication Technologies
Depositing User:	Prof. Dr. Ilija Hristoski
Date Deposited:	08 Mar 2023 09:08
Last Modified:	08 Mar 2023 09:08
URI:	https://eprints.uklo.edu.mk/id/eprint/7772

Actions (login required)

View Item