Performance analysis and simulation of an equivalent model of serial queues for computer networks

Authors:
Chung-Ping Chen;Ying-Wen Bai
Affiliations:
Graduate Institute of Applied Science and Engineering, Fu Jen Catholic University, Taiwan, Republic of China and Department of Electronic Engineering, National Taipei University of Technology, Tai ...;Department of Electronic Engineering, Fu Jen Catholic University, Taiwan, Republic of China
Venue:
MOAS'07 Proceedings of the 18th conference on Proceedings of the 18th IASTED International Conference: modelling and simulation
Year:
2007

Citing 8
Cited 0

Computer Networks and Systems: Queueing Theory and Performance Evaluation

Computer Networks and Systems: Queueing Theory and Performance Evaluation
Capacity Planning for Web Services: metrics, models, and methods

Capacity Planning for Web Services: metrics, models, and methods
Web Services: A Technical Introduction

Web Services: A Technical Introduction
Dynamic Load Balancing on Web-Server Systems

IEEE Internet Computing
Analysis of 100Mb/s Ethernet for the Whitney Commodity Computing Testbed

FRONTIERS '99 Proceedings of the The 7th Symposium on the Frontiers of Massively Parallel Computation
The Queuing Network Analysis Tool (QNAT)

MASCOTS '00 Proceedings of the 8th International Symposium on Modeling, Analysis and Simulation of Computer and Telecommunication Systems
Queueing Networks and Markov Chains

Queueing Networks and Markov Chains
An Analysis and Measurement of the Equivalent Model of Serial Queues for a Load Balancer and a Web Server of a Web Cluster with a Low Rejection Rate

ECBS '06 Proceedings of the 13th Annual IEEE International Symposium and Workshop on Engineering of Computer Based Systems

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper we propose a serial queues model to estimate the performance of computer networks with a serial stage. First, based on the relationship between the arrival rate and the service rate, which may be equal or not, we have altogether 13 kinds of working regions. For each region or case we deduce the system average response time from the equivalent serial queues model according to the arrival rate and the service rate. The average response time has something to do with arrival rates, service rates and buffer size and the relation between these and with the working regions. For the verification of the serial queues module of the equivalent models, we use the Queueing Network Analysis Tool (QNAT) to simulate a couple of serial queues and to compare model and simulation. The error between model and simulation is 4.18% or 7.06% depending on whether the arrival rate is equal to the service rate or not. Hence the model of the serial queues can predict the average system response time for computer networks with serial multi-stage switches.