Operational Scheme for Large Scale Web Server Cluster Systems

Abstract

Web server cluster systems are widely used, where a large number of PC level servers are interconnected via network. This paper focuses on forecasting an appropriate number of web servers which can serve four different classes of user requests, simple web page viewing, knowledge query, motion picture viewing and motion picture uploading. Two ways of serving different classes of web service requests are considered, commonly used web servers and service dedicated web servers. Computer simulation experiments are performed in order to find a good way of allocating web servers among different classes of web service requests, maintaining certain levels of resource utilization and response time.

Optimal Server Allocation to Parallel Queueing Systems by Computer Simulation

“…The web server allocation was optimally performed when the traffic intensities are balanced among the services, called the CETI (Close or Equal Traffic Intensity) rule, which is allocating      servers in a way that              or as close as possible. This paper is an extension and generalization of the result obtained in Park [1] . The CETI rule is not analytically proven to be optimal but can be shown optimal by computation when the arrival process is Poisson and the service times are exponential.…”

“…The problem stems from the allocation of servers in large scale web server cluster systems, where simple web page viewing and knowledge query services are performed separately by a fixed number of web servers [1][2] . With a given number of web severs, the weighted system time of customer's request was examined by computer simulation experiments.…”

“…With a given number of web severs, the weighted system time of customer's request was examined by computer simulation experiments. This paper is the generalization of the previous study for finding a general rule for allocating servers in an optimal way with respect to minimizing the weighted system time [1][2] . The queueing network models can be applied to various types of working environment, but its numerical analysis is restrictive because of its complexity.…”