Improving Performance of Dynamic Load Balancing among Web Servers by Using Number of Effective Parameters

Web application is being challenged to develop methods and techniques for large data processing at optimum response time. There are technical challenges in dealing with the increasing demand to handle vast traffic on these websites. As number of users' increases, several problems are faced by web servers like bottleneck, delayed response time, load balancing and density of services. The whole traffic cannot reside on a single server and thus there is a fundamental requirement of allocating this huge traffic on multiple load balanced servers. Distributing requests among servers in the web server clusters is the most important means to address such challenge, especially under intense workloads. In this paper, we propose a new request distribution algorithm for load balancing among web server clusters. The Dynamic Load Balancing among web servers take place based on user's request and dynamically estimating server workload using multiple parameters like processing and memory requirement, expected execution time and various time intervals. Our simulation results show that, the proposed method dynamically and efficiently balance the load to scale up the services, calculate average response time, average waiting time and server's throughput on different web servers. At the end of the paper, we presented an experimentation of running proposed system which proves the proposed algorithm is efficient in terms of speed of processing, response time, server utilization and cost efficiency.