Optimizing Request Routing in Heterogeneous Web Computation Environments
Prudhvi Rao Shedimbi · OhioLink ETD Center (Ohio Library and Information Network) · 2016
With the increase in popularity of web applications, load balancers have become a vital instrument in the day to day operations of almost every web application.They become more critical when underlying servers have heterogeneous capabilities.This thesis focuses on developing an effective load balancing algorithm for a system with heterogeneous server capabilities.We develop an algorithm, Prum, that balances requests based on the servers' latest response time.We then compare the performance of this algorithm with Round Robin, which is the default load balancing function for various load balancers.Detailed analysis of response time and request routing of both the algorithms show that Prum outperforms Round Robin when underlying servers have diverse configurations.