QoS-aware load balancing algorithm for globally distributed web systems

The use of Web servers is growing tremendously, but their performance and reliability haven´t been improved at the same rate. Users of highly popular Web sites may experience long. delays when accessing information. In this paper, we present a two-level distributed Web architecture that uses our QoS-aware load-balancing algorithm to significantly reduce the mean response times by taking into account both stale load information and content-based scheduling when choosing a server for an incoming request for globally distributed web systems. To achieve better end-to-end Quality of Service (QoS), different Classes of Service (CoS) of the customers are considered as one parameter in the load-balancing algorithm.