ADAPTIVE LOAD BALANCING FOR CLUSTER USING CONTENT AWARENESS WITH TRAFFIC MONITORING

With the rapid growth of both information and userson the Internet, how to effectively improve thequality of network service becomes an urgentproblem to be addressed. Load balancing is a solutionto this problem in an effective way. Differentadaptive load balancing methods have beendeveloped to estimate servers load performance.However, they suffer from either increasedprocessing load on the servers, or additional traffic onthe network and the servers, or are impractical in realtime scenario.In this paper, we used the concept of content basedqueues, and have used RTT passive measurementtechnique to get an adaptive load balancing algorithminside the cluster and used the round robin loadbalancing algorithm outside the cluster. Using thesame queue for each type of request results intooverload on server, so we use the concept of differentqueue for different type of request. The whole loadbalancing task is performed by a webproxy, a proxythat has access to all servers so it also removes thedrawback of dynamic algorithm in which most of theload balancing task is performed by server itself.