Abstrato

Decentralized Approach for Balancing Load in Dynamic Cloud Environment

Karthick Smiline Britto.J , PrittoPaul.P

Cloud computing is a domain which may hosts large amount of applications in its servers that are accessed by real time networks. It relies on distributed environment, but the nodes are monitored by central server which is a critical task as large no of systems needs to be monitored. Moreover in cloud, files are created dynamically which adds further load imbalance problem. Even though a load balancing algorithm was proposed to balance the load without the knowledge of the central server it did not take into its account the threshold load on nodes and the weight of the system. In terms of global knowledge is inefficient as the knowledge of the entire network is impossible to predict. As threshold is not clearly mentioned it transfers even in constant load. It also creates additional messages which add more traffic. It increases the time and movement cost. So this approach is also inefficient. Hence the objective is to propose a weight based load balancing algorithm which further rebalances the load to be uniform at each node and assigns the weight to each node considering its configuration. Thus it provides a solution to improve the performance in terms of reducing the imbalance factor and movement cost.