An Advanced Load Balancing Strategy for Cloud Environment

Abstract

The rapid development of cloud computing, bring great convenience to developers. Recently the resource management of the cloud platform has become a hot research topic, especially the load balancing problem in data center is very important for cloud provider. In this paper, a load balancing framework is proposed for cloud platform, it use the threshold window strategy and an advanced AR prediction model to reduce the migration of VMs. Experiments show that this method can effectively achieve load balancing, promote the utilization of the physical machines, and solve the frequent migration problem caused by high instantaneous peak values significantly.