Modelling Virtual Machine Workload in Heterogeneous Cloud Computing Platforms

Abstract

Cloud computing, as a trend technology, has stemmed from the concept of virtualization. Virtualization makes the resources available to the public to use without any concern for ownership or maintenance cost. In addition, the hosted applications in cloud computing platforms are highly interactive and require intensive resources. The new trend is to duplicate these applications in multiple virtual machines based on demand. Such a scheme needs an efficient resource provisioning to manage the resource assignment to multiple virtual machines properly. One of the issues in the current resource provisioning technique is assigning the resources proactively without predicting the workload of hosted applications, which cause load imbalance and resource wasting. Thus, this paper proposes a new model to predict the application workload. The experimental results show the ability of the proposed model to allocate more virtual machines and to balance the workload among the physical machines.