Predictable cloud provisioning using analysis of user resource usage patterns in virtualized environment

Abstract

Cloud computing is emerging paradigm based on the virtualization technology which supports easily accessing and integrating heterogeneous computing resources which are dispersed in the different locations. One of challenges which Cloud system has to solve is resource provisioning. In Cloud system, users can run diverse applications and require varying amount of resources. Therefore it is imperative of implementing the execution environment that can guarantee the performance of various applications balancing resource requests and the amount of resources provisioned. In this paper, we propose a prediction-based resource provisioning model with which Cloud system can analyze the resource usage history and predict the needed resource amount in advance before applications start requesting new/additional resources. In our model, we define several resource usage patterns and we employ resource usage history to find out the best-fit usage pattern at the given time window. The best-fit patterns determine whether Cloud system allocates additional resources to guarantee performance or release resources to prevent resource over-provisioning. As a result, our approach successfully predicts the amount of needed resources, and hence reduces the time to prepare the needed resources. In addition, our experiments show our model can utilize resources effectively while providing high level of services. © 2010 Springer-Verlag.