THE USE OF TIME SERIES AS A PREDICTION TOOL FOR PROVISIONING IN CLOUD COMPUTING ENVIRONMENTS

Abstract

Cloud Computing systems needs to ensure the distribution of resources for processing services within acceptable levels of Quality of Service - QoS. One problem that exists in this context is how to use available computational resources avoiding bottlenecks and wastage. This paper aims to determine if there is a statistical method able to predict the amount of resources that should be used for new virtual servers. Thus, we used a time series model named ARIMA (autoregressive moving average) as underlying technique for provisioning of new virtual machines in Cloud Computing environments. The results evidence the feasibility of use of such models in this context.