A Decision Support Model for Cloud Bursting

The cloud computing market divides into public (commercial) and private (self-provisioned) clouds. The concept of cloud bursting combines public and private clouds: The private cloud (internal resources) provides the computational capacity, but a part of the demand is offloaded onto public clouds. This article proposes an easy-to-apply economic decision support model for determining on the one hand the optimal size of the internal capacity for cloud bursting technology, and on the other hand the cost savings. The model uses an expected value approach that considers stochastic workload and is flexible with respect to the distribution choice. Two empirical examples demonstrate the applicability of the model.