In cloud computing paradigm, virtual resource autoscaling approaches havebeen intensively studied recent years. Those approaches dynamically scalein/out virtual resources to adjust system performance for saving operationcost. However, designing the autoscaling algorithm for desired performance withlimited budget, while considering the existing capacity of legacy networkequipment, is not a trivial task. In this paper, we propose a Deadline andBudget Constrained Autoscaling (DBCA) algorithm for addressing thebudget-performance tradeoff. We develop an analytical model to quantify thetradeoff and cross-validate the model by extensive simulations. The resultsshow that the DBCA can significantly improve system performance given thebudget upper-bound. In addition, the model provides a quick way to evaluate thebudget-performance tradeoff and system design without wide deployment, savingon cost and time.
展开▼