Online prediction-based dynamic cluster configuration for energy conservation

Dynamic cluster configuration is dynamically adjusting the servers scale based on the network load in order to achieve optimal service performance under the minimum system power consumption. In previous methods, they adopted adjustment methods based on the specific physical experimental model without the description of mathematical models. This paper presents a prediction-based dynamic clusters configuration strategy, it uses least mean square (LMS) algorithm to predict service requests in the future according to the historical information of service requests, and then on the basis of the load requests and the cluster processing power to decide the servers scale and dynamically adjust the opening and shutdown of the computers in the clusters. Simulation verifies the feasibility and superiority of the scheduling strategy.