Research Based on Multiple Integrated methods for Prediction on Resource Capacity

This paper constructs a method by combining the multiple prediction model on IT resource capacity. By using the result of this integrated forecasting model based on the historical data, we could forecast our IT resource utilities like the usage of CPU, hard disk, memory in few days or weeks. It could help optimize capacity to manage the limited resource to keep stable and reliable forecasting system. Moreover, this new method could promote the precision of prediction result comparing with the base single model. Therefore, this conclusion could be used in resource capacity management to guide the reasonable allocation of resource expansion and shrinkage.