Perspectives of data-driven LPV modeling of cloud computing systems

This paper presents an LPV approach to dynamic modeling of a web-server hosted on a private cloud. The cloud hosting web-server is a variable capacity system with two control inputs, first is the number of virtual machines which is indicative of the capacity of the cloud and second is the admission control used for regulating the workload. As the workload and the hosting conditions change frequently, the linear parameter varying (LPV) framework is well suited to derive the model. For the hosted web-server, we obtain a MIMO LPV model with performance metrics such as the response time and the throughput. The identification and validation experiments are performed on the open source Eucalyptus Cloud platform.