Performance Measurement and Configuration Optimization of Virtual Machines Based on the Bayesian Network

It is significant to accurately measure the performance of virtual machines (VMs) and reasonably allocate resources according to users’ requirements for both users and cloud resource providers in IaaS cloud computing. In this paper, we propose a Bayesian network based model, called PPBN, to describe uncertain relationships among properties and performance of VMs and then measure VM performance in the form of probabilities. Further, we design a linear optimization approach to minimize resource cost and improve host resource utilization at the same time. Experimental results show that our method can measure VM performance accurately and the achieved configuration can meet users’ performance requirements well.