Multiple Virtual Machines Resource Scheduling for Cloud Computing

Abstract

Cloud computing emerges as a new computing paradigm concerned by both academia and industry. Resource management of multiple virtual machines is the core of Infrastructure as a Service. Fo cusing on the CPU resources, the purpose of this paper is to increase the QoS of web service by properly scheduling the CPU resource across the virtual machines. We formulate the CPU scheduling of multiple virtual machines into an integer programming problem. Then, a global regulation algorithm based on utility optimization theory is proposed. Experimental result s show that the regulation syste m of CPUs can significantly improve the performance of Web applications.