Quality-of-service aware resource allocation for virtual machines

Abstract

We propose a quality of service (QoS) aware resource allocation algorithm for virtual machines (VMs) in cloud servers. The proposed scheme consists of two parts: a service profiler that describes the relationship between QoS and resource utilization, and a resource controller that determines an appropriate VM resource configuration. The performance of the proposed VM resource allocation scheme is evaluated on an OpenStack-based testbed. The experimental results show that the proposed scheme successfully identifies the VM with the highest resource demand, and increases the amount of the scarce resources to improve the QoS provided by the VM.