Affinity-Aware CPU Modeling for Virtual Machines Using KVM
The IUP Journal of Information Technology, Vol. XIV, No. 1, March 2018, pp. 7-15
Posted: 10 Oct 2018
Date Written: March 25, 2018
Abstract
In this era of technological advancements, cloud computing has become a buzzword. It increases the machine potential in terms of computing, using virtualization technology. Virtualization is the core of cloud computing, in which creation of Virtual Machines (VMs) provides scalability and portability by hosting the components of different applications. Since requirements in the cloud environment are dynamic, there is always a need to move VMs within the same cloud or among different clouds. It increases the resource utilization and computational power. As there are more than one VM that can be created on one physical host, internal scheduling is needed so that a host can handle the requirements of all the VMs that are handling different applications. The paper focuses on the concept of processor affinity and discusses its usefulness in improving the performance of Infrastructure-as-a-Service (IaaS) with the help of an experiment. The main purpose of this paper is to analyze the suitability of a VM scheduling technique so that it may be decided whether the given situation is suitable for VM scheduling using processor affinity or not. The paper also compares the performance of VM in different scenarios.
Keywords: Virtualization, Processor Affinity, Virtual Machine (VM), Elasticity
Suggested Citation: Suggested Citation