The Effect of Machine Virtualization on the Environmental Impact of Desktop Environments

Abstract

Through the course of this paper two important and recently prominent ideas in computing science are assessed in terms of their combined potential to provide responsive and environmentally responsible alternatives to traditional desktop computing deployments. The ideas of virtualization and green computing are combined to develop a production model for virtual desktop environments. This virtual model is then studied to determine possible concurrencies when varying desktop operating system and secondary storage input/output performance variables. The potential for concurrency in the production model is based on the use of modern equipment, hypervisor processing hosts, virtual machines and thin clients as end user devices. The virtual model, as a final presentation, is then compared to equal numbers of traditional desktop computers. Concurrency of the model studied is determined by monitoring static resources consumed at the hypervisor as it hosts increasing numbers of virtual machines. Cumulative performance of the hosted virtual machines is measured as well, but separate from the hypervisor. The environmental potential or “greenness” of the model studied is indicated after determining concurrency under each variable pair of operating system and secondary storage. Once concurrency is determined, greenness is measured in terms of a net reduction of CO2 added to the environment due to power generation necessary for processing and cooling. The study is empirical, as it collects real data from instrumented power circuits and benchmark data at the hypervisor to be used in calculations. The model developed is comprehensive, as it ensures all equipment involved in provisioning both environment types is included in data collection. It can therefore make confident statements about the requirements of implementation in terms of resources and electrical power. Throughout the study, conservative methods of calculation and estimation are used to ensure that real world implementations of the model developed are at least as environmental and at least as responsive the results indicated herein.