Title :
Towards Performance and Power Management of Cloud Servers
Author_Institution :
Dept. of Comput. Eng., Kuwait Univ., Kuwait City, Kuwait
Abstract :
Information and communication technology (ICT) negatively influence the environment due to production of very high CO2 emissions. One server rack produces as much emissions in one year as 25 homes. Usually, a mid-size data center contains a couple of hundred server racks. Therefore, "Green" IT and low power consumption networking infrastructures become more and more important for service and network providers, as well as for manufacturer companies for equipment. Cloud Computing is a novel emerging technology which can increase the utilization and efficiency of hardware equipment, leading to reduction of global CO2 emissions. Cloud Computing has become a necessity in the modern resource allocation and storage technologies. This paper proposes a new methodology of assigning resources based on queuing models taking into consideration of power saving techniques. The basic principle is to add and remove virtual machines based on the proposed queuing model since these virtual machines are acting as servers, but also to check if such adding/removing will affect the power consumption of the overall system. Based on these performance investigations, the system will decide whether to add or remove a virtual machine or just leave the system as it is. Evaluation results show that the proposed system is efficient when allocating resources and at the same time considers power saving effectively.
Keywords :
air pollution; carbon compounds; cloud computing; computer centres; computer networks; energy conservation; energy management systems; green computing; information technology; resource allocation; virtual machines; Green IT; ICT; cloud computing; cloud servers; information and communication technology; low power consumption networking infrastructures; manufacturer companies; mid-size data center; network providers; power management; power saving techniques; queuing model; resource allocation; server rack; storage technologies; virtual machines; Cloud computing; Companies; Hardware; Power demand; Servers; Time factors; Virtual machining; cloud computing; on/off models; performance management; power consumption; queuing models;
Conference_Titel :
Information Technology: New Generations (ITNG), 2014 11th International Conference on
Conference_Location :
Las Vegas, NV
Print_ISBN :
978-1-4799-3187-3
DOI :
10.1109/ITNG.2014.70