International Journal of Computer Science Trends and Technology (IJCST) – Volume 3 Issue 3, May-June 2015 RESEARCH ARTICLE OPEN ACCESS A Survey on the Power and Energy Consumption of Computing Harmanpreet Kaur [1], Jasmeet Singh Gurm [2] Department of Computer Science and Engineering [1] & [2] PTU/RIMT Institute of Engineering and Technology Mandi Gobindgarh Punjab – India

ABSTRACT is dilating regularly over few years in the technology commerce. It is a multitenant habitat pledge the services and resources distracted from IT infrastructure which are served “on demand” and “at scale” in the cloud domain. In this survey paper various components assesses cloud, deployment models and various cloud service models are discussed. Many analysts have proposed discrete techniques to improve energy efficiency and reducing power consumption. This paper represents the survey on existing power consumption and energy efficiency in cloud computing. Keywords:- Cloud Computing, Deployment models, Cloud, Cloud Service Models, Power consumption.

I. INTRODUCTION

Cloud Computing is an appearing paradigm in the Distributed servers is a part of a cloud which present computer industry where the computing is moved to a throughout the hosting various applications. cloud of computers. Cloud Computing is appraised as one of the transpiring arena which incorporates B. Deployment Models technologies, concepts and generates a platform for IT framework and cost-effective business applications. The • Public Cloud embracing of Cloud computing is growing steadily for The resources of public cloud are being obtainable to prior few years in the technology market. Some main the public by the provider itself. Public clouds are those benefits of cloud computing includes: elasticity, self- which are mainly stand-alone clouds and are present off service provisioning and pay per use. premises and run by third party organizations. • Private Cloud A. Component Resources those are only obtainable for limited group Cloud computing system is made of three components of customers. This cloud infrastructure executes in the that are clients, datacenter and distributed servers organization’s physical or it might be third 1) Client: End users communicate with the clouds to party co-location. It is formed generally to provide direct information associated to the cloud. Clients mainly internal services to an organization. Private clouds are categories in three: formulated and organized by an IT segment within an • Mobile: windows mobile smart phone such as organization. blackberry or an I Phone. • Hybrid Cloud • Thin client: These clients do not do computation Hybrid cloud may be described as a cloud which work. Thin clients only show the information. incorporates of both private and public cloud. It may All the work is done by servers for them. This type of considered as private cloud for carrying out day to day clients does not have internal memory. Examples of thin operations and it may be considered as the public cloud clients are Cherry pal, . when need to scale out. • Thick client: Thin client use various browsers such • Community Cloud as internet explorer or Mozilla Firefox or Apple Safari to The cloud infrastructure is shared by several to the different cloud. organizations and supports a specific community that has 2) Datacenter: Datacenter is collection of servers shared concerns. It can be organized by the organizations hosting various applications. End user connects to the or a third party and may exist on premise or off premise. datacenter to subscribe various applications. C. Cloud Service Models 3) Distributed Servers: Distributed servers are those which actively check the services of their hosts. • Software (SaaS)

ISSN: 2347-8578 www.ijcstjournal.org Page 28

International Journal of Computer Science Trends and Technology (IJCST) – Volume 3 Issue 3, May-June 2015 SaaS is a model where applications are managed by a motivating an exponential increase in demand for vendor or service provider and build obtainable to clients energy, and, along with it, it is bearing the related cost across a network, typically the Internet. increases. • (PaaS) In this survey paper, SectionII presents power It is a way for renting hardware, operating system, consumption in cloud computing. SectionIII presents storage and network capacity over the Internet. The literature review of the existing methods for reducing service delivery model allows the client to rent power consumption and to use energy efficiently are virtualized servers and associated services for running discussed.Section1V presents the conclusion. prevailing applications or developing and testing new ones. II. POWER CONSUMPTION IN CLOUD • Infrastructure as a Service (IaaS) COMPUTING IaaS is a model in where an organization outsources the equipment used to support operations, including Modern data centers, operating under the , servers and hardware and networking computing model, are hosting a various applications components. The service provider owns the equipment ranging from those that run for a few seconds (e.g. and service provider is also responsible for housing, serving requests of web applications such as e-commerce running and conserving it. The client pays on a per-use and social networks portals) to those that run for longer basis. periods of time (e.g. simulations or large dataset processing). Cloud Data Centers consume excessive amount of energy. It is accountable for global increase in energy consumption, and energy cost additionally as a

proportion of IT costs. Now days the incipient software SaaS which are being used are devouring more and more power per year. Some of them require virtually steady

Google Docs, access to the hard drive which drains power more rapidly MobileMe, Zoho than precedent software did. POWER AND ENERGY MANAGEMENT FOR SERVER SYSTEM PaaS Power and energy consumption are key concerns for data

Microsoft Azure, centers. These centers abode thousands of server and App Engine support infrastructures for cooling as well. Researchers have now made resultant tread in making their effort to conserve energy in servers because they have been given IaaS these benefits, By calculating the greatest power utilizing HP’s power calculator the power consumption for each Amazon EC2, S3 etc., Rack space server can be found. Then we can follow the convention Mosso Offering, Suns Cloud Services, which average power usage either for midrange or for Terremark Cloud Offering high-end servers which is around 66% of the utmost potency. Hard disk arrays comprise fortifying the functions like cache recollections, disk array controllers,

disk enclosures and redundant power supplies. When we Fig. 1. Examples of Cloud Service Models. verbalize about cloud computing data centers the storage spaces which have in the data center is consolidated and D. Green Cloud Computing hard disk utilization is centrally harmonized. Multiple Green computing is an art in which efficient number of users can share a single server through server utilization of computing resources is advocated. The virtualization, which ultimately increases resource main goal of green computing is to promote the utilization and in turn reduces the total number of utilization of eco cordial products which are facilely server’s desideratum. Users do not need to aware the recycled and reused. It is of utmost paramount for operations being performed by other users and can Information Technology sector due to the sprawling and facilely utilize the server cerebrating themselves to be extensive utilization of IT accommodations these days. It the only utilize on that server. Wherein some servers will be a revolution if it is laid efficaciously and used enter into a sleep mode, when they are not in demand, efficiently. It is not only confined to research studies but which ultimately reduces energy consumption. withal to be adopted by organizations depending largely  Jonathan Koomey, suggested his viewpoint upon IT products. Green computing is utilizing after his research that the cloud is computing resources proficiently. . The goals are responsible for 2% of the world’s electricity analogous to green chemistry; reducing the utilization of used .A report by the CDP has found that a harmful materials, maximizing energy efficiency during company that accepts cloud computing can the product's lifetime. The IT function of business is

ISSN: 2347-8578 www.ijcstjournal.org Page 29

International Journal of Computer Science Trends and Technology (IJCST) – Volume 3 Issue 3, May-June 2015 reduce its energy consumption and carbon energy efficiency. emissions and additionally can reduce its Introduced an outline principal outlay on their resources hence of use of energy in cloud computing and ameliorating efficiency of operations being contrast it with performed. consumption of energy in the  Prioritization of energy-savings in lieu of traditional operating on a more diminutive solar PV computing. system utilizing energy vigilant cloud H. 2013 Novel resource Lessen the total cost computing lets us manage the computing Goudarzi allocation of energy of cloud load even in extreme weather conditions by et.al [2] algorithms to computing performance and administration. shutting off the activities which are not energy . essential. This will lead to Energy Cognizant effiiciency in Efficient heuristic Cloud Computing. cloud computing algorithm used to resolve the quandary  Cloud Computing Futures (CCF) is one of based on dynamic the organizational divisions of programming and Research which focuses on the reduction in convex optimization. the costs of the operation of the data centers. Virtualization mechanics assist in It also increases the adaptability to failure. improving the power efficiency of  Power Usage Effectiveness (PUE) is the datacenters metric that is used to calculate the energy consolidation. efficiency of a data center. D. 2010 GreenCloud: a Simulation is Kliazovich packet-level efficiency of the  Workload diversification: Since many et.al [4] simulator of simulator acquired diverse users will avail themselves of energy-aware for two tiers and different resources which are there on cloud three tiers with high – different applications and different usage speed data centers in employing the volumes different feature preferences set – scheme of power this will get better hardware use and management applied consequently make worthy usage of power to networking and that is used to keep a server up and computing running. components with the help of simulator.  Power-management flexibility: Virtual A. Horri 2014 Energy-Efficient Proposed algorithm servers are not difficult to manage though et.al [3] Cloud lessens the number Computing of VM migration, physical servers are if we think from a SLAV and total data power perspective. The load can transmitted in automatically be deployed somewhere else contrast to current if hardware failure occurs. Likewise algorithm. theoretically all virtual loads could be Y. Tian 2014 Managing DVFS lowers relocated to certain servers when they are et.al [6] performance and processor voltage power and frequency while under loaded or power-down and idle. consumption the load is light, tradeoff rates of feasible  Usage of solar PV array: Solar service are distinct Photovoltaic (PV) is a process that and bounded and transforms sunlight into electricity. Solar PV 50% power cost can modules can be merge together as parallel be saved compared connected modules and an array of series to to the servers without DVFS. facilitate any level of power requirements Without DVFS, from watts to kilowatt and megawatt size. service rate is a constant. III. LITERATURE SURVEY Verma 2011 Power aware Worked upon live

et.al [9] scheduling of migration to move Author’s Year Topic Explanation virtual machines VMs to their new in DVSF frame but did not J. Baliga 2011 SLA-based Emphasize on the enabled cluster work upon SLA. et.al [1] Optimization of managing in the data Static VM Power and centers the consolidation Migration Cost utilization of power focused on Semi in Cloud has prompted static VM Computing various remarkable consolidation. improvement of

ISSN: 2347-8578 www.ijcstjournal.org Page 30

International Journal of Computer Science Trends and Technology (IJCST) – Volume 3 Issue 3, May-June 2015 [3] Abbas Horri, Mohammad Sadegh Mozafari, Gholamhossein Dastghaibyfard·, "Novel IV. CONCLUSION resource allocation algorithms to performance Cloud computing have several benefits over and energy effiiciency in cloud computing," traditional (non-cloud) environment and have capability Springer Science+Business Media New York, to handle most immediate, temporary peaks in 24 june 2014.. application demand on cloud infrastructures. Increasing [4] Dzmitry Kliazovich, Pascal Bouvry, Samee power consumption by cloud computing is one of the Ullah Khan,"GreenCloud: a packet-level largest issues. Various techniques and algorithms are simulator of energy-aware cloud computing used to solve the problem. In this paper we survey data centers," Springer Science+Business Media, various existing energy efficient techniques in cloud LLC 2010, 9 November 2010. environment. In future large number of parameters and [5] Erol Gelenbe, Marco Di Girolama, Giovanni different types of soft computing techniques can be Giuliani, Andreas Beri, "Energy-Efficient Cloud included for better energy efficiency. Computing," Oxford University Press on behalf of The British Computer Society, 28 july 2009 ACKNOWLEDGMENT [6] Yuan Tian, Chuang Lin, Keqin Li, "Managing performance and power consumption tradeoff," This research paper is made possible through the Springer Science+Business Media New York constant encouragement and helps from my parents and 2014, 26 March 2014. guide. I am extremely thankful and indebted to them for [7] Gregor von Laszewski "Power aware sharing expertise, and sincere and valuable guidance scheduling of virtual machines in DVSF extended to me. enabled cluster," Springer Science+ Business REFERENCES Media New York 2014, 24 June 2014. [8] Mohsen Sharifi, Hadi Salimi, Mahsa [1] Jayant Baliga, Robert W. A. Ayre, Kerry Hinton, Najafzadeh, "Power-efficient distributed andRodney S. Tucker, “Green Cloud scheduling of virtual machinesusing workload- Computing:Balancing:Energy in aware consolidation techniques," Springer Processing,Storage, and Transport”,IEEE Vol. Science+Business Media LLC(2011), 2011. 99, No. 1, January 2011 [9] Akshat Verma ,Gargi Dasgupta, Tapan Kumar [2] Hadi Goudarzi, Mohammad Ghasemazar, and Nayak ,Pradipta De ,Ravi Kothari, "Server Massoud Pedram, "SLA-based Optimization of workload analysis for power minimization using Power and Migration Cost in Cloud consolidation," in USENIX annual technical Computing," supported in part by a grant from conference, San Diego, USA, 2009. National Science, p. NA, NA..

ISSN: 2347-8578 www.ijcstjournal.org Page 31