Abstract
Typical data centers house several powerful ICT (Information and Communication Technology) equipment such as servers, storage devices and network equipment that are high-energy consuming. The nature of these high-energy consuming equipment is mostly accountable for the very large quantities of emissions which are harmful and unfriendly to the environment. The costs associated with energy consumption in data centers increases as the need for more computational resources increases, so also the appalling effect of CO2 (Carbon IV Oxide) emissions on the environment from the constituent ICT facilities-Servers, Cooling systems, Telecommunication systems, Printers, Local Area Network etc. Energy related costs would traditionally account for about 42% (forty-two per cent) of the total costs of running a typical data center. There is a need to have a good balance between optimization of energy budgets in any data center and fulfillment of the Service Level Agreements (SLAs), as this ensures continuity/profitability of business and customer’s satisfaction. A greener computing from what used to be would not only save/sustain the environment but would also optimize energy and by implication saves costs. This paper addresses the challenges of sustainable (or green computing) in the cloud and proffer appropriate, plausible and possible solutions. The idle and uptime of a node and the traffic on its links (edges) has been a concern for the cloud operators because as the strength and weights of the links to the nodes (data centres) increases more energy are also being consumed by and large. It is hereby proposed that the knowledge of centrality can achieve the aim of energy sustainability and efficiency therefore enabling efficient allocation of energy resources to the right path. Mixed-Mean centrality as a new measure of the importance of a node in a graph is introduced, based on the generalized degree centrality. The mixed-mean centrality reflects not only the strengths (weights) and numbers of edges for degree centrality but it combines these features by also applying the closeness centrality measures while it goes further to include the weights of the nodes in the consideration for centrality measures. We illustrate the benefits of this new measure by applying it to cloud computing, which is typically a complex system. Network structure analysis is important in characterizing such complex systems.
Highlights
Energy-measurements were carried out manually from the nodes and possibly edges of a network and from the analysis of the data collected predictions can be made on future-energy consumption
The degree centrality has as part of its advantage that only local structure round the node could be known, for ease of calculation this is acceptable, but it becomes a concern if a node is central and not accessible to other nodes for one reason or another
On applying the same to a subset of the Freeman EIES (Electronic Information Exchange System) dataset as presented by Opsahl et al [3], the Table 8 was generated: The ranking positions of the mixed-mean weighted centrality above shows the ranking according to the mean of the weightedness of closeness and degree centrality measures at different level of α, it can be inferred that Gary Coombs ranked highest in terms of centralities for both 1 1 2 and 1 2
Summary
Energy-measurements were carried out manually from the nodes and possibly edges of a network and from the analysis of the data collected predictions can be made on future-energy consumption. Trends in the power consumption behavior of a data center and its nodes could be analyzed in such a manner that future behavior of energy-usage could be predicted with good level of accuracy This will go a long way to ensure energy-efficiency and sustainable processes as idle nodes when they are in power-on state still consumes energy and thereby wasting useful and scarce resources. There are three standard measures of centrality namely Degree Centrality, Closeness Centrality and Betweenness Centrality, all formalised by [2] Each of these centralities either concern itself with nodes or edges [4]. This work concerns itself only with the degree and closeness centralities
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
