Energy Efficient Cloud Data Center

Abstract

Cloud computing has quickly arrived like a deeply accepted computing model. Still,the exploration and investigation on cloud computing is at a premature phase. Cloud computing is facing distinct issues in the field of security, power consumption, software frameworks, QoS, and standardization.The anagement of efficient energy is one of the most challenging research issues. The key and central services of cloud computing system are the SaaS, PaaS, and IaaS. In this thesis, the model of energy efficient cloud data center is proposed. Cloud data center is the main part of the IaaS layer of a cloud computing system. It absorbs a big part of the aggregate energy of a cloud computing system. Our goal is to supply a better explaining of the design issues of energy manage-ment of the IaaS layer in the cloud computing system. Servers and processors are the main component of the data center. Virtualization technologies that are the key features of the cloud computing environment provide the ability for migration of VMs between physical servers of the cloud data centre to improve the energy efficiency. This is called dynamic server consolidation that has direct impact on service response time. Energy efficient cloud data center reduces the overall energy consumed by the data center. This results in, reduction of cost incurred by the data center, long life of hardware components, green IT environment, and making more user friendly. Many VM placement techniques, server consolidation techniques have been proposed. They do not show optimal solution in every circumstances. They show optimum result only for a certain data set. They did not consider both VM placement and its migration simultaneously. They did not attempt to minimize the VM migrations during server consolidation. Still, forceful consolidation can result in the performance degradation and may lead the SLA negligence. So, there is a trade-off between performance and energy. A number of heuristics, protocols and archi-tectures have explored and investigated for server consolidation using VM migration to reduce energy consumption. The primary objective is to minimize the overall energy consumption by servers without violating the SLA. Our proposed model and scheme show the better result at most of the data set. It is based on virtualization technique, VMs, their placement and their migration. Our study focuses on problems like huge amount of energy consumption by server and processor. So, here energy consumption is reduced without violating SLA and to meet certain level of QoS parameters. Server consolidation is performed with minimum number of VM migration. Here, maximum utilization of re-sources is tried to achieve, but utilization of resources is not compared with the existing scheme. Our scheme may show different better result for different configuration of the data center for the same data set. Problem is formulated as a knapsack problem. Pro-posed scheme inherits some feature from heuristics approach like BF, FF, BFD, and FFD.These are used for greedy-bin-packing problem. For simulation, input data set is taken as random value. These random values are general data set used in real scenario and by the existing scheme. From simulation, it is found that proposed model is achieving the desired objectives for a number of data set, and for another data set, some percentage loss of objectives is occurring.