International Journal of Engineering Science Invention Research & Development; Vol. II Issue IX March 2016 www.ijesird.com e-ISSN: 2349-6185 G.Anuprabavathi, R.Rajmohan, J.Nulyn Punitha, D.Dinagaran and S.G.Sandhya ijesird , Vol. II Issue IX March 2016/623 ENERGY-EFFICIENT AND COST-EFFECTIVE RESOURCE PROVISIONING FRAMEWORK FOR MAP REDUCE WORKLOADS USING DCC ALGORITHM G.Anuprabavathi 1 , R.Rajmohan 2 , J.Nulyn Punitha 3 , D.Dinagaran 4 , S.G.Sandhya 5 Research Scholar 1 , Assistant Professor 2, 3, 4, 5 , Department of Computer Science & Engineering, IFET College of Engineering, Villupuram, India anuprabavathig@gmail.com 1 , rjmohan89@gmail.com 2 , mailnulyn@gmail.com 3 ,ddinagaran@gmail.com 4 , sgsandhyadhas@gmail 5 Abstract: Resource Provisioning is an important and challenging problem in the large-scale distributed systems such as Cloud computing environments. The growing trend in cloud computing is the combination of Big Data and Big Data analytics that has been driven with rapid evolution of data centre technologies towards more cost-effective solutions. Map Reduce profiling is used to automatically create the best cluster configuration for the jobs scheduling. The existing models provide a globally efficient resource allocation scheme that significantly reduces the resource usage cost in the cloud resource optimization for the jobs. But the energy efficiency is at critical concern due to evolve of Map reduce workloads. The proposed Energy Efficient and Cost effective (EECE) resource management framework aim to minimize the infrastructure cost in the data centre and energy conservation for cloud clusters. The proposed system is to reduce the cost for allocating the resources using the virtual clusters globally. The proposed system reduces the cost by eliminating the reconfiguration based approach. The proposed cluster configuration is based on integer partitioning based approach which selects optimal nodes in a dynamic cloud environment to configure a cluster for running Map Reduce jobs. The proposed approach is cost optimized, adheres to global resource utilization and provides high performance to the clients. Experiment results shows that our algorithm performs better than the existing scheduling algorithms in terms of energy utilization and minimal cost. Keywords: Map reduce, Energy efficient, Cost effective, DCC algorithm I. INTRODUCTION The most popular approach is using Map Reduce for big data analytics. Existing per-job services that require VMs to be created a fresh for each submitted job [6], EECE deals with such interactive workloads using a secure instant VM allocation scheme that minimizes the job latency. Existing cloud solutions are largely optimized based on per-job and per- customer optimization which leads to poor resource utilization and higher cost [5]. The first operational model (immediate execution) is a completely customer managed model where each job and its resources are specified by the customer on a per-job basis and the cloud provider [9] only ensures that the requested resources are provisioned upon job arrival. This model has the lowest rewards since there is lack of global optimization across jobs as well as other drawbacks discussed earlier. It is difficult to apply access control enforcement while the workflow is being executed, if the access rights of jobs change dynamically [7]. The leased clusters are under-utilized for a large fraction of the time leading to higher costs. The techniques result in poor utilization lead to higher cost [8]. The proposed Energy Efficient and Cost effective (EECE) resource management framework aim to minimize the infrastructure cost in the data centre and energy conservation for cloud clusters. The proposed system is to reduce the cost for allocating the resources using the virtual clusters globally. The proposed system reduces the cost by eliminating the reconfiguration based approach. The proposed cluster configuration is based on integer partitioning based approach which selects optimal nodes in a dynamic cloud environment to configure a cluster for running Map Reduce jobs. The proposed approach is cost optimized, adheres to global resource utilization and provides high performance to the clients. Experiment results shows that our algorithm performs better than the existing scheduling algorithms in terms of energy utilization and minimal cost.