105 Genetic Algorithm for Energy Efficient Placement of Virtual Machines in MapReduce Based Cloud Environment
-
Published:2012
Download citation file:
In this paper a classification of MapReduce workloads(CPU bound or I/O bound) based on the characteristics of MapReduce jobs is proposed. Also the packing of mixes of CPU and I/O jobs on to nodes is formulated as an optimization problem based on total cost of migration, down time and operational cost in any dynamic MapReduce base cloud computing environment. A genetic algorithm is used to solve this optimization problem since the packing problem is NP-hard. The load balancer framework is implemented by using the proposed genetic algorithm. Experimental results show that proposed model outperforms existing models in terms of minimizing the number of active nodes, reduction in energy consumption and improving pool utilization.