An In-depth Study of Map Reduce in Cloud Environment

Cloud computing has introduced an utility computing model which offers an alternative to traditional servers and computing clusters. Due to fault tolerance and scalability feature, MapReduce distributed data processing architecture has now become the choice for data-intensive analysis in the clouds. Recently Mapruduce is used as a service or for setting up one's own Mapreduce cluster. In this paper we evaluate the architecture and performance of different MapReduce framework, such as AzureMapreduce built using the Microsoft Azure cloud infrastructure, Cloud MapReduce(CMR) built on top of Amazon cloud OS. We belief that the techniques we discussed in this paper are general enough that would encouraged others to use them in other applications. Our survey would definitely open an promising approach to improve MapReduce performance for Cloud Computing.