MULTI-CORE MAP-REDUCE IMPLEMENTATION FOR EFFICIENT RESOURCE MANAGEMENT
Bathina Narendra Kumar, Banda S. N. V. Ramanamurthy, Rayavarapu Sri Divya · IJITR International Journal of Innovative Technology and Research - IJITR International Journal of Innovative Technology and Research · 2014
Big data analysis is one of the prominent field and its requirement is also increasing due to internet wide spread e-commerce and e-governance.Most of the cases data is unstructured in nature.For example web logs, transaction logs, images, time series data etc.Analyzing such type of data with single system is not possible in general.So Map-Reduce based process is required to work with such big data streams.This paper proposed a solution which is an extension to the existing work proposed in [1] by utilizing modern multi-core functionality for Map-Reduce tasks.We simulate the project using Remote Method Invocation (RMI) using Java.Optimal way of placement of available jobs by reducing the number of jobs in the waiting queue is the major objective of this paper.This can be achieved if the number of jobs requested as well as number of resources they can be utilized are known prior to execution.But, this model will be failed on dynamic demand.For this purpose a novel framework is required to cater dynamic demands.Proposed system considers this objective also.