MAP Reduce for Big Data Applications Based on Improving the Network Traffic Performance.
Loading...
Files
Date
2018-10-17T13:37:53Z
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The Map-Reduce model exploits parallel map tasks and reduces tasks to simplify large-scale
data processing on commodity cluster.
The network traffic generated in the shuffle phase is ignored even if many efforts have
been made to improve the map reduce job's performance. this plays a critical role in
performance enhancement.
Traditionally, the intermediate data is partitioned among the reduce tasks by utilizing hash
function; this is not traffic efficient because the key is associated with network topology and
data size which are not taken into consideration.
By designing a novel intermediate data partition scheme, the network traffic cost for a map
reduce job can be reduced.
Additionally there might be a
Description
Keywords
Divya Bharathi B, 1NZ16MCA72, MAP Reduce for Big Data Applications Based on Improving the Network Traffic Performance., MCA Projects 2018