Emerging trends in Cloud Computing and Big Data

This talk looks at the processing requirements of Big Data and challenges in using Map-Reduce paradigm. It looks at some of the bench mark big data analytics problems using Map-Reduce on HDFS. We motivate the use of new paradigm called Generate Map Reduce (GMR) for doing big data analytics in the cloud. GMR supports shared data structure abstraction across map jobs and supports iterative and recursive map-reduce jobs. GMR performs quite well for Big Data Analytics on the cloud and the talk presents the results of using GMR for Big Data Analytics on the cloud.