The Research on Optimization Method of Hadoop MapReduce

Jian-bin Hu, Peng Wang

Abstract


Hadoop MapReduce, with characteristics of large volume, various modes, fast speed and so on, is widely used in large-scale data parallel processing. In recent years, the research on Big data is more and more hot and lots of researchers are interested in optimizing the performance of Hadoop framework from all aspects. This paper firstly introduces the related work of researchers, and then analyzes the execution flow of Hadoop MapReduce framework, including assignment of state transitions and the task of life cycle. It also analyzes the reasons that affect the efficiency of the system and propose a Hadoop strategy of inner workings based on MapReduce Task.


DOI
10.12783/dtetr/apetc2017/11280

Full Text:

PDF

Refbacks

  • There are currently no refbacks.