A feasible MapReduce peer-to-peer framework for distributed computing applications
| Main Authors: | Ha Manh Tran, Synh Viet Uyen Ha, Tu Kha Huynh, Son Thanh Le |
|---|---|
| Format: | Journal Book |
| Bahasa: | eng |
| Terbitan: |
Springer Science & Business Media B.V
, 2014
|
| Subjects: | |
| Online Access: |
http://opac.unila.ac.id/ucs/index.php?p=show_detail&id=41484 |
Daftar Isi:
- This article presents a MapReduce Peer-to-Peer(P2P) framework that enables aMapReduce implementationon P2P networks to support a class of MapReduce-basedcomputing applications. This framework can be useful forresearchers who cannot afford expensive and dedicated clustersfor infrequent demands of solving distributed computingproblems. The framework also allows Internet users fromsocial and P2P network communities to perform large dataprocessing experiments on distributed environment. The articledescribes the architecture and prototyping implementationof a MapReduce P2P system. The main features of thissystem include exploiting leisure computing resources onP2P networks efficiently for computation, providingMapReduceoperations using task management for various distributedcomputing problems, and supporting peer failure managementfor improving fault tolerance on Internet environment.We have evaluated and compared theMapReduce P2Pimplementation with a Hadoop MapReduce implementationon local and global-scale networks. The article also includesa discussion of applying the framework to a realistic distributedcase-based reasoning system