Back to Search Start Over

Effective Resource Allocation in a JXTA-Based Grid Computing Platform JXTPIA.

Authors :
Yi Pan
Daoxu Chen
Jiannong Cao
Dongarra, Jack J.
Sumitomo, Kenichi
Izaiku, Takato
Saitoh, Yoshihiro
Hui Wang
Minyi Guo
Jie Huang
Source :
Parallel & Distributed Processing & Applications (9783540297697); 2005, p804-815, 12p
Publication Year :
2005

Abstract

To extend a Peer-to-Peer (P2P) network system with the mechanisms of distributed/Grid computing, we developed a flexible JXTA- based P2P network interface and architecture, JXTPIA. The JXTPIA system provides the basic functionalities for Grid computing, such as resources allocation and sharing, task scheduling and assignment, network structure constructing and maintenance, etc. One of the main challenges in developing the JXTPIA system is efficient allocation of resources. We developed and evaluated algorithms for resource allocation to improve the efficiency of the JXTPIA system. The experimental results show that the efficiency of the JXTPIA system differs depending on the adopted algorithms. It indicates that scheduling based on limited information about peers has effects on the performance of the entire system. Though the adopted algorithms are specialized for the JXTPIA system only, the principle of the algorithms can be widely used on similar systems. Keywords: JXTPIA system, resource allocation, Grid computing, network structures. [ABSTRACT FROM AUTHOR]

Details

Language :
English
ISBNs :
9783540297697
Database :
Supplemental Index
Journal :
Parallel & Distributed Processing & Applications (9783540297697)
Publication Type :
Book
Accession number :
32965582
Full Text :
https://doi.org/10.1007/11576235_80