DocumentCode :
2286028
Title :
A Broker-Based Approach to Resource Discovery and Selection in Grid Environments
Author :
Malarvizhi, N. ; Uthariaraj, V. Rhymend
Author_Institution :
Anna Univ., Chennai
fYear :
2008
fDate :
20-22 Dec. 2008
Firstpage :
322
Lastpage :
326
Abstract :
The grid provides mechanisms to share dynamic, heterogeneous, distributed resources spanned across multiple administrative domains. The dynamic nature of resources on computational grids raises the question of an efficient scheduling process, which mainly consists of resource discovery and resource selection. So, appropriate resource discovery and selection mechanism is an important aspect of grid computing. In this paper a novel approach for resource discovery and resource selection is proposed. The approach is based on client agents which act on behalf of grid users and search for resources in a network of resource brokers that are registries for various grid resources. After suitable resources are discovered, client agents determines the willing resources and uses the current load information and memory availability of the resources to determine the optimum set of resources for the tasks in hand. The effectiveness of our algorithm is evaluated through simulation results, which mainly reduces the response time of the tasks.
Keywords :
grid computing; multi-agent systems; scheduling; broker-based approach; client agents; grid computing; multiple administrative domains; resource discovery; resource selection; scheduling process; Availability; Computational modeling; Delay; Distributed computing; Dynamic scheduling; Filtering; Grid computing; Parallel processing; Physics computing; Processor scheduling; grid computing; resource brokers; resource discovery; resource selection; scheduling;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computer and Electrical Engineering, 2008. ICCEE 2008. International Conference on
Conference_Location :
Phuket
Print_ISBN :
978-0-7695-3504-3
Type :
conf
DOI :
10.1109/ICCEE.2008.149
Filename :
4740999
Link To Document :
بازگشت