An alternative ranking scheme for computational grid resource selection

With widespread use of Internet and other communication technologies, it has become extremely popular to employ idle processing power of computers, which distributed around the world to achieve high performance computing. Grid computing is one the solutions to compute and process large amount of dat...

Full description

Saved in:

Bibliographic Details
Main Author:	Seraydashti, Mehran Ranjbar
Format:	Thesis
Language:	English
Published:	2013
Subjects:	TK7885-7895 Computer engineer Computer hardware
Online Access:	http://eprints.utm.my/id/eprint/37025/5/MehranRanjbarMFSKSM2013.pdf
Tags:	Add Tag No Tags, Be the first to tag this record!

id	my-utm-ep.37025
record_format	uketd_dc
spelling	my-utm-ep.370252017-07-04T00:31:47Z An alternative ranking scheme for computational grid resource selection 2013-08 Seraydashti, Mehran Ranjbar TK7885-7895 Computer engineer. Computer hardware With widespread use of Internet and other communication technologies, it has become extremely popular to employ idle processing power of computers, which distributed around the world to achieve high performance computing. Grid computing is one the solutions to compute and process large amount of data by using distributed systems. Find suitable resources among numerous resources to execute the submitted job is a critical problem in grid environment. The main idea of proposed algorithm for resource selection is finding reliable resources that can meet job requirements and then based on available resources characteristics and processing power decides about workload of each subtasks. The number and workload of each subtask directly related to the number of available resources and their specifications such as CPU, memory and the status of the network that they are connected. In addition, a ranking mechanism developed to assign a rank to all resources, which join the selection process. It assists reliable resources have more chance to be selected among resources with similar processing performance. Experimental results present that proposed algorithm achieved maximum 95.4% success rate among 1000 submitted job which is 3.3% better than current success rate achieved by Grid Resource Vector (GRV) . The proposed algorithm achieved maximum one second execution time which is showing 2.7s improvement among evaluated methods. In addition, it showed 251.12s completion time among a job with 100K length with 5000 resources where evaluated method achieved 376.45s that shows 125.33s improvement. Proposed algorithm succeeds to increase execution success rate and improve average execution and completion time in comparison of similar algorithms, which shows that alternative ranking scheme is effective for selection reliable resources. 2013-08 Thesis http://eprints.utm.my/id/eprint/37025/ http://eprints.utm.my/id/eprint/37025/5/MehranRanjbarMFSKSM2013.pdf application/pdf en public http://dms.library.utm.my:8080/vital/access/manager/Repository/vital:70093?site_name=Restricted Repository masters Universiti Teknologi Malaysia, Faculty of Computing Faculty of Computing
institution	Universiti Teknologi Malaysia
collection	UTM Institutional Repository
language	English
topic	TK7885-7895 Computer engineer Computer hardware
spellingShingle	TK7885-7895 Computer engineer Computer hardware Seraydashti, Mehran Ranjbar An alternative ranking scheme for computational grid resource selection
description	With widespread use of Internet and other communication technologies, it has become extremely popular to employ idle processing power of computers, which distributed around the world to achieve high performance computing. Grid computing is one the solutions to compute and process large amount of data by using distributed systems. Find suitable resources among numerous resources to execute the submitted job is a critical problem in grid environment. The main idea of proposed algorithm for resource selection is finding reliable resources that can meet job requirements and then based on available resources characteristics and processing power decides about workload of each subtasks. The number and workload of each subtask directly related to the number of available resources and their specifications such as CPU, memory and the status of the network that they are connected. In addition, a ranking mechanism developed to assign a rank to all resources, which join the selection process. It assists reliable resources have more chance to be selected among resources with similar processing performance. Experimental results present that proposed algorithm achieved maximum 95.4% success rate among 1000 submitted job which is 3.3% better than current success rate achieved by Grid Resource Vector (GRV) . The proposed algorithm achieved maximum one second execution time which is showing 2.7s improvement among evaluated methods. In addition, it showed 251.12s completion time among a job with 100K length with 5000 resources where evaluated method achieved 376.45s that shows 125.33s improvement. Proposed algorithm succeeds to increase execution success rate and improve average execution and completion time in comparison of similar algorithms, which shows that alternative ranking scheme is effective for selection reliable resources.
format	Thesis
qualification_level	Master's degree
author	Seraydashti, Mehran Ranjbar
author_facet	Seraydashti, Mehran Ranjbar
author_sort	Seraydashti, Mehran Ranjbar
title	An alternative ranking scheme for computational grid resource selection
title_short	An alternative ranking scheme for computational grid resource selection
title_full	An alternative ranking scheme for computational grid resource selection
title_fullStr	An alternative ranking scheme for computational grid resource selection
title_full_unstemmed	An alternative ranking scheme for computational grid resource selection
title_sort	alternative ranking scheme for computational grid resource selection
granting_institution	Universiti Teknologi Malaysia, Faculty of Computing
granting_department	Faculty of Computing
publishDate	2013
url	http://eprints.utm.my/id/eprint/37025/5/MehranRanjbarMFSKSM2013.pdf
_version_	1747816492739592192

An alternative ranking scheme for computational grid resource selection

Similar Items