A Novel Fault-tolerant Particle Swarm Optimization Scheduler for Scheduling Independent Task in Grid Computing Environment

Grid computing allows one to unite pools of servers, storage systems, and networks from different domain with their specific management policies, into a single large system. The Grid Environment is dynamic and its domains act autonomously. Unfortunately, in such an environment failure may occur occa...

Full description

Saved in:
Bibliographic Details
Main Authors: Nikkhah, M., Rahmani, A.M., Yektaie, M.H.
Format: Conference Proceeding
Language:English
Subjects:
Computational Grid
Computer network management
Costs
Delay effects
Fault tolerance
Genetic algorithms
Grid computing
Network servers
Particle swarm optimization
Processor scheduling
Scheduling
Simulated annealing
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Grid computing allows one to unite pools of servers, storage systems, and networks from different domain with their specific management policies, into a single large system. The Grid Environment is dynamic and its domains act autonomously. Unfortunately, in such an environment failure may occur occasionally or a volatile host can delay the entire execution for a long period of time, which in turn can fail taskpsilas execution. In this paper, a Novel Fault-tolerant Particle Swarm Optimization Scheduler (NFPSO) is suggested to schedule independent tasks. This approach aims to generate a scheduling plan to overcome the resource failure problem while it decreases total taskpsilas completion time, cost and the percentage of Unsuccessful scheduled task. The experimental results of NFPSO scheduler are compared with results of Genetic Algorithm, Simulated Annealing, mountain Climbing and a Random scheduler. NFPSO shows better result in cost and success rate criteria than all other, but in completion time criteria GA has better than our proposed algorithm which is better than other.
DOI:10.1109/ICIS.2009.71