Abstract | ||
---|---|---|
Tasks running in computational Grids may requiremultiple types of resources simultaneously. Thus, co-allocationof multiple types of resources is neededwhen scheduling tasks. Traditional list schedulingheuristics cannot address the problem of resourceco-allocation in computational Grids. This paperpresents models for computational Grids and paral-leltasks running in computational Grids. Based onthese models, a parallel task scheduling algorithm forresource co-allocation in computational Grids is proposed.The algorithm identifies the critical resource ofparallel tasks and tries to schedule a task as early aspossible when all the resources required by the taskare available. Simulation results show that the algorithmcan gain performance improvement, especiallyin the resource competitive computational Grid. |
Year | DOI | Venue |
---|---|---|
2003 | 10.1109/CLADE.2003.1210004 | CLADE |
Keywords | Field | DocType |
co-allocationof multiple type,early aspossible,neededwhen scheduling task,onthese model,computational grids,paperpresents model,parallel tasks,parallel task scheduling,resource competitive computational,algorithmcan gain performance improvement,resource co-allocation,critical resource ofparallel task,computational modeling,high performance computing,computer science,grid computing,distributed computing,resource management,scheduling algorithm,sun,concurrent computing | Resource management,Grid computing,Supercomputer,Scheduling (computing),Computer science,Parallel computing,Real-time computing,Concurrent computing,Grid,Computational resource,Performance improvement,Distributed computing | Conference |
ISBN | Citations | PageRank |
0-7695-1984-9 | 9 | 0.66 |
References | Authors | |
5 | 5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Lizhe Wang | 1 | 2973 | 191.46 |
Wentong Cai | 2 | 1928 | 197.81 |
Bu-Sung Lee | 3 | 2119 | 140.18 |
Simon See | 4 | 196 | 27.70 |
Wei Jie | 5 | 71 | 12.25 |