Scheduling divisible workloads on heterogeneous platforms

被引:43
作者
Beaumont, O
Legrand, A
Robert, Y
机构
[1] CNRS, UMR 5800, LaBRI, F-33405 Talence, France
[2] Ecole Normale Super Lyon, CNRS, INRIA, LIP, F-69364 Lyon 07, France
关键词
scheduling; divisible tasks; multi-round algorithms; asymptotical optimality; PROCESSORS; COSTS;
D O I
10.1016/S0167-8191(03)00095-4
中图分类号
TP301 [理论、方法];
学科分类号
081202 ;
摘要
In this paper, we discuss several algorithms for scheduling divisible workloads on heterogeneous systems. Our main contributions are (i) new optimality results for single-round algorithms and (ii) the design of an asymptotically optimal multi-round algorithm. This multi-round algorithm automatically performs resource selection, a difficult task that was previously left to the user. Because it is periodic, it is simpler to implement, and more robust to changes in the speeds of the processors and/or communication links. On the theoretical side, to the best of our knowledge, this is the first published result assessing the absolute performance of a multi-round algorithm. On the practical side, extensive simulations reveal that our multi-round algorithm outperforms existing solutions on a large variety of platforms, especially when the communication-to-computation ratio is not very high (the difficult case). (C) 2003 Elsevier B.V. All rights reserved.
引用
收藏
页码:1121 / 1152
页数:32
相关论文
共 25 条
[1]  
Altilar DT, 2002, LECT NOTES COMPUT SC, V2400, P197
[2]  
[Anonymous], 1999, COMPLEXITY APPROXIMA, DOI DOI 10.1007/978-3-642-58412-1
[3]  
[Anonymous], IEEE INT C MULT COMP
[4]  
[Anonymous], MEAS MOD COMP SYST S
[5]   CLOSED-FORM SOLUTIONS FOR BUS AND TREE NETWORKS OF PROCESSORS LOAD SHARING A DIVISIBLE JOB [J].
BATAINEH, S ;
HSIUNG, TY ;
ROBERTAZZI, TG .
IEEE TRANSACTIONS ON COMPUTERS, 1994, 43 (10) :1184-1196
[6]  
BEAUMONT O, 2002, INT PAR DISTR PROC S
[7]  
Bharadwaj V., 1996, Scheduling Divisible Loads in Parallel and Distributed Systems
[8]   Divisible task scheduling - Concept and verification [J].
Blazewicz, J ;
Drozdowski, M ;
Markiewicz, M .
PARALLEL COMPUTING, 1999, 25 (01) :87-98
[9]  
CASANOVA H, 2002, PARAMETER SWEEPS GRI
[10]  
CASANOVA H, 2001, P IEEE S CLUST COMP