A load balancing framework for adaptive and asynchronous applications

被引:43
作者
Barker, K [1 ]
Chernikov, A
Chrisochoides, N
Pingali, K
机构
[1] Coll William & Mary, Dept Comp Sci, Williamsburg, VA 23185 USA
[2] Cornell Univ, Dept Comp Sci, Ithaca, NY 14853 USA
基金
美国国家科学基金会;
关键词
dynamic load balancing; adaptive and irregular applications; runtime support software; multithreading; message passing; parallel; distributed; and grid computing; scientific computing; parallel mesh generation;
D O I
10.1109/TPDS.2004.1264800
中图分类号
TP301 [理论、方法];
学科分类号
081202 [计算机软件与理论];
摘要
This paper describes the design of a flexible load balancing framework and runtime software system for supporting the. development of adaptive applications on distributed-memory parallel computers. The runtime system supports a global namespace, transparent object migration, automatic message forwarding and routing, and automatic load balancing. These features can be used at the discretion of the application developer in order to simplify program development and to eliminate complex bookkeeping associated with mobile data objects. An evaluation of this system in the context of a three-dimensional tetrahedral advancing front parallel mesh generator shows that overall runtime improvements of 15 percent compared to common stop-and-repartition load balancing methods, 30 percent compared to explicit intrusive load balancing methods, and 42 percent compared to no load balancing are possible on large processor configurations. At the same time, the overheads attributable to the runtime system are a fraction of 1 percent of the total runtime., The parallel advancing front method is a coarse-grained and highly adaptive application and therefore exercises all of the features of the runtime system.
引用
收藏
页码:183 / 192
页数:10
相关论文
共 38 条
[1]
AMARAL P, 1992, P ECOOP 92 WORKSH W2, P72
[2]
TreadMarks: Shared memory computing on networks of workstations [J].
Amza, C ;
Cox, AL ;
Dwarkadas, S ;
Keleher, P ;
Lu, HH ;
Rajamony, R ;
Yu, WM ;
Zwaenepoel, W .
COMPUTER, 1996, 29 (02) :18-&
[3]
ORCA - A LANGUAGE FOR PARALLEL PROGRAMMING OF DISTRIBUTED SYSTEMS [J].
BAL, HE ;
KAASHOEK, MF ;
TANENBAUM, AS .
IEEE TRANSACTIONS ON SOFTWARE ENGINEERING, 1992, 18 (03) :190-205
[4]
Date Movement and Control Substrate for parallel adaptive applications [J].
Barker, K ;
Chrisochoides, N ;
Dobbelaere, J ;
Nave, D ;
Pingali, K .
CONCURRENCY AND COMPUTATION-PRACTICE & EXPERIENCE, 2002, 14 (02) :77-101
[5]
Dynamic load-balancing of finite element applications with the DRAMA library [J].
Basermann, A ;
Clinckemaillie, J ;
Coupez, T ;
Fingberg, J ;
Digonnet, H ;
Ducloux, R ;
Gratien, JM ;
Hartmann, U ;
Lonsdale, G ;
Maerten, B ;
Roose, D ;
Walshaw, C .
APPLIED MATHEMATICAL MODELLING, 2000, 25 (02) :83-98
[6]
BELGUELIN A, 1993, SUPERCOMPUTING 93 P, P765
[7]
BLUMOFE R, 1995, P 5 S PRINC PRACT PA, P55
[8]
CHANG C, 1996, PARALLEL PROGRAMMING
[9]
CHASE JS, 1989, P 12 ACM S OP SYST P, P147
[10]
CHERNIKOV A, 2002, P 8 INT C NUM GRID G