A probabilistic model to analyse workflow performance on production grids

被引:14
作者
Glatard, Tristan [1 ]
Montagnat, Johan [3 ]
Pennec, Xavier [2 ]
机构
[1] UNSA, CNRS, INRIA I3S, F-06903 Sophia, France
[2] INRIA, F-06903 Sophia, France
[3] CNRS, UNSA, Sophia, France
来源
CCGRID 2008: EIGHTH IEEE INTERNATIONAL SYMPOSIUM ON CLUSTER COMPUTING AND THE GRID, VOLS 1 AND 2, PROCEEDINGS | 2008年
关键词
D O I
10.1109/CCGRID.2008.123
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 [计算机科学与技术];
摘要
Production grids are complex and highly variable systems whose behavior is not well understood and difficult to anticipate. The goal of this study is to estimate the impact of the variability of those infrastructures on the performance of workflow-based applications. A probabilistic model of workflows execution time is proposed and evaluated. Results show that the variability of the EGEE grid infrastructure impacts the execution time of a particular medical image analysis application by a factor 2. The model gives interesting insights on the grid behavior for different application parallelization modes.
引用
收藏
页码:510 / +
页数:2
相关论文
共 19 条
[1]
[Anonymous], 2003, J GRID COMPUT
[2]
[Anonymous], Journal of Grid Computing, DOI [DOI 10.1007/S10723-005-9012-6, 10.1007/s10723-005-9012-6]
[3]
Feitelson DG, 2002, LECT NOTES COMPUT SC, V2459, P114
[4]
Metric and workload effects on computer systems evaluation [J].
Feitelson, DG .
COMPUTER, 2003, 36 (09) :18-+
[5]
GAUTAMA H, 2003, PARALLEL PROCESSING, V13, P513
[6]
GAUTAMA H, 1998, THESIS DELFT U TECHN
[7]
GELENBE E, 1986, INT WORKSH PAR ALG A, P127
[8]
GLATARD T, 2006, LNCS
[9]
GLATARD T, 2007, INT S CLUST COMP GRI, P100
[10]
GLATARD T, 2008, INT J HIGH IN PRESS