-
Holly Dail,
Otto Sievert,
Fran Berman,
Henri Casanova,
Asim YarKhan,
Sathish Vadhiyar,
Jack Dongarra,
Chuang Liu,
Lingyun Yang,
Dave Angulo,
and Ian Foster.
Scheduling in the Grid Application Development Software Project.
In Jarek Nabrzyski,
Jennifer Schopf,
and Jan Weglarz, editors,Grid Resource Management.
Kluwer,
2003.
@InCollection{ InCollectionDail.DSBCYVDLYAF_03,
author = {Holly Dail and Otto Sievert and Fran Berman and Henri Casanova and Asim YarKhan and Sathish Vadhiyar and Jack Dongarra and Chuang Liu and Lingyun Yang and Dave Angulo and Ian Foster },
title = {Scheduling in the Grid Application Development Software Project},
booktitle = {Grid Resource Management},
publisher = {Kluwer},
year = {2003},
editor = {Jarek Nabrzyski and Jennifer Schopf and Jan Weglarz}
}
-
Patrick R. Amestoy,
Iain S. Duff,
Jean-Yves L'Excellent,
and Xiaoye S. Li.
Impact of the Implementation of MPI Point-to-Point Communications on the Performance of Two General Sparse Solvers.
Parallel Computing,
29(7):833--847,
2003.
@Article{ ArticleAmestoy.ADLL_PC03,
title = {Impact of the Implementation of {MPI} Point-to-Point Communications on the Performance of Two General Sparse Solvers},
author = {Patrick R. Amestoy and Iain S. Duff and Jean-Yves L'Excellent and Xiaoye S. Li},
journal = {Parallel Computing},
volume = {29},
number = {7},
pages = {833--847},
year = {2003}
}
-
Olivier Beaumont,
Arnaud Legrand,
Loris Marchal,
and Yves Robert.
Scheduling strategies for mixed data and task parallelism on heterogeneous clusters.
Parallel Processing Letters,
13(2):225-244,
2003.
@Article{ ArticleBeaumont.BLMR_PPL03,
author = {Olivier Beaumont and Arnaud Legrand and Loris Marchal and Yves Robert},
journal = {Parallel Processing Letters},
title = {Scheduling strategies for mixed data and task parallelism on heterogeneous clusters},
volume = {13},
number = {2},
pages = {225-244},
year = {2003}
}
-
Olivier Beaumont,
Arnaud Legrand,
and Yves Robert.
The master-slave paradigm with heterogeneous processors.
IEEE Trans. Parallel Distributed Systems,
14(9):897-908,
2003.
@Article{ ArticleBeaumont.BLR_ITPDS03,
author = {Olivier Beaumont and Arnaud Legrand and Yves Robert},
journal = {IEEE Trans. Parallel Distributed Systems},
title = {The master-slave paradigm with heterogeneous processors},
year = {2003},
pages = {897-908},
volume = {14},
number = {9}
}
-
Olivier Beaumont,
Arnaud Legrand,
and Yves Robert.
Scheduling divisible workloads on heterogeneous platforms.
Parallel Computing,
29:1121-1152,
2003.
@Article{ ArticleBeaumont.BLR_PC03,
author = {Olivier Beaumont and Arnaud Legrand and Yves Robert},
journal = {Parallel Computing},
title = {Scheduling divisible workloads on heterogeneous platforms},
year = {2003},
pages = {1121-1152},
volume = {29}
}
-
F. Berman,
R. Wolski,
H. Casanova,
W. Cirne,
H. Dail,
M. Faerman,
S. Figueira,
J. Hayes,
G. Obertelli,
J. Schopf,
G. Shao,
S. Smallen,
N. Spring,
A. Su,
and D. Zagorodnov.
Adaptive Computing on the Grid Using AppLeS.
IEEE Transactions on Parallel and Distributed Systems (TPDS),
14(4):369--382,
April 2003.
@Article{ ArticleBerman.BWCCDFFHOSSSSSZ_ITPDST03,
author = {F. Berman and R. Wolski and H. Casanova and W. Cirne and H. Dail and M. Faerman and S. Figueira and J. Hayes and G. Obertelli and J. Schopf and G. Shao and S. Smallen and N. Spring and A. Su and D. Zagorodnov},
title = {Adaptive Computing on the Grid Using AppLeS},
journal = {IEEE Transactions on Parallel and Distributed Systems (TPDS)},
month = apr,
volume = {14},
number = {4},
pages = {369--382},
year = {2003}
}
-
Eddy Caron,
Frédéric Desprez,
and Frédéric Suter.
Parallel Extension of a Dynamic Performance Forecasting Tool.
Parallel and Distributed Computing Practice (PDCP),
6(1):57-69,
March 2003.
Note: Special Isue: Internet-Based Computing.
@Article{ ArticleCaron.CDS_PDCPP03,
author = {Caron, Eddy and Desprez, Fr{\'e}d{\'e}ric and Suter, Fr{\'e}d{\'e}ric},
journal = {Parallel and Distributed Computing Practice (PDCP)},
month = mar,
note = {Special Isue: Internet-Based Computing},
number = {1},
pages = {57-69},
title = {Parallel Extension of a Dynamic Performance Forecasting Tool},
volume = {6},
year = {2003}
}
-
Holly Dail,
Fran Berman,
and Henri Casanova.
A Decoupled Scheduling Approach for Grid Application Development Environments.
Journal of Parallel and Distributed Computing,
63(5):505--524,
May 2003.
@Article{ ArticleDail.DBC_JPDC03,
author = {Holly Dail and Fran Berman and Henri Casanova},
title = {A Decoupled Scheduling Approach for Grid Application Development Environments},
journal = {Journal of Parallel and Distributed Computing},
month = may,
volume = {63},
number = {5},
pages = {505--524},
year = {2003}
}
-
Abdou Guermouche,
Jean-Yves L'Excellent,
and Gil Utard.
Impact of reordering on the Memory of a Multifrontal Solver.
Parallel Computing,
29(9):1191--1218,
2003.
@Article{ ArticleGuermouche.GLU_PC03,
title = {Impact of reordering on the Memory of a Multifrontal Solver},
author = {Abdou Guermouche and Jean-Yves L'Excellent and Gil Utard},
journal = {Parallel Computing},
year = {2003},
volume = {29},
number = {9},
pages = {1191--1218}
}
-
Frédéric Vivien.
On the Optimality of Feautrier's Scheduling Algorithm.
Concurrency and Computation: Practice and Experience,
15(11-12):1047-1068,
September 2003.
Note: Special issue on Euro-Par 2002.
@Article{ ArticleVivien.V_CCPE03,
author = {Fr\'ed\'eric Vivien},
title = {On the {O}ptimality of {F}eautrier's {S}cheduling {A}lgorithm},
journal = {Concurrency and Computation: Practice and Experience},
volume = 15,
number = {11-12},
pages = {1047-1068},
month = sep,
year = 2003,
note = {Special issue on Euro-Par 2002}
}
-
Olivier Beaumont,
Arnaud Legrand,
and Yves Robert.
Scheduling strategies for mixed data and task parallelism on heterogeneous clusters and grids.
In PDP'2003, 11th Euromicro Workshop on Parallel, Distributed and Network-based Processing,
pages 209-216,
2003.
IEEE Computer Society Press.
@InProceedings{ InProceedingsBeaumont.BLR_03,
author = "Olivier Beaumont and Arnaud Legrand and Yves Robert",
booktitle = "{PDP'2003}, 11th Euromicro Workshop on Parallel, Distributed and Network-based Processing",
publisher = "IEEE Computer Society Press",
title = "Scheduling strategies for mixed data and task parallelism on heterogeneous clusters and grids",
pages = "209-216",
year = "2003"
}
-
Olivier Beaumont,
Arnaud Legrand,
and Yves Robert.
Optimal algorithms for scheduling divisible workloads on heterogeneous systems.
In HCW'2003, the 12th Heterogeneous Computing Workshop,
2003.
IEEE Computer Society Press.
@InProceedings{ InProceedingsBeaumont.BLR_03*1,
author = {Olivier Beaumont and Arnaud Legrand and Yves Robert},
booktitle = {{HCW'2003}, the 12th Heterogeneous Computing Workshop},
publisher = {IEEE Computer Society Press},
title = {Optimal algorithms for scheduling divisible workloads on heterogeneous systems},
year = {2003}
}
-
Anne Benoit,
Leonardo Brenner,
Paulo Fernandes,
and Brigitte Plateau.
Aggregation of Stochastic Automata Networks with replicas.
In A.N. Langville and W.J. Stewart, editors,
4th International Conference on the Numerical Solution of Markov Chains NSMC'03,
Urbana, Illinois, USA,
pages 145--166,
September 2003.
@InProceedings{ InProceedingsBenoit.BBFP_03,
author = {Benoit, Anne and Brenner, Leonardo and Fernandes, Paulo and Plateau, Brigitte},
title = {{Aggregation of Stochastic Automata Networks with replicas}},
booktitle = {4th International Conference on the Numerical Solution of Markov Chains NSMC'03},
editor = {A.N. Langville and W.J. Stewart},
pages = {145--166},
address = {Urbana, Illinois, USA},
year = 2003,
month = sep
}
-
Anne Benoit,
Leonardo Brenner,
Paulo Fernandes,
Brigitte Plateau,
and William J. Stewart.
The PEPS Software Tool.
In P. Kemper and W.H. Sanders, editors,
13th International Conference on Modelling Techniques and Tools for Computer Performance Evaluation TOOLS 2003,
Urbana, Illinois, USA,
pages 98--115,
September 2003.
@InProceedings{ InProceedingsBenoit.BBFPS_03,
author = {Benoit, Anne and Brenner, Leonardo and Fernandes, Paulo and Plateau, Brigitte and Stewart, William J.},
title = {{The PEPS Software Tool}},
booktitle = {13th International Conference on Modelling Techniques and Tools for Computer Performance Evaluation TOOLS 2003},
editor = {P. Kemper and W.H. Sanders},
pages = {98--115},
address = {Urbana, Illinois, USA},
year = 2003,
month = sep
}
-
Anne Benoit,
Brigitte Plateau,
and William J. Stewart.
Memory Efficient Kronecker algorithms with applications to the modelling of parallel systems.
In Proceedings of the International Workshop on Performance Modelling, Evaluation, and Optimization of Parallel and Distributed Systems (PMEO-PDS'03), at the International Parallel and Distributed Processing Symposium (IPDPS'03),
Nice, France,
April 2003.
@InProceedings{ InProceedingsBenoit.BPS_03,
author = {Benoit, Anne and Plateau, Brigitte and Stewart, William J.},
title = {{Memory Efficient Kronecker algorithms with applications to the modelling of parallel systems}},
booktitle = {Proceedings of the International Workshop on Performance Modelling, Evaluation, and Optimization of Parallel and Distributed Systems (PMEO-PDS'03), at the International Parallel and Distributed Processing Symposium (IPDPS'03)},
year = 2003,
month = apr,
address = {Nice, France}
}
-
Vincent Boudet,
Frédéric Desprez,
and Frédéric Suter.
One-Step Algorithm for Mixed Data and Task Parallel Scheduling Without Data Replication.
In Proceedings of the 17th International Parallel and Distributed Processing Symposium (IPDPS'03),
April 2003.
@InProceedings{ InProceedingsBoudet.BDS_03,
author = {Boudet, Vincent and Desprez, Fr\'ed\'eric and Suter, Fr\'ed\'eric},
title = {{One-Step Algorithm for Mixed Data and Task Parallel Scheduling Without Data Replication}},
booktitle = {Proceedings of the 17th International Parallel and Distributed Processing Symposium (IPDPS'03)},
year = 2003,
month = apr,
pdf = {http://dlib.computer.org/conferen/ipdps/1926/pdf/19260041b.pdf}
}
-
Yves Caniou and Emmanuel Jeannot.
New Dynamic Heuristics in the Client-Agent-Server Model.
In IEEE 13th Heteregeneous Computing Workshop - HCW'03,
Nice, France,
April 2003.
Keywords:
time-shared resources,
dynamic scheduling heuristics,
historical trace manager,
perturbation,
mct.
@InProceedings{ InProceedingsCaniou.CJ_03,
author = {Yves Caniou and Emmanuel Jeannot},
booktitle = {IEEE 13th Heteregeneous Computing Workshop - HCW'03},
title = {New Dynamic Heuristics in the Client-Agent-Server Model},
year = {2003},
month = apr,
address = {Nice, France},
keywords = {time-shared resources, dynamic scheduling heuristics, historical trace manager, perturbation, mct}
}
-
Eddy Caron.
Client-Server Approach for Simulation over the GRID.
In Korea-France joint Workshop on GRID computing, editor,
Application and Middleware systems,
December 2003.
Seoul National University.
Keywords:
Grid Computing,
DIET.
@Conference{ Con-Caron.C_03*1,
author = {Caron, Eddy},
booktitle = {Application and Middleware systems},
editor = {Korea-France joint Workshop on GRID computing},
keywords = {Grid Computing, DIET},
month = dec,
publisher = {Seoul National University},
title = {Client-Server Approach for Simulation over the GRID},
year = {2003}
}
-
Eddy Caron,
Frédéric Desprez,
Franck Petit,
and Vincent Villain.
A Hierarchical Resource Reservation Algorithm for Network Enabled Servers.
In IPDPS'03. The 17th International Parallel and Distributed Processing Symposium,
Nice - France,
April 2003.
Keywords:
Grid Computing,
DIET,
Scheduling.
@InProceedings{ InProceedingsCaron.CDPV_03,
address = {Nice - France},
author = {Caron, Eddy and Desprez, Fr{\'e}d{\'e}ric and Petit, Franck and Villain, Vincent},
booktitle = {IPDPS'03. The 17th International Parallel and Distributed Processing Symposium},
keywords = {Grid Computing, DIET, Scheduling},
month = apr,
title = {A Hierarchical Resource Reservation Algorithm for Network Enabled Servers},
year = 2003
}
-
Eddy Caron.
Activities around Client-Server Computing over the Grid.
In French/Japan Workshop on Grid Computing,
March 2003.
CNRS.
Keywords:
Grid Computing,
DIET.
@Conference{ Con-Caron.C_03,
author = {Caron, Eddy},
booktitle = {French/Japan Workshop on Grid Computing},
keywords = {Grid Computing, DIET},
month = mar,
publisher = {CNRS},
title = {Activities around Client-Server Computing over the Grid},
year = {2003}
}
-
Henri Casanova,
Arnaud Legrand,
and Loris Marchal.
Scheduling Distributed Applications: the SimGrid Simulation Framework.
In Proceedings of the third IEEE International Symposium on Cluster Computing and the Grid (CCGrid'03),
May 2003.
@InProceedings{ InProceedingsCasanova.CLM_03,
author = {Henri Casanova and Arnaud Legrand and Loris Marchal},
title = {Scheduling Distributed Applications: the SimGrid Simulation Framework},
booktitle = {Proceedings of the third IEEE International Symposium on Cluster Computing and the Grid (CCGrid'03)},
month = may,
year = {2003}
}
-
Olivier Cozette,
Cyril Randriamaro,
and Gil Utard.
READ$^2$: put disks at network level.
In CCGRID'03, Workshop on Parallel IO,
Tokyo (Japan),
May 2003.
@InProceedings{ InProceedingsCozette.CRU_03,
author = {Olivier Cozette and Cyril Randriamaro and Gil Utard},
title = {{READ}$^2$: put disks at network level},
booktitle = {{CCGRID'03, Workshop on Parallel IO}},
year = 2003,
address = {Tokyo (Japan)},
month = may
}
-
Nicolas Fugier,
Éric Lemoine,
Marc Herbert,
and Bernard Tourancheau.
MPI for the Clint Gb/s interconnect, A hardware/software design experience.
In Recent Advances in Parallel Virtual Machine and Message Passing Interface,
LNCS,
Venice, Italy,
September 2003.
Springer-Verlag.
@InProceedings{ InProceedingsFugier.FLHT_03,
address = {Venice, Italy},
author = {Fugier, Nicolas and Lemoine, \'Eric and Herbert, Marc and Tourancheau, Bernard},
booktitle = {Recent Advances in Parallel Virtual Machine and Message Passing Interface},
month = sep,
publisher = {Springer-Verlag},
series = {LNCS},
title = {{MPI} for the {Clint} {Gb/s} interconnect, {A} hardware/software design experience},
year = {2003}
}
-
Stéphane Genaud,
Arnaud Giersch,
and Frédéric Vivien.
Load-Balancing Scatter Operations for Grid Computing.
In Proceedings of the 12th Heterogeneous Computing Workshop (HCW'2003),
April 2003.
IEEE Computer Society Press.
@InProceedings{ InProceedingsGenaud.GGV_03,
author = {St\'ephane Genaud and Arnaud Giersch and Fr\'ed\'eric Vivien},
title = {Load-Balancing Scatter Operations for Grid Computing},
booktitle = {Proceedings of the 12th Heterogeneous Computing Workshop (HCW'2003)},
year = 2003,
month = apr,
publisher = {IEEE Computer Society Press}
}
-
Arnaud Giersch,
Yves Robert,
and Frédéric Vivien.
Scheduling tasks sharing files on heterogeneous clusters.
In Euro-PVM-MPI-2003: Recent Advances in Parallel Virtual Machine and Message Passing Interface,
LNCS 2840,
pages 657-660,
2003.
Springer Verlag.
@InProceedings{ InProceedingsGiersch.GRV_03,
author = {Arnaud Giersch and Yves Robert and Fr\'ed\'eric Vivien},
booktitle = {{Euro-PVM-MPI-2003}: Recent Advances in Parallel Virtual Machine and Message Passing Interface},
publisher = {Springer Verlag},
series = {LNCS 2840},
title = {Scheduling tasks sharing files on heterogeneous clusters},
pages = {657-660},
year = {2003}
}
-
A. Guermouche.
Impact de l'ordonnancement sur l'occupation mémoire d'un solveur multifrontal parallèle.
In M. Auguin,
F. Baude,
D. Lavenier,
and M. Riveill, editors,
15e Rencontres Francophones en Parallélisme, La Colle sur Loup, France,
pages 37-45,
2003.
@InProceedings{ InProceedingsGuermouche.G_03,
title = {Impact de l'ordonnancement sur l'occupation m\'emoire d'un solveur multifrontal parall\`ele},
author = {A. Guermouche},
pages = {37-45},
editor = {M. Auguin and F. Baude and D. Lavenier and M. Riveill},
booktitle = {15e Rencontres Francophones en Parall\'elisme, La Colle sur Loup, France},
year = 2003
}
-
Abdou Guermouche,
Jean-Yves L'Excellent,
and Gil Utard.
On the memory Usage of a Parallel Multifrontal Solver.
In Proceedings of the 17th International Parallel and Distributed Processing Symposium (IPDPS'03),
2003.
@InProceedings{ InProceedingsGuermouche.GLU_03,
title = {On the memory Usage of a Parallel Multifrontal Solver},
author = {Abdou Guermouche and Jean-Yves L'Excellent and Gil Utard},
booktitle = {Proceedings of the 17th International Parallel and Distributed Processing Symposium (IPDPS'03)},
year = {2003}
}
-
Arnaud Legrand,
Hélène Renard,
Yves Robert,
and Frédéric Vivien.
Load-balancing iterative computations on heterogeneous clusters with shared communication links.
In PPAM-2003: Fifth International Conference on Parallel Processing and Applied Mathematics,
LNCS 3019,
pages 930-937,
2003.
Springer Verlag.
@InProceedings{ InProceedingsLegrand.LRRV_03*1,
author = {Arnaud Legrand and H\'el\`ene Renard and Yves Robert and Fr\'ed\'eric Vivien},
booktitle = {{PPAM-2003}: Fifth International Conference on Parallel Processing and Applied Mathematics},
publisher = {Springer Verlag},
series = {LNCS 3019},
title = {Load-balancing iterative computations on heterogeneous clusters with shared communication links},
pages = {930-937},
year = {2003}
}
-
Arnaud Legrand,
Hélène Renard,
Yves Robert,
and Frédéric Vivien.
Mapping and load-balancing iterative computations on heterogeneous clusters.
In Euro-PVM-MPI-2003: Recent Advances in Parallel Virtual Machine and Message Passing Interface,
LNCS 2840,
pages 586-594,
2003.
Springer Verlag.
@InProceedings{ InProceedingsLegrand.LRRV_03,
author = {Arnaud Legrand and H\'el\`ene Renard and Yves Robert and Fr\'ed\'eric Vivien},
booktitle = {{Euro-PVM-MPI-2003}: Recent Advances in Parallel Virtual Machine and Message Passing Interface},
publisher = {Springer Verlag},
series = {LNCS 2840},
title = {Mapping and load-balancing iterative computations on heterogeneous clusters},
pages = {586-594},
year = {2003}
}
-
V. Portigliatti and L. Philippe.
Java components with constraints and preferences in automatic administration of execution and placement.
In 11th Euromicro Conference on Parallel, Distributed and Network based Processing (Euro PDP 2003),
Genova, Italy,
pages 266--273,
February 2003.
IEEE computer society press.
@InProceedings{ InProceedingsPortigliatti.PP_03,
author = {Portigliatti, V. and Philippe, L.},
title = {Java components with constraints and preferences in automatic administration of execution and placement},
booktitle = {11th Euromicro Conference on Parallel, Distributed and Network based Processing (Euro PDP~2003)},
pages = {266--273},
address = {Genova, Italy},
publisher = {IEEE computer society press},
month = feb,
year = 2003
}
-
Martin Quinson and Antoine Vernois.
Getting to know the grid to use it better.
In Korea-France joint Workshop on GRID computing, editor,
Application and Middleware systems,
December 2003.
Seoul National University.
@Conference{ Con-Quinson.QV_03,
author = {Quinson, Martin and Vernois, Antoine},
title = {Getting to know the grid to use it better},
booktitle = {Application and Middleware systems},
year = {2003},
editor = {Korea-France joint Workshop on GRID computing},
publisher = {Seoul National University},
month = dec
}
-
Hélène Renard,
Yves Robert,
and Frédéric Vivien.
Static load-balancing techniques for iterative computations on heterogeneous clusters.
In Euro-Par-2003: International Conference on Parallel Processing,
LNCS 2790,
pages 148-159,
2003.
Springer Verlag.
Note: Distinguished paper.
@InProceedings{ InProceedingsRenard.RRV_03,
author = {H\'el\`ene Renard and Yves Robert and Fr\'ed\'eric Vivien},
booktitle = {{Euro-Par-2003}: International Conference on Parallel Processing},
publisher = {Springer Verlag},
series = {LNCS 2790},
title = {Static load-balancing techniques for iterative computations on heterogeneous clusters},
pages = {148-159},
year = {2003},
note = {Distinguished paper}
}
-
Hélène Renard.
Placement et équilibrage de charge sur plates-formes hétérogènes.
In 15e Rencontres Francophones du Parallélisme des Architectures et des Systèmes,
La Colle sur Loup, France,
15-17 Octobre 2003.
| Abstract: |
Cet article est consacr\'e \`a la mise en \oe uvre d'algorithmes it\'eratifs sur plate-formes h\'et\'erog\`enes. Les donn\'ees sont r\'eparties sur l'ensemble des processeurs, qui sont organis\'es en anneau virtuel. \`A chaque it\'eration, des calculs ind\'ependants sont effectu\'es en parall\`ele et des communications ont lieu entre les processeurs cons\'ecutifs de l'anneau. Le probl\`eme est de d\'eterminer comment partitionner les donn\'ees et comment les r\'epartir pour que le temps total d'ex\'ecution soit minimal. Une difficult\'e majeure est d'inclure un anneau dans un r\'eseau qui n'est pas forc\'ement un graphe complet, de telle sorte que certains liens de communication soient partag\'es par plusieurs couples de ressources. Nous avons d\'emontr\'e un r\'esultat de complexit\'e qui \'etablit la difficult\'e de ce probl\`eme, et nous proposons une heuristique polynomiale qui fournit des sch\'emas efficaces d'allocation, de routage et de distribution de donn\'ees. |
@InProceedings{ InProceedingsRenard.R_03,
author = {H\'el\`ene Renard},
title = {Placement et \'equilibrage de charge sur plates-formes h\'et\'erog\`enes},
booktitle = {15e Rencontres Francophones du Parall\'elisme des Architectures et des Syst\`emes},
year = {2003},
address = {La Colle sur Loup, France},
month = {15-17 Octobre},
abstract = {Cet article est consacr\'e \`a la mise en \oe uvre d'algorithmes it\'eratifs sur plate-formes h\'et\'erog\`enes. Les donn\'ees sont r\'eparties sur l'ensemble des processeurs, qui sont organis\'es en anneau virtuel. \`A chaque it\'eration, des calculs ind\'ependants sont effectu\'es en parall\`ele et des communications ont lieu entre les processeurs cons\'ecutifs de l'anneau. Le probl\`eme est de d\'eterminer comment partitionner les donn\'ees et comment les r\'epartir pour que le temps total d'ex\'ecution soit minimal. Une difficult\'e majeure est d'inclure un anneau dans un r\'eseau qui n'est pas forc\'ement un graphe complet, de telle sorte que certains liens de communication soient partag\'es par plusieurs couples de ressources. Nous avons d\'emontr\'e un r\'esultat de complexit\'e qui \'etablit la difficult\'e de ce probl\`eme, et nous proposons une heuristique polynomiale qui fournit des sch\'emas efficaces d'allocation, de routage et de distribution de donn\'ees.}
}
-
A. Su,
F. Berman,
and H. Casanova.
Performance Modeling for Entity-Level Simulations.
In Proceedings of Parallel and Distributed Scientific and Engineering Computing Applications (PDSECA 2003),
pages (9 pages),
April 2003.
Note: In conjunction with IPDPS 2003.
@InProceedings{ InProceedingsSu.SBC_03,
author = {A. Su and F. Berman and H. Casanova},
title = {{Performance Modeling for Entity-Level Simulations}},
booktitle = {{Proceedings of Parallel and Distributed Scientific and Engineering Computing Applications (PDSECA 2003)}},
note = {{in conjunction with IPDPS 2003}},
year = 2003,
month = apr,
pages = {(9 pages)}
}
-
Thierry Trochet,
Laurent Comas,
Jean-Marc Nicod,
Oleg Blagosklonov,
Jean-Patrick Daspet,
Josette Verdenet,
Michel Baud,
and Jean-Claude Cardot.
Interface conviviale pour le modèle MCAT appliqué à la correction des artefacts en tomographie cardiaque.
In 41ème Colloque de Médecine Nucléaire de la Langue Française,
volume 27 of M/'edecine nucl/'eaire,
pages 461,
2003.
@InProceedings{ InProceedingsTrochet.TCNBDVBC_03,
author = {Trochet, Thierry and Comas, Laurent and Nicod, Jean-Marc and Blagosklonov, Oleg and Daspet, Jean-Patrick and Verdenet, Josette and Baud, Michel and Cardot, Jean-Claude},
title = {Interface conviviale pour le mod\`ele MCAT appliqué à la correction des artefacts en tomographie cardiaque},
booktitle = {41\`eme Colloque de M\'edecine Nucl\'eaire de la Langue Française},
pages = 461,
year = 2003,
volume = 27,
number = 9,
series = {M/'edecine nucl/'eaire}
}
-
Bora Uçar and Cevdet Aykanat.
Minimizing Communication Cost in Fine-Grain Partitioning of Sparse Matrices.
In Computer and Information Sciences - ISCIS 2003,
volume 2869 of Lecture Notes in Computer Science,
pages 926--933,
2003.
Springer Berlin / Heidelberg.
@InProceedings{ InProceedingsUcar.UA_03,
author = {Bora U\c{c}ar and Cevdet Aykanat},
booktitle = {Computer and Information Sciences - ISCIS 2003},
number = {A. Yazici and C. \c{S}ener},
pages = {926--933},
publisher = {Springer Berlin / Heidelberg},
series = {Lecture Notes in Computer Science},
title = {Minimizing Communication Cost in Fine-Grain Partitioning of Sparse Matrices},
volume = {2869},
year = {2003}
}
-
Antoine Vernois.
Pérennité dans les systèmes de stockage pair à pair.
In M. Auguin,
F. Baude,
D. Lavenier,
and M. Riveill, editors,
15e Rencontres Francophones en Parallélisme, La Colle sur Loup, France,
pages 153-160,
October 2003.
@InProceedings{ InProceedingsVernois.V_03,
author = {Antoine Vernois},
title = {P{\'e}rennit{\'e} dans les syst{\`e}mes de stockage pair {\`a} pair},
booktitle = {15e Rencontres Francophones en Parall{\'e}lisme, La Colle sur Loup, France},
pages = {153-160},
editor = {M. Auguin and F. Baude and D. Lavenier and M. Riveill},
month = oct,
year = {2003}
}
-
Patrick R. Amestoy,
Iain S. Duff,
Jacko Koster,
and Jean-Yves L'Excellent.
MUltifrontal Massively Parallel Solver (MUMPS Version 4.3) Users' guide.
Users' guide,
July 2003.
Note: Available from rlhttp://graal.ens-lyon.fr/MUMPS/doc.html.
@TechReport{ Tec-Amestoy.ADKL_03,
title = {{MU}ltifrontal Massively Parallel Solver ({MUMPS} Version 4.3) Users' guide},
author = {Patrick R. Amestoy and Iain S. Duff and Jacko Koster and Jean-Yves L'Excellent},
year = 2003,
month = jul,
note = {Available from \url{http://graal.ens-lyon.fr/MUMPS/doc.html}},
type = {Users' guide}
}
-
Olivier Beaumont,
Henri Casanova,
Arnaud Legrand,
Yves Robert,
and Yang Yang.
Scheduling Divisible Loads on Star and Tree Networks: Results and Open Problems.
Technical report 2003-41,
LIP,
September 2003.
| Abstract: |
Many applications in scientific and engineering domains are structured as large numbers of independent tasks with low granularity. These applications are thus amenable to straightforward parallelization, typically in master-worker fashion, provided that efficient scheduling strategies are available. Such applications have been called \emph{divisible loads} because a scheduler may \emph{divide} the computation among worker processes arbitrarily, both in terms of number of tasks and of task sizes. Divisible load scheduling has been an active area of research for the last twenty years. A vast literature offers results and scheduling algorithms for various models of the underlying distributed computing platform. Broad surveys are available that report on accomplishments in the field. By contrast, in this paper we propose a unified theoretical perspective that synthesizes previously published results, several novel results, and open questions, in a view to foster novel divisible load scheduling research. Specifically, we discuss both one-round and multi-round algorithms, and we restrict our scope to the popular star and tree network topologies, which we study with both linear and affine cost models for communication and computation. |
@TechReport{ TechReportRR-LIP-2003-41,
author = {Olivier Beaumont and Henri Casanova and Arnaud Legrand and Yves Robert and Yang Yang},
title = {Scheduling Divisible Loads on Star and Tree Networks: Results and Open Problems},
institution = {LIP},
year = {2003},
number = {2003-41},
month = sep,
postscript = {http://www.ens-lyon.fr/LIP/Pub/Rapports/RR/RR2003/RR2003-41.ps.gz},
abstract = {Many applications in scientific and engineering domains are structured as large numbers of independent tasks with low granularity. These applications are thus amenable to straightforward parallelization, typically in master-worker fashion, provided that efficient scheduling strategies are available. Such applications have been called \emph{divisible loads} because a scheduler may \emph{divide} the computation among worker processes arbitrarily, both in terms of number of tasks and of task sizes. Divisible load scheduling has been an active area of research for the last twenty years. A vast literature offers results and scheduling algorithms for various models of the underlying distributed computing platform. Broad surveys are available that report on accomplishments in the field. By contrast, in this paper we propose a unified theoretical perspective that synthesizes previously published results, several novel results, and open questions, in a view to foster novel divisible load scheduling research. Specifically, we discuss both one-round and multi-round algorithms, and we restrict our scope to the popular star and tree network topologies, which we study with both linear and affine cost models for communication and computation.}
}
-
Olivier Beaumont,
Arnaud Legrand,
Loris Marchal,
and Yves Robert.
Optimizing the steady-state throughput of Broadcasts on heterogeneous platforms.
Technical report 2003-34,
LIP,
June 2003.
| Abstract: |
In this paper, we consider the communications involved by the execution of a complex application, deployed on a heterogeneous ``grid'' platform. Such applications extensively use macro-communication schemes, for example to broadcast data items. Rather than aiming at minimizing the execution time of a single broadcast, we focus on the steady-state operation. We assume that there is a large number of messages to be broadcast in pipeline fashion, and we aim at maximizing the throughput, i.e. the (rational) number of messages which can be broadcast every time-step. We target heterogeneous platforms, modeled by a graph where resources have different communication and computation speeds. Achieving the best throughput may well require that the target platform is used in totality: we show that neither spanning trees nor DAGs are as powerful as general graphs. We show how to compute the best throughput using linear programming, and how to exhibit a periodic schedule, first when restricting to a DAG, and then when using a general graph. The polynomial compactness of the description comes from the decomposition of the schedule into several broadcast trees that are used concurrently to reach the best throughput. It is important to point out that a concrete scheduling algorithm based upon the steady-state operation is asymptotically optimal, in the class of all possible schedules (not only periodic solutions). |
@TechReport{ TechReportRR-LIP-2003-34,
author = {Olivier Beaumont and Arnaud Legrand and Loris Marchal and Yves Robert},
title = {Optimizing the steady-state throughput of Broadcasts on heterogeneous platforms},
institution = {LIP},
year = {2003},
number = {2003-34},
month = jun,
postscript = {http://www.ens-lyon.fr/LIP/Pub/Rapports/RR/RR2003/RR2003-34.ps.gz},
abstract = {In this paper, we consider the communications involved by the execution of a complex application, deployed on a heterogeneous ``grid'' platform. Such applications extensively use macro-communication schemes, for example to broadcast data items. Rather than aiming at minimizing the execution time of a single broadcast, we focus on the steady-state operation. We assume that there is a large number of messages to be broadcast in pipeline fashion, and we aim at maximizing the throughput, i.e. the (rational) number of messages which can be broadcast every time-step. We target heterogeneous platforms, modeled by a graph where resources have different communication and computation speeds. Achieving the best throughput may well require that the target platform is used in totality: we show that neither spanning trees nor DAGs are as powerful as general graphs. We show how to compute the best throughput using linear programming, and how to exhibit a periodic schedule, first when restricting to a DAG, and then when using a general graph. The polynomial compactness of the description comes from the decomposition of the schedule into several broadcast trees that are used concurrently to reach the best throughput. It is important to point out that a concrete scheduling algorithm based upon the steady-state operation is asymptotically optimal, in the class of all possible schedules (not only periodic solutions).}
}
-
Olivier Beaumont,
Arnaud Legrand,
Loris Marchal,
and Yves Robert.
Steady-state scheduling of task graphs on heterogeneous computing platforms.
Technical report 2003-29,
LIP,
May 2003.
| Abstract: |
In this paper, we consider the execution of a complex application on a heterogeneous "grid" computing platform. The complex application consists of a suite of identical, independent problems to be solved. In turn, each problem consists of a set of tasks. There are dependences (precedence constraints) between these tasks. A typical example is the repeated execution of the same algorithm on several distinct data samples. We use a non-oriented graph to model the grid platform, where resources have different speeds of computation and communication. We show how to determine the optimal steady-state scheduling strategy for each processor (the fraction of time spent computing and the fraction of time spent communicating with each neighbor) and how to build such as schedule. This result holds for a quite general framework, allowing for cycles and multiple paths in the platform graph. |
@TechReport{ TechReportRR-LIP-2003-29,
author = {Olivier Beaumont and Arnaud Legrand and Loris Marchal and Yves Robert},
title = {Steady-state scheduling of task graphs on heterogeneous computing platforms},
institution = {LIP},
year = {2003},
number = {2003-29},
month = may,
postscript = {http://www.ens-lyon.fr/LIP/Pub/Rapports/RR/RR2003/RR2003-29.ps.gz},
abstract = {In this paper, we consider the execution of a complex application on a heterogeneous "grid" computing platform. The complex application consists of a suite of identical, independent problems to be solved. In turn, each problem consists of a set of tasks. There are dependences (precedence constraints) between these tasks. A typical example is the repeated execution of the same algorithm on several distinct data samples. We use a non-oriented graph to model the grid platform, where resources have different speeds of computation and communication. We show how to determine the optimal steady-state scheduling strategy for each processor (the fraction of time spent computing and the fraction of time spent communicating with each neighbor) and how to build such as schedule. This result holds for a quite general framework, allowing for cycles and multiple paths in the platform graph. }
}
-
Eddy Caron,
Pushpinder Kaur Chouhan,
and Arnaud Legrand.
Automatic Deployment for Hierarchical Network Enabled Server.
Research report 2003-51,
Laboratoire de l'Informatique du Parallélisme (LIP),
November 2003.
Note: Also available as INRIA Research Report RR-5146.
Keywords:
Deployment,
Grid Computing,
Network Enabled Servers,
Steady-state scheduling,
Resource localization and selection,
DIET.
| Abstract: |
This paper focus on the deployment of grid infrastructures, more specifically Problem Solving Environments (PSE) for numerical applications on the grid. Even if the deployment of such an architecture is forced by physical constraints (firewall, access permission, security,...) its efficiency heavily depends on the quality of the mapping between its different components and the grid resources. This paper proposes a new model based on linear programming to estimate the performance of a deployment of a hierarchical PSE. The advantages of the modeling approach in this case are multiple: evaluate a virtual deployment before an actual deployment, provide a decision builder tool (i.e., designed to compare different architectures or buy new resource), take into account the platform scalability. Using this modeling, it is possible to determine the bottleneck of the platform and thus to know whether a given deployment can be improved or not. We illustrate this modeling by applying this results to an existing hierarchical PSE called DIET. |
@TechReport{ TechReportRR-Laboratoire-2003-51,
abstract = {This paper focus on the deployment of grid infrastructures, more specifically Problem Solving Environments (PSE) for numerical applications on the grid. Even if the deployment of such an architecture is forced by physical constraints (firewall, access permission, security,...) its efficiency heavily depends on the quality of the mapping between its different components and the grid resources. This paper proposes a new model based on linear programming to estimate the performance of a deployment of a hierarchical PSE. The advantages of the modeling approach in this case are multiple: evaluate a virtual deployment before an actual deployment, provide a decision builder tool (i.e., designed to compare different architectures or buy new resource), take into account the platform scalability. Using this modeling, it is possible to determine the bottleneck of the platform and thus to know whether a given deployment can be improved or not. We illustrate this modeling by applying this results to an existing hierarchical PSE called DIET. },
author = {Caron, Eddy and Chouhan, Pushpinder Kaur and Legrand, Arnaud},
institution = {Laboratoire de l'Informatique du Parall{\'e}lisme (LIP)},
keywords = {Deployment, Grid Computing, Network Enabled Servers, Steady-state scheduling, Resource localization and selection,DIET},
month = nov,
note = {Also available as INRIA Research Report RR-5146},
number = {2003-51},
postscript = {http://www.ens-lyon.fr/LIP/Pub/Rapports/RR/RR2003/RR2003-51.ps.gz},
title = {Automatic Deployment for Hierarchical Network Enabled Server},
type = {Research report},
year = 2003
}
-
Eddy Caron,
Frédéric Desprez,
Franck Petit,
and Vincent Villain.
A Hierarchical Resource Reservation Algorithm for Network Enabled Servers.
Research report RR-4701,
Institut National de Recherche en Informatique et en Automatique (INRIA),
January 2003.
Note: Also available as LIP Research Report 2003-03 and LaRIA Research Report LaRIA-2003-14.
Keywords:
Grid Computing,
DIET.
@TechReport{ TechReportRR-Institut-RR-4701,
author = {Caron, Eddy and Desprez, Fr{\'e}d{\'e}ric and Petit, Franck and Villain, Vincent},
institution = {Institut National de Recherche en Informatique et en Automatique (INRIA)},
keywords = {Grid Computing, DIET},
month = jan,
note = {Also available as LIP Research Report 2003-03 and LaRIA Research Report LaRIA-2003-14},
number = {RR-4701},
postscript = {ftp://ftp.inria.fr/INRIA/publication/publi-ps-gz/RR/RR-4701.ps.gz},
title = {A Hierarchical Resource Reservation Algorithm for Network Enabled Servers},
type = {Research report},
year = 2003
}
-
Eddy Caron,
Frédéric Desprez,
Franck Petit,
and Vincent Villain.
A Hierarchical Resource Reservation Algorithm for Network Enabled Servers.
Research report 2003-03,
Laboratoire de l'Informatique du Parallélisme (LIP),
January 2003.
Note: Also available as INRIA Research Report RR-4701 and LaRIA Research Report LaRIA-2003-14.
Keywords:
Grid Computing,
DIET,
Scheduling.
@TechReport{ TechReportRR-Laboratoire-2003-03,
author = {Caron, Eddy and Desprez, Fr{\'e}d{\'e}ric and Petit, Franck and Villain, Vincent},
institution = {Laboratoire de l'Informatique du Parall{\'e}lisme (LIP)},
keywords = {Grid Computing, DIET, Scheduling},
month = jan,
note = {Also available as INRIA Research Report RR-4701 and LaRIA Research Report LaRIA-2003-14},
number = {2003-03},
postscript = {ftp://ftp.ens-lyon.fr/pub/LIP/Rapports/RR/RR2003/RR2003-03.ps.gz},
title = {A Hierarchical Resource Reservation Algorithm for Network Enabled Servers},
type = {Research report},
year = 2003
}
-
Eddy Caron,
Frédéric Desprez,
Franck Petit,
and Vincent Villain.
A Hierarchical Resource Reservation Algorithm for Network Enabled Servers.
Research report LaRIA-2003-14,
November 2003.
Note: Also available as INRIA Research Report RR-4701 and LIP Research Report 2003-03.
Keywords:
Grid Computing,
DIET.
@TechReport{ Tec-Caron.CDPV_03,
author = {Caron, Eddy and Desprez, Fr{\'e}d{\'e}ric and Petit, Franck and Villain, Vincent},
keywords = {Grid Computing, DIET},
month = nov,
note = {Also available as INRIA Research Report RR-4701 and LIP Research Report 2003-03},
number = {LaRIA-2003-14},
title = {A Hierarchical Resource Reservation Algorithm for Network Enabled Servers},
type = {Research report},
year = 2003
}
-
Stéphane Genaud,
Arnaud Giersch,
and Frédéric Vivien.
Load-Balancing Scatter Operations for Grid Computing.
Research Report 4770,
INRIA,
March 2003.
Note: Also available as LIP, ENS Lyon, reserach report 2003-17.
@TechReport{ TechReportRR-INRIA-4770,
author = {St\'ephane Genaud and Arnaud Giersch and Fr\'ed\'eric Vivien},
title = {Load-Balancing Scatter Operations for Grid Computing},
institution = {INRIA},
year = 2003,
month = mar,
number = 4770,
note = {Also available as LIP, ENS Lyon, reserach report 2003-17},
type = {Research Report}
}
-
Arnaud Giersch,
Yves Robert,
and Frédérc Vivien.
Scheduling tasks sharing files from distributed repositories.
Research Report 4976,
INRIA,
October 2003.
Note: Also available as LIP, ENS Lyon, research report 2003-49.
@TechReport{ TechReportRR-INRIA-4976,
author = {Arnaud Giersch and Yves Robert and Fr\'ed\'erc Vivien},
title = {Scheduling tasks sharing files from distributed repositories},
institution = {INRIA},
year = 2003,
number = 4976,
month = oct,
note = {Also available as LIP, ENS Lyon, research report 2003-49},
type = {Research Report}
}
-
Arnaud Giersch,
Yves Robert,
and Frédérc Vivien.
Scheduling tasks sharing files on heterogeneous clusters.
Research Report 4819,
INRIA,
May 2003.
Note: Also available as LIP, ENS Lyon, research report 2003-28.
@TechReport{ TechReportRR-INRIA-4819,
author = {Arnaud Giersch and Yves Robert and Fr\'ed\'erc Vivien},
title = {Scheduling tasks sharing files on heterogeneous clusters},
institution = {INRIA},
year = 2003,
number = 4819,
month = may,
note = {Also available as LIP, ENS Lyon, research report 2003-28},
type = {Research Report}
}
-
Abdou Guermouche,
Jean-Yves L'Excellent,
and Gil Utard.
Analysis and Improvements of the Memory Usage of a Multifrontal Solver.
Research report RR-4829,
INRIA,
2003.
Note: Also LIP report 2003-08.
@TechReport{ TechReportRR-INRIA-RR-4829,
title = {Analysis and Improvements of the Memory Usage of a Multifrontal Solver},
author = {Abdou Guermouche and Jean-Yves L'Excellent and Gil Utard},
institution = {INRIA},
number = {RR-4829},
year = {2003},
type = {Research report},
note = "Also LIP report 2003-08"
}
-
Abdou Guermouche,
Jean-Yves L'Excellent,
and Gil Utard.
Analysis and Improvements of the Memory Usage of a Multifrontal Solver.
Research report RR-2003-08,
LIP,
2003.
Note: Also INRIA report RR-4829.
@TechReport{ TechReportRR-LIP-RR-2003-08,
title = {Analysis and Improvements of the Memory Usage of a Multifrontal Solver},
author = {Abdou Guermouche and Jean-Yves L'Excellent and Gil Utard},
institution = {LIP},
number = {RR-2003-08},
year = {2003},
type = {Research report},
note = "Also INRIA report RR-4829"
}
-
Laurent Lefèvre,
Eric Lemoine,
Cong-Duc Pham,
and Bernard Tourancheau.
Fast forwarding with network processors.
Technical report RR-4710,
INRIA,
January 2003.
@TechReport{ TechReportRR-INRIA-RR-4710,
author = {Lef\`evre, Laurent and Lemoine, Eric and Pham, Cong-Duc and Tourancheau, Bernard},
institution = {INRIA},
month = jan,
number = {RR-4710},
title = {Fast forwarding with network processors},
year = {2003}
}
-
Arnaud Legrand,
Loris Marchal,
and Yves Robert.
Optimizing the steady-state throughput of scatter and reduce operations on heterogeneous platforms.
Technical report 2003-33,
LIP,
June 2003.
| Abstract: |
In this paper, we consider the communications involved by the execution of a complex application, deployed on a heterogeneous ``grid'' platform. Such applications intensively use collective macro-communication schemes, such as scatters, personalized all-to-alls or gather/reduce operations. Rather than aiming at minimizing the execution time of a single macro-communication, we focus on the steady-state operation. We assume that there is a large number of macro-communication to perform in pipeline fashion, and we aim at maximizing the throughput, i.e. the (rational) number of macro-communications which can be initiated every time-step. We target heterogeneous platforms, modeled by a graph where resources have different communication and computation speeds. The situation is simpler for series of scatters or personalized all-to-alls than for series of reduces operations, because of the possibility of combining various partial reductions of the local values, and of interleaving computations with communications. In all cases, we show how to determine the optimal throughput, and how to exhibit a concrete periodic schedule that achieves this throughput. |
@TechReport{ TechReportRR-LIP-2003-33,
author = {Arnaud Legrand and Loris Marchal and Yves Robert},
title = {Optimizing the steady-state throughput of scatter and reduce operations on heterogeneous platforms},
institution = {LIP},
year = {2003},
number = {2003-33},
month = jun,
postscript = {http://www.ens-lyon.fr/LIP/Pub/Rapports/RR/RR2003/RR2003-33.ps.gz},
abstract = {In this paper, we consider the communications involved by the execution of a complex application, deployed on a heterogeneous ``grid'' platform. Such applications intensively use collective macro-communication schemes, such as scatters, personalized all-to-alls or gather/reduce operations. Rather than aiming at minimizing the execution time of a single macro-communication, we focus on the steady-state operation. We assume that there is a large number of macro-communication to perform in pipeline fashion, and we aim at maximizing the throughput, i.e. the (rational) number of macro-communications which can be initiated every time-step. We target heterogeneous platforms, modeled by a graph where resources have different communication and computation speeds. The situation is simpler for series of scatters or personalized all-to-alls than for series of reduces operations, because of the possibility of combining various partial reductions of the local values, and of interleaving computations with communications. In all cases, we show how to determine the optimal throughput, and how to exhibit a concrete periodic schedule that achieves this throughput. }
}
-
Arnaud Legrand,
Frédéric Mazoit,
and Martin Quinson.
An Application-Level Network Mapper.
Technical report 2003-09,
LIP,
February 2003.
| Abstract: |
This paper presents a tool to automatically discover the network topology. The goal is to evaluate the performance of concurrent transfers (for example to improve collective communications) and not to discover the physical machines interconnection scheme (for administration purposes). The problems encountered, preliminary algorithms to solve them, as well as theoretical proofs of their validity (under some conditions) are presented. |
@TechReport{ TechReportRR-LIP-2003-09,
author = {Arnaud Legrand and Fr\'ed\'eric Mazoit and Martin Quinson},
title = {An Application-Level Network Mapper},
institution = {LIP},
year = {2003},
number = {2003-09},
month = feb,
postscript = {http://www.ens-lyon.fr/LIP/Pub/Rapports/RR/RR2003/RR2003-09.ps.gz},
abstract = { This paper presents a tool to automatically discover the network topology. The goal is to evaluate the performance of concurrent transfers (for example to improve collective communications) and not to discover the physical machines interconnection scheme (for administration purposes). The problems encountered, preliminary algorithms to solve them, as well as theoretical proofs of their validity (under some conditions) are presented. }
}
-
Arnaud Legrand and Martin Quinson.
Automatic deployment of the Network Weather Service using the Effective Network View.
Research report 2003-42,
LIP, ENS Lyon,
September 2003.
| Abstract: |
The monitoring infrastructure constitutes a key component of any Grid middleware. The Network Weather Service (NWS) is the most commonly used tool to fulfill this need. Unfortunately, users have to deploy the NWS manually, which can be very tedious and error-prone. This paper introduces a method based on the Effective Network View (ENV) network mapper to automatically deploy of NWS using the deployment on our lab's LAN as lead. |
@TechReport{ TechReportRR-LIP-2003-42,
author = {Arnaud Legrand and Martin Quinson},
institution = {LIP, ENS Lyon},
title = {Automatic deployment of the Network Weather Service using the Effective Network View},
number = {2003-42},
month = sep,
type = {Research report},
year = {2003},
abstract = {The monitoring infrastructure constitutes a key component of any Grid middleware. The Network Weather Service (NWS) is the most commonly used tool to fulfill this need. Unfortunately, users have to deploy the NWS manually, which can be very tedious and error-prone. This paper introduces a method based on the Effective Network View (ENV) network mapper to automatically deploy of NWS using the deployment on our lab's LAN as lead.}
}
-
Arnaud Legrand,
Hélène Renard,
Yves Robert,
and Frédérc Vivien.
Load-balancing iterative computations on heterogeneous clusters with shared communication links.
Research Report 4800,
INRIA,
April 2003.
Note: Also available as LIP, ENS Lyon, research report 2003-23.
@TechReport{ TechReportRR-INRIA-4800,
author = {Arnaud Legrand and H\'el\`ene Renard and Yves Robert and Fr\'ed\'erc Vivien},
title = {Load-balancing iterative computations on heterogeneous clusters with shared communication links},
institution = {INRIA},
year = 2003,
number = 4800,
month = apr,
note = {Also available as LIP, ENS Lyon, research report 2003-23},
type = {Research Report}
}
-
Arnaud Legrand,
Hélène Renard,
Yves Robert,
and Frédéric Vivien.
Load-balancing iterative computations in heterogeneous clusters with shared communication links.
Technical report 2003-23,
LIP,
April 2003.
| Abstract: |
This paper is devoted to mapping iterative algorithms onto heterogeneous clusters. The application data is partitioned over the processors, which are arranged along a virtual ring. At each iteration, independent calculations are carried out in parallel, and some communications take place between consecutive processors in the ring. The question is to determine how to slice the application data into chunks, and to assign these chunks to the processors, so that the total execution time is minimized. One major difficulty is to embed a processor ring into a network that typically is not fully connected, so that some communication links have to be shared by several processor pairs. We establish a complexity result that assesses the difficulty of this problem, and we design a practical heuristic that provides efficient mapping, routing, and data distribution schemes. |
@TechReport{ TechReportRR-LIP-2003-23,
author = {Arnaud Legrand and H\'el\`ene Renard and Yves Robert and Fr\'ed\'eric Vivien},
title = {Load-balancing iterative computations in heterogeneous clusters with shared communication links},
institution = {LIP},
year = {2003},
number = {2003-23},
month = apr,
postscript = {http://www.ens-lyon.fr/LIP/Pub/Rapports/RR/RR2003/RR2003-23.ps.gz},
abstract = { This paper is devoted to mapping iterative algorithms onto heterogeneous clusters. The application data is partitioned over the processors, which are arranged along a virtual ring. At each iteration, independent calculations are carried out in parallel, and some communications take place between consecutive processors in the ring. The question is to determine how to slice the application data into chunks, and to assign these chunks to the processors, so that the total execution time is minimized. One major difficulty is to embed a processor ring into a network that typically is not fully connected, so that some communication links have to be shared by several processor pairs. We establish a complexity result that assesses the difficulty of this problem, and we design a practical heuristic that provides efficient mapping, routing, and data distribution schemes.}
}
-
Hélène Renard,
Yves Robert,
and Frédérc Vivien.
Static load-balancing techniques for iterative computations on heterogeneous clusters.
Research Report 4745,
INRIA,
February 2003.
Note: Also available as LIP, ENS Lyon, research report 2003-12.
@TechReport{ TechReportRR-INRIA-4745,
author = {H\'el\`ene Renard and Yves Robert and Fr\'ed\'erc Vivien},
title = {Static load-balancing techniques for iterative computations on heterogeneous clusters},
institution = {INRIA},
year = 2003,
number = 4745,
month = feb,
note = {Also available as LIP, ENS Lyon, research report 2003-12},
type = {Research Report}
}