Maio 2020 vol. 3 num. 1 - XIX Simpósio de Pesquisa Operacional & Logística da Marinha

Artigo Completo - Open Access.

Idioma principal | Segundo idioma

METODOLOGIA DE AVALIAÇÃO OPERACIONAL PARA SISTEMAS DE COMPUTAÇÃO CIENTÍFICA DISTRIBUIDA DE ALTO DESEMPENHO

METODOLOGIA DE AVALIAÇÃO OPERACIONAL PARA SISTEMAS DE COMPUTAÇÃO CIENTÍFICA DISTRIBUIDA DE ALTO DESEMPENHO

Ferro, Mariza ; Schulze, Bruno ; Mury, Antônio R. ;

Artigo Completo:

Scientic Computing typically requires large computational needs which have been addressed with High Performance Distributed Computing. It is essential to eciently deploy a number of complex scientic applications, which have dierent characteristics, and so require distinct computational resources too. However, in many research laboratories, this high performance architecture is not dedicated. So, the architecture must be shared to execute a set of scientic applications, with so many dierent execution times and relative importance to research. Also, the high performance architectures have dierent characteristics and costs. When a new infrastructure has to be acquired to meet the needs of this scenario, the decision-making is hard and complex. In this work, we present a Gain Function as a model of an utility function, with which it is possible a decisionmaking with condence. With the function is possible to evaluate the best architectural option taking into account aspects of applications and architectures, including the executions time, cost of architecture, the relative importance of each application and also the relative importance of performance and cost on the nal evaluation. This paper presents the Gain Function, examples, and a real case showing their applicabilities.

Artigo Completo:

Scientic Computing typically requires large computational needs which have been addressed with High Performance Distributed Computing. It is essential to eciently deploy a number of complex scientic applications, which have dierent characteristics, and so require distinct computational resources too. However, in many research laboratories, this high performance architecture is not dedicated. So, the architecture must be shared to execute a set of scientic applications, with so many dierent execution times and relative importance to research. Also, the high performance architectures have dierent characteristics and costs. When a new infrastructure has to be acquired to meet the needs of this scenario, the decision-making is hard and complex. In this work, we present a Gain Function as a model of an utility function, with which it is possible a decisionmaking with condence. With the function is possible to evaluate the best architectural option taking into account aspects of applications and architectures, including the executions time, cost of architecture, the relative importance of each application and also the relative importance of performance and cost on the nal evaluation. This paper presents the Gain Function, examples, and a real case showing their applicabilities.

Download (PDF)

Palavras-chave: Operational Analysis; High Performance Computing; Scientic Computing; Decision Making.,

Palavras-chave: Operational Analysis; High Performance Computing; Scientic Computing; Decision Making.,

DOI: 10.5151/spolm2019-119

Referências bibliográficas
  • [1] STEVENS, R. T. Operational Test and Evaluation: A Systems Enginnering Process. Malabar, Florida: Robert E. Krieger Publisshing Company, 1986. 2 [2] JAIN, A. K.; MURTY, M. N.; FLYNN, P. J. Data clustering: A review. ACM Comput. Surv., ACM, New York, NY, USA, v. 31, n. 3, p. 264{323, 1999. ISSN 0360-0300. Disponvel em: . 2 [3] DANTAS, M. Computa??o Distribu?da de Alto Desempenho: redes, grids e clusters computacionais. [S.l.]: Axcel Books, 2005. 207{208 p. 3 [4] HEY, T.; TANSLEY, S.; TOLLE, K. (Ed.). The Fourth Paradigm: Data-Intensive Scientic Discovery. Redmond, Washington: Microsoft Research, 2009. Disponvel em: . 3 [5] JAIN, R. The Art of Computer Systems Performance Analysis - Techniques for experimental design, measurement, simulation, and modeling. [S.l.]: Wiley, 199 I-XXVII, 1-685 p. (Wiley Professional Computing). ISBN 978-0-471-50336- 4, 5 [6] COLELLA, P. Dening software requirements for scientic computing. DARPA HPCS presentation. 2004. 4 [7] ASANOVIC, K. et al. A View of the Parallel Computing Landscape. Commun. ACM, ACM, New York, NY, USA, v. 52, n. 10, p. 56{67, oct 2009. ISSN 0001-0782. Disponvel em: . 4 [8] PATTERSON, D. Orgins and Vision of the UC Berkeley Parallel Computing Laboratory. In: . ed. [S.l.]: Microsoft Corporation, 2013. cap. 1, p. 11{42. 4 [9] CHOUDHARY, A. et al. The DAMSEL Project - A Data Model Storage Library for Exascale Science. 2013. Http://science.energy.gov/damsel-11022pdf. 4 [10] JOHN, L. K.; EECKHOUT, L. (Ed.). Performance Evaluation and Benchmarking. [S.l.]: CRC Press, 2006. 5 [11] BUZEN, J. P. Fundamental operational laws of computer system performance. Acta Informatica, Springer Berlin / Heidelberg, v. 7, p. 167{182, 1976. ISSN 0001-5903. 10.1007/BF00265769. Disponvel em: . 5 [12] DENNING, P. J.; BUZEN, J. P. The operational analysis of queueing network models. ACM Comput. Surv., ACM, New York, NY, USA, v. 10, n. 3, p. 225{261, set. 1978. ISSN 0360-0300. Disponvel em: . 5 [13] BUZEN, J. P. Operational Derivations of some Basic Results in Queueing Theory. In: BOKSENBAUM, M. et al. (Ed.). Int. CMG Conference. [S.l.]: Computer Measurement Group, 1982. p. 356{36 5 [14] BUZEN, J. P. Queueing Models: Beyond the Basics. In: MORLEY, D. et al. (Ed.). Twelfth International Computer Measurement Group Conference,Las Vegas, NV, USA, December 8-12, 1986, Proceedings. [S.l.: s.n.], 1986. p. 727{732. 5 [15] KELLY, T. et al. Operational analysis of parallel servers. In: MILLER, E. L.; WILLIAMSON, C. L. (Ed.). MASCOTS. [S.l.]: IEEE Computer Society, 2008. p. 227{236. ISBN 978-1-4244-2818-2. 5 [16] SHEN, K. et al. Operational analysis of processor speed scaling. In: Proceedings of the twentieth annual symposium on Parallelism in algorithms and architectures. New York, NY, USA: ACM, 2008. (SPAA '08), p. 179{18 ISBN 978-1-59593-973-9. Disponvel em: . 5 [17] BUZEN, J. P. Ubiquity symposium: Computation, Uncertainty and Risk. Ubiquity, ACM, New York, NY, USA, v. 2011, jan 201 ISSN 1530-2180. Disponvel em: . 5 [18] FERRO, M. Avaliac~ao de Sistemas de Computac~ao Massivamente Paralela e Distribu da: Uma metodologia voltada aos requisitos das aplicac~oes cientcas. Tese (Tese de Doutorado) | Laboratorio Nacional de Computac~ao Cientca, Petropolis - RJ, Maio 2015. 8, 11 [19] FERRO, M.; MURY, A. R.; SCHULZE, B. Manual de Metodologia de Analise Operacional de Sistemas de Computac~ao Cientca Distribuda de Alto Desempenho. Petropolis - RJ, 2015. Disponvel em: . 14
Como citar:

Ferro, Mariza; Schulze, Bruno; Mury, Antônio R.; "METODOLOGIA DE AVALIAÇÃO OPERACIONAL PARA SISTEMAS DE COMPUTAÇÃO CIENTÍFICA DISTRIBUIDA DE ALTO DESEMPENHO", p. 1629-1643 . In: Anais do XIX Simpósio de Pesquisa Operacional & Logística da Marinha. São Paulo: Blucher, 2020.
ISSN 2175-6295, DOI 10.5151/spolm2019-119

últimos 30 dias | último ano | desde a publicação


downloads


visualizações


indexações