Implementação e Avaliação Preliminar de um novo Sistema Software DSM para Cluster de Computadores
Resumo
Neste trabalho apresentamos resultados preliminares da implementação e de desempenho de Clik, um novo sistema software DSM (SDSM) multithread para clusters de computadores. Em contraste com sistemas SDSM tradicionais que suportam a programação SPMD, Clik implementa o modelo de programação multithread proposto pela linguagem Cilk, o qual pemlite explorar paralelismo assíncrono e dinâmico freqüentemente encontrado em aplicações paralelas, mas que o modelo SPMD não consegue expressar de maneira eficiente. Como resultado, além de oferecer um sistema SDSM multithread, Clik também introduz um mecanismo eficiente e transparente de distribuição de tarefas entre os nós de um cluster, facilitando o balanceamento de carga. Nossos resultados preliminares de desempenho de Clik para três aplicações paralelas executadas em um cluster de 8 processadores mostraram que Clik é potencialmente capaz de superar o desempenho de sistemas SDSM estado-da-arte para essa classe importante de aplicações.
Referências
R. Blumofe, C. Joerg, B. Kuszmaul, C. Leiserson, K. Randall, and Y. Zhou. Cilk: An efficient multithreaded runtime system. In Proceedings of the 5th Symposium on Principies and Practice of Parallel Programming, 1995.
R. D. Blumofe, M. Frigo, C. F. Joerg, C. E. Leiserson, and K. H. Randall. Dag-consistent distributed shared memory. In Proc. of the 10th lnt'l Parallel Processing Symp. (IPPS'96), pages 132-141, 1996.
R. Farias, C. Bentes, A. Coelho, S. Guedes, and L. G. calves. Work distribution for parallel zsweep algorithm. In XI Brazilian Symposium on Computer Graphics and Image Processing, pages 107-114, October 2003.
M. Frigo, C. E. Leiserson, and K. H. Randall. The implementation of the Cilk-5 multithreaded language. In Proceedings of the ACM SIGPLAN '98 Conference on Programming Language Design and Implementation, pages 212-223. Montreal, Canada, June 1998.
L. lftode, C. Dubnicki, E. W. Felten, and K. Li. Improving Release-Consistent Shared Virtual Memory Using Automatic Update. In Proc. of the Symposium on High-Performance Computer Architecture, pages 14-25, February 1996.
P. Keleher. The Relative Imponance of Concurrent Writers and Weak Consistency Models. In Proceedings of the 16th International Conference on Distributed Computing Systems, May 1996.
P. Keleher, A. Cox, and W. Zwaenepoel. Lazy Release Consistency for Software Distributed Shared Memory. In Proceedings of the 19th Annual International Symposium on Computer Architecture, pages 13-21, May 1992.
P. Keleher, S. Dwarkadas, A. Cox, and W. Zwaenepoel. TreadMarks: Distributed Shared Memory on Standard Workstations and Operating Systems. In Proceedings of the 1994 Wimer Usenix Conference, pages 115-131, January 1994.
R. Mendes, L. Whately, M. Lobosco, and C. L. Amorim. Memória compartilhada distribuída para redes udp/ip: Implementação e avaliação. In Workshop em Sistemas Computacionais de Alto Desempenho, October 2003.
L. R. Monnerat and R. Bianchini. Efficiently Adapting to Sharing Pattems in Software DSMs. In Proceedings of the 4th IEEE Symposium on High-Performance Computer Architecture, pages 289-299, February 1998.
L. Peng, W. F. Wong, M. D. Feng, and C. K. Yuen. SilkRoad: A multithreaded runtime system with software distributed shared memory for SMP clusters. In IEEE International Conference on Cluster Computing, pages 243-249, Nov. 2000.
L. Peng, W.-F. Wong, and C.-K. Yuen. Silkroad ii: mixed paradigm cluster computing with rc-dag consistency. Parallel Comput., 29(8): 1091-1115, 2003.
K. H. Randall. Cilk: Efficient Multithreaded Computing. PhD thesis, Department of Electrical Engineering and Computer Science, Massachusetts Institute of Technology, May 1998.
C. B. Seidel, R. Bianchini, and C. L. Amorim. The Affinity Entry Consistency Protocol. In Proceedings of the 1997 International Conference on Parallel Processing, pages 65-78, August 1997.
W. E. Speight and J. K. Bennett. Brazos: A Third Generation DSM System. In Proceedings of the 1997 USENIX Windows/NT Workshop, pages 95-106, August 1997.
L. Whately, R. Pinto, M. Ragarjan, L. Iftode, R. Bianchini, and C. L. Amorim. Adaptative Techniques for Home-Based Software DSMs. In Symposium on Computer Architecture and High-Performance Computing, pages 164-171, September 2001.
Y. Zhou, L. Iftode, and K. Li. Performance Evaluation of Two Home-Based Lazy Release Consistency Protocols for Shared Memory Virtual Memory Systems. In Proceedings of the 2nd Symposium on Operating Systems Design and Implementation, pages 75-88, Oct. 1996.
R. D. Blumofe, M. Frigo, C. F. Joerg, C. E. Leiserson, and K. H. Randall. Dag-consistent distributed shared memory. In Proc. of the 10th lnt'l Parallel Processing Symp. (IPPS'96), pages 132-141, 1996.
R. Farias, C. Bentes, A. Coelho, S. Guedes, and L. G. calves. Work distribution for parallel zsweep algorithm. In XI Brazilian Symposium on Computer Graphics and Image Processing, pages 107-114, October 2003.
M. Frigo, C. E. Leiserson, and K. H. Randall. The implementation of the Cilk-5 multithreaded language. In Proceedings of the ACM SIGPLAN '98 Conference on Programming Language Design and Implementation, pages 212-223. Montreal, Canada, June 1998.
L. lftode, C. Dubnicki, E. W. Felten, and K. Li. Improving Release-Consistent Shared Virtual Memory Using Automatic Update. In Proc. of the Symposium on High-Performance Computer Architecture, pages 14-25, February 1996.
P. Keleher. The Relative Imponance of Concurrent Writers and Weak Consistency Models. In Proceedings of the 16th International Conference on Distributed Computing Systems, May 1996.
P. Keleher, A. Cox, and W. Zwaenepoel. Lazy Release Consistency for Software Distributed Shared Memory. In Proceedings of the 19th Annual International Symposium on Computer Architecture, pages 13-21, May 1992.
P. Keleher, S. Dwarkadas, A. Cox, and W. Zwaenepoel. TreadMarks: Distributed Shared Memory on Standard Workstations and Operating Systems. In Proceedings of the 1994 Wimer Usenix Conference, pages 115-131, January 1994.
R. Mendes, L. Whately, M. Lobosco, and C. L. Amorim. Memória compartilhada distribuída para redes udp/ip: Implementação e avaliação. In Workshop em Sistemas Computacionais de Alto Desempenho, October 2003.
L. R. Monnerat and R. Bianchini. Efficiently Adapting to Sharing Pattems in Software DSMs. In Proceedings of the 4th IEEE Symposium on High-Performance Computer Architecture, pages 289-299, February 1998.
L. Peng, W. F. Wong, M. D. Feng, and C. K. Yuen. SilkRoad: A multithreaded runtime system with software distributed shared memory for SMP clusters. In IEEE International Conference on Cluster Computing, pages 243-249, Nov. 2000.
L. Peng, W.-F. Wong, and C.-K. Yuen. Silkroad ii: mixed paradigm cluster computing with rc-dag consistency. Parallel Comput., 29(8): 1091-1115, 2003.
K. H. Randall. Cilk: Efficient Multithreaded Computing. PhD thesis, Department of Electrical Engineering and Computer Science, Massachusetts Institute of Technology, May 1998.
C. B. Seidel, R. Bianchini, and C. L. Amorim. The Affinity Entry Consistency Protocol. In Proceedings of the 1997 International Conference on Parallel Processing, pages 65-78, August 1997.
W. E. Speight and J. K. Bennett. Brazos: A Third Generation DSM System. In Proceedings of the 1997 USENIX Windows/NT Workshop, pages 95-106, August 1997.
L. Whately, R. Pinto, M. Ragarjan, L. Iftode, R. Bianchini, and C. L. Amorim. Adaptative Techniques for Home-Based Software DSMs. In Symposium on Computer Architecture and High-Performance Computing, pages 164-171, September 2001.
Y. Zhou, L. Iftode, and K. Li. Performance Evaluation of Two Home-Based Lazy Release Consistency Protocols for Shared Memory Virtual Memory Systems. In Proceedings of the 2nd Symposium on Operating Systems Design and Implementation, pages 75-88, Oct. 1996.
Publicado
24/10/2005
Como Citar
MENDES, Rafael; WHATELY, Lauro; BENTES, Cristiana; AMORIM, Claudio Luis.
Implementação e Avaliação Preliminar de um novo Sistema Software DSM para Cluster de Computadores. In: SIMPÓSIO EM SISTEMAS COMPUTACIONAIS DE ALTO DESEMPENHO (SSCAD), 6. , 2005, Rio de Janeiro.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2005
.
p. 33-40.
DOI: https://doi.org/10.5753/wscad.2005.18973.
