Usando aprendizado supervisionado para composição de políticas de escalonamento
Resumo
Este artigo propõe a utilização de aprendizado supervisionado para melhorar as políticas de escalonamento em ambientes de computação de alto desempenho, utilizando dados que incorporam políticas existentes (algoritmos tutores). O modelo treinado revela-se adaptável, superando, em alguns casos, as limitações dos algoritmos tutores.Referências
Carastan-Santos, D. and de Camargo, R. Y. (2017). Obtaining dynamic scheduling policies with simulation and machine learning. In Proceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis, SC ’17, New York, NY, USA. Association for Computing Machinery.
Carastan-Santos, D., De Camargo, R. Y., Trystram, D., and Zrigui, S. (2019). One can only gain by replacing easy backfilling: A simple scheduling policies case study. In 2019 19th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (CCGRID), pages 1–10.
Feitelson, D. G. and Rudolph, L. (1998). Metrics and benchmarking for parallel job scheduling. In Workshop on Job Scheduling Strategies for Parallel Processing, pages 1–24. Springer.
Feitelson, D. G., Tsafrir, D., and Krakov, D. (2014). Experience with using the parallel workloads archive. Journal of Parallel and Distributed Computing, 74(10):2967–2982.
Koslovski, G. P., Pereira, K., and Albuquerque, P. R. (2024). Dag-based workflows scheduling using actor–critic deep reinforcement learning. Future Generation Computer Systems, 150:354–363.
Ontivero-Ortega, M., Lage-Castellanos, A., Valente, G., Goebel, R., and Valdes-Sosa, M. (2017). Fast gaussian naı̈ve bayes for searchlight classification analysis. NeuroImage, 163:471–479.
Carastan-Santos, D., De Camargo, R. Y., Trystram, D., and Zrigui, S. (2019). One can only gain by replacing easy backfilling: A simple scheduling policies case study. In 2019 19th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (CCGRID), pages 1–10.
Feitelson, D. G. and Rudolph, L. (1998). Metrics and benchmarking for parallel job scheduling. In Workshop on Job Scheduling Strategies for Parallel Processing, pages 1–24. Springer.
Feitelson, D. G., Tsafrir, D., and Krakov, D. (2014). Experience with using the parallel workloads archive. Journal of Parallel and Distributed Computing, 74(10):2967–2982.
Koslovski, G. P., Pereira, K., and Albuquerque, P. R. (2024). Dag-based workflows scheduling using actor–critic deep reinforcement learning. Future Generation Computer Systems, 150:354–363.
Ontivero-Ortega, M., Lage-Castellanos, A., Valente, G., Goebel, R., and Valdes-Sosa, M. (2017). Fast gaussian naı̈ve bayes for searchlight classification analysis. NeuroImage, 163:471–479.
Publicado
24/04/2024
Como Citar
DIEL, Guilherme; KRAUS, Ana Eloina Nascimento; KOSLOVSKI, Guilherme Piêgas.
Usando aprendizado supervisionado para composição de políticas de escalonamento. In: ESCOLA REGIONAL DE ALTO DESEMPENHO DA REGIÃO SUL (ERAD-RS), 24. , 2024, Florianópolis/SC.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2024
.
p. 73-76.
ISSN 2595-4164.
DOI: https://doi.org/10.5753/eradrs.2024.238657.
