Calisto: Sistema para processamento rápido de pacotes com baixa latência em centro de dados
Resumo
A demanda por suporte a baixa latência em ambientes de alta vazão tem aumentado nosúltimos anos. Sistemas como Shenango permitem que servidores processem pacotes rapidamente, mantendo a eficiência na utilização de CPU através de um algoritmo de detecção de congestionamento e o IOKernel, componente dedicado a orquestrar a alocação de novos núcleos. O sistema porém aloca núcleos utilizando a política de escalonamento round-robin. Neste trabalho, apresenta-se Calisto, um novo sistema que implementa uma política eficiente de distribuição de pacotes entre núcleos que mantém afinidade de fluxos, reduzindo a latência em até 10%.
Referências
Barbette, T., Katsikas, G. P., Maguire, G. Q., and Kostíc, D. (2019). Rss++: Load and state-aware receive side scaling. In Proceedings of the 15th International Conference on Emerging Networking Experiments And Technologies, CoNEXT ’19, page 318–333, New York, NY, USA. Association for Computing Machinery.
Belay, A., Prekas, G., Primorac, M., Klimovic, A., Grossman, S., Kozyrakis, C., , and Bugnion, E. (2014). Ix operating system: Combining low latency, high throughput, and efficiency in a protected dataplane. In Proceedings of the USENIX Annual Technical Conference (ATC), pages 1–14.
Bosshart, P., Daly, D., Gibb, G., Izzard, M., McKeown, N., Rexford, J., Schlesinger, C., Talayco, D., Vahdat, A., Varghese, G., and Walker, D. (2014). P4: Programming protocol-independent packet processors. SIGCOMM Comput. Commun. Rev., 44(3):87–95.
DPDK (2021 (acessado em Janeiro de 2021)). Data Plane Development Kit.
Duplyakin, D., Ricci, R., Maricq, A., Wong, G., Duerig, J., Eide, E., Stoller, L., Hibler, M., Johnson, D., Webb, K., Akella, A., Wang, K., Ricart, G., Landweber, L., Elliott, C., Zink, M., Cecchet, E., Kar, S., and Mishra, P. (2019). The design and operation of CloudLab. In Proceedings of the USENIX Annual Technical Conference (ATC), pages 1–14.
Hood, R., Jin, H., Mehrotra, P., Chang, J., Djomehri, J., Gavali, S., Dennis Jespersen, K. T., and Biswas, R. (2010). Performance impact of resource contention in multicore In IEEE International Symposium on Parallel & Distributed Processing, systems. pages 1–3.
Moon, Y., Lee, S., Jamshed, M. A., and Park, K. (2020). Acceltcp: Accelerating network In 17th USENIX Symposium on Networapplications with stateful TCP ofoading. ked Systems Design and Implementation (NSDI 20), pages 77–92, Santa Clara, CA. USENIX Association.
Ousterhout, A., Fried, J., Behrens, J., Belay, A., and Balakrishnan, H. (2019). Shenango: In 16th Achieving high CPU efficiency for latency-sensitive datacenter workloads. USENIX Symposium on Networked Systems Design and Implementation (NSDI 19), pages 361–378, Boston, MA. USENIX Association.
Pantuza, G., Bleme, L. A. C., Vieira, M. A. M., and Vieira, L. F. M. (2021). Danian: Tail latency reduction of networking application through an o(1) scheduler. In 26th IEEE Symposium on Computers and Communications (ISCC).
Pantuza, G., Sampaio, F., Vieira, L. F. M., Guedes, D., and Vieira, M. A. M. (2014). Network management through graphs in software defined networks. In 10th International Conference on Network and Service Management (CNSM) and Workshop, pages 400–405.
Peter, S., Li, J., Zhang, I., Ports, D. R. K., Woos, D., Krishnamurthy, A., and Anderson, T. (2014). Arrakis: The operating system is the control plane. In Proceedings of the USENIX Annual Technical Conference (ATC), pages 1–14.
Prekas, G., Kogias, M., Kogias, M., and Bugnion, E. (2017). Zygos: Achieving low tail latency for microsecond-scale networked tasks. In Proceedings of the USENIX Annual Technical Conference (ATC), pages 1–14.
Qin, H. (2019). The arachne distributed operating system. In Proceedings of the USENIX Annual Technical Conference (ATC), pages 1–14.
Rushing’s, S. (2021 (acessado em Março de 2021)). Coroutine library threads.
Vieira, M. A. M., Castanho, M. S., Pacífico, R. D. G., Santos, E. R. S., Júnior, E. P. M. C., and Vieira, L. F. M. (2020). Fast packet processing with ebpf and xdp: Concepts, code, challenges, and applications. ACM Comput. Surv., 53(1).
Belay, A., Prekas, G., Primorac, M., Klimovic, A., Grossman, S., Kozyrakis, C., , and Bugnion, E. (2014). Ix operating system: Combining low latency, high throughput, and efficiency in a protected dataplane. In Proceedings of the USENIX Annual Technical Conference (ATC), pages 1–14.
Bosshart, P., Daly, D., Gibb, G., Izzard, M., McKeown, N., Rexford, J., Schlesinger, C., Talayco, D., Vahdat, A., Varghese, G., and Walker, D. (2014). P4: Programming protocol-independent packet processors. SIGCOMM Comput. Commun. Rev., 44(3):87–95.
DPDK (2021 (acessado em Janeiro de 2021)). Data Plane Development Kit.
Duplyakin, D., Ricci, R., Maricq, A., Wong, G., Duerig, J., Eide, E., Stoller, L., Hibler, M., Johnson, D., Webb, K., Akella, A., Wang, K., Ricart, G., Landweber, L., Elliott, C., Zink, M., Cecchet, E., Kar, S., and Mishra, P. (2019). The design and operation of CloudLab. In Proceedings of the USENIX Annual Technical Conference (ATC), pages 1–14.
Hood, R., Jin, H., Mehrotra, P., Chang, J., Djomehri, J., Gavali, S., Dennis Jespersen, K. T., and Biswas, R. (2010). Performance impact of resource contention in multicore In IEEE International Symposium on Parallel & Distributed Processing, systems. pages 1–3.
Moon, Y., Lee, S., Jamshed, M. A., and Park, K. (2020). Acceltcp: Accelerating network In 17th USENIX Symposium on Networapplications with stateful TCP ofoading. ked Systems Design and Implementation (NSDI 20), pages 77–92, Santa Clara, CA. USENIX Association.
Ousterhout, A., Fried, J., Behrens, J., Belay, A., and Balakrishnan, H. (2019). Shenango: In 16th Achieving high CPU efficiency for latency-sensitive datacenter workloads. USENIX Symposium on Networked Systems Design and Implementation (NSDI 19), pages 361–378, Boston, MA. USENIX Association.
Pantuza, G., Bleme, L. A. C., Vieira, M. A. M., and Vieira, L. F. M. (2021). Danian: Tail latency reduction of networking application through an o(1) scheduler. In 26th IEEE Symposium on Computers and Communications (ISCC).
Pantuza, G., Sampaio, F., Vieira, L. F. M., Guedes, D., and Vieira, M. A. M. (2014). Network management through graphs in software defined networks. In 10th International Conference on Network and Service Management (CNSM) and Workshop, pages 400–405.
Peter, S., Li, J., Zhang, I., Ports, D. R. K., Woos, D., Krishnamurthy, A., and Anderson, T. (2014). Arrakis: The operating system is the control plane. In Proceedings of the USENIX Annual Technical Conference (ATC), pages 1–14.
Prekas, G., Kogias, M., Kogias, M., and Bugnion, E. (2017). Zygos: Achieving low tail latency for microsecond-scale networked tasks. In Proceedings of the USENIX Annual Technical Conference (ATC), pages 1–14.
Qin, H. (2019). The arachne distributed operating system. In Proceedings of the USENIX Annual Technical Conference (ATC), pages 1–14.
Rushing’s, S. (2021 (acessado em Março de 2021)). Coroutine library threads.
Vieira, M. A. M., Castanho, M. S., Pacífico, R. D. G., Santos, E. R. S., Júnior, E. P. M. C., and Vieira, L. F. M. (2020). Fast packet processing with ebpf and xdp: Concepts, code, challenges, and applications. ACM Comput. Surv., 53(1).
Publicado
16/08/2021
Como Citar
BLEME, Lucas A. C.; PANTUZA, Gustavo; VIEIRA, Marcos Augusto M.; VIEIRA, Luiz Filipe M..
Calisto: Sistema para processamento rápido de pacotes com baixa latência em centro de dados. In: WORKSHOP DE GERÊNCIA E OPERAÇÃO DE REDES E SERVIÇOS (WGRS), 26. , 2021, Uberlândia.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
2021
.
p. 28-39.
ISSN 2595-2722.
DOI: https://doi.org/10.5753/wgrs.2021.17183.
