Detectando Heavy Hitters globalmente em dispositivos programáveis multi-pipes
Abstract
A way to contribute to network management involves detecting high-impact traffic flows, known as “Heavy Hitters”. Heavy Hitters are flows that account for the majority of bytes transmitted over the network, consequently consuming more resources. The use of programmable hardware, such as switches and DPUs, enables the detection of these flows in-line rate, meaning direct detection in the network’s data plane. While the literature reveals extensive analysis of detection in single-pipe switches, this study presents two approaches to identify Heavy Hitters in programmable switches with multiple pipes. One approach features an accumulator in the switch that centralizes data from all pipes and communicates with the control plane. In the other approach, communications with the control plane are independent for each pipe. Both approaches were developed and validated through an emulator, demonstrating effectiveness and improvement in detection in multi-pipe switches compared to single-pipe switches.
References
Basat, R. B., Chen, X., Einziger, G., and Rottenstreich, O. (2020). Designing heavy-hitter detection algorithms for programmable switches. IEEE/ACM Transactions on Networking, 28(3):1172–1185.
Basat, R. B., Einziger, G., Feibish, S. L., Moraney, J., and Raz, D. (2018a). Network-wide routing-oblivious heavy hitters. In Proceedings of the 2018 Symposium on Architectures for Networking and Communications Systems, pages 66–73.
Basat, R. B., Einziger, G., Keslassy, I., Orda, A., Vargaftik, S., and Waisbard, E. (2018b). Memento: Making sliding windows efficient for heavy hitters. In Proceedings of the 14th International Conference on Emerging Networking EXperiments and Technologies, pages 254–266.
CAIDA (2019). The caida ucsd anonymized internet traces - 20190117.
Chiesa, M. and Verdi, F. L. (2023). Network monitoring on multi-pipe switches. Proc. ACM Meas. Anal. Comput. Syst., 7(1).
Ding, D., Savi, M., Antichi, G., and Siracusa, D. (2020). An incrementally-deployable p4-enabled architecture for network-wide heavy-hitter detection. IEEE Transactions on Network and Service Management, 17(1):75–88.
Harrison, R., Cai, Q., Gupta, A., and Rexford, J. (2018). Network-wide heavy hitter detection with commodity switches. In Proceedings of the Symposium on SDN Research, pages 1–7.
Hofstede, R., Čeleda, P., Trammell, B., Drago, I., Sadre, R., Sperotto, A., and Pras, A. (2014). Flow monitoring explained: From packet capture to data analysis with netflow and ipfix. IEEE Communications Surveys & Tutorials, 16(4):2037–2064.
Lin, Y.-B., Huang, C.-C., and Tsai, S.-C. (2019). Sdn soft computing application for detecting heavy hitters. IEEE Transactions on Industrial Informatics, 15(10):5690–5699.
Machado, D., de Castro, A. G., Vogt, F., and Luizelli, M. C. (2019). Avaliação de desempenho de heavy hitters utilizando p4 e xdp. In Anais da XVII Escola Regional de Redes de Computadores, pages 122–123. SBC.
Silva, M. V. B. d. (2019). Prevendo e identificando fluxos elefantes em redes de ponto de troca de tráfego com suporte à programabilidade.
Sivaraman, V., Narayana, S., Rottenstreich, O., Muthukrishnan, S., and Rexford, J. (2017). Heavy-hitter detection entirely in the data plane. In Proceedings of the Symposium on SDN Research, pages 164–176.
Tang, L., Huang, Q., and Lee, P. P. (2020). A fast and compact invertible sketch for network-wide heavy flow detection. IEEE/ACM Transactions on Networking, 28(5):2350–2363.
Turkovic, B., Oostenbrink, J., and Kuipers, F. (2019). Detecting heavy hitters in the data-plane. arXiv preprint arXiv:1902.06993.
Vilela, G. S. (2006). Caracterização de tráfego utilizando classificação de fluxos de comunicação. Mestre em ciências em engenharia de sistemas e computação, Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brasil.
Wheeler, B. (2019). Tomahawk 4 switch first to 25.6 tbps. Microprocessor Report.
