Flow-aware explicit congestion notification for datacenter networks

Springer Science and Business Media LLC - Tập 22 - Trang 1431-1446 - 2019
Pan Zhou1, Hongfang Yu1, Gang Sun1, Long Luo1, Shouxi Luo2, Zilong Ye3
1University of Electronic Science and Technology of China, Chengdu, People’s Republic of China
2Southwest Jiaotong University, Chengdu, People’s Republic of China
3California State University, Los Angeles, USA

Tóm tắt

Explicit congestion notification (ECN) has been widely adopted by recent proposals to build up high-throughput and low-latency datacenter network transport. In these ECN-based proposals, when the queue length of a switch exceeds a pre-defined threshold, the switch would mark all arriving packets with ECN to explicitly notify their senders to slow down the rates. Such a design enables the network to eliminate congestions quickly. However, it marks packets without considering the flow state, which may overkill flows, especially those only send a few packets, thus resulting in significant throughput loss and long flow completion times. In this paper, we propose a novel flow-aware ECN marking approach (FECN), which can improve the throughput and flow completion time by taking flow states into consideration. By selectively marking packets respecting to their flow rates, FECN enables the network to precisely slow down the high-speed flows to avoid congestions without killing low-speed short flows. Moreover, FECN does not require switches to maintain per-flow state, which yields low overhead and thus makes FECN to be easily implemented and deployed in commodity switches. Simulations show that FECN can shorten the flow completion time by up to 44.7% and reduce the throughput loss by up to 40.3%, compared with prior flow-agnostic ECN marking approach.

Tài liệu tham khảo

Sun, G., Zhu, G., Yu, H., et al.: Cost-efficient service function chain orchestration for low-latency applications in NFV networks. IEEE Syst. J. (2018) Alizadeh, M., Yang, S., Sharif, M., Katti, S. et al.: pFabric: minimal near-optimal datacenter transport. In: Proc. SIGCOMM, pp. 435446 (2013) Hoff, T.: Latency is everywhere and it costs you sales how to crush it. http://highscalability.com/blog/2009/7/25/latency- is-everywhere-and-it-costs-you-sales-how-to-crush-it.html (2009) Sun, G., Li, Y., Vasilakos, A., Guizani, M.: Energy-efficient and traffic-aware service function chaining orchestration in multi-domain networks. Future Gener. Comput. Syst. 91, 347–360 (2019) Sun, G., Yu, H.: A new technique for efficient live migration of multiple virtual machines. Future Gener. Comput. Syst. 55, 74–86 (2016) Sun, G., Liao, D., Yu, H.: Live migration for multiple correlated virtual machines in cloud-based data centers. IEEE Trans. Serv. Comput. 11(2), 279–291 (2018) Munir, A., Qazi, I.: Minimizing flow completion times in data centers, INFOCOM. Proc. IEEE IEEE 2013, 2157–2165 (2013) Alizadeh, M., Greenberg, A., Maltz, D. et al.: Data center TCP (DCTCP). In: Proc. SIGCOMM, pp. 6374 (2010) Luo, S., Hongfang, Y., Zhao, Y., Wang, S., Shui, Y., Li, L.: Towards practical and near-optimal coflow scheduling for data center networks. IEEE Trans. Parallel Distrib. Syst. 27(11), 3366–3380 (2016) Zhu, Y., Eran, H., Firestone, D., Guo, C. et al.: Congestion Control for Large-Scale RDMA Deployments. In: Proc. SIGCOMM (2015) Wu, H., Ju, J., Lu, G., Guo, C., Xiong, Y., Zhang, Y.: Tuning ECN for data center networks. In: CoNEXT (2012) Bai, W., Chen, L., Chen, K., Wu, H.: Enabling ECN in multi-service multi-queue data centers. In: Usenix Conference on Networked Systems Design and Implementation USENIX Association, pp. 537–549 (2016) Floyd, S., Jacobson, V.: Random early detection gateways for congestion avoidance. IEEE/ACM Trans. Netw. 4, 397–413 (1993) Shan, D., Ren, F.: Improving ECN marking scheme with micro-burst traffic in data center networks. In: INFOCOM (2017) The Network Simulator NS-3. https://www.nsnam.org/ Lin, D., Morris, R.: Dynamics of random early detection. In: Proc. SIGCOMM, pp. 127–137 (1997) Zhao, Z., Jiang, Z., Lu, C. et al.: Towards coordinated congestion control and load balancing in datacenter networks. In: Global Communications Conference (GLOBECOM), IEEE (2013) Alizadeh, M., Kabbani, A., et al.: Less is more: trading a little bandwidth for ultra-low latency in the data center. In: Usenix Conference on Networked Systems Design and Implementation pp. 19–19 (2012) Rong, P., Prabhakar, B., Psounis, K.: CHOKe—a stateless active queue management scheme for approximating fair bandwidth allocation. In: INFOCOM (2000) Lakshman, T., Wong, L.: SRED: stabilized RED. In: Proceedings of INFOCOM pp. 1346–1355 (1999) Mittal, R., Radhika, V., et al.: TIMELY: RTT-based congestion control for the datacenter. In: ACM Conference on Special Interest Group on Data Communication ACM, pp. 537–550 (2015) Lee, C., Park, C.: DX: latency-based congestion control for datacenters. IEEE/ACM Trans. Netw. 25(1), 335–348 (2017) Zhao, Z., Li, Q., et al.: Reduce completion time and guarantee throughput by transport with slight congestion. In: IEEE International Conference on Communications IEEE pp. 1–6 (2016) Bai, W., Chen, K., et al.: Enabling ECN over Generic Packet Scheduling. In: International on Conference on Emerging NETWORKING Experiments and Technologies ACM, pp. 191–204 (2016) Wilson, C., Ballani, H.: Better never than late: meeting deadlines in datacenter networks. Acm Sigcomm Comput. Commun. Rev. 41(4), 50–61 (2011) Hong, C., Caesar, M., Godfrey, P.: Finishing flows quickly with preemptive scheduling. Acm Sigcomm Comput. Commun. Rev. 42(4), 127–138 (2012) RFC 791. https://tools.ietf.org/html/rfc791 Nichols, K., Jacobson, V.: Controlling queue delay. Commun. ACM 55, 1–7 (2012) Yuanwei, L., et al.: Multi-Path Transport for RDMA in Datacenters. In: 15th USENIX Symposium on Networked Systems Design and Implementation (NSDI) (2018) Alizadeh, M., Yang, S., Sharif, M., Katti, S., McKeown, N., Prabhakar, B., Shenker, S.: pFabric: minimal near-optimal datacenter transport. In: ACM SIGCOMM (2013) Perry, J., Ousterhout, A., Balakrishnan, H., Shah, D., Fugal, H.: Fastpass: A centralized zero-queue datacenter network. In: Proc. ACM SIGCOMM (2014) Perry, J., Balakrishnan, H., Shah, D.: Flowtune: flowlet control for datacenter networks. In: NSDI (2017) Vamanan, B., Hasan, J., Vijaykumar, T. N.: Deadline-aware datacenter TCP (D2TCP). In: Proc. ACM SIGCOMM (2012) Gao, Chengxi, Lee, Victor C.S., Li, Keqin: DemePro: DEcouple packet marking from enqueuing for multiple services with PROactive congestion control. IEEE Trans. Cloud Comput. 1, 1–1 (2017) David, Z., et al.: DeTail: reducing the flow completion time tail in datacenter networks. In: Proceedings of the ACM SIGCOMM 2012 conference on Applications, technologies, architectures, and protocols for computer communication. ACM (2012) Sun, G., Liao, D., Zhao, D., Sun, Z., Chang, V.: Towards provisioning hybrid virtual networks in federated cloud data centers. Future Gener. Comput. Syst. 87, 457–469 (2018) Alizadeh, M., Kabbani, A., Atikoglu, B., Prabhakar, B.: Stability analysis of QCN: the averaging principle. In: SIGMETRICS (2011) Alizadeh, M., Javanmard, A., Prabhakar, B.: Analysis of DCTCP: Stability, convergence and fairness. In: SIGMETRICS (2011) Cisco White Paper: Intelligent Buffer Management on Cisco Nexus 9000 Series Switches. https://www.cisco.com/c/en/us/products/collateral/switches/nexus-9000-series-switches/white-paper-c11-738488.html Lee, C., Nakagawa, Y., Hyoudou, K., Kobayashi, S., Shiraki, O., Shimizu, T.: Control, flow-aware congestion, to improve throughput under TCP incast in datacenter networks. In: IEEE 39th Annual Computer Software and Applications Conference. Taichung, pp. 155–162 (2015) Sivaraman A., et al.: Programmable packet scheduling at line rate. In: Proc. ACMSIGCOMM Conf., pp. 4457 (2016) Sharma, N. et al.: Approximating fair queueing on reconfigurable switches. In: USENIX Symposium on Networked Systems Design and Implementation (2018)