Lavanya Jose
Nick McKeown
ABSTRACT
As datacenter speeds scale to 100 Gb/s and beyond, traditional congestion control algorithms like TCP and RCP converge slowly to steady sending rates, which leads to poorer and less predictable user performance. These reactive algorithms use congestion signals to perform gradient descent to approach ideal sending rates, causing poor convergence times. In this paper, we propose a proactive congestion control algorithm called PERC, which explicitly computes rates independently of congestion signals in a decentralized fashion. Inspired by message-passing algorithms with traction in other fields (e.g., modern Low Density Parity Check decoding algorithms), PERC improves convergence times by a factor of 7 compared to reactive explicit rate control protocols such as RCP. This fast convergence reduces tail flow completion time (FCT) significantly in high speed networks; for example, simulations of a realistic workloads in a 100 Gb/s network show that PERC achieves up to 4x lower 99th percentile FCT compared to RCP.
Supplemental Material
- M. Alizadeh, A. Greenberg, D. A. Maltz, J. Padhye, P. Patel, B. Prabhakar, S. Sengupta, and M. Sridharan. Data center TCP (DCTCP). In Proc. ACM SIGCOMM 2010 Conference, pages 63--74, Aug 2010. Google Scholar
Digital Library
- M. Alizadeh, A. Kabbani, T. Edsall, B. Prabhakar, A. Vahdat, and M. Yasuda. Less is more: trading a little bandwidth for ultra-low latency in the data center. In Proc. USENIX Conference on Networked Systems Design and Implementation (NSDI 2012), April 2012. Google ScholarDigital Library
- M. Alizadeh, S. Yang, M. Sharif, S. Katti, N. McKeown, B. Prabhakar, and S. Shenker. pFabric: minimal near-optimal datacenter transport. In Proc. ACM SIGCOMM 2013 Conference, pages 435--446, Aug 2013. Google ScholarDigital Library
- H. Ballani, P. Costa, T. Karagiannis, and A. Rowstrong. Towards predictable datacenter networks. In Proc. ACM SIGCOMM 2011 Conference, pages 242--253, Aug 2011. Google ScholarDigital Library
- D. Bertsekas and R. Gallager. Data Networks, pages 524--527. Prentice Hall, 2nd edition, 1992.Google Scholar
- P. Bosshart, D. Daly, G. Gibb, M. Izzard, N. McKeown, J. Rexford, C. Schlesinger, D. Talayco, A. Vahdat, G. Varghese, and D. Walker. P4: Programming protocol-independent packet processors. SIGCOMM Computer Communication Review, pages 87--95, Jul 2014. Google ScholarDigital Library
- P. Bosshart, G. Gibb, H.-S. Kim, G. Varghese, N. McKeown, M. Izzard, F. Mujica, and M. Horowitz. Forwarding metamorphosis: fast programmable match-action processing in hardware for SDN. In Proc. ACM SIGCOMM 2013 Conference, pages 99--110, Aug 2013. Google ScholarDigital Library
- A. Charny, D. D. Clark, and R. Jain. Congestion control with explicit rate indication. In Proc. IEEE International Conference on Communications (ICC), 1995.Google ScholarCross Ref
- M. Chiang, S. H. Low, J. C. Doyle, et al. Layering as optimization decomposition: A mathematical theory of network architectures. Proceedings of the IEEE, 95(1):255--312, 2007.Google ScholarCross Ref
- M. Chowdhury, Y. Zhong, and I. Stoica. Efficient coflow scheduling with varys. In Proc. ACM SIGCOMM 2014 Conference, pages 443--454, Aug 2014. Google ScholarDigital Library
- F. R. Dogar, T. Karagiannis, H. Ballani, and A. Rowstron. Decentralized task-aware scheduling for data center networks. In Proc. ACM SIGCOMM 2014 Conference, pages 431--442, Aug 2014. Google ScholarDigital Library
- N. Dukkipati. Rate Control Protocol (RCP): Congestion control to make flows complete quickly. PhD thesis, Citeseer, 2007. Google ScholarDigital Library
- C. V. Hollot, V. Misra, D. Towsley, and W.-B. Gong. A control theoretic analysis of red. In Proc. IEEE INFOCOM, pages 1510--1519, Apr 2001.Google ScholarDigital Library
- C.-Y. Hong, M. Caesar, and P. B. Godfrey. Finishing flows quickly with preemptive scheduling. In Proc. ACM SIGCOMM 2012 Conference, pages 127--138, Aug 2012. Google ScholarDigital Library
- T. Issariyakul and E. Hossain. Introduction to Network Simulator NS2. Springer Publishing Company, 1st edition, 2008. Google ScholarDigital Library
- V. Jacobson and M. J. Karels. Congestion avoidance and control. In Proc. ACM SIGCOMM 1988 conference, Aug. 1988. Google ScholarDigital Library
- V. Jeyakumar, M. Alizadeh, D. Mazières, B. Prabhakar, A. Greenberg, and C. Kim. EyeQ: Practical network performance isolation at the edge. In Proc. USENIX Conference on Networked Systems Design and Implementation (NSDI 2013), Apr 2013. Google ScholarDigital Library
- D. Katabi, M. Handley, and C. Rohrs. Congestion control for high bandwidth-delay product networks. In Proc. ACM SIGCOMM 2002 Conference, Aug 2002. Google ScholarDigital Library
- F. P. Kelly, A. K. Maulloo, and D. K. H. Tan. Rate control for communication networks: shadow prices, proportional fairness, and stability. In The Journal of the Operational Research Society, pages 237--252, Mar 1998.Google Scholar
- S. H. Low and L. L. H. Andrew. Understanding xcp: equilibrium and fairness. In Proc. IEEE INFOCOM, pages 1025--1036, Mar 2005.Google ScholarCross Ref
- D. J. MacKay. Information theory, inference and learning algorithms. Cambridge University Press, 2003. Google ScholarDigital Library
- J. Perry, A. Ousterhout, H. Balakrishnan, D. Shah, and H. Fugal. Fastpass: a centralized "zero-queue" datacenter network. In Proc. ACM SIGCOMM 2014 Conference, pages 307--318, Aug 2014. Google ScholarDigital Library
- T. J. Richardson and R. L. Urbanke. The capacity of low-density parity-check codes under message-passing decoding. In Proc. IEEE Transactions on Information Theory, pages 599--618, Feb 2001. Google ScholarDigital Library
- R. Srikant. The mathematics of Internet congestion control. Springer Science & Business Media, 2012. Google ScholarDigital Library
- B. Vamanan, J. Hasan, and T. N. Vijaykumar. Deadline-aware datacenter TCP (D2TCP). In Proc. ACM SIGCOMM 2012 Conference, pages 115--126, Aug 2012. Google ScholarDigital Library
- A. Varga and R. Hornig. An overview of the OMNeT++ simulation environment. In Proc. SIMUTools, Mar 2008. Google ScholarDigital Library
- C. Wilson, H. Ballani, T. Karagiannis, and A. Rowtron. Better never than late: meeting deadlines in datacenter networks. In Proc. ACM SIGCOMM 2011 Conference, pages 50--61, Aug 2011. Google ScholarDigital Library
Index Terms
- High Speed Networks Need Proactive Congestion Control
Recommendations
Unreliable transport protocol using congestion control for high-speed networks
Currently there is no control for the real-time traffic of multimedia applications using UDP (User Datagram Protocol) in high-speed networks. Therefore, although a number of high-speed TCP (Transmission Control Protocol) protocols have been developed ...
An efficient and fair explicit congestion control protocol for high bandwidth-delay product networks
XCP and VCP can achieve excellent performance under high bandwidth-delay product networks, but they all have some defects. In XCP, router needs to calculate a feedback for each departing packet, the cost will be un-negligible in high-speed networks. In ...
Congestion control for high bandwidth-delay product networks
Proceedings of the 2002 SIGCOMM conferenceTheory and experiments show that as the per-flow product of bandwidth and latency increases, TCP becomes inefficient and prone to instability, regardless of the queuing scheme. This failing becomes increasingly important as the Internet evolves to ...
Comments