|
|
- 2018
数据中心网络中一种基于ECN的TCP慢启动拥塞控制策略
|
Abstract:
在数据中心网络中，高带宽、低时延的链路和多对一的网络结构造成了TCP Incast吞吐率崩溃问题。现有的改进方法都关注于在TCP拥塞避免阶段改进其拥塞控制算法，而忽视了慢启动阶段中激进的指数增窗方式是造成Incast问题的关键原因。因此，该文提出了一种基于ECN的TCP慢启动拥塞控制策略（gentle slow sTart，GST），利用已有的拥塞标志位动态反馈缓存拥塞状况，以调节慢启动阶段的增窗速率。实验结果表明，该方法帮助现有的数据中心TCP协议有效地避免了并发传输中的吞吐率崩溃现象，将并发度和吞吐率分别提升了3.4倍和85倍。
| [1] | CHENG Peng, REN Feng-yuan, SHU Ran, et al. Catch the whole lot in an action:Rapid precise packet loss notification in data centers[C]//Proceedings of USENIX NSDI. Berkeley, USA:USENIX Press, 2014:17-28. |
| [2] | MEISNER D, SADLER C M, BARROSO L A, et al. Power management of online data-intensive services[C]//Proceedings of ISCA. New York, USA:ACM, 2011:319-330. |
| [3] | REN Yong-mao, ZHAO Yu, LIU Pei, et al. A survey on TCP incast in data center networks[J]. International Journal of Communication Systems, 2014, 27(8):1160-1172. |
| [4] | BAI Wei, CHEN Kai, WU Hai-tao, et al. PAC:Taming TCP incast congestion using proactive ACK control[C]//Proceedings of ICNP. Piscataway, USA:IEEE, 2014:385-396. |
| [5] | SHUKLA S, CHAN S, TAM A S W, et al. TCP PLATO:Packet labelling to alleviate time-out[J]. IEEE Journal on Selected Areas in Communications, 2014, 32(1):65-76. |
| [6] | ZHANG Jiao, REN Feng-yuan, TANG Li, et al. Taming TCP incast throughput collapse in data center networks[C]//Proceedings of ICNP. Piscataway, USA:IEEE, 2013:1-10. |
| [7] | SHAN Dan-feng, JIANG Wan-chun, REN Feng-yuan, et al. Absorbing micro-burst traffic by enhancing dynamic threshold policy of data center switches[C]//Proceedings of INFOCOM. Piscataway, USA:IEEE, 2015:118-126. |
| [8] | BENSON T, AKELLA A, MALTZ D A. Network traffic characteristics of data centers in the wild[C]//Proceeding of IMC. New York, USA:ACM, 2010:267-280. |
| [9] | BORTHAKUR D. The hadoop distributed file system:Architecture and design[EB/OL].[2016-10-16]. https://svn.eu.apache.org/repos/asf/hadoop/common/tags/release-0.16.3/docs/hdfs_design.pdf. |
| [10] | BAI W, CHEN L, CHEN K, et al. Enabling ECN in multi-service multi-queue data centers[C]//Proceedings of NSDI. Berkeley, USA:USENIX Press, 2016:537-549. |
| [11] | JUDD G. Attaining the promise and avoiding the pitfalls of TCP in the datacenter[C]//Proceedings of NSDI. Berkeley, USA:USENIX Press, 2015:145-157. |
| [12] | WU Hai-tao, FENG Zhen-qian, GUO Chuan-xiong, et al. ICTCP:Incast congestion control for TCP in data center networks[C]//Proceedings of CoNEXT. New York, USA:ACM, 2010. |
| [13] | ALIZADEH M, GREENBERG A, MALTZ D A, et al. Data center TCP (DCTCP)[C]//Proceedings of ACM SIGCOMM. New York, USA:ACM, 2010:63-74. |
| [14] | ZHANG Jun, WEN Jiang-tao, WANG Jing-yuan, et al. TCP-FITDC:an adaptive approach to TCP incast avoidance for data center applications[C]//Proceedings of ICNC. Piscataway, USA:IEEE, 2013:1048-1052. |
| [15] | HUANG Jia-wei, HE Tian, HUANG Yi, et al. ARS:Cross-layer adaptive request scheduling to mitigate TCP incast in data center networks[C]//Proceedings of INFOCOM. Piscataway, USA:IEEE, 2016:1-9. |
| [16] | CHEN Wen, REN Feng-yuan, XIE Jing, et al. Comprehensive understanding of TCP incast problem[C]//Proceedings of INFOCOM. Piscataway, USA:IEEE, 2015:1688-1696. |
| [17] | XIA Yong, SUBRAMANIAN L, STOICA I, et al. One more bit is enough[C]//Proceedings of ACM SIGCOMM. New York, USA:ACM, 2005:37-48. |
| [18] | LIU Fang-ming, GUO Jian, HUANG Xiao-meng. EBA:Efficient bandwidth guarantee under traffic variability in datacenters[C]//IEEE/ACM Transactions on Networking. Piscataway:IEEE, 2017, 25(1):506-519. |
| [19] | MITTAL R, LAM V T, DUKKIPATI N, et al. TIMELY:RTT-based congestion control for the datacenter[C]//Proceedings of ACM SIGCOMM. New York, USA:ACM, 2015:537-550. |
| [20] | LEE C, PARK C, JANG K, et al. Accurate Latency-based congestion feedback for datacenters[C]//Proceedings of USENIX ATC. Berkeley, USA:USENIX Press, 2015:403-415. |
| [21] | VAMANAN B, HASAN J, VIJAYKUMAR T N. Deadline-aware datadcenter TCP (D2TCP)[C]//Proceedings of ACM SIGCOMM. New York, USA:ACM, 2012:115-126. |
| [22] | MUNIR A, QAZI I A, UZMI Z A, et al. Minimizing flow completion times in data centers[C]//Proceedings of INFOCOM. Piscataway, USA:IEEE, 2013:2157-2165. |
| [23] | CHEN Yan-pei, GRIFFITH R, LIU Jun-da, et al. Understanding TCP incast throughput collapse in datacenter networks[C]//Proceedings of WREN Workshop. New York, USA:ACM, 2009:73-82. |
| [24] | VASUDEVAN V, PHANISHAYEE A, SHAH H, et al. Safe and effective fine-grained TCP retransmissions for datacenter communication[C]//Proceedings of ACM SIGCOMM. New York, USA:ACM, 2009:303-314. |
| [25] | KANDULA S, SENGUPTA S, GREENBERG A, et al. The nature of data center traffic:Measurements and analysis[C]//Proceedings of ACM SIGCOMM. New York, USA:ACM, 2009. |
