Size-based scheduling to improve the performance of short TCP flows

The Internet today carries different types of traffic that have different service requirements. A large fraction of the traffic is either Web traffic requiring low response time or peer-to-peer traffic requiring high throughput. Meeting both performance requirements in a network where routers use droptail or RED for buffer management and FIFO as the service policy is an elusive goal. It is therefore worthwhile to investigate alternative scheduling and buffer management policies for bottleneck links. We propose to use the least attained service (LAS) policy to improve the response time of Web traffic. Under LAS, the next packet to be served is the one belonging to the flow that has received the least amount of service. When the buffer is full, the packet dropped belongs to the flow that has received the most service. We show that under LAS, as compared to FIFO with droptail, the transmission time and loss rate for short TCP flows are significantly reduced, with only a negligible increase in transmission time for the largest flows. The improvement seen by short TCP flows under LAS is mainly due to the way LAS interacts with the TCP protocol in the slow start phase, which results in shorter round-trip times and zero loss rates for short flows.

[1]  Liang Guo,et al.  The war between mice and elephants , 2001, Proceedings Ninth International Conference on Network Protocols. ICNP 2001.

[2]  Bill Lin,et al.  Fast and scalable priority queue architecture for high-speed network switches , 2000, Proceedings IEEE INFOCOM 2000. Conference on Computer Communications. Nineteenth Annual Joint Conference of the IEEE Computer and Communications Societies (Cat. No.00CH37064).

[3]  Eric J. Friedman,et al.  Fairness and efficiency in web server protocols , 2003, SIGMETRICS '03.

[4]  John S. Heidemann,et al.  Preferential treatment for short flows to reduce web latency , 2003, Comput. Networks.

[5]  Sally Floyd,et al.  Wide-area traffic: the failure of Poisson modeling , 1994 .

[6]  Hui Zhang,et al.  Service disciplines for guaranteed performance service in packet-switching networks , 1995, Proc. IEEE.

[7]  Konstantin Avrachenkov,et al.  Differentiation between short and long TCP flows: predictability of the response time , 2004, IEEE INFOCOM 2004.

[8]  Nsf Ncr,et al.  A Generalized Processor Sharing Approach to Flow Control in Integrated Services Networks: The Single Node Case* , 1991 .

[9]  Ibrahim Matta,et al.  Differentiated control of web traffic: a numerical analysis , 2002, SPIE ITCom.

[10]  J. W. Roberts,et al.  Internet traffic, QoS, and pricing , 2004, Proceedings of the IEEE.

[11]  Chienhua Chen,et al.  Service disciplines for guaranteed performance service , 1997, Proceedings Fourth International Workshop on Real-Time Computing Systems and Applications.

[12]  Robert B. Cooper,et al.  Queueing systems, volume II: computer applications : By Leonard Kleinrock. Wiley-Interscience, New York, 1976, xx + 549 pp. , 1977 .

[13]  Abhay Parekh,et al.  A generalized processor sharing approach to flow control in integrated services networks-the single node case , 1992, [Proceedings] IEEE INFOCOM '92: The Conference on Computer Communications.

[14]  Helen J. Wang,et al.  Server-based inference of Internet link lossiness , 2003, IEEE INFOCOM 2003. Twenty-second Annual Joint Conference of the IEEE Computer and Communications Societies (IEEE Cat. No.03CH37428).

[15]  QUTdN QeO,et al.  Random early detection gateways for congestion avoidance , 1993, TNET.

[16]  Carey L. Williamson,et al.  A case for context-aware TCP/IP , 2002, PERV.

[17]  Sally Floyd,et al.  Wide area traffic: the failure of Poisson modeling , 1995, TNET.

[18]  Xipeng Xiao,et al.  Internet QoS: a big picture , 1999, IEEE Netw..

[19]  Guillaume Urvoy-Keller,et al.  Analysis of LAS scheduling for job size distributions with high variance , 2003, SIGMETRICS '03.

[20]  kc claffy,et al.  The nature of the beast: Recent traffic measurements from an Internet backbone , 1998 .