Low-complexity multi-resource packet scheduling for network function virtualization

Network functions are widely deployed in modern networks, providing various network services ranging from intrusion detection to HTTP caching. Various virtual network function instances can be consolidated into one physical middlebox. Depending on the type of services, packet processing for different flows consumes different hardware resources in the middlebox. Previous solutions of multi-resource packet scheduling suffer from high computational complexity and memory cost for packet buffering and scheduling, especially when the number of flows is large. In this paper, we design a novel low-complexity and space-efficient packet scheduling algorithm called Myopia, which supports multi-resource environments such as network function virtualization. Myopia is developed based upon the fact that most Internet traffic is contributed by a small fraction of elephant flows. Myopia schedules elephant flows with precise control and treats mice flows using FIFO, to achieve simplicity of packet buffering and scheduling. We will demonstrate, via theoretical analysis, prototype implementation, and simulations, that Myopia achieves multi-resource fairness at low cost with short packet delay.

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

[2]  Eddie Kohler,et al.  The Click modular router , 1999, SOSP.

[3]  Minlan Yu,et al.  Software Defined Traffic Measurement with OpenSketch , 2013, NSDI.

[4]  M. Newman.  Power laws, Pareto distributions and Zipf's law , 2005 .

[5]  Scott Shenker,et al.  Analysis and simulation of a fair queueing algorithm , 1989, SIGCOMM '89.

[6]  David A. Maltz,et al.  Network traffic characteristics of data centers in the wild , 2010, IMC '10.

[7]  Xin Li,et al.  Distributed and collaborative traffic monitoring in software defined networks , 2014, HotSDN.

[8]  Abhay Parekh,et al.  A generalized processor sharing approach to flow control in integrated services networks: the single-node case , 1993, TNET.

[9]  Lixia Zhang,et al.  VirtualClock: a new traffic control algorithm for packet-switched networks , 1991, TOCS.

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

[11]  Vyas Sekar,et al.  Multi-resource fair queueing for packet processing , 2012, CCRV.

[12]  Amin Vahdat,et al.  A scalable, commodity data center network architecture , 2008, SIGCOMM '08.

[13]  Walter Willinger,et al.  Self-similarity through high-variability: statistical analysis of Ethernet LAN traffic at the source level , 1997, TNET.

[14]  Michalis Faloutsos,et al.  On power-law relationships of the Internet topology , 1999, SIGCOMM '99.

[15]  Azer Bestavros,et al.  Self-similarity in World Wide Web traffic: evidence and possible causes , 1996, SIGMETRICS '96.

[16]  Ming Zhang,et al.  An untold story of middleboxes in cellular networks , 2011, SIGCOMM.

[17]  Neelam Goyal,et al.  Signature Matching in Network Processing using SIMD / GPU Architectures , 2007 .

[18]  Baochun Li,et al.  Low complexity multi-resource fair queueing with bounded delay , 2014, IEEE INFOCOM 2014 - IEEE Conference on Computer Communications.

[19]  Xin Li,et al.  Distributed Collaborative Monitoring in Software Defined Networks , 2014, ArXiv.

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

[21]  Vyas Sekar,et al.  Design and Implementation of a Consolidated Middlebox Architecture , 2012, NSDI.

[22]  Srinivasan Seshan,et al.  Packet caches on routers: the implications of universal redundant traffic elimination , 2008, SIGCOMM '08.

[23]  Benjamin Hindman,et al.  Dominant Resource Fairness: Fair Allocation of Multiple Resource Types , 2011, NSDI.

[24]  Chen Qian,et al.  ROME: Routing on metropolitan-scale Ethernet , 2012, 2012 20th IEEE International Conference on Network Protocols (ICNP).

[25]  George Varghese,et al.  New directions in traffic measurement and accounting , 2002, CCRV.

[26]  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.

[27]  Wenzhi Cui,et al.  DiFS: Distributed flow scheduling for adaptive routing in hierarchical data center networks , 2014, 2014 ACM/IEEE Symposium on Architectures for Networking and Communications Systems (ANCS).

[28]  Graham Cormode,et al.  An improved data stream summary: the count-min sketch and its applications , 2004, J. Algorithms.

[29]  Harrick M. Vin,et al.  Start-time fair queueing: a scheduling algorithm for integrated services packet switching networks , 1996, SIGCOMM '96.

[30]  Michel L. Goldstein,et al.  Problems with fitting to the power-law distribution , 2004, cond-mat/0402322.

[31]  Ratul Mahajan,et al.  Measuring ISP topologies with Rocketfuel , 2004, IEEE/ACM Transactions on Networking.

[32]  Sylvia Ratnasamy,et al.  A Survey of Enterprise Middlebox Deployments , 2012 .

[33]  T. V. Lakshman,et al.  Design considerations for supporting TCP with per-flow queueing , 1998, Proceedings. IEEE INFOCOM '98, the Conference on Computer Communications. Seventeenth Annual Joint Conference of the IEEE Computer and Communications Societies. Gateway to the 21st Century (Cat. No.98.

[34]  2015 IEEE Conference on Computer Communications, INFOCOM 2015, Kowloon, Hong Kong, April 26 - May 1, 2015 , 2015, IEEE Conference on Computer Communications.

[35]  Ye Yu,et al.  Space Shuffle: A Scalable, Flexible, and High-Bandwidth Data Center Network , 2014, ICNP.

[36]  M. E. J. Newman,et al.  Power laws, Pareto distributions and Zipf's law , 2005 .

[37]  Baochun Li,et al.  Multi-Resource Round Robin: A low complexity packet scheduler with Dominant Resource Fairness , 2013, 2013 21st IEEE International Conference on Network Protocols (ICNP).