Performance Engineering E2EpiPEs and FastTCP Internet2 member meeting - Indianapolis World Telecom 2003 - Geneva October 15, 2003

Slides:



Advertisements
Similar presentations
Helping TCP Work at Gbps Cheng Jin the FAST project at Caltech
Advertisements

TCP transfers over high latency/bandwidth network & Grid TCP Sylvain Ravot
Ahmed El-Hassany CISC856: CISC 856 TCP/IP and Upper Layer Protocols Slides adopted from: Injong Rhee, Lisong Xu.
Presentation by Joe Szymanski For Upper Layer Protocols May 18, 2015.
Advanced Computer Networking Congestion Control for High Bandwidth-Delay Product Environments (XCP Algorithm) 1.
Maximizing End-to-End Network Performance Thomas Hacker University of Michigan October 5, 2001.
Fair queueing and congestion control Jim Roberts (France Telecom) Joint work with Jordan Augé Workshop on Congestion Control Hamilton Institute, Sept 2005.
Congestion Control on High-Speed Networks
Efficient Network Protocols for Data-Intensive Worldwide Grids Seminar at JAIST, Japan 3 March 2003 T. Kelly, University of Cambridge, UK S. Ravot, Caltech,
Congestion Control Tanenbaum 5.3, /12/2015Congestion Control (A Loss Based Technique: TCP)2 What? Why? Congestion occurs when –there is no reservation.
High speed TCP’s. Why high-speed TCP? Suppose that the bottleneck bandwidth is 10Gbps and RTT = 200ms. Bandwidth delay product is packets (1500.
TCP on High-Speed Networks Sangtae Ha and Injong Rhee North Carolina State University.
High-performance bulk data transfers with TCP Matei Ripeanu University of Chicago.
TCP Congestion Control TCP sources change the sending rate by modifying the window size: Window = min {Advertised window, Congestion Window} In other words,
1 Emulating AQM from End Hosts Presenters: Syed Zaidi Ivor Rodrigues.
Transport Level Protocol Performance Evaluation for Bulk Data Transfers Matei Ripeanu The University of Chicago Abstract:
Congestion Control for High Bandwidth-delay Product Networks Dina Katabi, Mark Handley, Charlie Rohrs.
Congestion Control for High Bandwidth-Delay Product Environments Dina Katabi Mark Handley Charlie Rohrs.
Introduction 1 Lecture 14 Transport Layer (Congestion Control) slides are modified from J. Kurose & K. Ross University of Nevada – Reno Computer Science.
KEK Network Qi Fazhi KEK SW L2/L3 Switch for outside connections Central L2/L3 Switch A Netscreen Firewall Super Sinet Router 10GbE 2 x GbE IDS.
Courtesy: Nick McKeown, Stanford 1 TCP Congestion Control Tahir Azim.
Transport Layer 4 2: Transport Layer 4.
Transport Layer3-1 Chapter 3 outline r 3.1 Transport-layer services r 3.2 Multiplexing and demultiplexing r 3.3 Connectionless transport: UDP r 3.4 Principles.
Transport Layer3-1 Chapter 3 outline 3.1 Transport-layer services 3.2 Multiplexing and demultiplexing 3.3 Connectionless transport: UDP 3.4 Principles.
Experiences in Design and Implementation of a High Performance Transport Protocol Yunhong Gu, Xinwei Hong, and Robert L. Grossman National Center for Data.
Large File Transfer on 20,000 km - Between Korea and Switzerland Yusung Kim, Daewon Kim, Joonbok Lee, Kilnam Chon
J. Bunn, D. Nae, H. Newman, S. Ravot, X. Su, Y. Xia California Institute of Technology High speed WAN data transfers for science Session Recent Results.
J. Bunn, D. Nae, H. Newman, S. Ravot, X. Su, Y. Xia California Institute of Technology State of the art in the use of long distance network International.
Maximizing End-to-End Network Performance Thomas Hacker University of Michigan October 26, 2001.
Masaki Hirabaru Internet Architecture Group GL Meeting March 19, 2004 High Performance Data transfer on High Bandwidth-Delay Product Networks.
Technology for Using High Performance Networks or How to Make Your Network Go Faster…. Robin Tasker UK Light Town Meeting 9 September.
Principles of Congestion Control Congestion: informally: “too many sources sending too much data too fast for network to handle” different from flow control!
UDT: UDP based Data Transfer Yunhong Gu & Robert Grossman Laboratory for Advanced Computing University of Illinois at Chicago.
UDT: UDP based Data Transfer Protocol, Results, and Implementation Experiences Yunhong Gu & Robert Grossman Laboratory for Advanced Computing / Univ. of.
NORDUnet 2003, Reykjavik, Iceland, 26 August 2003 High-Performance Transport Protocols for Data-Intensive World-Wide Grids T. Kelly, University of Cambridge,
Congestion Control - Supplementary Slides are adapted on Jean Walrand’s Slides.
Data transfer over the wide area network with a large round trip time H. Matsunaga, T. Isobe, T. Mashimo, H. Sakamoto, I. Ueda International Center for.
Comparison of Public End-to-End Bandwidth Estimation tools on High-Speed Links Alok Shriram, Margaret Murray, Young Hyun, Nevil Brownlee, Andre Broido,
Experience with Loss-Based Congestion Controlled TCP Stacks Yee-Ting Li University College London.
27th, Nov 2001 GLOBECOM /16 Analysis of Dynamic Behaviors of Many TCP Connections Sharing Tail-Drop / RED Routers Go Hasegawa Osaka University, Japan.
FAST TCP in Linux Cheng Jin David Wei Steven Low California Institute of Technology.
High TCP performance over wide area networks Arlington, VA May 8, 2002 Sylvain Ravot CalTech HENP Working Group.
High-speed TCP  FAST TCP: motivation, architecture, algorithms, performance (by Cheng Jin, David X. Wei and Steven H. Low)  Modifying TCP's Congestion.
Masaki Hirabaru Tsukuba WAN Symposium 2005 March 8, 2005 e-VLBI and End-to-End Performance over Global Research Internet.
Rate Control Rate control tunes the packet sending rate. No more than one packet can be sent during each packet sending period. Additive Increase: Every.
Masaki Hirabaru NICT Koganei 3rd e-VLBI Workshop October 6, 2004 Makuhari, Japan Performance Measurement on Large Bandwidth-Delay Product.
First of ALL Big appologize for Kei’s absence Hero of this year’s LSR achievement Takeshi in his experiment.
TERENA Networking Conference, Zagreb, Croatia, 21 May 2003 High-Performance Data Transport for Grid Applications T. Kelly, University of Cambridge, UK.
Transport Layer 3-1 Chapter 3 Transport Layer Computer Networking: A Top Down Approach 6 th edition Jim Kurose, Keith Ross Addison-Wesley March
CS640: Introduction to Computer Networks Aditya Akella Lecture 15 TCP – III Reliability and Implementation Issues.
CS640: Introduction to Computer Networks Aditya Akella Lecture 15 TCP – III Reliability and Implementation Issues.
Transport Layer3-1 Chapter 3 outline r 3.1 Transport-layer services r 3.2 Multiplexing and demultiplexing r 3.3 Connectionless transport: UDP r 3.4 Principles.
Transport Layer 3- Midterm score distribution. Transport Layer 3- TCP congestion control: additive increase, multiplicative decrease Approach: increase.
TCP transfers over high latency/bandwidth networks Internet2 Member Meeting HENP working group session April 9-11, 2003, Arlington T. Kelly, University.
Winter 2008CS244a Handout 71 CS244a: An Introduction to Computer Networks Handout 7: Congestion Control Nick McKeown Professor of Electrical Engineering.
Chapter 11.4 END-TO-END ISSUES. Optical Internet Optical technology Protocol translates availability of gigabit bandwidth in user-perceived QoS.
TCP continued. Discussion – TCP Throughput TCP will most likely generate the saw tooth type of traffic. – A rough estimate is that the congestion window.
TCP transfers over high latency/bandwidth networks & Grid DT Measurements session PFLDnet February 3- 4, 2003 CERN, Geneva, Switzerland Sylvain Ravot
Final EU Review - 24/03/2004 DataTAG is a project funded by the European Commission under contract IST Richard Hughes-Jones The University of.
S. Ravot, J. Bunn, H. Newman, Y. Xia, D. Nae California Institute of Technology CHEP 2004 Network Session September 1, 2004 Breaking the 1 GByte/sec Barrier?
An Analysis of AIMD Algorithm with Decreasing Increases Yunhong Gu, Xinwei Hong, and Robert L. Grossman National Center for Data Mining.
Masaki Hirabaru (NICT) and Jin Tanaka (KDDI) Impact of Bottleneck Queue on Long Distant TCP Transfer August 25, 2005 NOC-Network Engineering Session Advanced.
@Yuan Xue A special acknowledge goes to J.F Kurose and K.W. Ross Some of the slides used in this lecture are adapted from their.
@Yuan Xue A special acknowledge goes to J.F Kurose and K.W. Ross Some of the slides used in this lecture are adapted from their.
Karn’s Algorithm Do not use measured RTT to update SRTT and SDEV Calculate backoff RTO when a retransmission occurs Use backoff RTO for segments until.
Chapter 3 outline 3.1 transport-layer services
TCP Performance over a 2.5 Gbit/s Transatlantic Circuit
High-Performance Data Transport for Grid Applications
Presentation transcript:

Performance Engineering E2EpiPEs and FastTCP Internet2 member meeting - Indianapolis World Telecom Geneva October 15, 2003

Agenda è High TCP performance over wide area networks : r TCP at Gbps speed r MTU bias r RTT bias r TCP fairness è How to use 100% of the link capacity with TCP Reno r Network buffers impact è New Internet2 Land Speed record

Single TCP stream performance under periodic losses Loss rate =0.01%: è LAN BW utilization= 99% è WAN BW utilization=1.2% Bandwidth available = 1 Gbps u TCP throughput is much more sensitive to packet loss in WANs than in LANs r TCP’s congestion control algorithm (AIMD) is not suited to gigabit networks r Poor limited feedback mechanisms r The effect of packets loss is disastrous u TCP is inefficient in high bandwidth*delay networks u The future performance of computational grids looks bad if we continue to rely on the widely-deployed TCP RENO

Responsiveness (I)  The responsiveness  measures how quickly we go back to using the network link at full capacity after experiencing a loss if we assume that the congestion window size is equal to the Bandwidth Delay product when the packet is lost.  C. RTT 2. MSS 2 C : Capacity of the link

Responsiveness (II) CaseC RTT (ms) MSS (Byte) Responsiveness Typical LAN today 1 Gb/s 2 (worst case) ms WAN Geneva Chicago 1 Gb/s min WAN Geneva Sunnyvale 1 Gb/s min WAN Geneva Tokyo 1 Gb/s h 04 min WAN Geneva Sunnyvale 2.5 Gb/s min Future WAN CERN Starlight 10 Gb/s h 32 min Future WAN link CERN Starlight 10 Gb/s (Jumbo Frame) 15 min The Linux kernel 2.4.x implements delayed acknowledgment. Due to delayed acknowledgments, the responsiveness is multiplied by two. Therefore, values above have to be multiplied by two !

Single TCP stream TCP connection between Geneva and Chicago: C=1 Gbit/s; MSS=1,460 Bytes; RTT=120ms u Time to increase the throughout from 100Mbps to 900Mbps = 35 minutes u Loss occurs when the bandwidth reaches the pipe size u 75% of bandwidth utilization (assuming no buffering) u Cwnd<BDP : è Throughput < Bandwidth è RTT constant è Throughput = Cwnd / RTT 35 minutes

Measurements with Different MTUs TCP connection between Geneva and Chicago: C=1 Gbit/s; RTT=120ms u In both cases: 75% of the link utilization u Large MTU accelerate the growth of the window u Time to recover from a packet loss decreases with large MTU u Larger MTU reduces overhead per frames (saves CPU cycles, reduces the number of packets)

Starlight (Chi) CERN (GVA) MTU and Fairness u Two TCP streams share a 1 Gbps bottleneck u RTT=117 ms u MTU = 1500 Bytes; Avg. throughput over a period of 4000s = 50 Mb/s u MTU = 9000 Bytes; Avg. throughput over a period of 4000s = 698 Mb/s u Factor 14 ! u Connections with large MTU increase quickly their rate and grab most of the available bandwidth RR GbE Switch Host #1 POS 2.5 Gbps 1 GE Host #2 Host #1 Host #2 1 GE Bottleneck

Sunnyvale Starlight (Chi) CERN (GVA) RTT and Fairness RR GbE Switch Host #1 POS 2.5 Gb/s 1 GE Host #2 Host #1 Host #2 1 GE Bottleneck R POS 10 Gb/s R 10GE u Two TCP streams share a 1 Gbps bottleneck u CERN Sunnyvale RTT=181ms ; Avg. throughput over a period of 7000s = 202Mb/s u CERN Starlight RTT=117ms; Avg. throughput over a period of 7000s = 514Mb/s u MTU = 9000 bytes u Connection with small RTT increases quickly there rate and grab most of the available bandwidth

How to use 100% of the bandwidth? Bandwidth delay product u Single TCP stream GVA - CHI u MSS=8960 Bytes; Throughput = 980Mbps u Cwnd > BDP => Throughput = Bandwidth u RTT increase u Extremely Large buffer at the bottleneck u Network buffers have an important impact on performance u Have buffers to be well dimensioned in order to scale with the BDP? u Why not use the end-to-end delay as congestion indication.

Single stream TCP performance Date From Geneva to Size of transfer Duration(second)RTT(ms)MTU(Bytes) IP version ThroughputRecordAward Feb 27 Sunnyvale 1,1 TByte IPv Gbps u Internet2 LSR u CENIC award u Guinness World Record May 27 Tokyo 65.1 GByte IPv4 931 Mbps May 2 Chicago 385 GByte IPv6 919 Mbps May 2 Chicago 412 GByte IPv6 983 Mbps u Internet2 LSR NEW Submission (Oct-11): 5.65 Gbps from Geneva to Los Angeles across the LHCnet, Starlight, Abilene and CENIC.

Early 10 Gb/s 10,000 km TCP Testing u Single TCP stream at 5,65 Gbps u Transferring a full CD in less than 1s u Un-congestioned network u No packet loss during the transfer u Probably qualifies as new Internet2 LSR Monitoring of the Abilene traffic in LA

Conclusion u The future performance of computational grids looks bad if we continue to rely on the widely-deployed TCP RENO u How to define the fairness? è Taking into account the MTU è Taking into account the RTT u Larger packet size (Jumbogram : payload larger than 64K) è Is standard MTU the largest bottleneck? è New Intel 10GE cards : MTU=16K è J. Cain (Cisco): “It’s very difficult to build switches to switch large packets such as jumbogram” u Our vision of the network: “The network, once viewed as an obstacle for virtual collaborations and distributed computing in grids, can now start to be viewed as a catalyst instead. Grid nodes distributed around the world will simply become depots for dropping off information for computation or storage, and the network will become the fundamental fabric for tomorrow's computational grids and virtual supercomputers”