Considerations for Controlling TCP’s Fairness on End Hosts Michael Welzl Capacity Sharing Workshop Stuttgart 13. 10. 2011.

Slides:



Advertisements
Similar presentations
Network Resource Broker for IPTV in Cloud Computing Lei Liang, Dan He University of Surrey, UK OGF 27, G2C Workshop 15 Oct 2009 Banff,
Advertisements

TCP--Revisited. Background How to effectively share the network? – Goal: Fairness and vague notion of equality Ideal: If N connections, each should get.
A Transport Protocol for Content-Centric Networking with Explicit Congestion Control Feixiong Zhang, Yanyong Zhang (Rutgers Univ.), Alex Reznik (InterDigital),
One control to rule them all Michael Welzl IAB/IRTF CC
TCP Congestion Control Dina Katabi & Sam Madden nms.csail.mit.edu/~dina 6.033, Spring 2014.
Confused, Timid and Unstable: Picking a Video Rate is Hard Te-Yuan (TY) Huang Stanford University Nov 15 th, 2012 Joint work with Nikhil Handigol, Brandon.
Rice Networks Group Aleksandar Kuzmanovic & Edward W. Knightly TCP-LP: A Distributed Algorithm for Low Priority Data Transfer.
1 EE 627 Lecture 11 Review of Last Lecture UDP & Multimedia TCP & UDP Interaction.
Transport Layer 3-1 Fast Retransmit r time-out period often relatively long: m long delay before resending lost packet r detect lost segments via duplicate.
Explicit Congestion Notification ECN Tilo Hamann Technical University Hamburg-Harburg, Germany.
1 Is the Round-trip Time Correlated with the Number of Packets in Flight? Saad Biaz, Auburn University Nitin H. Vaidya, University of Illinois at Urbana-Champaign.
1 EE 689 Lecture 3 Review of Last Lecture UDP & Multimedia TCP & UDP Interaction.
High-performance bulk data transfers with TCP Matei Ripeanu University of Chicago.
Efficient Internet Traffic Delivery over Wireless Networks Sandhya Sumathy.
1 Manpreet Singh, Prashant Pradhan* and Paul Francis * MPAT: Aggregate TCP Congestion Management as a Building Block for Internet QoS.
Lecture 5: Congestion Control l Challenge: how do we efficiently share network resources among billions of hosts? n Last time: TCP n This time: Alternative.
Transport Level Protocol Performance Evaluation for Bulk Data Transfers Matei Ripeanu The University of Chicago Abstract:
P2P File Sharing Systems
Data Transfer Case Study: TCP  Go-back N ARQ  32-bit sequence # indicates byte number in stream  transfers a byte stream, not fixed size user blocks.
MulTFRC: TFRC with weighted fairness draft-welzl-multfrc-00 Michael Welzl, Dragana Damjanovic 75th IETF Meeting Stockholm, Sweden 29 July 2009.
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 r 3.1 Transport-layer services r 3.2 Multiplexing and demultiplexing r 3.3 Connectionless transport: UDP r 3.4 Principles.
TeamViewer. TeamViewer is an easy and friendly way for desktop sharing.
Experiences in Design and Implementation of a High Performance Transport Protocol Yunhong Gu, Xinwei Hong, and Robert L. Grossman National Center for Data.
Potential Applications of Shared Bottleneck Detection (SBD) Michael Welzl NNUW-1 Simula, Oslo
Omer Boyaci, Andrea Forte and Henning Schulzrinne December 16,
Presented by Rajan Includes slides presented by Andrew Sprouse, Northeastern University CISC 856 TCP/IP and Upper Layer Protocols Date:May 03, 2011.
Uni Innsbruck Informatik - 1 We Don‘t Need No Control Plane Michael Welzl, Kashif Munir Michael Welzl DPS NSG Team
Requirements for Simulation and Modeling Tools Sally Floyd NSF Workshop August 2005.
MulTFRC: TFRC with weighted fairness draft-welzl-multfrc-01 Michael Welzl, Dragana Damjanovic 76th IETF Meeting Hiroshima, Japan 10 November2009.
Field Study # 2 #1 - WHAT AM I LEARNING ABOUT THE TEACHING AND LEARNING PROCESS? “I am learning that engagement and learning increases as familiarity.
Transport Layer3-1 TCP throughput r What’s the average throughout of TCP as a function of window size and RTT? m Ignore slow start r Let W be the window.
DCCP: Issues From the Mailing List Sally Floyd, Eddie Kohler, Mark Handley, et al. DCCP WG March 4, 2004.
Self-generated Self-similar Traffic Péter Hága Péter Pollner Gábor Simon István Csabai Gábor Vattay.
Uni Innsbruck Informatik - 1 Doing away with TCP-friendliness: is there a realistic option? Panelists: Michael Welzl (UiO, Norway) Matt Mathis (PSC, USA)
Transport Layer 3-1 Chapter 3 Transport Layer Computer Networking: A Top Down Approach 6 th edition Jim Kurose, Keith Ross Addison-Wesley March
David Wetherall Professor of Computer Science & Engineering Introduction to Computer Networks Fairness of Bandwidth Allocation (§6.3.1)
Random Early Detection (RED) Router notifies source before congestion happens - just drop the packet (TCP will timeout and adjust its window) - could make.
TCP continued. Discussion – TCP Throughput TCP will most likely generate the saw tooth type of traffic. – A rough estimate is that the congestion window.
XCP: eXplicit Control Protocol Dina Katabi MIT Lab for Computer Science
RMCAT architectural overview Michael Welzl 1 RMCAT, 85 th IETF Meeting
Safety in the digital world. Survey Thank you for completing the Information Security Survey!
Peer-to-Peer Networks 13 Internet – The Underlay Network
Receiver-based Management of Low-bandwidth Access Links INFOCOM 2000 March 28, 2000 Neil Spring, Maureen Chesire, Mark Berryman, Vivek Sahasranaman, Thomas.
The Benefits to Applications of using Explicit Congestion Notification (ECN) draft-welzl-ecn-benefits-00 89th IETF Meeting London, UK 4 March 2014 Michael.
Uni Innsbruck Informatik th IETF, PMTUD WG: Path MTU Discovery Using Options draft-welzl-pmtud-options-01.txt Michael Welzl
IT 424 Networks2 IT 424 Networks2 Ack.: Slides are adapted from the slides of the book: “Computer Networking” – J. Kurose, K. Ross Chapter 3: Transport.
Tightly Coupled Congestion Control in WebRTC Michael Welzl Upperside WebRTC conference Paris
Date of Presentation Name of Presenter Insert image _________ Toolkit.
Module 7 Computers Unit 2 When do you use a computer?
@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.
温州市实验中学 陈玫月. Give opinions in different ways. I think students should be allowed to …. I don’t think students should …. I agree / disagree that … I think.
Dr. Adil Yousif University of Alneelian – Master of CS - IT Electronic Mail.
@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.
CIS679: UDP and Multimedia r Review of last lecture r UDP and multimedia.
TCP over Wireless PROF. MICHAEL TSAI 2016/6/3. TCP Congestion Control (TCP Tahoe) Only ACK correctly received packets Congestion Window Size: Maximum.
Lecture (2).
TCP Vegas Congestion Control Algorithm
TCP-in-UDP draft-welzl-irtf-iccrg-tcp-in-udp-00.txt
Muhammad Murtaza Yousaf, Michael Welzl
Generalizing The Network Performance Interference Problem
TCP-LP Distributed Algorithm for Low-Priority Data Transfer
TCP-LP: A Distributed Algorithm for Low Priority Data Transfer
Michael Welzl University of Oslo
Runa Barik, Simone Ferlin, Michael Welzl University of Oslo
Jiyong Park Seoul National University, Korea
TCP Congestion Control
Congestion Michael Freedman COS 461: Computer Networks
Lecture 6, Computer Networks (198:552)
Presentation transcript:

Considerations for Controlling TCP’s Fairness on End Hosts Michael Welzl Capacity Sharing Workshop Stuttgart

2 The problem

3 TCP’s fairness...has been criticized a lot. Hi Bob! Many good reasons –e.g., depending on RTT = technical artifact Here: a very pragmatic, practical view of the problem, and what to do about it

4 How we use the Internet today: 2 stories 1.I clean our flat while listening to Spotify via my wife’s laptop –in parallel, downloading files via my own –Suddenly I begin to think: “please, dear downloads, don’t make the music stop!” 2.I am in a hotel room, using Skype with video to see my daughter –Quality barely good enough –I avoid clicking on anything –Note: that’s different when I talk to my mother...

5 A major problem We may have become used to this, but that doesn’t mean it’s good?! –Would like to specify: do not interrupt Spotify / Skype (or know: do downloads disturb Spotify / Skype or not?) These were just two examples –Downloads can also have different priorities –When I download two files, I try to guess whether the downloads slow each other down

6 So you care more about “performance”? What is it to you?

7 Opinions: 139 of my work colleagues, students, and Facebook “friends”

8

9

10 The solution

11 NOT queue management! (e.g. Linux gateway with tc or GUI tools like NetLimiter) Your access link may not be the bottleneck –Even if the access is likely, it can also be the other side (e.g. P2P, Skype,..) We want TCP to maintain priorities at all times Two cases, both relevant to end users, and separate but interoperating solutions needed: 1.Uploads2. Downloads

12 Uploads Exact control over fairness between N flows across one bottleneck requires cwnd sharing –but need “aggression” of N to avoid being disadvantageous => a good MulTCP-like mechanism –We have PA-MulTCP, CP, MulTFRC, and some more Share cwnd if flows use different paths: very inappropriate behavior –Do this only when traversing the same bottleneck –Need shared bottleneck detection

13 Downloads Need to control the sender –Need signaling extension to TCP Do this only for flows that share bottlenecks –Need shared bottleneck detection

14 Conclusion: Ingredients of the fairness soup Shared bottleneck detection –for the user: know about mutual influence of transfers –for upload and download: control fairness only among flows that share a bottleneck –Solutions exist; have been critized for not being 100% reliable – not a problem for this application?! cwnd sharing –Solutions exist (CM, TCB interdependence (RFC 2140)) Tunable-aggression-TCP –Solutions exist E2E-signaling of fairness requirements –Doesn’t exist?! (... and a GUI that shows transfers by application; existing tools can do that)

15 Thank you! Questions?