From vinayakh at gmail.com Wed Sep 1 06:49:50 2010 From: vinayakh at gmail.com (Vinayak Hegde) Date: Wed, 1 Sep 2010 19:19:50 +0530 Subject: [e2e] References for TCP congestion algorithms Message-ID: Are there any good references online (or research papers) that compare and contrast current (or recent implementations) of congestion algorithms in Linux/freebsd such as TCP Vegas, Reno, bic, cubic etc. Also what works best for long haul high latency networks would be helpful. -- Vinayak From faber at isi.edu Wed Sep 1 09:16:11 2010 From: faber at isi.edu (Ted Faber) Date: Wed, 1 Sep 2010 09:16:11 -0700 Subject: [e2e] References for TCP congestion algorithms In-Reply-To: References: Message-ID: <20100901161611.GB6796@zod.isi.edu> On Wed, Sep 01, 2010 at 07:19:50PM +0530, Vinayak Hegde wrote: > Are there any good references online (or research papers) that compare > and contrast current (or recent implementations) of congestion > algorithms in Linux/freebsd such as TCP Vegas, Reno, bic, cubic etc. > Also what works best for long haul high latency networks would be > helpful. Some of this discussion goes on in the ICCRG (Internet Congestion Control Research Group). I've only been skimming the list lately, but most of the players have some representation there. The mailman link is http://oakham.cs.ucl.ac.uk/mailman/listinfo/iccrg -- Ted Faber http://www.isi.edu/~faber PGP: http://www.isi.edu/~faber/pubkeys.asc Unexpected attachment on this mail? See http://www.isi.edu/~faber/FAQ.html#SIG -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 196 bytes Desc: not available Url : http://mailman.postel.org/pipermail/end2end-interest/attachments/20100901/0a369427/attachment.bin From pganti at gmail.com Wed Sep 1 13:07:58 2010 From: pganti at gmail.com (Paddy Ganti) Date: Wed, 1 Sep 2010 13:07:58 -0700 Subject: [e2e] References for TCP congestion algorithms In-Reply-To: References: Message-ID: Some links that can be perused to answer your question http://www.hamilton.ie/net/eval.htm http://research.microsoft.com/en-us/um/people/padhye/ -Paddy On Wed, Sep 1, 2010 at 6:49 AM, Vinayak Hegde wrote: > Are there any good references online (or research papers) that compare > and contrast current (or recent implementations) of congestion > algorithms in Linux/freebsd such as TCP Vegas, Reno, bic, cubic etc. > Also what works best for long haul high latency networks would be > helpful. > > -- Vinayak > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.postel.org/pipermail/end2end-interest/attachments/20100901/fd93ce09/attachment.html From Doug.Leith at nuim.ie Thu Sep 2 08:55:34 2010 From: Doug.Leith at nuim.ie (Douglas Leith) Date: Thu, 02 Sep 2010 16:55:34 +0100 Subject: [e2e] References for TCP congestion algorithms Message-ID: Here are a few links to recent papers by the group here at the hamilton institute: http://www.hamilton.ie/net/ton_eval.pdf (hs-tcp, scalable, bic, htcp, fast) http://www.hamilton.ie/net/pfldnet2007_cubic_final.pdf (cubic) http://www.hamilton.ie/net/delay_tests_final.pdf (compound, illinois) also some by the folks at Swinburne: http://dx.doi.org/10.1007/978-3-642-01399-7_31 and by NIST: http://www.antd.nist.gov/Measurement%20Science%20for%20Complex%20Information%20Systems_files/ICC-study-Draft.htm Doug > From: Vinayak Hegde > Subject: [e2e] References for TCP congestion algorithms > To: End-to-End > Message-ID: > > Content-Type: text/plain; charset=ISO-8859-1 > > Are there any good references online (or research papers) that compare > and contrast current (or recent implementations) of congestion > algorithms in Linux/freebsd such as TCP Vegas, Reno, bic, cubic etc. > Also what works best for long haul high latency networks would be > helpful. > > -- Vinayak > From sangtaeh at princeton.edu Thu Sep 2 16:11:18 2010 From: sangtaeh at princeton.edu (Sangtae Ha) Date: Thu, 2 Sep 2010 19:11:18 -0400 Subject: [e2e] References for TCP congestion algorithms In-Reply-To: References: Message-ID: All testing results related to CUBIC are available at http://netsrv.csc.ncsu.edu/wiki/index.php/TCP_Testing. You can click each of testing run. Some of highlights are: Internet2 testing results (BIC, CUBIC, HTCP, HSTCP and STCP) http://netsrv.csc.ncsu.edu/wiki/index.php/Testing_between_Korea_and_Japan Intra-protocol testing results (CUBIC, HTCP, HSTCP and TCP-SACK) http://netsrv.csc.ncsu.edu/wiki/index.php/Intra_protocol_fairness_testing_with_kernel_2.6.18-rc4 TCP-friendliness testing results (CUBIC, HTCP, and HSTCP) http://netsrv.csc.ncsu.edu/wiki/index.php/TCP_friendliness_testing_with_kernel_2.6.18-rc4 Spare capacity testing results (CUBIC, HTCP, HSTCP and TCP-SACK) http://netsrv.csc.ncsu.edu/wiki/index.php/Spare_capacity_usage_testing_with_kernel_2.6.18-rc4 Impact of background traffic results (CUBIC, HTCP, HSTCP and TCP-SACK) http://netsrv.csc.ncsu.edu/wiki/index.php/Impact_of_background_traffic_testing_with_kernel_2.6.18-rc4 Stability of high-speed protocols (CUBIC, BIC, HTCP, HSTCP, STCP, FAST, TCP-SACK and Westwood) Paper: http://www.csc.ncsu.edu/faculty/rhee/export/stochastic-ordering.pdf Results: http://netsrv.csc.ncsu.edu/convex-ordering/ Fairness, convergence time, RTT fairness, and TCP friendliness with and without background traffic (CUBIC, BIC, HTCP, HSTCP, STCP, FAST and TCP-SACK) Tech Report: http://www.csc.ncsu.edu/faculty/rhee/export/bitcp/asteppaper.htm Results: http://netsrv.csc.ncsu.edu/highspeed Rebuttal to CUBIC analysis paper at PFLDnet 2007 http://www.csc.ncsu.edu/faculty/rhee/Rebuttal-LSM-new.pdf Sangtae On Wed, Sep 1, 2010 at 9:49 AM, Vinayak Hegde wrote: > Are there any good references online (or research papers) that compare > and contrast current (or recent implementations) of congestion > algorithms in Linux/freebsd such as TCP Vegas, Reno, bic, cubic etc. > Also what works best for long haul high latency networks would be > helpful. > > -- Vinayak > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://mailman.postel.org/pipermail/end2end-interest/attachments/20100902/ce52ceeb/attachment.html From jatinder at stanford.edu Sun Sep 19 12:37:21 2010 From: jatinder at stanford.edu (Jatinder Pal Singh) Date: Sun, 19 Sep 2010 12:37:21 -0700 (PDT) Subject: [e2e] Call for Workshop Proposals - IEEE INFOCOM 2011 Message-ID: The 30th IEEE International Conference on Computer Communications (IEEE INFOCOM 2011) Call for Workshop Proposals The IEEE INFOCOM 2011 conference will be held in Shanghai, China, April 10-15, 2011.The program will feature technical sessions, tutorials, panel discussions and workshops. The aims of the workshops are to explore special topics, and provide a forum for researchers to present early research results and share experiences focused on more specific research areas than the main conference. There were four workshops at Infocom 2010, and the organizers of Infocom 2011 welcome more quality workshops in 2011, and encourage new workshop proposals on emerging topics including Cloud Computing, Green Computing/Communications, Cyber-Physical Systems, Smart Grid, Wireless/mobile networking for the Developing World, etc. Each workshop will have 4 sessions of 1 1/2 hours each. The dates for paper submission, notification and camera-ready version can vary with each workshop, but in general, papers are due in December or January, with camera-ready version due in late February or early March. All workshop papers will appear in IEEE Xplore (and are EI Indexed). The workshop proposal, in plain text or PDF, should provide the following information: 1. Title, scope and topics of the workshop 2. Names, addresses, and affiliation of Workshop organizers 3. Tentative committee lists (organizers, steering...) 4. Contribution format (short papers, long papers, demos, panels...) 5. Past history of the workshop (if any) 6. Rationale - reasons why the workshop would attract enough submissions, and why is the topic important. Please send a draft call for papers, if it is available. Workshop organizers should send their workshop proposals to the workshop co-chairs at the email addresses given below by October 1, 2010. Notification of acceptance for workshop proposals will be on October 15, 2010. Workshop Co-Chairs Chunming Qiao - State University of New York, Buffalo, USA Ivan Stojmenovic - University of Ottawa, Canada, ------------------------------------------------- Organizing Committee General Co-Chairs: Lionel Ni ( Hong Kong University of Science and Technology, Hong Kong ) Wenjun Zhang ( Shanghai Jiao Tong University , P.R. China) General Vice Chair: Minglu Li ( Shanghai Jiao Tong University , P.R. China) Technical Program Co-Chairs: Byrav Ramamurthy ( University of Nebraska-Lincoln , USA ) Jie Wu ( Temple University , USA ) Qian Zhang ( Hong Kong University of Science and Technology, Hong Kong ) Standing Committee Chair: Harvey Freeman (HAF Consulting, Inc., USA ) -------------------------------------------- Jatinder Pal Singh Vice President of Research Deutsche Telekom, Inc. Consulting Assistant Professor Stanford University Homepage: http://www.stanford.edu/~jatinder From nspring at cs.umd.edu Mon Sep 20 08:16:33 2010 From: nspring at cs.umd.edu (Neil Spring) Date: Mon, 20 Sep 2010 11:16:33 -0400 Subject: [e2e] PAM 2011 CFP Message-ID: Call for Papers Passive and Active Measurement Conference (PAM) 2011 Atlanta, GA, USA Conference Web: http://pam2011.gatech.edu/ Submissions: http://pam11.cs.umd.edu/ The 12th Passive and Active Measurement conference will be held March 20-22nd, 2011 in Atlanta. PAM focuses on research and practical applications of network measurement and analysis techniques. The conference's goal is to provide a forum for current work in its early stages. Original papers are invited from the research and operations communities on topics including, but not limited to: * Active Network Measurements * Passive Network Measurements * Performance Metrics * Traffic Statistics * Measurement Visualization * New Measurement Approaches & Techniques * Deployment of Measurement Infrastructure * New Measurement Initiatives * Applications of Network Measurements * Network Measurements and Security * Network Troubleshooting using Measurements * Reproduce (or Refute) Previous Measurement Results The PAM steering committee believes that releasing measurement data allows for better science to be conducted in the field of network measurement. Therefore, an award will be given at PAM 2011 for the best paper based on a new dataset that the authors are releasing for community use in subsequent research. To qualify, a paper must significantly utilize a dataset that has been collected for the work presented in the paper. Further, the dataset must be freely available to any researcher; wireless data sets may, for instance, be published through CRAWDAD. Authors should flag papers they wish to be considered for this award with a footnote on the first page of the paper. Novel datasets are especially encouraged. The awarded paper will be chosen from the set of qualifying papers accepted for the conference by a committee made up of a subset of the program and steering committees. Important Dates * Abstract Registration: October 2, 2010 23:59pm GMT (UK Time) * Paper Submission: October 9, 2010 23:59pm GMT (UK Time) * Conference: March 20-22, 2011 Conference Organizers General Chair: George Riley, Georgia Institute of Technology Program Chair: Neil Spring, University of Maryland Program Committee Nevil Brownlee, University of Auckland kc claffy, CAIDA Benoit Donnet, Laboratoire d'Informatique de Paris 6 Constantine Dovrolis, Georgia Tech Mehmet Gunes, University of Nevada, Reno Gianluca Iannaccone, Intel Research Arvind Krisnamurthy, University of Washington Aleksandar Kuzmanovic, Northwestern University Sridhar Machiraju, Google Alan Mislove, Northeastern University Ricardo Oliveira, Thousandeyes Dina Papagiannaki, Intel Research George Riley, Georgia Tech Eve Schooler, Intel Research Rob Sherwood, Deutsche Telekom Inc. R&D Lab Fernando Silveira, Technicolor Joel Sommers, Colgate University Oliver Spatschek, AT&T Labs -- Research Nina Taft, Intel Research Arun Venkataramani, University of Massachusetts Walter Willinger, AT&T Labs -- Research _______________________________________________ IMRG mailing list IMRG at irtf.org https://www.irtf.org/mailman/listinfo/imrg From touch at isi.edu Mon Sep 20 14:05:15 2010 From: touch at isi.edu (Joe Touch) Date: Mon, 20 Sep 2010 14:05:15 -0700 Subject: [e2e] CCW 2010 Call for Participation Message-ID: <4C97CC8B.1060905@isi.edu> We cordially invite you to join us: IEEE Computer Communications Workshop (CCW) 2010 Lake Arrowhead Resort Lake Arrowhead, CA http://www.ieee-ccw.org/ CCW is the annual flagship workshop of the IEEE Communications Society's Technical Committee on Computer Communications (TCCC). It is a panel-based workshop with informal, interactive sessions exploring emerging issues and trends in networking and computer communications. This year's panels include: - a keynote by Mostafa Ammar, Recipient of the 2009 IEEE TCCC Service Award - two industry panels on smart grid issues - panels on satellite networking, quantum crypto, and social networking - panels on network architecture, including future networks Please kindly note: - The workshop program spans 2.5 days from Monday to Wednesday noon - Early registration deadline: Oct 3, 2010 - Hotel reservation cut-off due: Sep 24, 2010 TCCC welcomes new members and encourages anyone interested to join and participate in our technical activities. Joining TCCC only requires that you subscribe to the TCCC mailing list. Instructions on how to join our mailing list can be found under the Mailing List heading of the TCCC Home Page: http://www.comsoc.org/~tccc/ In case you need more information on TCCC, feel free to contact any of the current TCCC officers. We look forward to seeing you soon in Lake Arrowhead! Joe Touch, USC/ISI and Gene Tsudik, UCI CCW 2010 Chairs From rhee at ncsu.edu Wed Sep 22 08:39:31 2010 From: rhee at ncsu.edu (Injong Rhee) Date: Wed, 22 Sep 2010 11:39:31 -0400 Subject: [e2e] Call for papers -- PFLDNeT Deadline 10/8/2010 Message-ID: <4C9A2333.2060403@ncsu.edu> Hi Folks, I would like to bring your attention to PFLDNeT 2010. This year, we are looking for short position papers in topics related to transport protocols for wired and wireless networks including new protocols, enhancements to existing ones and measurement studies and interesting observations. It is a great way to get feedback on your work-in-progress and proof-of-concept work. Note: PFLDNeT will be held immediately before CoNEXT 2010, and its venue is located only a short drive or train ride from Philadelphia, making it easy to attend both events in one trip. It will also be co-located with ICCRG. Injong ------------------------------------------------------------------------------------------------------------------------------------------------- CALL FOR PAPERS PFLDNeT 2010 The 8th International Workshop on Protocols for Future, Large-Scale and Diverse Network Transports (PFLDNeT) Franklin & Marshall College, Lancaster, Pennsylvania, USA November 28-29, 2010 Web page: http://pfld.net/2010 Scope: The Internet continues to evolve along several dimensions, allowing more and more end systems to communicate in increasingly diverse ways. At one end of the performance spectrum, the Internet protocols provide communication facilities for extremely-high-speed special-use networks. At the other end of the performance spectrum, the Internet contains very low-power and low-bandwidth networks that cater to infrequent, bursty communication. Enabling efficient and high-performance end-to-end communication across such a diverse internetwork is a difficult problem, which is not solved by current transport layer protocols. The need to support an application base that grows more and more dissimilar adds additional challenges. The 8th International Workshop on Protocols for Future, Large-scale & Diverse Network Transports (PFLDNeT) brings together researchers and practitioners from all continents to exchange their ideas and experiences in the area of transport issues for modern communication networks. The workshop provides theorists, experimentalists and technologists with a focused, highly interactive opportunity to present, discuss and exchange experience on leading research, development and future directions in transport and application protocols for networks that are increasingly growing in size, heterogeneity and dynamicity of interaction. PFLDNeT 2010 solicits papers that further the research on end-to-end communication protocols for todays and tomorrows Internet in all its diversity along the continuum from specialized grid networks, optical transports, wireless connections, to lossy and low-power networks. A specific focus of the workshop lies on transport protocols for the efficient end-to-end transfer of data for a diverse set of applications and application-layer protocols. Now approaching its eighth instantiation, the PFLDneT workshop has broadened its focus over the years from protocols targeted at specific fast, long-distance networks (the original expansion of the PFLDNeT acronym) into a venue where all kinds of new ideas relating to end-to-end transport protocols for diverse network scenarios are being discussed first. The previous International Workshops on Protocols for Fast, Long-Distance Networks held at CERN (2003), Argonne (2004), Lyon (2005), Nara (2006), Marina del Rey (2007), Manchester (2008), and Tokyo (2009) were very successful in bringing together many researchers from all over the world, including North America, Europe and Asia, who are working on these problems. PFLDNeT 2010 will continue this tradition, and provide a perfect forum for researchers in this area to exchange ideas and experience. As in previous years, a meeting of the IRTF Internet Congestion Control Research Group (ICCRG) will be co-located with PFLDNeT, on November 29, 2010. Important Dates and Relevant Event Information: Abstract submission: October 4, 2010 Position paper submission: October 8, 2010 Notification of acceptance: October 27, 2010 Final camera ready submission: November 14, 2010 Workshop: November 28-29, 2010 IRTF ICCRG meeting (co-located): November 29, 2010 Note: PFLDNeT 2010 will be immediately before CoNEXT 2010, and its venue is located only a short drive or train ride from Philadelphia, making it easy to attend both events in one trip. Topics: PFLDNeT 2010 covers all aspects related to transport protocols for the current and future Internet, including, but not limited to: - Transport protocol development - Enhancements to TCP and other transports - Innovative congestion control mechanisms - Novel data transport protocols designed for new networks and applications - Transport services for data center networks and grids - Transport services for wireless and sensor networks - Explicit signaling protocols: optimization criteria and deployment strategies - Pacing and shaping of traffic - Parallel transfers and multi-streaming - Performance evaluation - Modeling and simulation-based results - Interaction of transport protocols and network equipment - Experiments on real networks and live measurements - Transport protocol benchmarking - Transport over optical networks - Transport implementation and hardware issues - End system performance - Data replication and striping - Applications with demanding or unusual network performance requirements - Bulk-data transfer applications - Quality-of-service and scalability issues - Multicast Workshop Organizers: Program Committee Chairs: Bryan Ford, Yale University, USA Injong Rhee, North Carolina State University, USA Steering Committee: Lachlan Andrew, Swinburne University of Technology, Australia Lars Eggert, Nokia Research Center, Finland Richard Hughes-Jones, Univ. of Manchester, UK Katsushi Kobayashi, AIST, Japan Doug Leith, Hamilton Institute, Ireland Injong Rhee, North Carolina State University, USA Pascale Vicat-Blanc, INRIA, France Michael Welzl, University of Oslo, Norway Technical Program Committee (preliminary): Mark Allman, ICIR, USA Scott Brim, Cisco, USA Bob Briscoe, BT, UK Dirceu Cavendish, Kyushu Institute of Technology, Japan Stuart Cheshire, Apple, USA Larry Dunn, University of Minnesota, USA Lars Eggert, Nokia Research Center, Finland Ted Faber, USC/ISI, USA Saikat Guha, Microsoft Research, India Sangtae Ha, Princeton University, USA Janardhan Iyengar, Franklin & Marshall College, USA Katsushi Kobayashi, AIST, Japan Aleksandar Kuzmanovic, Northwestern University, USA Preethi Natarajan, Cisco, USA Joerg Ott, TKK, Finland Narasimha Reddy, Texas A&M University, USA Medy Sanadidi, UCLA, USA Pasi Sarolahti, Aalto University, Finland Michael Scharf, Alcatel-Lucent, Germany Hideyuki Shimonishi, NEC, Japan Murari Sridharan, Microsoft, USA Joe Touch, USC/ISI, USA Michael Welzl, University of Oslo, Norway Lisong Xu, University of Nebraska, USA Local Arrangements: Janardhan Iyengar, Franklin & Marshall College, USA -- Professor Computer Science North Carolina State University From v13 at v13.gr Thu Sep 30 13:24:47 2010 From: v13 at v13.gr (Stefanos Harhalakis) Date: Thu, 30 Sep 2010 23:24:47 +0300 Subject: [e2e] Early FIN/close for query-response TCP connections like DNS's Message-ID: <201009302324.47717.v13@v13.gr> Hello, I'm somehow new to the list and I just read a large part of "TCP improved closing strategies?" thread that took place about a year ago (Aug 2009). After that I tested DNSSEC queries against some local DNS servers. I found out this (and please, correct me if I'm wrong): * Client connects * Client queries * Server responds * Client closes * Server closes (Last two lines may be reversed (?)) >From what I understand, the FINs are problematic for this kind of traffic so I was wondering: Why not transform this to: * Client connects * Client queries * Client closes * Server responds * Server closes IOW: Since we know that this connection will have a Query and a Response, why not half-close the connection after the query is sent? This requires no modifications to TCP's behavior and will save at least RTT/2 because the server side will not wait for a FIN before closing. (no?) Of course this approach can be generalized for all query/response sessions and more generaly, for cases where there is a "last query".