To:    TCPSAT@...

NEW SPEAKERS ADDED

PLEASE FORWARD THIS TO YOUR MANAGER OF SYSTEMS SECURITY AND SECURITY
TECHNOLOGIES GROUP.

Market*Access International is proud to present ....

NETWORK AND INFORMATION SYSTEMS SECURITY - TRAINING CONFERENCE

AGENCY INTRUSION DETECTION REQUIREMENTS and TECHNOLOGY SOLUTIONS

Date: July 16, 2001

Ronald Reagan Building and International Trade Center
1300 Pennsylvania Avenue
Washington D.C.
Atrium Ballroom

Time: 7:30 AM Registration and Continental Breakfast
Program Starts: 8:30 AM
Wrap-up: 4:15 PM

About this Conference:

As government becomes more dependent on e-business, databases, information
storehouses, mission critical information storage and information sharing, new
risks are posed. These risks may come from student hackers, foreign governments
or foreign military, terrorist organizations or even internal users.

With the new e-Government applications and communications, computer security has
emerged as a top issue and challenge. This conference will highlight the risks,
opportunities and innovative management and technical approaches to the
detection and response to unauthorized intrusions into agency data. The
conference will focus on business and military applications and agency plans and
programs for securing these applications.

A highlight of this conference will be a section on AGENCY REQUIREMENTS AND NEW
TECHNOLOGIES and strategies of coping with unauthorized attempts to compromise
agency and DoD systems and data.

SPEAKERS WILL REPRESENT THE FOLLOWING AGENCIES/COMPANIES:

Navy NMCI – Mark Lavoie – Communications Officer, CDR, USNR PEO IT (Program
Executive Office of Information Technology).

Department of Transportation - Bonnie Fisher – Director, TASC Millennium
Solutions Center

National Science Foundation - Dara Murry – Director, ADP Security

FBI - James Burrell – Acting Unit Chief, Computer Investigation Unit

DARPA - Jim Webster – Manager, Information Assurance Office

GSA/FedCIRC - Larry Hale –  Liaison Director, Federal Computer Incident Response
Center

DOJ – TBD

DISA – TBD

Gartner – Keynote –  TBD

Guardant – Robert Lee – Director, Senior Principal Consultant

Global Integrity - Errol Weiss, Bill Morgan

Verizon, Federal Network Systems - Char Sample –

Raytheon – Barton Abbott – Director, Information Assurance, Navy /USMC Intranet,
Information Strike Force

Symantec - TBD


For more information on speakers and the conference agenda, please visit our web
site at www.marketaccess.org.

Who should attend:

* Agency IT Executives, Managers, and Staff
* Agency Security Executives, Managers and Staff
* Agency information systems program managers
* Agency Telecommunications Executives, Managers and Staff
* Tele-work and Telecomm Directors, Managers, and Staff
* Functional area managers
* Systems integrators that support federal agency security requirements
* Hardware and software solutions providers

What you will learn:

* Agency plans, programs and priorities
* Successes and Lessons learned
* Innovative government intrusion detection security approaches and applications
* New technologies and strategies - what is on the drawing boards
* How the military approaches network security and intrusion detection
* Security of remote database management
* How to structure intrusion detection solutions
* Risks, sources of attacks... the internal risk
* Commercial and government best practices

Corporate Sponsors:

* Symantec
* Verizon
* Market*Access

.... others to be announced

Organizational Sponsors:

* Department of Transportation
* INPUT Government

....other sponsors to be announced.

Please register early. The conference area has limited seating available and we
anticipate a sell out.

Points of Contact:

* For technical support with this web site, please contact Mr. Parrish Knight,
703/807-2748
* For general information about this event, please contact Ms. Kristen Brooks,
703/807-2745
* For information on sponsorship opportunities & exhibitor arrangements, please
contact Ms. Cara Lombardi at 703/807-2743

The registration fee for this important training conference is:

*  Government Credit Card or Check in Advance: $395
*  Government Training Forms/Invoice: $445
*  Industry and Federal Contractors, Payment in Advance: $595
*  Industry and Federal Contractors/Invoice: $645

We accept government training forms and government and commercial credit cards
(VISA, MC, American Express).

Options:

[1] Phone: 703-807-2745 and speak with Ms. Kristen Brooks.
[2] Email: kbrooks@...
[3] Register online: Use our online booking form to register and pay by credit
card electronically.  To register, go to www.marketaccess.org.
[4] Fax: registration form to 301-652-0914.
[5] Mail: registration form to:

Market*Access International, Inc.
4301 Wilson Blvd. #1003
Arlington, VA 22203

Sponsorships Available! For sponsorship information, please contact:

Cara Lombardi
Market*Access International
4301 Wilson Blvd. #1003
Arlington, VA 22203
Phone (703) 807-2743
Fax (703) 807-2728
clombardi@...

If you wish to be REMOVED from this list, please REPLY and place REMOVE in the
SUBJECT line.

Thank you

Corporation, HNS is the largest and most established competitor in the satellite
industry.  HNS currently has over 120,000 VSAT installations in the United
States, and over 250,000 worldwide the most of any company by a large margin. 
HNS has created Wide Area Networks for companies such as Wal Mart, Holiday Inn,
Cracker Barrel, A.G. Edwards & Co., Amoco, and many others.  Additionally, HNS
is the parent company of DirecTV and DirecPC.  HNS currently offers the widest
range of proven satellite technology available.

Broadcast access to consumers. 


 Internet Via Satellite Report 1999 
 During
the research for this latest edition it was found that the market for
satellite-based Internet link capacity has grown four fold in the last year.
Forward looking companies are taking advantage of this growth phenomena.

SoftCOM 2001
9th IEEE Conference on Software,
Telecommunications and Computer Networks

Split, Dubrovnik (Croatia), Ancona, Bari (Italy)
October 09-12, 2001 (aboard the ship "Marko Polo")


Dear member of the SoftCOM mailing list,

Please note that the deadline for submission of papers
has been extended. The final deadline is July 14, 2001.

Call for papers and Author kit can be found at

            www.fesb.hr/SoftCOM


Sincerely yours

Conference secretary
Hrvoje Dujmic

On Wed, Jun 27, 2001 at 06:43:35PM -0700, Don Sullivan wrote:
> ...but we all knew that, right ?

never noticed.

mutt doesn't seem to see these virus things.

--
[ Jim Mercer        jim@...         +1 416 410-5633 ]
[ Now with more and longer words for your reading enjoyment. ]

...but we all knew that, right ?

Don

P.S.
Yes, I copy the original mail below, but deleted the virus attachment.
-----------------------------------------------
        Don Sullivan
        NASA Ames Research Center
        MS 242-4
        Moffett Field, CA 94035-1000
        Office: +1.650.604.0526
        Lab:    +1.650.604.0595
        Fax:    +1.650.604.4680
        email: dsullivan@...
-----------------------------------------------

---------- Original message ----------
Date: Wed, 27 Jun 101 05:45:51 CDT
From: Jeremy McDevitt <jeremy@...>
Subject: HABITAT

Be a year in December since I started here….It quite honestly feels like 1
month.  Time flies and honestly, I’ve learned a lot!  I’ve never worked a
commission based job before; I wasn’t too sure how it was going to work out. 
Dealing with meeting goals not only monthly, but daily.  Through all of this, I
think the most important thing I’ve learned is team work.

Be a year in December since I started here….It quite honestly feels like 1
month.  Time flies and honestly, I’ve learned a lot!  I’ve never worked a
commission based job before; I wasn’t too sure how it was going to work out. 
Dealing with meeting goals not only monthly, but daily.  Through all of this, I
think the most important thing I’ve learned is team work.

ANNOUNCEMENT AND CALL FOR PAPERS

	 Please Accept Our Sincere Apologies Should you Receive
                       Multiple Copies of this CFP



     The Society for Computer Simulation International (SCS) presents:

              COMMUNICATION NETWORKS AND DISTRIBUTED SYSTEMS
                    MODELING AND SIMULATION CONFERENCE
                                   2002


                           January 27-31, 2002

                 Four Points Sheraton Hotel, Riverwalk North
                            San Antonio, Texas

        Please Visit the Conference Web Site for Additional Information

	           http://www.ece.neu.edu/conf/CNDS/2002/


Part of the 2002 SCS Western Multiconference on Computer Simulation

	 The objective of this conference is to serve as a forum for the
exchange of ideas among a talented group of international researchers on
advances in the design and performance analysis of communication networks
and distributed systems.  The conference emphasizes modeling, technical
advances and unique applications in the areas of high speed networking,
wireless communications, multimedia applications, computer architecture
and distributed, parallel and mobile computing systems. The structure of
the conference is designed to facilitate collegiality and continuity of
discussion among the conference speakers and attendees on the topics
covered during the week.

	 Paper sessions are designed to promote discussion of concepts and
methodologies based on the presentation of results and innovative
modeling, simulation or analysis techniques. The conference grew out of
the need to improve the efficiency and quality of models of increasingly
complex networks and distributed systems. In addition to the technical
sessions, the conference includes invited speakers, vendor presentations
and exhibits arranged in conjunction with the 2002 SCS Western Simulation
Multiconference (WMC).

	 Original, high-quality technical papers, addressing significant
aspects of the design, modeling, or performance analysis of communication
networks and distributed systems are solicited for review, possible
presentation and subsequent publication in the conference proceedings.
Papers that include well-conceived or innovative simulation modeling
techniques with strong results and analysis are especially sought.

Papers are solicited from not limited to the following topics:

	 * Performance Modeling and Analysis
	 * Mobile and Nomadic Computing
	 * Distributed and Wireless Multimedia Systems
	 * QoS Support for Wireless Ad Hoc Network Applications
	 * Routing and Mobility Management in Wireless Networks
	 * Power Control in Wireless Computer Systems
	 * Distributed and Intelligent Sensor Networks
	 * Wearable Computer Systems
	 * Resource Allocation in High Speed Networks
	 * Switch and Processor Architectures for High Speed Routers
	 * Advances in Computer Architecture and Memory Systems
         * Network and Parallel Computing
	 * Storage Area Networks
	 * Content Based Image Retrieval Systems
	 * Real-Time and Mobile Web-Based Computing
	 * Reliable Multicast Routing and Group Communications
	 * Network and Application Security and Availability
	 * Web Caching and Switching
         * Scalable and Reliable Distributed Servers
         * Load Balancing in Distributed Systems


IMPORTANT DATES:

September 15, 2001  Paper Submission Deadline.
October 15, 2001  Acceptance Notification.
November 15, 2001  Camera Ready Paper Deadline.


Authors are requested to submit manuscripts electronically to the Program
Chair by 11:00 pm (Eastern Time) of the due date via Anonymous ftp
directly: ftp.ece.neu.edu:/pub/conf/CNDS/2002/Submissions. Papers can
also be uploaded using the Web Site: www.ece.neu.edu/conf/CNDS/2002/.
Papers must be in English and submitted files must be in Postscript or PDF
format with all figures and tables included. Files names must be unique
and reflect the name of the primary author, his or her primary
affiliation, the version number of the file and an extension indicating
the file type. Multiple versions may be uploaded; however, only the most
recent will be reviewed. Filenames are not readable once the file transfer
has been completed. Confirmations will be sent to the contact author when
the file has been retrieved and successfully viewed or printed. Only
original unpublished material will be considered for publication.

This year's International TPC is outstanding. Its members represent a
broad spectrum of interrelated interests and technical expertise relevent
to CNDS including performance analysis, simulation, communications
networks, both wired and wireless, mobile computing, computer
architecture, distributed systems and more. The TPC consists of newcomers
and veterans of CNDS from Universities and Research Labs around the
world. Each member is committed to ensure that all submissions are fully
refereed for accuracy, technical content, and relevance, and to make this
year's conference an exciting International forum for the presentation of
high quality work and the exchange of ideas. Papers should be limited to
20 double-spaced pages (camera-ready versions are limited to 6
single-spaced pages). Authors must include a single page which includes
all the authors' names, titles and affiliations, a paper abstract, list of
top three relevant topics and contact person information (complete
address, email, phone, and fax). Accepted papers will appear in the
Conference's Proceedings to be published by the Society for Computer
Simulation. Instructions for submitting camera ready papers will be sent
to authors following paper acceptance. An award will be presented to the
best paper presented at WMC'02. Authors of top papers will also be
encouraged to submit a follow-on paper to the International Journal in
Computer Simulation. Authors must obtain employer, client, or government
releases prior to submittal of the final manuscript.

Please do not hesitate to contact the Program Chair, Prof. McDonald via
email (mcdonald@...), or telephone ((617) 373-3028) with any
questions concerning the paper submission and review process, or questions
regarding the relevance of a paper to the conference.


    General Chair                       Program Chair

Dr. Taieb F. Znati                 Prof. A. Bruce McDonald
Senior Program Director            Dept. of Electrical and Computer Eng.
National Science Foundation        417 Dana Research Center
ANIR and  			   Northeastern University
Computer Science Department        360 Huntington Avenue
University of Pittsburgh           Boston, MA 15221
Pittsburgh, PA 15260               U.S.A.
U.S.A.

Phone:  +1(412)624-8417            Phone:  +1(617)328-3028
FAX:    +1(412)624-8854            FAX:    +1(617)328-8970
E_Mail: znati@...          E_Mail: mcdonald@...


Please visit the conference Web Site for up-to-date information and to see
a list of the distinguished  members of the International Technical
Program Commmittee:

	     http://www.ece.neu.edu/conf/CNDS/2002/



------------------------------------------------------------------------
---------------------------------------------------------------------------
Saad Biaz
Al Akhawayn University
B.P. 2146
53000 Ifrane, Morocco

Hi Everyone,

I'm Fauzan. I'm looking for references, papers, etc about
TCP/IP Spoofing to improve TCP performance over Satellite
links. Can anybody tell me where can I find it? Specially
papers or references about the techniques? Thanks.

Regards,
-Fauzan-

PLEASE FORWARD TO THE PERSON
RESPONSIBLE FOR PURCHASING
YOUR LASER PRINTER SUPPLIES

**** VORTEX  SUPPLIES ****

LASER PRINTER TONER CARTRIDGES,
COPIER AND FAX CARTRIDGES

SAVE UP TO 30% FROM RETAIL

ORDER BY PHONE:1-888-288-9043
ORDER BY FAX: 1-888-977-1577
CUSTOMER SERVICE: 1-888-248-2015
E-MAIL REMOVAL LINE: 1-888-248-4930

UNIVERSITY AND/OR SCHOOL PURCHASE ORDERS WELCOME. (NO CREDIT APPROVAL REQUIRED)
ALL OTHER PURCHASE ORDER REQUESTS REQUIRE CREDIT APPROVAL.
PAY BY CHECK (C.O.D), CREDIT CARD OR PURCHASE ORDER (NET 30 DAYS).

IF YOUR ORDER IS BY CREDIT CARD PLEASE LEAVE YOUR CREDIT CARD # PLUS EXPIRATION
DATE.
IF YOUR ORDER IS BY PURCHASE ORDER LEAVE YOUR SHIPPING/BILLING ADDRESSES AND
YOUR P.O. NUMBER


FOR THOSE OF YOU WHO REQUIRE MORE INFORMATION ABOUT OUR COMPANY
INCUDING FEDERAL TAX ID NUMBER, CLOSEST SHIPPING OR CORPORATE ADDRESS IN THE
CONTINENTAL
U.S. OR  FOR CATALOG  REQUESTS PLEASE CALL OUR CUSTOMER SERVICE LINE 
1-888-248-2015


OUR NEW , LASER PRINTER TONER CARTRIDGE, PRICES ARE  AS FOLLOWS:
(PLEASE ORDER BY PAGE NUMBER AND/OR ITEM NUMBER)

HEWLETT PACKARD: (ON PAGE 2)

ITEM #1  LASERJET SERIES  4L,4P (74A)------------------------$44
ITEM #2  LASERJET SERIES  1100 (92A)-------------------------$44
ITEM #3  LASERJET SERIES  2 (95A)----------------------------$39
ITEM #4  LASERJET SERIES  2P (75A)---------------------------$54
ITEM #5  LASERJET SERIES  5P,6P,5MP, 6MP (3903A)----------  -$44
ITEM #6  LASERJET SERIES  5SI, 8000 (09A)--------------------$95
ITEM #7  LASERJET SERIES  2100 (96A)-------------------------$74
ITEM #8  LASERJET SERIES  8100 (82X)------------------------$145
ITEM #9  LASERJET SERIES  5L/6L (3906A)----------------------$35
ITEM #10 LASERJET SERIES  4V---------------------------------$95
ITEM #11 LASERJET SERIES 4000 (27X)--------------------------$72
ITEM #12 LASERJET SERIES 3SI/4SI (91A)-----------------------$54
ITEM #13 LASERJET SERIES 4, 4M, 5,5M-------------------------$49
ITEM #13A LASERJET SERIES 5000 (29X)-------------------------$95

HEWLETT PACKARD FAX (ON PAGE 2)

ITEM #14 LASERFAX 500, 700 (FX1)----------$49
ITEM #15  LASERFAX 5000,7000 (FX2)--------$54
ITEM #16  LASERFAX (FX3)------------------$59
ITEM #17  LASERFAX (FX4)------------------$54


LEXMARK/IBM (ON PAGE 3)

OPTRA 4019, 4029 HIGH YIELD---------------$89
OPTRA R, 4039, 4049 HIGH YIELD-----------$105

OPTRA E-----------------------------------$59
OPTRA N----------------------------------$115
OPTRA S----------------------------------$165


EPSON (ON PAGE 4)

ACTION LASER 7000,7500,8000,9000----------$105
ACTION LASER 1000,1500--------------------$105


CANON PRINTERS (ON PAGE 5)

PLEASE CALL FOR MODELS AND UPDATED PRICES
FOR CANON PRINTER CARTRIDGES

PANASONIC (0N PAGE 7)

NEC SERIES 2 MODELS 90 AND 95----------$105

APPLE (0N PAGE 8)

LASER WRITER PRO 600 or 16/600------------------$49
LASER WRITER SELECT 300,320,360-----------------$74
LASER WRITER 300 AND 320------------------------$54
LASER WRITER NT, 2NT----------------------------$54
LASER WRITER 12/640-----------------------------$79

CANON FAX (ON PAGE 9)

LASERCLASS 4000 (FX3)---------------------------$59
LASERCLASS 5000,6000,7000 (FX2)-----------------$54
LASERFAX 5000,7000 (FX2)------------------------$54
LASERFAX 8500,9000 (FX4)------------------------$54

CANON COPIERS (PAGE 10)

PC 3, 6RE, 7 AND 11 (A30)---------------------$69
PC 300,320,700,720 and 760 (E-40)-------------$89

IF YOUR CARTRIDGE IS NOT LISTED CALL CUSTOMER SERVICE AT 1-888-248-2015

90 DAY UNLIMITED WARRANTY INCLUDED ON ALL PRODUCTS.

ALL TRADEMARKS AND BRAND NAMES LISTED ABOVE ARE PROPERTY OF THE
RESPECTIVE HOLDERS AND USED FOR DESCRIPTIVE PURPOSES ONLY.

!!!    PLEASE ACCEPT OUR SICNERE APOLOGIES FOR MULTIPLE COPIES. !!!

>
>
>          INFOCOM 2002 CALL FOR PAPERS
>
>                 *************************************************
>                 *               IEEE INFOCOM 2002               *
>                 *   The Conference on Computer Communications   *
>                 *                                               *
>                 *       June 23-27, 2002,   New York City       *
>                 *                                               *
>                 *    The 21st Annual Joint Conference of the    *
>                 *   IEEE Computer and Communications Societies  *
>                 *                                               *
>                 *       http://www.ieee-infocom.org/2002        *
>                 *************************************************
>
> SCOPE
> =====
>
> The major conference on computer communications and networking is
> celebrating its 21st anniversary at the Hilton Hotel in New York
> during the week of June 23-27, 2002. The conference will bring
> researchers and practitioners of every aspect of data
> communications and networks together to present the most
> up-to-date results and achievements in the field.
>
> Original papers are invited on recent advances in computer
> communications and networking. Topics of interest include,
> but are not limited to, the following:
>
>         * Broadband & ATM
>         * Congestion & admission control
>         * Flow control
>         * Internetworking & IP
>         * Mobile & wireless networks
>         * Multicast
>         * Multimedia
>         * Network applications & services
>         * Network architectures
>         * Network design & planning
>         * Network management & control
>         * Network  measurements & testbeds
>         * Network modeling
>         * Network pricing
>         * Network protocols
>         * Network system software
>         * Optical networks
>         * Quality of service
>         * Routing
>         * Scheduling
>         * Security & privacy
>         * Storage area networks
>         * Switches & switching
>         * Traffic engineering
>         * Web caching
>         * Web performance
>
> SCHEDULE (STRICTLY ENFORCED)
>         Full paper due                   July 31, 2001
>         Notification of acceptance       November 14, 2001
>         Final version due                December 20, 2001
>
> CONFERENCE SCHEDULE
>         Tutorials                        June 23-24, 2002
>         Conference                       June 25-27, 2002
>
> INFORMATION ON:
>
>
>     * Submission instructions        * Program and registration
>     * Tutorials                              * Workshops
>     * Local arrangements             * Student and travel grants
>
>
>     appears at http://www.ieee-infocom.org/2002.
>
> For general information please contact the Genaral Chair, Parviz Kermani
> <parviz@...>
>
> PROGRAM COMMITTEE CO-CHAIRS
> ===========================
> David Lee, Bell Labs Research China (lee@...)
> Ariel Orda, Technion - Israel Inst. of Technology
(ariel@...)
>
>
>

Please send mail to my new email as: dnguyen@...

Thanks

I've been sucessful running ssh2 and even X terminals over ssh over a
satellite link.  However, I'm having a significant performance problem
when trying to do file transfers over ssh.  It reduces my throughput
from about 400 kbps to about 40 kbps.  When I look at the packet
transfers it looks like the secure FTP server will only allow about 10
kBytes of unacknowledged data in the pipe even though the window size is
set to 64 kBytes (standard FTP uses the entire window and achieves a
much higher throughput).  Any ideas or experience with this problems
would be greatly appreciated.  Thanks in advance.

Ed Olechna

Mike Jagdis wrote:
>
> > Some download managers such as NetAnts claim to be designed to maximize
> > throughput by making multiple HTTP connections, each of which downloads
> > a separate part of a single big file.
> >
> > I wonder why doing this is helpful.
>
> By violating the slow start principle. With one TCP connection
> you send one packet, get an ACK, send two packets, get two ACKs
> etc. With short lived connections, such as HTTP/1.0, you may never
> actually get out of slow start and thus use significantly less
> then bandwidth than is actually available.
>
>   HTTP/1.1 addresses this at the application level by by reusing
> the same TCP connection. With a longer lived connection you have
> the chance to ramp up further and get out of slow start.

There are two reasons this helps:

	 1. the sequence of items in a single page ramp
	    up the window, rather than restarting it,
	    as Mike indicated


	 2. the sender-driven request/response interaction
	    defeats slow-start restart, which should occur
	    during the idle times between the pages.

We had an old ID on this, which we started implementing
and got put aside during some hectic times. We're hoping
to finish it off this summer. Here's a personally archived
version, that describes the situation further:

	 http://www.isi.edu/touch/pubs/tcpimpl-restart-01.txt

>   I seem to remember early versions of Netscape used multiple
> connections. I believe you'll find it mentioned in the tcpsat archives
> if you search for netscape and marco.

Most still do - up to 4 or 8 connections between the same endpoint
pairs.

Joe

CALL FOR PAPERS

SoftCOM 2001
IEEE International Conference on Software, Telecommunications
and Computer Networks

Workshop on Future Wireless Communications

       October 09-12, 2001
       Split, Dubrovnik (Croatia)
       Ancona, Bari (Italy)

Workshop on Future Wireless Communications will be held aboard the
   ship Marko Polo travelling on the route Split (Croatia)
   Ancona (Italy)  Bari (Italy)  Dubrovnik (Croatia) in the frame
   of the 9 th International Conference on Software, Telecommunications
   and Computer Networks  SoftCOM 2001.

The increasing popularity of wireless networks indicates that most
   of future networking structures will be based on wireless networking
   technologies (Satellite, Ad Hoc, UMTS, ..). Wireless communication
   poses some problems that do not exist in wired networks, such as
   limited resources and fading due to mobility. This has generated an
   enormous interest for new communication protocols and enhanced
   traffic/resource management mechanisms.

This Workshop on Future Wireless Communications will cover both
theoretical
   and pratical aspects and will provide an opportunity for researchers
   to meet and discuss aspects of their work in an informal setting.

Papers presenting new results in (but not limited to) the following areas
are invited:
     * Emerging Satellite Systems
     * Protocols for Ad Hoc Wireless Networks
     * Transport/Data Link Protocols for Mobile Satellite Communications
     * QoS in Wireless Networks
     * Heterogeneous Wireless Networks
     * Traffic & Resource Management in Satellite Networks
     * Integration between Wired and Wireless (Satellite, Ad Hoc,
       UMTS,..) Networks
     * Services and Applications for Wireless Communications

Each day of the conference the ship will be anchored in one of the ports
   on the route, and overnight it will sail towards the next destination.
   This provides participants with the opportunity to share ideas in close
   contact with their coleagues and to enjoy the pleasant and inspiring
   ambience while visiting the ports along the beautiful Adriatic coast.
   During the conference the car deck will serve as an exhibition arena for
   exhibitors of software and telecommunication products.

Schedule
   Complete manuscript to be received by July 1, 2001
   Notification of acceptance mailed out by September 1, 2001
   Camera-ready manuscript to be received by September 15, 2001

More information about the Conference including details on the submission
   process and authors kit are available on the website

        http://www.fesb.hr/SoftCOM

Workshop Chair: Mario De Blasi, University of Lecce, Italy
                 (mario.deblasi@...)

Conference Secretary: Hrvoje Dujmic, University of Split, Croatia
                       (softcom@...)

> Some download managers such as NetAnts claim to be designed to maximize
> throughput by making multiple HTTP connections, each of which downloads
> a separate part of a single big file.
>
> I wonder why doing this is helpful.

By violating the slow start principle. With one TCP connection
you send one packet, get an ACK, send two packets, get two ACKs
etc. With short lived connections, such as HTTP/1.0, you may never
actually get out of slow start and thus use significantly less
then bandwidth than is actually available.

   HTTP/1.1 addresses this at the application level by by reusing
the same TCP connection. With a longer lived connection you have
the chance to ramp up further and get out of slow start.

   The multiple connections brigade reduces the slow start effect
by using multiple connections. With four connections rather than
one you send four packets, get four ACKs, send eight packets,
get eight ACKs etc. Sounds good? Well, kind of, yeah...

> Now suppose the network is the performance bottleneck (maybe network
> congestion), then no matter how many simultanous connections are used,
> the overall transfer speed is limited by the network. How can the
> multiple connections help?

That's where things go wrong. If you are using four connections to
do the work of one then under congestion you will get four times
your fair share of bandwidth. This is good for you. It's bad for
everyone else who happens to be using the same link. Unless they
happen to be using one of these multiple connection atrocities as
well in which case you are heading for an arms race with people
increasing the number of connections they use to try and grab more
of the bandwidth for themselves.

   Once you are in to an arms race things go from bad to worse. Use of
multiple connections like that creates bursts of SYN, SYN-ACK and
data packets that disturb TCPs attempts to adapt to the available
bandwidth. This tends to result in active connections staying
backed off so there is always free capacity to absorb these bursts.
i.e. the link owner doesn't get to make efficient use of the
bandwidth they are paying for.

> However, by real experience, this kind of download manager is helpful
> and reduce the overall respond time. Why? Is it related to the HTTP 1.1?

Yes. It *appears* helpful to the end user. But it's snake oil. It helps
the end user by screwing over every one else trying to use any of the
links between you and the remote system you are accessing.

   I seem to remember early versions of Netscape used multiple
connections. I believe you'll find it mentioned in the tcpsat archives
if you search for netscape and marco.

   Incidentally there is ongoing TCP research to avoid slow start
limitations without being so massively anti-social. Things like
large initial windows, shared state, congestion managers etc.

				 Mike

--
Chief Network Architect  Mobile: +44 7780 608 368
Kokua Communications Ltd Office: +44 20 7292 1680
52-53 Conduit Street  Fax:  +44 20 7292 1681
London W1S 2YX

This is especially relevant to satellite IP networks where the
latency affects each TCP connection by limiting the maximum
throughput to an un-modified TCP stack on a client.

Opening multiple parallel TCP connections is a way "pull" more
bandwidth on "one session".

You're right though, it probably wouldnt improve things over low
latency 56k or T1 lines.

jm



>?Some download managers such as NetAnts claim to be designed to maximize
>throughput by making multiple HTTP connections, each of which downloads
>a separate part of a single big file.
>
>I wonder why doing this is helpful.
>
>Suppose a user connects to the Internet via a 56kbps line. His first
>HTTP connection should consume all the bandwidth of his line, how can
>the multiple simultanous connections help?
>
>Assume the Web server haven't limited the transfer rate for the requests
>to it. If the user had a 1.5Mbps line now, then his first HTTP
>connection could still transfer the file at 1.5Mbps (assume the network
>is not the bottleneck), how can the multiple connections help?
>
>Now suppose the network is the performance bottleneck (maybe network
>congestion), then no matter how many simultanous connections are used,
>the overall transfer speed is limited by the network. How can the
>multiple connections help?
>
>However, by real experience, this kind of download manager is helpful
>and reduce the overall respond time. Why? Is it related to the HTTP 1.1?
>
>Thank you for your attention.
>
>Angus Wong

Some download managers such as NetAnts claim to be designed to maximize
throughput by making multiple HTTP connections, each of which downloads
a separate part of a single big file.

I wonder why doing this is helpful.

Suppose a user connects to the Internet via a 56kbps line. His first
HTTP connection should consume all the bandwidth of his line, how can
the multiple simultanous connections help?

Assume the Web server haven't limited the transfer rate for the requests
to it. If the user had a 1.5Mbps line now, then his first HTTP
connection could still transfer the file at 1.5Mbps (assume the network
is not the bottleneck), how can the multiple connections help?

Now suppose the network is the performance bottleneck (maybe network
congestion), then no matter how many simultanous connections are used,
the overall transfer speed is limited by the network. How can the
multiple connections help?

However, by real experience, this kind of download manager is helpful
and reduce the overall respond time. Why? Is it related to the HTTP 1.1?

Thank you for your attention.

Angus Wong

> 2. I read in a book ("TCP/IP Illustrated" which surely referes to
> an older version of linux kernel)

Actually, the above book covers BSD networking code.  A very
different animal than the linux code, I am sure.

> that the tcp_time_stamp is practically an integer number which
> counts "ticks" of 500msec and, as a consequence, it is updated
> every 500 msec... is it possible?

Under BSD that is true.  I think linux has a finer-grained clock.

> how could i have a good evaluation of RTT if i can estimate only
> ticks of such a big interval?

TCP does not try to obtain a "good evaluation of RTT".  TCP attempts
to determine an appropriate **retransmission timeout** (RTO), which
is a much different problem.  See the following paper for an
evaluation of the impact of clock granularity (among other things)
on the standard RTO algorithm:

     Mark Allman, Vern Paxson. On Estimating End-to-End Network Path
     Properties.  ACM SIGCOMM, September 1999, Cambridge, MA.
     http://roland.grc.nasa.gov/~mallman/papers/estimation.ps

(The conclusion from the paper is that finer-grained timers provide
better RTO performance when used in conjunction with a fairly
healthy lower bound on the RTO.)

RFC 2988 outlines the standard RTO algorithm.

allman


---
Mark Allman -- BBN/NASA GRC -- http://roland.grc.nasa.gov/~mallman/

Good morning to everyone,
I'm introducing some modifications to the
implementation of the TCP (Linux kernel 2.2.17)
in order to optimize performance in satellite
networks.

I have some problems and I hope you can help me:

1. I need the RTT value immediately after the
sync phase between the two TCP entities.
Is the RTT evaluated during the sync phase? If so,
can you
suggest me in which variable is it stored (infact
the
RTT values should be stored in "tp->srtt", which is
zero after the sync phase)

2. I read in a book ("TCP/IP Illustrated" which
surely
referes to an older version of linux kernel) that
the
tcp_time_stamp is practically an integer number
which
counts "ticks" of 500msec and, as a consequence, it
is
updated every 500 msec... is it possible? how could
i
have a good evaluation of RTT if i can estimate only
ticks of such a big interval?

I hope someone of You can kindly help me...

thank You in advance

Renato Narcisi


__________________________________________________
Do You Yahoo!?
Get personalized email addresses from Yahoo! Mail - only $35
a year!  http://personal.mail.yahoo.com/

Hi,

To enable large window size in Win2000: create a reg key (DWORD):
HKEY_LOCALMACHINE\system\currentcontrolset\services\tcpip\parameters\tcpwindowsi\
ze
and set it to something larger than 65535 (I use 200000 for a RTT of approx
500ms).

On Linux: (2.2)
for send window
echo 200000 > /proc/sys/net/core/wmem_default
echo 200000 > /proc/sys/net/core/wmem_max
for receive window
echo 200000 > /proc/sys/net/core/rmem_default
echo 200000 > /proc/sys/net/core/rmem_max
Sack and timestamping are enabled by default on Linux.

note that you have to change this on both http server and client.

The TT card will download with reasonable stability up to 3 Mbit/s.

Robert



                     Dusan Statelov
                     <statelov@maind        To:     tcpsat@...
                     ata.sk>                cc:
                     Sent by:               Subject:     How to reach Large TCP
window in Win (2000) or
                     owner-tcpsat@gr        Linux
                     c.nasa.gov


                     05/28/01 05:00
                     PM
                     Please respond
                     to statelov






Hello:

As we are now in time pressure I would like ask you for advices how to
properly configure Win2000 or Linux Http data server in order to achieve
bigger TCP window and improve Single TCP connection from poor 20-30 kbyte/s
(for RTT 240 ms) to according to us achievable 1 mbps.
We have changed "Default Receive Window", but TCP is running at only 20-30
kbyte/s what would corresponde to 8 kbyte Window rather then to 64 kbyte.
I assume there are other TCP registry entities which influence Sender
Window.

Here is a configuration emulating satellite internet:

Http Data Server [Linux or Win2000]  -> DVB/ip -> DVB-S Modulator -> L band
output -> PC DVB board [SkyMedia - Telemann, TT DVB Technotrend] ->
client PC [Win98] o---------------o  connected to DVB/ip GW in return via
LAN
LAN RTT 20 ms
Artefficial delay in DVB/ip GW 220 ms (in order to simulate long RTT for
satellite)

Windows98
Default Receive Window to max.values 65535
Data Sever Win2000
Default Receive Window to max.values 65535 (we tried even 0,5 Mbyte with no
effect)
RTT 240 ms
Single TCP connection 20 - 30 kbyte/s

Thanks in advance.

Dusan Statelov
Managing Director
Main Data,s.r.o.
mailto:statelov@...

Tel:     +421 7 54789 586
Fax:     +421 7 54789 585
GSM:     +421 905 606 027
http://www.maindata.sk

(See attached file: statelov.vcf)





--
DISCLAIMER:
This e-mail contains proprietary information some or all of which may be
legally privileged. It is for the intended recipient only. If an addressing
or transmission error has misdirected this e-mail, please notify the author
by replying to this e-mail. If you are not the intended recipient you must
not use, disclose, distribute, copy, print, or rely on this e-mail.

I normally wouldn't ask these questions (and I do apologize),
but since your data source is labeled as being an http server:

Are you convinced that you are passing enough data during the life
of the measured TCP connections to allow your congestion window to
fully open?

If yes, are you passing enough *additional* data to make the
time spent in slow-start negligible when calculating the average
throughput of the connection?

How large are your test data flows?

Tuning the TCP connection parameters is important, but that
can only remove artificial bounds on performance. You still
need to allow for TCP to probe for the capacity of the path;

It could be that you simply aren't transferring enough data...

Eric

Eric Travis
travis@...



On Mon, 28 May 2001, Dusan Statelov wrote:

> Hello:
>
> As we are now in time pressure I would like ask you for advices how to
> properly configure Win2000 or Linux Http data server in order to achieve
> bigger TCP window and improve Single TCP connection from poor 20-30
> kbyte/s (for RTT 240 ms) to according to us achievable 1 mbps.
> We have changed "Default Receive Window", but TCP is running at only
> 20-30 kbyte/s what would corresponde to 8 kbyte Window rather then to 64
> kbyte.
> I assume there are other TCP registry entities which influence Sender
> Window.
>
> Here is a configuration emulating satellite internet:
>
> Http Data Server [Linux or Win2000]  -> DVB/ip -> DVB-S Modulator -> L
> band output -> PC DVB board [SkyMedia - Telemann, TT DVB Technotrend] ->
>
> client PC [Win98] o---------------o  connected to DVB/ip GW in return
> via LAN
> LAN RTT 20 ms
> Artefficial delay in DVB/ip GW 220 ms (in order to simulate long RTT for
> satellite)
>
> Windows98
> Default Receive Window to max.values 65535
> Data Sever Win2000
> Default Receive Window to max.values 65535 (we tried even 0,5 Mbyte with
> no effect)
> RTT 240 ms
> Single TCP connection 20 - 30 kbyte/s
>
> Thanks in advance.
>
> Dusan Statelov
> Managing Director
> Main Data,s.r.o.
> mailto:statelov@...
>
> Tel:     +421 7 54789 586
> Fax:     +421 7 54789 585
> GSM:     +421 905 606 027
> http://www.maindata.sk
>
>

You need to increase the client's RECEIVE window, the server normally
can scale up to quite large send windows by default.

Make sure you enable SACK and RFC1323. Use TCPTune for this

http://moat.nlanr.net/Software/TCPtune/

jm


At 5:00 PM +0200 5/28/01, Dusan Statelov wrote:
>Hello:
>
>As we are now in time pressure I would like ask you for advices how
>to properly configure Win2000 or Linux Http data server in order to
>achieve bigger TCP window and improve Single TCP connection from
>poor 20-30 kbyte/s (for RTT 240 ms) to according to us achievable 1
>mbps.
>We have changed "Default Receive Window", but TCP is running at only
>20-30 kbyte/s what would corresponde to 8 kbyte Window rather then
>to 64 kbyte.
>I assume there are other TCP registry entities which influence Sender Window.
>
>Here is a configuration emulating satellite internet:
>
>Http Data Server [Linux or Win2000]  -> DVB/ip -> DVB-S Modulator ->
>L band output -> PC DVB board [SkyMedia - Telemann, TT DVB
>Technotrend] ->
>client PC [Win98] o---------------o  connected to DVB/ip GW in return via LAN
>LAN RTT 20 ms
>Artefficial delay in DVB/ip GW 220 ms (in order to simulate long RTT
>for satellite)
>
>Windows98
>Default Receive Window to max.values 65535
>Data Sever Win2000
>Default Receive Window to max.values 65535 (we tried even 0,5 Mbyte
>with no effect)
>RTT 240 ms
>Single TCP connection 20 - 30 kbyte/s
>
>Thanks in advance.
>
>Dusan Statelov
>Managing Director
>Main Data,s.r.o.
><mailto:statelov@...>mailto:statelov@...
>
>Tel:     +421 7 54789 586
>Fax:     +421 7 54789 585
>GSM:     +421 905 606 027
><http://www.maindata.sk>http://www.maindata.sk
>
>
>Content-Type: text/x-vcard; charset=iso-8859-2;
>  name="statelov.vcf"
>Content-Transfer-Encoding: 7bit
>Content-Description: Card for Dusan Statelov
>Content-Disposition: attachment;
>  filename="statelov.vcf"
>
>Attachment converted: Macintosh HD:statelov.vcf 1 (TEXT/ttxt) (000CAF94)

As we are now in time pressure I would like ask you for advices how to properly configure Win2000 or Linux Http data server in order to achieve bigger TCP window and improve Single TCP connection from poor 20-30 kbyte/s (for RTT 240 ms) to according to us achievable 1 mbps.
We have changed "Default Receive Window", but TCP is running at only 20-30 kbyte/s what would corresponde to 8 kbyte Window rather then to 64 kbyte.
I assume there are other TCP registry entities which influence Sender Window.

Here is a configuration emulating satellite internet:

Http Data Server [Linux or Win2000]  -> DVB/ip -> DVB-S Modulator -> L band output -> PC DVB board [SkyMedia - Telemann, TT DVB Technotrend] ->
client PC [Win98] o---------------o  connected to DVB/ip GW in return via LAN
LAN RTT 20 ms
Artefficial delay in DVB/ip GW 220 ms (in order to simulate long RTT for satellite)

Windows98
Default Receive Window to max.values 65535
Data Sever Win2000
Default Receive Window to max.values 65535 (we tried even 0,5 Mbyte with no effect)
RTT 240 ms
Single TCP connection 20 - 30 kbyte/s

Thanks in advance.

Dusan Statelov
Managing Director
Main Data,s.r.o.
mailto:statelov@...

Tel:     +421 7 54789 586
Fax:     +421 7 54789 585
GSM:     +421 905 606 027
http://www.maindata.sk
 

Can you please advice URL to some TCP analyser - monitor SW for Linux or Windows ?

We have installed Linux PC for TCP monitoring purposes and installed some TCP monitor to watch traffic from  http://www.ethereal.com/ but SW does not provide us information about real "congestion window" influencing the TCP performance.

We can see some values of window ranging between 63-64-65 kbyte, what is meaningless.
I expect (due to TCP slow start alghorithm) to see congestion window growing from bottom almost from 0 to TCP window size so from 0-64 kbyte for any old TCP and this should correlate to real TCP performance.

Thanks in advance.

Best regards

Dusan Statelov
Managing Director
Main Data,s.r.o.
mailto:statelov@...

Tel:     +421 7 54789 586
Fax:     +421 7 54789 585
GSM:     +421 905 606 027
http://www.maindata.sk
 

You make look into the issues of the satellite (uplink) bandwidth. There is
always a maximum threshold of average queue size in the egress router which will
limit the performace of incremental window size.



-----Original Message-----
From:    pambosc pambosc@...
Sent:    Fri, 18 May 2001 10:15:49 -0600
To:      statelov@..., tcpsat@...
Subject: Re: TCP performance


Take a look at my reasearch and papers at
www.ittc.ukans.edu/~ccharala/research.html

Pambos
----- Original Message -----
From: "Dusan Statelov" <statelov@...>
To: <tcpsat@...>
Sent: Thursday, May 17, 2001 9:13 AM
Subject: TCP performance


> Hello:
>
> We are investigating reasons of poor TCP/ip performance over satellite.
>
> We expect to get single TCP/ip performance of Windows size/RTT but in a
> fact windows size does not influence performance at all.
> We are working ni Windows NT platform.
> In a LAN we get 180 kbyte/s the same system over sat falls to 25 kbyte/s
>
> RTT is 220 ms windows size set to 64 kbyte.
>
> I would appreciate your response and assistance.
>
>
>
> --
> Best regards
>
> Dusan Statelov
> Managing Director
> Main Data,s.r.o.
> mailto:statelov@...
>
> Tel:     +421 7 54789 586
> Fax:     +421 7 54789 585
> GSM:     +421 905 606 027
> http://www.maindata.sk
>
>




___________________________________________________________________________
Visit http://www.visto.com/info, your free web-based communications center.
Visto.com. Life on the Dot.