<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; charset=ks_c_5601-1987">
<META content="MSHTML 6.00.2900.2963" name=GENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=#ffffff>
<DIV><FONT face=Arial size=2>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt">This
is a resend with fixed web links. The links were broken in my previous email --
sorry about multiple transmissions.</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt">---------------------------------------------------------------------------------</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt">Hi
Doug,</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt">Thanks for sharing your paper. Also congratulations to
the acceptance of your journal paper to TONs. But I am wondering what's new in
this paper. At first glance, I did not find many new things that are different
from your previously publicized reports. How much is this different from the
ones you put out in this mail list a year or two ago and also the one publicized
in PFLDnet February this year <A title=http://www.hpcc.jp/pfldnet2006/
onclick="return top.js.OpenExtLink(window,event,this)"
href="http://www.hpcc.jp/pfldnet2006/"
target=_blank>http://www.hpcc.jp/pfldnet2006/</A>? In that same workshop, we
also presented our experimental results that shows significant discrepancy from
yours but i am not sure why you forgot to reference our experimental work
presented in that same PFLDnet. Here is a link to a more detailed version of
that report accepted to COMNET <A
title=http://netsrv.csc.ncsu.edu/highspeed/comnet-asteppaper.pdf
onclick="return top.js.OpenExtLink(window,event,this)"
href="http://netsrv.csc.ncsu.edu/highspeed/comnet-asteppaper.pdf"
target=_blank>http://netsrv.csc.ncsu.edu/highspeed/comnet-asteppaper.pdf</A></SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt">The
main point of contention [that we talked about in that PFLDnet workshop] is the
presence of background traffic and the method to add them. Your report mostly
ignores the effect of background traffic. Some texts in this paper state that
you added some web traffic (10%), but the paper shows only the results from NO
background traffic scenarios. But our results differ from yours in many aspects.
Below are the links to our results (the links to them have been available in our
BIC web site for a long time and also mentioned in our PFLDnet paper; this
result is with the patch that corrects HTCP bugs). </SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt">[Convergence and intra protocol fairness]
</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT><FONT face="Courier New" size=2><SPAN
lang=EN-US style="FONT-SIZE: 10pt">without background traffic: <A
title=http://netsrv.csc.ncsu.edu/highspeed/1200/nobk/intra_protocol/intra_protocol.htm
onclick="return top.js.OpenExtLink(window,event,this)"
href="http://netsrv.csc.ncsu.edu/highspeed/1200/nobk/intra_protocol/intra_protocol.htm"
target=_blank>http://netsrv.csc.ncsu.edu/highspeed/1200/nobk/intra_protocol/intra_protocol.htm</A></SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt">with
background traffic: <A
title=http://netsrv.csc.ncsu.edu/highspeed/1200/bk/intra_protocol/intra_protocol.htm
onclick="return top.js.OpenExtLink(window,event,this)"
href="http://netsrv.csc.ncsu.edu/highspeed/1200/bk/intra_protocol/intra_protocol.htm"
target=_blank>http://netsrv.csc.ncsu.edu/highspeed/1200/bk/intra_protocol/intra_protocol.htm</A></SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt">[RTT
fairness]: </SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt">w/o
background traffic: <A
title=http://netsrv.csc.ncsu.edu/highspeed/1200/nobk/rtt_fairness/rtt_fairness.htm
onclick="return top.js.OpenExtLink(window,event,this)"
href="http://netsrv.csc.ncsu.edu/highspeed/1200/nobk/rtt_fairness/rtt_fairness.htm"
target=_blank>http://netsrv.csc.ncsu.edu/highspeed/1200/nobk/rtt_fairness/rtt_fairness.htm</A></SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt">with
background traffic: <A
title=http://netsrv.csc.ncsu.edu/highspeed/1200/bk/rtt_fairness/rtt_fairness.htm
onclick="return top.js.OpenExtLink(window,event,this)"
href="http://netsrv.csc.ncsu.edu/highspeed/1200/bk/rtt_fairness/rtt_fairness.htm"
target=_blank>http://netsrv.csc.ncsu.edu/highspeed/1200/bk/rtt_fairness/rtt_fairness.htm</A></SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt">[TCP
friendliness]</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt">without background traffic: <A
title=http://netsrv.csc.ncsu.edu/highspeed/1200/nobk/tcp_friendliness/tcp_friendliness.htm
onclick="return top.js.OpenExtLink(window,event,this)"
href="http://netsrv.csc.ncsu.edu/highspeed/1200/nobk/tcp_friendliness/tcp_friendliness.htm"
target=_blank>http://netsrv.csc.ncsu.edu/highspeed/1200/nobk/tcp_friendliness/tcp_friendliness.htm</A></SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt">with
background traffic: <A
title=http://netsrv.csc.ncsu.edu/highspeed/1200/bk/tcp_friendliness/tcp_friendliness.htm
onclick="return top.js.OpenExtLink(window,event,this)"
href="http://netsrv.csc.ncsu.edu/highspeed/1200/bk/tcp_friendliness/tcp_friendliness.htm"
target=_blank>http://netsrv.csc.ncsu.edu/highspeed/1200/bk/tcp_friendliness/tcp_friendliness.htm</A></SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt">After our discussion in that PFLDnet, I puzzled why we
get different results. My guess is that the main difference between your
experiment and ours is the inclusion of mid-sized flows with various RTTs -- our
experience tells that the RTT variations of mid size flows play a very important
role in creating significant dynamics in testing environments. The same point
about the importance of mid size flows with RTT variations has been raised in
several occasions by Sally Floyd as well, including in this year's E2E research
group meeting. You can find some reference to the importance of RTT variations
in her paper too [ <A title=http://www.icir.org/models/hotnetsFinal.pdf
onclick="return top.js.OpenExtLink(window,event,this)"
href="http://www.icir.org/models/hotnetsFinal.pdf"
target=_blank>http://www.icir.org/models/hotnetsFinal.pdf</A>]. Just having
web-traffic (all with the same RTTs) does not create a realistic environment as
it does not do anything about RTTs and also flow sizes tend to be highly skewed
with the Pareto distribution-- but I don't know exactly how you create your
testing environment with web-traffic -- I can only guess from the description
you have about the web traffic in your paper. </SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt">Another puzzle in this difference seems that even under
no background traffic, we also get different results from
yours..hmm...especially with FAST because under no background traffic, FAST
seems to work fairly well with good RTT fairness in our experiment. But your
results show FAST has huge RTT-unfairness. That is very strange. Is that because
we have different bandwidth and buffer sizes in the setup? I think we need to
compare our notes more. Also in the journal paper of FAST experimental results [
<A title=http://netlab.caltech.edu/publications/FAST-ToN-final-060209-2007.pdf
onclick="return top.js.OpenExtLink(window,event,this)"
href="http://netlab.caltech.edu/publications/FAST-ToN-final-060209-2007.pdf"
target=_blank>http://netlab.caltech.edu/publications/FAST-ToN-final-060209-2007.pdf</A>
], FAST seems to work very well under no background traffic. We will verify our
results again in the exact same environment as you have in your report, to make
sure we can reproduce your results....but here are some samples of our results
for FAST. </SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt"><A
title=http://netsrv.csc.ncsu.edu/highspeed/1200/nobk/rtt_fairness/1200--2.4_FAST-2.4_FAST-NONE--400-3-1333--1000-76-3-0-0-0-5-500--200000-0.6-1000-10-1200-64000-150--1/
onclick="return top.js.OpenExtLink(window,event,this)"
href="http://netsrv.csc.ncsu.edu/highspeed/1200/nobk/rtt_fairness/1200--2.4_FAST-2.4_FAST-NONE--400-3-1333--1000-76-3-0-0-0-5-500--200000-0.6-1000-10-1200-64000-150--1/"
target=_blank>http://netsrv.csc.ncsu.edu/highspeed/1200/nobk/rtt_fairness/1200--2.4_FAST-2.4_FAST-NONE--400-3-1333--1000-76-3-0-0-0-5-500--200000-0.6-1000-10-1200-64000-150--1/</A></SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt">In
this experiment, FAST flows are just perfect. Also the same result is confirmed
inthe FAST journal paper [ <A
title=http://netlab.caltech.edu/publications/FAST-ToN-final-060209-2007.pdf
onclick="return top.js.OpenExtLink(window,event,this)"
href="http://netlab.caltech.edu/publications/FAST-ToN-final-060209-2007.pdf"
target=_blank>http://netlab.caltech.edu/publications/FAST-ToN-final-060209-2007.pdf</A></SPAN></FONT><FONT
face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt">-- please
look at Section IV.B and C. But your results show really bad RTT
fairness.]</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt">Best
regards,</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt">Injong</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt">---</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt">Injong Rhee</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt">NCSU</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt">On
Sep 22, 2006, at 10:22 AM, Douglas Leith wrote:</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt">For
those interested in TCP for high-speed environments, and perhaps also people
interested in TCP evaluation generally, I'd like to point you towards the
results of a detailed experimental study which are now available at:
</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT> </P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt"><A
title=http://www.hamilton.ie/net/eval/ToNfinal.pdf
onclick="return top.js.OpenExtLink(window,event,this)"
href="http://www.hamilton.ie/net/eval/ToNfinal.pdf"
target=_blank>http://www.hamilton.ie/net/eval/ToNfinal.pdf
</A></SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT> </P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt">This
study consistently compares Scalable-TCP, HS-TCP, BIC-TCP, FAST-TCP and H-TCP
performance under a wide range of conditions including with mixes of long and
short-lived flows. This study has now been subject to peer review (to hopefully
give it some legitimacy) and is due to appear in the Transactions on Networking.
</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT> </P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt">The
conclusions (see summary below) seem especially topical as BIC-TCP is currently
widely deployed as the default algorithm in Linux.</SPAN> </FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT> </P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt">Comments appreciated. Our measurements are publicly
available - on the web or drop me a line if you'd like a copy.</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT> </P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt">Summary:</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt">In
this paper we present experimental results evaluating the</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt">performance of the Scalable-TCP, HS-TCP, BIC-TCP,
FAST-TCP and</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt">H-TCP proposals in a series of benchmark
tests.</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT> </P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt">We
find that many recent proposals perform surprisingly poorly in</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt">even
the most simple test, namely achieving fairness between two</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt">competing flows in a dumbbell topology with the same
round-trip</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt">times and shared bottleneck link. Specifically, both
Scalable-TCP</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt">and
FAST TCP exhibit very substantial unfairness in this test.</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT> </P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt">We
also find that Scalable-TCP, HS-TCP and BIC-TCP induce significantly greater RTT
unfairness between competing flows with different round-trip times. The
unfairness can be an order of magnitude greater than that with standard TCP and
is such that flows with longer round-trip times </SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt">can
be completely starved of bandwidth.</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT> </P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt">While the TCP proposals studied are all successful at
improving</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt">the
link utilisation in a relatively static environment with</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt">long-lived flows, in our tests many of the proposals
exhibit poor</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt">responsiveness to changing network conditions. We
observe that</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt">Scalable-TCP, HS-TCP and BIC-TCP can all suffer from
extremely</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt">slow
(>100s) convergence times following the startup of a new</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt">flow. We also observe that while FAST-TCP flows
typically converge</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt">quickly initially, flows may later diverge again to
create</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt">significant and sustained unfairness.</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT> </P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt">--Doug</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt"></SPAN></FONT> </P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US
style="FONT-SIZE: 10pt">Hamilton Institute</SPAN></FONT></P>
<P><FONT face="Courier New" size=2><SPAN lang=EN-US style="FONT-SIZE: 10pt"><A
title=http://www.hamilton.ie/
onclick="return top.js.OpenExtLink(window,event,this)"
href="http://www.hamilton.ie/"
target=_blank>www.hamilton.ie</A></SPAN></FONT></P></FONT></DIV></BODY></HTML>