<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; charset=us-ascii">
<META content="MSHTML 6.00.2900.3132" name=GENERATOR></HEAD>
<BODY>
<DIV dir=ltr align=left><SPAN class=486442423-26072007><FONT face=Arial
color=#0000ff size=2><FONT face="Times New Roman" color=#000000>The sounds a bit
like an extnesion of "</FONT><A
href="http://www.itu.int/rec/T-REC-G/recommendation.asp?lang=en&parent=T-REC-G.1040"><I><FONT
face="Times New Roman">ITU-T Rec.G1040 "Network contribution to transaction
time"</FONT></I></A><FONT face="Times New Roman" color=#000000> which
calculates the network contribution to transaction time. The contribution
depends on the <I>RTT</I>, loss probability (<I>p</I>), the Retransmission Time
Out (<I>RTO</I>) and the number of round trips involved (<I>n</I>) in a
transaction. The Network Contribution to Transation Time (<I>NCTT</I>) is given
as: <BR></FONT>
<CENTER><FONT size=-1><I>Average(NCTT) = (n * RTT) + (p * n * RTO)
</I></FONT></CENTER><FONT size=-1><BR>In our case (PingER) typical values for
<I>n</I> are 8, for <I>RTO</I> we take 2.5 seconds, we take the <I>RTT</I> and
loss probability (<I>p</I>) from the PingER measurements. The main
difference is that you seem to ignore the losses.</FONT></FONT></SPAN></DIV><BR>
<DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left>
<HR tabIndex=-1>
<FONT face=Tahoma size=2><B>From:</B> end2end-interest-bounces@postel.org
[mailto:end2end-interest-bounces@postel.org] <B>On Behalf Of </B>Paddy
Ganti<BR><B>Sent:</B> Thursday, July 26, 2007 2:44 PM<BR><B>To:</B>
end2end-interest@postel.org<BR><B>Subject:</B> [e2e] Analytic Model of Download
Time as a Function of TCP ConnectTime<BR></FONT><BR></DIV>
<DIV></DIV>I am thinking of an approach to analytically determine the download
time as a function of RTT given a few initial real world samples. Say, I
measured a web page from 4 locations around the globe. Knowing this sample, what
can I infer anything about the population of download times as a function of
RTT. <BR><BR>If I assume that Download time (dt)can be expressed as
follows:<BR><BR>dt = n* RTT + c<BR><BR>where n is the number of round trips (RTT
ping pongs, includes one burst of data which can be multiple packets) with c
being the server stall time between sending the data or server processing time
plus some random noise all factored into once constant. <BR><BR>The above
equation is of the form y=mx +c and I can equate the slope with that of number
of round trips (makes sense as the lesser the number of round trips the lower
the response time) while x is RTT.<BR><BR>So if I take enough sampls, say 10,
and perform a regression analysis on those to generate the equation wouldnt that
classify the population. If I have such an equation then I would plug in various
RTT(s) and asuming the R-squared value is high wouldnt that be representative of
real performance. A few initial measurements showed encouraging results but a
few measurements didnt converge and a few had negative valus,etc. <BR><BR>Before
I go further and present this to an internal audience I want to poll this group
for any feedback/remarks/comments on using this method and its
pitfalls.<BR><BR>-Paddy Ganti<BR><BR><BR><BR></BODY></HTML>