<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7653.38">
<TITLE>RE: [arin-ppml] "Millions of Internet Addresses Are Lying Idle"(slashdot)</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/plain format -->
<BR>
<P><FONT SIZE=2>But like many political statements....it's not necessarily the substance that is important.<BR>
This spark, if fanned could be an opening to do reclaimation...not that I am in favor of this, but it would preserve the status quo.<BR>
<BR>
<BR>
<BR>
-----Original Message-----<BR>
From: arin-ppml-bounces@arin.net on behalf of Leo Bicknell<BR>
Sent: Sat 10/18/2008 4:22 PM<BR>
To: ppml@arin.net<BR>
Subject: Re: [arin-ppml] "Millions of Internet Addresses Are Lying Idle"(slashdot)<BR>
<BR>
In a message written on Sat, Oct 18, 2008 at 09:01:17PM +0000, Paul Vixie wrote:<BR>
> "The most comprehensive scan of the entire internet for several decades<BR>
> shows that millions of allocated addresses simply aren't being<BR>
> used. Professor John Heidemann from the University of Southern California<BR>
<BR>
Unfortunately while I might even give him that this is the most<BR>
comprehensive, I believe there are more than a few severe holes in<BR>
it that mean it may not be representative.<BR>
<BR>
A large problem is that many hosts will not respond to unsolicited<BR>
ICMP, TCP, or other packets. Indeed, many hosts today are protected<BR>
by multiple levels of this, service providers helpfully dropping<BR>
some packets before a organization based firewall dropping some<BR>
more, followed by a host based firewall dropping the rest. I have<BR>
no doubt that millions of computers were missed as a result.<BR>
<BR>
However, the larger problem is that this is the entirely wrong<BR>
definition of "in use". Let's take a simple example of a University.<BR>
They may have wired dorm rooms, students with laptops, and wifi<BR>
enabled classrooms. If you ping at 11AM, the classroom IP's respond,<BR>
and the dorm ones do not, if you ping at 11PM, the dorm rooms<BR>
respond, the classrooms do not. And if you ping at 12 noon when<BR>
they are all walking to lunch almost none of either respond.<BR>
<BR>
This survey equates in use to be having a computer actively using<BR>
it and responding at the time of the probe. I would submit that<BR>
is a very poor definition. The dorm room block and classroom blocks<BR>
are both "in use" every day, and unfortunately the technology<BR>
involved does not make it possible to move back and forth.<BR>
<BR>
In fact, the interesting thing to do here is simple. Cross reference<BR>
the "pingable" IP's in a block of space with requests over the past<BR>
month to google, or akamai, or updates.microsoft.com and see what<BR>
percentage of the "unpingable" IP's are in fact in use on a regular<BR>
basis. We could haggle over if "in use" means active in the last<BR>
day, week, or month, if it really matters.<BR>
<BR>
A pretty chart for sure; in terms of being helpful for planning the<BR>
future, it's borderline chart junk.<BR>
<BR>
--<BR>
Leo Bicknell - bicknell@ufp.org - CCIE 3440<BR>
PGP keys at <A HREF="http://www.ufp.org/~bicknell/">http://www.ufp.org/~bicknell/</A><BR>
<BR>
</FONT>
</P>
</BODY>
</HTML>