RE: [RRG] Renumbering...

["Fleischman, Eric" <eric.fleischman@boeing.com>]

>From: Tony Li [mailto:tony.li@tony.li] 
>Hello yet again Eric,
 
|I concur with you that ISPs will usually be strongly motivated to 
|support their largest customers so that the PI space will be supported 
|indefinitely for both IPv4 and IPv6. If this would not be the case, 
|then the largest end users may become motivated to take action 
|themselves -- such as to decree ourselves to be our own ISPs (i.e., 
|this is just one of many scenarios that large end users have toyed with

|since 2000).
|Regardless, however you cut it, I believe that the PI space will exist 
|and I recommend that you emotionally accept that probability. More 
|pertinently, I recommend that RRG accepts PI support as a requirement 
|for advancing any RRG variant.

>Thanks, but I'm not going to be accepting that anytime soon.  When 
>true PI is a requirement, then the 'net cannot possible scale.  
>Period.  Full stop.

>Thus, this means that NAT is necessarily a requirement.

Tony,

I feel embarrassed at the RRG bandwidth that my postings have recently
consumed. I am eager to resume my traditional near-silence. However, I
am very concerned at your reply. I reluctantly feel compelled to
respectfully state a position that I believe must be heard.

I hope that you know how greatly I esteem your consistently brilliant
insights and honor your many contributions and leadership. I have felt
this way ever since our first meeting in the early 1990s. Nevertheless,
I believe that your latest posting reveals a shockingly inaccurate view
of the Internet and therefore a seriously flawed understanding of the
nature of worldwide network scaling.

From your postings in recent years, I am confident that you distinguish
between tiers of ISPs and that you recognize that not all ISPs perform
identical roles in servicing the Internet:  Some perform a core network
backbone function and others solely perform a relaying function at the
backbone's edge. I also, perhaps incorrectly, perceive that you don't
seem to realize that the difference between ISPs is trivial when
compared to the extreme differences between end users. I state that any
model that treats all end users as equivalent is inherently broken and
any Internet solution based on such a faulty notion is unlikely to
succeed because it does not recognize reality.

Quick: what is the largest network in the world? Is it AT&T? No. Sprint?
No. Deutsche Telecom? No. It is the US Government. The US Government's
networks are the largest network in the world. 

My employer's internal corporate network is small and insignificant when
compared to the US Government's. Nevertheless, there are few ISP
networks of equivalent size to ours using whatever metric you choose to
use. Our corporate network includes a PoP in every airport of the world
in which one of our products can physically land -- there are
exceedingly few countries, therefore, that our internal network does not
touch. True, we leverage SITA, SATCOM, local Telcos and a host of
network services to accomplish that. But compare our coverage of
continents like Africa or countries like Papua New Guinea with any ISP
anywhere and you will begin to understand the nature of our corporate
network. In April 2008, 65 Pbytes (i.e., 65 * 10**15) of data was
conveyed across our corporate *backbone* alone. This is less than was
conveyed in the previous month (March) but is fairly typical of recent
months. Certain readers of this posting will recall the large number and
huge size of our North American data pipes as they existed in 1994
before that information became proprietary. Compare the relative number
of routers (thousands in our case), number of users directly served
(hundreds of thousands in our case), number of computers (vast), or
whatever metric you choose and you will see that our network is
comparable to that of the largest ISPs -- and we are by no means the
largest end user network, not even close.

When we talk about PI versus PA space we are really talking about
whether network addresses are owned or leased. We are also talking about
business dependencies. You can look up our public PI address space and
see its scope. What I want you to realize is that when we became part of
the Internet we owned our own addresses -- that is the model that we
bought into. Certain forces are trying to re-define that model, trying
to compel us to lease addresses and to become dependent upon outside
ISPs that are smaller than us. How dare they!! Who are they to try to
put our business at risk? Whether they realize it or not, the new model
that they are trying to foist upon us resembles blackmail -- switch ISPs
and it will cost you an arm and a leg. This is unacceptable. The PI
space is a non-negotiable fact. Any model that does not accept PI for
large end users for *both* IPv4 and IPv6 is inherently broken. 

Concerning your equation that PI equals NAT, I can only say Bah! (or
whatever the English equivalent is for the German doch!). That is
ridiculous. (I.e., it is a function of the large multiplicity of
different PI spaces, not the existence of PI itself -- 2000 different PI
spaces in IPv6 will not harm the Internet.) However, let's pretend that
it is accurate. Given that, then NATs are ****vastly**** preferable to
losing PI. Vastly. Incomparably so.

--Eric



--
to unsubscribe send a message to rrg-request@psg.com with the
word 'unsubscribe' in a single line as the message text body.
archive: <http://psg.com/lists/rrg/> & ftp://psg.com/pub/lists/rrg