Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 3 Feb 2001 23:56:17 -0800
From:      "Ted Mittelstaedt" <tedm@toybox.placo.com>
To:        "Cliff Sarginson" <cliff@raggedclown.net>
Cc:        "Tim McMillen" <timcm@umich.edu>, "G D McKee" <freebsd@gdmckee.com>, "freebsd-questions" <freebsd-questions@FreeBSD.ORG>
Subject:   RE: IP6
Message-ID:  <006801c08e7f$f3e907a0$1401a8c0@tedm.placo.com>
In-Reply-To: <20010203215937.A5437@raggedclown.net>

next in thread | previous in thread | raw e-mail | index | archive | help
Hmmm...  Now why do you want one for your cat?  Don't
roaming users generally use DHCP? ;-)

I'm not holding my breath to see IPv6 implemented.  It seems
that every decade has some fantastic networking technology
that everyone is going to switch to.  20 years ago it was
the ill-fated ISO protocols.  A lot of people put a lot of
impressive work into it, they all use the technology as great
material to write doctoral dissertations and thesises with,
we generate a bunch of PhD's, and then after 10 years it
becomes obvious that the new technology isn't going to
happen and people all lose interest and start working on
something else.

People tend to forget that the strength of backwards compatibility
in large networks is so powerful that you can still take a
telephone that was produced 100 years ago, attach it to a
modular jack, and plug it into the phone network and get
dialtone.

IPv6 has a tremendous achilles heel that is so large most
people forget it is there.  It is simply that the single
most major argument to moving to it is that it will allow
everything, such as your toaster, microwave, coffee mug, and
cat, to have an IP number assigned to it.  There is an implicit
assumption by it's designers that everyone _with_ and IP number
just automatically wants a _public_ IP number.  Thus, a crisis
exists because the supply of IPv4 numbers is obviously finite.

However, if you look at the assumption that everything with an IP
number actually needs a public, reachable IP number, I think you
will find that it's very, very shaky.  With current trends in
security, people are actively working out ways to _avoid_ this
very thing.  From running NAT overloaded onto a single public IP
number, to PPP-over-TCP in the cable modem crowd, to PPP-over-ATM
on the DSL crowd, the trend is to either assign devices private
numbers or to not permanently assign them numbering at all.

Once you take away the assumption that everything that's TCP/IP
needs a public IP number, it almost totally guts all reasoning to
switch to a new IP numbering system.  Instead, there are tremendous
compelling arguments for organizations to implement IP number
conservation.

Consider that companies like Microsoft have entire class B public
subnets assigned to them.  Yet, do you really think that most
of the 60 thousand IP numbers that this subnet represents are actually
reachable from the Internet?  Of course they aren't.  It's highly 
unlikely that Microsoft has more than 200-300 devices that are
authorized to accept incoming TCP connections initiated from hosts
on the Internet.  All they really need is a couple of legal class C
subnets assigned to them and they can put the rest of their systems
behind a NAT/firewall.  And considering how embarassing some of their
recent break-in's are, they better be thinking like this if they know
what's good for them.

Today, ARIN is charging the equivalent of 15 cents a year per IP number
for organizations like Microsoft.  For the average Internet user
at home behind a DSL line that's a static IP number, this is nothing.
There's a tremendous amount of room for improvement here.  If Arin
raised the cost to $10.00 per year per IP number, and the ISP's all
passed the cost along to their users, this would still represent
nothing to that DSL user with a static IP number.  But, it would
jump the cost of holding on to that class B from a miserable $5K
a year to over $600,000.00 a year.  This would result in tremendous
IP number conservation by the organizations that are the worst abusers,
and it would solve the problem of a shortage of IPv4 numbers.

Switching to IPv6 will mandate a costly renumber by everyone.  Raising
the price of IP numbers will only mandate a costly renumber by the
worst abusers, the people that efficiently utilize IP numbers won't
be punished.  (because they will be able to pass along the cost to the
end users)  Either way you do it, the abusers are going to have to
renumber, if they aren't on private IP numbering. 

The problem with these engineers that dreamed up IPv6 is they are living
with a 20-year old image of the Internet where IP numbering was free
and anyone who wanted to plug in could do so, and everyone wanted to
make love to everyone else.  Well, the Internet was like this back in
the days of NSF-net when the US Government was paying the bills but it
ain't like that today.  The Internet is a commercial network now and
you can't make engineering decisions on it that don't take into account
business and commercial issues, such as Return on Investment and
least cost solutions.  We aren't going to see IPv6 stuffed down our
throats until the cost of IP numbering becomes a significant part of
doing business, and at the rate that Arin (and the other numbering
registries) are going, I'm going to be an old man with my beard down
into my boots before that happens.

Ted Mittelstaedt                      tedm@toybox.placo.com
Author of:          The FreeBSD Corporate Networker's Guide
Book website:         http://www.freebsd-corp-net-guide.com


> -----Original Message-----
> From: owner-freebsd-questions@FreeBSD.ORG
> [mailto:owner-freebsd-questions@FreeBSD.ORG]On Behalf Of Cliff Sarginson
> Sent: Saturday, February 03, 2001 1:00 PM
> To: Ted Mittelstaedt
> Cc: Tim McMillen; G D McKee; freebsd-questions
> Subject: Re: IP6
> 
> 
> On Sat, Feb 03, 2001 at 12:44:58PM -0800, Ted Mittelstaedt wrote:
> > I support IPv6 wholeheartedly!  I just got word that
> > the parties responsible for it have agreed to renumber the
> > ten thousand some-odd IP addresses that I'm responsible for!
> > 
> > Ted Mittelstaedt                      tedm@toybox.placo.com
> 
> Yippee ! I want one for my toaster, my microwave, my cat,
> my coffee mug .. pleeze..
> 
> Cliff
> 
> p.s. I am reading (and enjoying) your book, a few constructive
> comments will be winging their way when I finish it...
> 
> 
> 
> 
> To Unsubscribe: send mail to majordomo@FreeBSD.org
> with "unsubscribe freebsd-questions" in the body of the message
> 


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-questions" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?006801c08e7f$f3e907a0$1401a8c0>