[ppml] NANOG IPv4 Exhaustion BoF

michael.dillon at bt.com michael.dillon at bt.com
Fri Mar 7 06:06:41 EST 2008

> Tom, as Scott has already requested from you, we haven't seen 
> any other options from you yet! Once the unallocated space is 
> exhausted then if you still want IPv4 addresses you are going 
> to have get addresses from someone who already has them.

Not so. You can also NOT GET ANY new IP addresses, and suffer
the consequences. That could mean bankruptcy or it could mean
increased costs to transfer to IPv6 in a hurry. The increased
costs would mainly be lost opportunities while you fix the 
problem, and increased churn rates as some customers treat your
company like a leper.

The fact is that nobody needs to be dependent on new IPv4
addresses after the runout date. There is plenty of time for
companies to make their businesses work with a mix of IPv6 and
IPv4. There have been some very public demonstrations of this
mixture at the last ARIN and NANOG meetings. There will be 
another demo at the upcoming IETF meeting. So far, if you
examine the results of these demonstrations, there are no 
serious problems that could not be fixed within a two year
timeframe. And we do have two years, probably more, to fix
these issues.

Once it is demonstrably possible to run a fully mixed IPv4
and IPv6 network, the need for transfers disappears. Note
that I am not referring here to dual-stack networks. When
I say MIXED I mean that you have some endpoints that are
IPv4 and some endpoints that are IPV6 and that they can
both communicate with each other over infrastructure that
has at least one pure IPv6 section in it.

> Yes, the situation the this industry 
> has driven itself into is messy, but I'm pretty sure that 
> global collapse is not exactly on the agenda.

Indeed. The situation is more like the telecom collapse era
when companies that were not sufficiently robust went out
of business or were bought out or shrunk. This is a good
thing for the economy since it is part of the process of
natural selection of business models.

> What I would claim is that this transition is just another 
> risk factor in an industry that has its fair share already 
> and doubtless will continue to have its fair share in the future.

Exactly. There is no need for RIRs to take any special actions
to prepare for IPv4 runout. No transfer policy is needed. Just
keep running the RIR as normal and keep reporting status so that
we can see how much time is left.

--Michael Dillon

Don't forget that ARIN has a wiki with info (and pointers to resources)
for network providers who want to add IPv6 to their networks.

More information about the ARIN-PPML mailing list