[arin-ppml] IPv4 is depleted today - unrealistic statements about IPv6 inevitability

[michael.dillon at bt.com]

> I don't think they will "need" IPv6, because IPv6 does not 
> give them what their customers need: IPv4 connectivity.

> A major limitation is that this service would not allow the 
> end-user's PCs to grab a port on the NAT boxes public 
> address, as I understand is often done today for P2P 
> applications and perhaps some games.

You seem to be assuming that everything else stays the
same, and that IPv6 users have to adapt to the reality
of the IPv4 network or do without. But that is not a
reasonable view. Everything is in flux and as IPv4 exhaustion
becomes better publicized, the developers of IPv4 software
are beginning to make it IPv6 compatible. 

So yes, there are some things that won't work well or won't
work at all, if IPv6 users have to go through NAT gateways
to access the IPv4 network. But that only means that there
is greater impetus for people to put these things on the
IPv6 Internet, and then the problem goes away.

Transition to IPv6 is an evolutionary process, not some kind
of big bang product launch.

> I think the costs of running dual stack and making it work 
> would be pretty high.  

So do many others. That's why we are looking at using 6PE over
an IPv4 MPLS core network. Minimize the impact on the existing
network. Any ISP with routers in more than one city really should
be considering migration to MPLS as a possible transition to 
the IPv6 Internet.

> There are plenty of apps which don't 
> work with IPv6 and there are few servers on IPv6.

This is just a time argument. Nobody is saying that today is the
time to start hooking up IPv6-only customers. We are saying that
today is the time to start trialing IPv6 in the lab, and to start
the careful process of deploying it on your live network. This
will all take a couple of years to get right, and then there will be
apps and servers on IPv6.

> It could get this bad if there is no widely deployed 
> Core-Edge Scheme to solve the routing scaling problem (LISP, 
> APT, Ivip, TRRP or Six/One Router). 

There are other non-technology ways to solve this problem.

> There's always a point in an "IPv6 will start to be widely 
> adopted real soon now, really" story where the logic takes 
> leave of this Earth.

Agreed. But the fact is that IPv6 has been deployed in production
for many years in research and academic networks. And in all
continents, government and military organizations are increasing
their deployment. People who remember the early Internet understand
how exponential growth works. A precondition for 1500% per annum
growth (that was 1995), is many, many years of fractional percentage
growth. Slow and steady wins the race.

> The 100% connectivity the IPv4 Internet offers to all other 
> Internet users is not at all provided by IPv6

It's not provided by IPv4 either. There are vast parts of the IPv4
Internet that you cannot get to because of language barriers and
because of routing issues at the edge of the big cloud.

> Pure IPv6 connectivity won't be any value to most end-users 
> until some very high fraction of all other end-users - 
> including those at home, in offices, and all web servers, 
> game servers etc. - have native IPv6 connectivity too.  

You don't understand the network effect and how it drives 
that long slow tail that leads to a sudden leap in growth.
It's not flashy, but its essential and it is well under
way for IPv6.

--Michael Dillon