IPv6 and Firewall traversal

[bogstad-e+AXbWqSrlAAvxtiuMwx3w at public.gmane.org (Bill Bogstad)]

On Wed, Mar 30, 2011 at 12:32 PM, Chuck Anderson <cra-WCkJK2/AXBA at public.gmane.org> wrote:

>....
>> if it weren't for NAT, consumers would be paying their ISPs "per-node"
>> connection fees. If things move in that direction in a mostly-IPv6 world, we
>> could see a resurgence of NAT.
>
> That is not supposed to happen. ?There is no reason for an ISP to hand
> out IPv6 addresses one /128 at a time. ?They would be creating a
> management nightmare for themselves and hurting their customers. ?The
> latest recommendations on prefix sizing are here:
>
> http://tools.ietf.org/html/rfc5375
> http://tools.ietf.org/html/rfc6177
>
> There is rough consensus that residential customers will be provided a
> /56 prefix, enough for 256 /64 subnets, but there is no hard
> requirement on that specific size.

It probably won't happen, but you aren't thinking about this the right way.
If it's more valuable to you, EVEN IF IT DOESN'T COST THEM A NICKEL
EXTRA, then they would like to charge for it.  That's just the way that
business works.   Since residential broadband is basically a monoply
in the USA, competition won't stop them from doing this.   Now it may be
too hard, but if it isn't they will certainly think about trying.

There is also the argument that you will use more bandwidth if you have
multiple devices attached then if you are restricted to just one
which means it WILL cost them extra.  That's why you can't tether
your laptop to your cellphone when it has an unlimited data
plan.  Wireless companies know you will probably use 10x the bandwidth
if you are allowed to do this.

Residential broadband companies took a different path.   Customers could
NAT, but 'unlimited' really wasn't unlimited.  Over time,
they even acknowledged this and Comcast even has a meter you
can check to see how close you are are to the limit this month.

Bill Bogstad