> What I am nervous about on the other hand, is that the lack of IPv6
> adoption is making the alternative growth scenarios using Carrier
> Grade NAT ever more attractive for some ISPs. CGN breaks the key end
> to end principle that we have on the Internet and many mobile networks
> are using CGN. Some non-mobile TelCos are now saying "hey we do not
> need to roll out IPv6 - just look how happy customers are in using
> their mobile networks with CGN!"
thanks for understanding what the enemy is.
at the dublin meeting, cgn could be seen on the horizon, and i was
sufficiently unhappy to stand up in plenary and rant against it. russ
and a few others said "that's nice, put your money where your mouth is.
what's the alternative?" this was the impetus for a+p, as the document
states right up front.
for a broader view than 6346, see
Nejc Skoberne, Olaf Maennel, Iain Phillips, Randy Bush, Jan Zorz,
and Mojca Ciglaric, "IPv4 Address Sharing Mechanism Classification
and Tradeoff Analysis," IEEE/ACM Transactions On Networking April
2014. http://archive.psg.com/130419.ton-v4-sharing.pdf
some of the alternatives sure make stateless a+p look pretty good, eh?
my favorite may be ds-blight; nat in the core and you get to fork-lift
your cpe, a double win there.
ipv4 has pretty much run out; the iana free pool ran out long ago, along
with apnic and ripe. arin will very soon. like it or not, ipv6 is
depressingly poorly deployed; and our major forward strategy seems to be
denial as so well demonstrated in this thread. the stonewalling and
denial have distracted from reducing the friction of deploying ipv6, so
have helped drive cgn deployment; so it *really* pisses me off [0].
the paragraph to which bob hinden objected
We are facing the exhaustion of the IANA IPv4 free IP address pool.
Unfortunately, IPv6 is not yet deployed widely enough to fully
replace IPv4, and it is unrealistic to expect that this is going to
change before the depletion of IPv4 addresses. Letting hosts
seamlessly communicate in an IPv4 world without assigning a unique
globally routable IPv4 address to each of them is a challenging
problem.
is tragically true; please point out a single untruth. get over it.
and i like the situation less than most folk here, and with good reason.
my employer has thrown money at real ipv6 deployment for over 15 years,
wrote a lot of the stack many implementations still use today, ...
randy
---
[0] - i still feel the bruises and broken bones of the multiple years it
took to get the tla/nla crap removed. and that was just one of
the battles.