On Mon, Feb 13, 2012 at 03:42:58PM -0500, Noel Chiappa wrote:
> From: Doug Barton <dougb(_at_)dougbarton(_dot_)us>
> If the RIRs do not deny these requests there is likely to be a revolt.
On what grounds? The ISPs will come along and say 'I have X new customers,
please give me more space for them'. The former being true, on what ground can
the RIRs refuse (modulo cases like RIPE)?
RIPE NCC is not unique; IIRC there has been extensive synchronisation between
RIRen
on the subject of sunset allocations.
I'd expect any RIR to more or less laugh at the request of a /10 (the
size of network we've been told (without any proof beyond handwaving)
is required to do CGN nicely) at this stage of address depletion.
APNIC:
As of Friday, 15 April 2011, each new or existing APNIC account holder
is only eligible to request and receive delegations totalling a maximum
/22 worth of address space from the APNIC IPv4 address pool.
ARIN:
Has 3-month consumption cap on allocations. Apparently not so low on
space as the others.
LACNIC:
Has a /22 cap per LIR en route in its PDP.
RIPE:
Has a /22 rule, as discussed earlier.
AFRINIC:
I can't at the moment locate their policy, more than that they have
incorporated the "last five /8's" document as governing text in their
policy set. Pointers to text welcome.
To sum things up, we are at the stage where a /10 is a laughable
proposition. The largest usable block (bar squatting on DOD space)
for CGN is 10/8. I'd suggest planning for it. That last /22 will probably
end up as P router addresses in MPLS core, since all vendors have dragged
their feet in implementing v6 MPLS.
> If the IETF rightly denies this request then the ISPs are going to
> be forced to use the proper option, 1918 space.
How? The IETF has neither police, nor an army.
It is either 10/8 or squat. No other alternatives exist. I'd expect
squatting on the "wrong" /8 to be punishable as a Homeland Security
Department -related crime, if one takes the US-centric view.
Folks, we've run out.
--
Måns.
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf