Re: [v6ops] Chair decision on WGLC for draft-ietf-v6ops-dhcp-pd-per-device-04

[Gert Doering <[e-mail]>]

Hi,

On Sun, Nov 05, 2023 at 03:36:44PM +0100, Martin Hun??k wrote:
> To be honest, I don't know from the top of my head.
> 
> However, we don't see approx. 3.6 billion Android smartphones are all asking for their own /64 yet, do we? If all of those were in a single network, we would need /32 just for them. If Apple is to join the club, we will be on approx. /30. In reality, where there are multiple networks and where every single one of them had to somehow solve the higher demand, how much address space would this draft cost?
> 
> Most of the time, /64 is useless for the phone. So much lost for a very little gain ...

Actually, having /64s handed out to phones *in 3G/LTE/etc.* networks is
a much less problematic use case, and this is due to the address distribution
hierarchy.

Over here in RIPE land, the smallest a RIPE member (=LIR) can get is a
/32, or really a /29 nowadays, and (mobile) telcos are really expected to
be a member of their respective RIR.

In a /29, there's space for 34 billion (2^35) /64s, so even adding a
few internal distribution/aggregation layers ("one /32 per country, 
one /40 per region") there are still sufficient /64s for each phone
that could connect to "a region" - Telcos are good at this sort of
sizing.

Of course, if you get that wrong ("a /60 per cell, nobody would connect
more than 16 IPv6-enabled phones here") it won't work.


Now, with networks that are *not* LIRs - and most end user networks are
expected to *not* become a RIR member, due to scaling on that side - and
thus receiving a /56 or /48 from their address-distributing ISP, being
able to hand out /64s "like they are free" on every wired or wifi segment
is just not going to work out, thus creating avoidable annoyance on the
customer-to-ISP side.

RIPE originally had "one /48 per customer", and this was reduced to
"oh, well, maybe make this a /56?" when Geoff Huston did a bit of math
and figured out that "one /48 per end user, plus 1-2 extra for the other
ISP, plus internal hierarchy loss in the ISP networks" might actually
make us run out of /48s.  Thus, /56s were introduced.

(Lorenzo might have been at that RIPE meeting...)


Now (this is actually part of a response to a different mail) - in RIPE
land, ISPs get a /29, so we could loosen up the end user policies a bit,
but ISPs generally at least get the "do a /56 or more" part *right*.  In
ARIN land, there is an incentive for ISPs to only get a /36, because,
/32 costs more money, so there is an incentive for ISPs to more tighly
conserve space and give all their customers only a /64.  I think that 
this sort of charging policy is extremely short-sighted, but ARIN is not
my home RIR... I'm busy enough with RIPE NCC charging "ideas".

Gert Doering
        -- NetMaster
-- 
have you enabled IPv6 on something today...?

SpaceNet AG                      Vorstand: Sebastian v. Bomhard, Michael Emmer
Joseph-Dollinger-Bogen 14        Aufsichtsratsvors.: A. Grundner-Culemann
D-80807 Muenchen                 HRB: 136055 (AG Muenchen)
Tel: [phone]         USt-IdNr.: DE813185279