[address-policy-wg] PI concept in general, was IPv6 PI resource question!

Chipping in 1.5 cents,

On Wed, Feb 16, 2011 at 8:55 AM, Shane Kerr <shane at time-travellers.org> wrote:
> Daniel,
>
> On Tue, 2011-02-15 at 20:41 +0100, Daniel Roesen wrote:
>> We're allowing PIv6 for "multihomed" (whatever that means really) sites.
>> Those could "just simply" use PA space from any ISP they connect to. Why
>> do we make Internet multihoming special compared to Internet-plus-others?
>> With the "multiple IPs per device" argument there cannot be PIv6. I
>> thought we've left that behind by now.
>
> So I was wondering "why do we allow PI for IPv6 for multihoming?", and
> realized "that's how we do things in IPv4". AIUI people don't consider
> it reasonable to get one ISP to carry routes from another on behalf of
> their customers. Fair enough.

Routing slots notwithstanding, the hurdles the RIPE community puts up
for PIv6 is *directly* contra-productive to roll-out of IPv6 in
general, and the "end-to-end"-roll-out specifically.  In the absence
of a solution to the renumbering problem that your random IT-staffed
medium sized company can actually understand and use, NAT:ing is a lot
simpler, and in many cases fits the v4 model anyway.

Forgetting for a moment that having a chunk of your LIRs space is "PA"
and not PI, this whole idea that you should route a more-specific of a
LIR's allocation is also contra-productive to the ease of
understanding and implementing filtering in v6-land (which is key to
the future of the 2000::/3 trial, if you will). It makes the PI holder
dependent on the LIR, not only for the lease of the address block, but
in the face of routing policy, also its reachability.
Irregardless of the address block being PI or not, it is the same
number of routing slots in an unfiltered DFZ.  And I see only clear
downsides to "de-aggregating" the LIRs blocks. It also make route
lookups deeper.

LIRs (in terms of LIRs) are short-sighted to think easening up PI is a
threat to their business anyway IMO, since the potential customer
base, the size of the market, would grow considerably.

> In this IPv6 PI discussion people say it is too expensive to be an LIR.
> This strikes me as weird... even cheap network gear costs way more than
> the LIR fee, and certainly the personnel costs are many times that, even
> in countries where technical folks are poorly paid. But anyway....

I find it hard to motivate a difference between PIv4 and PIv6 in this regard.

> Is there another reason to want to limit PI assignments? There may be.
> Certainly the folks over in abuse-land think we should have super strict
> controls over all addresses, and revoke them at the first hint of
> naughtiness. Or there may be administrative reasons - an entirely
> different kind of infrastructure is needed to manage hundreds of
> thousands of low-fee (well, NO fee now) PI-holders than thousands of
> high-fee LIRs. I don't know.

Do you mean there is a superlinear growth of IPRA needs at the RIR if
PI scales up? Why? I don't get why that would be the case at all.
Isn't it just more of the same? The monetary cost for an application
should cover the true costs of the RIR in handling it of course. (The
fee for a end-user direct contractual relationship with RIPE is
comparative to LIR status today anyway)

> So... do you think we should simply remove all restrictions for IPv6 PI
> and issue space if someone wants it, period? If not, what possible
> restriction would there be?

Clue-filter seems appropriate and sufficient, ie, no change from v4,
except the need to be creative is inherently smaller with PIv6 than
PIv4 which is a good thing.

> I guess the "LIR middleman" model comes from the mists of time, long ago
> in history. It certainly reduces the burden on the RIPE NCC, but it
> never really struck me as sensible. After all, isn't the whole point of
> PI space that you don't want to have to depend on an LIR?

Indeed.  You go to the IP shop and leave your name and number, and a
pay small fee and then you walk out with your leased space. Then you
go to the Internet field and start playing with it.  The only function
of a LIR in this sense is to increase the number of shops where you
can get the PI, making it more available, so you won't have to travel
to the head quarters, if it is far away. :) There is little reason,
except for consulting services, for you to be in contract with your
random local PI shop for this, since it is only a service mediator for
the head quarters.

/Martin
IPv6 to the people!