[ppml] Proposed Policy: PI assignments for V6 (and v6 fees)

[Howard, W. Lee]

> -----Original Message-----
> From: Owen DeLong [mailto:owen at delong.com] 
> Sent: Monday, December 06, 2004 1:02 PM
> To: Howard, W. Lee; 'Stephen Sprunk'; ARIN Policy
> Subject: RE: [ppml] Proposed Policy: PI assignments for V6 
> (and v6 fees)
> 
> 
> > If a site isn't connected, why would PI space be necessary? RFC1918 
> > made it so that private organizations who interconnected 
> would always 
> > have overlapping assignments.  Maybe there's a Better Practice 
> > available for IPv6, to reduce the likelihood of conflict (by 
> > increasing randomization).
> >
> For one thing, RFC1918 doesn't apply to v6.  For another, the 

True, but we have operational experience based on it.

> V6WG killed the Site Local proposal (which I think should be 
> resurrected, but, that's a separate issue).  The current 
> proposal is ULA, which, for a  variety of reasons is 
> operationally a very bad idea in my opinion, but, meets your 
> criteria in your last sentence).

It overkills my criterion.  

> I think a better approach is to recognize that networks fall into 3
> categories:
> 
> 	1.	Connected -- need unique v6 addresses among a 
> common namespace known as "the internet".

"The Internet."

> 	2.	Disconnected -- might as well run their own 
> separate registry
> 		and distribute another copy of the entire v6 
> space in whatever
> 		way they see fit, no need for IETF to set 
> policy in any way.

Agree.

> 	3.	Semi-Connected -- This is what most people are calling a
> 		disconnected network.  The reality is that 
> these hosts are
> 		connected, directly, or, indirectly to the 
> larger internet,
> 		and, as such, need unique addresses.  A simple 
> recognition
> 		that these hosts are connected, but, don't want global
> 		reachability would allow us to allocate space to them in
> 		a rational manner.  I see no need for PI space for such
> 		hosts in an organization that doesn't otherwise qualify
> 		for PI space.  I'd like to see site-local 
> resurrected for
> 		these sites.  Another alternative would be for 
> these hosts
> 		to receive part of a PA block and simply filter 
> at the borders.

In a significant portion of these cases, they may be connected 
to "an internet" but not the Internet.  An internet is two or more
networks under separate authority which connect to each other.

> Either way, I think ARINs involvement should be limited to category 1.

I agree that ARIN's policy involvement is limited to numbering on
the Internet, but we sure (the public) sure have some useful expertise
on internets.

> > Start with:  
> http://www.arin.net/library/minutes/bot/bot200> 4_1020.html
> > 
> "The ARIN Board of Trustees sets the fees for 
> IPv6 allocations as 
> > follows:
> >
> > XS/Micro /48 $1,250
> > Small  /32 $2,250
> > Medium /30-31 $4,500
> > Large  /27-29 $9,000
> > X-Large /22-26 $18,000
> > XX-Large /21 or greater $36,000
> >
> > This fee structure is effective January 1, 2005."
> >
> Lee, what about non-subscriber prices?

Multiple choice:
A.  There's no such thing.  Policy requires 200 sub-assignments within 5
years (IIRC), therefore anyone meeting those criteria is a subscriber,
receiving an allocation.  http://www.arin.net/policy/index.html#six5
"To qualify for an initial allocation of IPv6 address space, an 
organization must . . . not be an end site."

B.  Except micro-allocations, which have a category all their own (above).

C.  In case an assignment policy is ever established, the Board passed 
a motion setting initial fees to be the same as allocations, with 
maintenance fees consistent with other maintenance fees (currently $100):
http://www.arin.net/library/minutes/bot/bot2004_0803.html item 12D.


 
> > Arguably, you could put this proposal together with the 
> newly adopted 
> > fee schedule and waiver and say that there's a two-year 
> trial period; 
> > after two years, end-user sites will have to decide whether to pay 
> > $2,250 annually, or return their space and renumber into 
> > provider-assigned space.
> >
> Or, during that two years, we can work with the board and 
> membership to try and achieve a more useful fee structure for 
> the new policies.

Fair enough.

Lee

> 
> Owen
> 
> -- 
> If it wasn't crypto-signed, it probably didn't come from me.
>