[arin-ppml] IPv6 Non-connected networks

Barbara Roseman barbara.roseman at icann.org
Thu Feb 4 19:21:31 EST 2010


Bill,

I think that's a fair restatement of what I was saying, except you left out the part about uniqueness vs. routeability. My points were twofold:

1) ARIN can only offer unique registration of IP addresses, they cannot offer unique use of addresses -- nor can anyone else because Bad People tend to ignore the rules, and Ignorant People are, well, ignorant.

2) The ARIN community and the NANOG community have discussed repeatedly that it's not simple cause and effect from ARIN's allocation/assignment policies to operator's routing policies, though the influence is extreme. If the connectivity provider is large enough, as CJ notes, than they can do whatever they want.

-Barb
On Feb 4, 2010, at 2:37 PM, William Herrin wrote:

> On Thu, Feb 4, 2010 at 4:14 PM, Barbara Roseman
> <barbara.roseman at icann.org> wrote:
>> I think there's a subtle but important distinction
>> that's being lost, and that Cathy and Marla tried to clarify.
> 
> Hi Barbara,
> 
> If I understood the distinction that Cathy and Marla drew, the point
> they're making is that nowhere does ARIN say what routes an ISP has to
> carry. ARIN doesn't even say what routes an ISP ought to carry. Nor is
> there any other legal obligation which compels ISPs to carry any
> particular routes for addresses ARIN allocates. Like Verizon has done
> with IPv6 /48's, ISPs are welcome to disregard any routes they don't
> like for any reason at all. That they rarely exercise that choice
> means nothing more sinister than that routing the address ARIN
> allocates is good for business.
> 
> Do you feel that's a reasonable re-statement of what they said? I'd
> hate to be obtuse about it, so if I'm missing the their point please
> help me understand.
> 
> 
> I think I understand the claim, but I dispute it.
> 
> Cause and effect is no less potent for being indirect. The
> practicality of altering BGP routing activity based on a second feed
> of assignments matched to assignment classes is doubtful. Even if it
> could be done from a technical perspective, ARIN doesn't publish the
> criteria by which each registrant qualified for their addresses, so
> such a feed has no source for its content.
> 
> Realistically, functionally, technically, the tools I as an operator
> have for implementing routing policy can only paint with a broad
> brush. Discard small announcements in this /8. Depreference that one.
> Unless ARIN classifies registrants in a manner consistent with those
> tools, my choices for routing policy devolve to two basic options:
> accept routes announcing ARIN IP addresses or don't.
> 
> In other words, ARIN sets the routing policy for North America. Not
> explicitly, cleanly and openly, but haphazardly, indirectly and
> silently.
> 
> The unspoken and unintentional yet cumulative net effect of ARIN's
> allocation practices make ISPs a mob-style offer they can't refuse:
> carry all of the addresses we approve or carry none of them and lose
> your business.
> 
> At this point we're stuck with that for IPv4. The addresses are
> already deployed. With some mild adjustments to our allocation
> practices, we don't have to be stuck with that effect for IPv6.
> 
> Regards,
> Bill Herrin
> 
> 
> -- 
> William D. Herrin ................ herrin at dirtside.com  bill at herrin.us
> 3005 Crane Dr. ...................... Web: <http://bill.herrin.us/>
> Falls Church, VA 22042-3004




More information about the ARIN-PPML mailing list