[ppml] Resurrecting ULA Central [was: Re: Policy Proposal 2006-2: Micro-allocations for Internal Infrastructure - to be revised ]

[Owen DeLong]

--On April 20, 2006 4:36:37 PM -0400 Thomas Narten <narten at us.ibm.com> 
wrote:

> On 4/14/06, Randy Bush <randy at psg.com> wrote:
>
>> fwiw, after discussion with jason, i would support a more simple, direct,
>> and clear proposal to the same end.
>>
>> randy
>
> Question:
>
> I gather that resurrecting
>
> http://tools.ietf.org/html?draft=draft-ietf-ipv6-ula-central
>
> would also solve the technical problem at hand (since the technical
> requirement seems to be globally-unique address space, with no
> need/desire to have it be globally routable).
>
> I understand that RFC 4193 style addresses are not "unique enough" for
> that purpose.
>
> Would there be interest in resurrecting the ula-central document?
>
> Pros:
>
> 1) globally-unique space would be available to everyone, including end
>    sites. I.e., for pretty much any purpose. Even during the ARIN
>    meeting, it was pointed out that anyone with an ASN could/would
>    presumably want something like this.
>
> Cons:
>
> 1) ARIN pretty vocally shot down the document a year or more ago, and
>    the IETF basically decided "we don't need this so badly as to have
>    a showdown with the ARIN community". Having said that, I (and
>    others) still think the idea has some merit and would be willing to
>    push on it on the IETF end, assuming we wouldn't get a repeat
>    reaction at future meetings for our efforts...
>
>    Note: AFAIK, no such reaction seemed to come out of APNIC or RIPE.
>
I remember most of the ARIN opposition to this being opposition to
ULA-Random which, unfortunately, was preserved, and, concern that
somehow ULA-Central would end up getting routed if there was no PI
policy.  Since a PI policy appears to be imminent, I would not have
nearly such an objection to ULA-central now.

> 2) Does solve Jason's problem, but perhaps there is no desire to fight
>    the larger battle at the expense of just solving the narrow/simple
>    problem (i.e., just for ISPs). Note, however, that it will
>    presumably take at least another 6 months (until the St. Louis
>    meeting) to make progress on this. (Realistically, it would
>    probably also take 6 months to get the ula-central document through
>    the IETF, assuming there was no significant opposition from ARIN,
>    so I'm not sure either approach is necessarily longer).
>
I think it is 5.999999999999999999998 of one 5.999999999999999999 of the
other.

> 3) Would make such address space available to everyone, including all
>    end sites, not just ISPs. Not sure this is necessarily bad, but it
>    will result in orders of magnitude more such addresses in use. And
>    the concerns raised in the past centered around the fear that ISPs
>    would be  asked/forced to route them...
>
Given the ability for sites to get PI space (assuming 2005-1 moves
from last call to policy), I don't see this as being the major issue
I perceived it to be when ULA-central was proposed and it looked
like PI was impossible.

> I know that there is at least one person willing to resurrect the
> ula-central document, but I (personally) don't want to invest cycles
> in it if it's going to get a frosty reception in ARIN again. Been
> there, done that.
>
As one of the coldest shoulders presented to ULA (in general), I can
say that assuming 2005-1 does not go down in some last-ditch set of
unanticipated flames, I would not oppose resurrection of ULA-central
because I believe the potential for economic-based abuse is much
less now.

Owen


-- 
If this message was not signed with gpg key 0FE2AA3D, it's probably
a forgery.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 186 bytes
Desc: not available
URL: <https://lists.arin.net/pipermail/arin-ppml/attachments/20060421/64b1370e/attachment.sig>