[arin-ppml] ULA-C

[William Herrin]

On Fri, Apr 2, 2010 at 7:08 PM, Durand, Alain
<alain_durand at cable.comcast.com> wrote:
>> That's essentially the same as saying that routing an IPv4 /32 in the
>> Internet DFZ is "essentially a matter of how much an organization is
>> willing to pay its service provider to announce the prefix." It's an
>> intellectually dishonest statement.
>
> Do you remember the efforts from Randy Bush & others to limit the length of
> the prefixes advertized in the DMZ to /21s? Well, how many /24 (or longer)
> do you see now?

/24? Lots. Longer? Zero that I see announced from more than one of my transits.

And that's my point. Netops already know to exclude fc00::/7 from the
DFZ. More precisely, netops know to include only 2000::/3 in the
Internet DFZ (http://www.iana.org/assignments/ipv6-address-space/).
Too many people at too many organizations would have to choose to
exceptionally accept a given ULA prefix from within fc00::/8 for it to
be usefully routed on the Internet. this would be -effectively-
impossible.

As for Randy's efforts, you never had a strong consensus to limit IPv4
announcements to /21. A weak consensus sure, but "weak consensus" is
just a fancy way of saying "boisterous minority." The strong consensus
to filter IPv4 at /24 is all but set in stone these days and the
consensus that the IPv6 Internet consists of 2000:/3 until IETF says
otherwise seems likely to stick.

Likely enough to stick, at any rate, that you can tweak fc00::/8 with
the reasonable expectation that there won't be any impact on the IPv6
Internet DFZ.


As you are not an idiot, you already knew all this. That's why I
called your claim that getting a ULA into the DFZ would just be a
question of money intellectually dishonest.

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