[arin-ppml] Encouraging IPv6 Transition

paul vixie paul at redbarn.org
Sat May 12 12:03:21 EDT 2012

On 5/12/2012 1:30 PM, Mike Burns wrote:
> No offense, Paul, but this discussion IS about market and stewardship.

no offense taken, mike.

> And forgive me if I doubt your level of libertarianism, any
> libertarian I know would have simply pointed out that Enron went
> bankrupt, taking many investors and management down. Hardly a shining
> example of the faults of a market, and there is no doubt that the
> market Enron operated in was far from free and open.

i think we'd be far afield if we get into those details. my point in
mentioning libertarianism was that i'm a knowledgeable amateur when it
comes to the austrian school, but am not a knee-jerk anarcho-lib as some
in this audience occasionally seem (to me). so i forgive you noting that
we're at cross purposes here.

> Instead of treating ASNs as the valuable and scarce resources they
> are, and creating policies to engender their recycling and re-use
> through the profit motive, they languish like IPv4 addresses did
> before we began to create a market for them.

on the merits of the specific policy you mention (asn reassignment),
i've heard no credible benefit to it and one credible cost. i call
"4-byte asn's are less desireable" a canard, because many operators
around the world use them successfully and these asn's represent the
inevitable future once the 2-byte pool is exhausted. if i thought we
could have "steady state forever" on 2-byte then i'd think differently.
there is a cost in recycling, brought up in the vancouver meeting by
bill woodcock, which is that old network traces stored on disk are
pretty stable in the asn:identity mapping. some asn's have churned, like
174 used to be psi and is now cogent. most have not churned. someone
sitting on a 20-year pile of netflow or bgp traces may not thank us for
making them go back and put a time-domain on every asn reference. that's
a small matter but it's more credible than the best example yet given as
to why asn's should be recycled more freely.

(yes, i wish i had asn 42 instead of woody... because of douglas adams.
that doesn't make it good policy.)

> You will note that my policy prescriptions always separate the free
> pool from the pool of already allocated resources. I have no problem
> with a needs test and other central control of the free pool, but
> markets work best for allocations of resources already held by
> disparate parties.

you could argue that the goodness of stewardship depends on the
efficiency of utilization, and that an asn sitting on the beach unable
to be recycled is therefore a bad thing. i don't think anyone could
argue. if you do this, you should quote arin's mission statement and the
relevant parts of rfc 2050.

but if you argue on the basis that markets are good, i predict very
little traction.

"I suspect I'm not known as a font of optimism." (VJS, 2012)

More information about the ARIN-PPML mailing list