[arin-ppml] Policy Proposal 122: Reserved Pool for Future Policy Development
In a message written on Thu, Nov 18, 2010 at 08:21:37PM -0500, Hannigan, Martin wrote:
> I agree, but the way that this is written incents [you] to participate and
> to become a part of the solution. Right now, I see [you] as part of the
> problem. You have a comfortable landing, more so than the rest of us. I
> think this is another demonstration of how broken this is.
You're implying there is some pain to me, but that is not the case.
If we can't get IPv4 space to deploy more sites we will deploy IPv6
only instances. That doesn't cause me any pain, indeed it cuts my
work in half. If anything it makes my life easier. There are
already plenty of root servers (not just ours) to handle the load
well into the future, what is being limited here is the ability to
reduce latency to root servers (in my case, or the nearest exchange
What it does mean is that there will be less incentive for our
partners and sponsors to deploy more root nodes. You might think
that was "pain" for us, but we don't sell instances and many end
up costing us money.
So the folks that are being hurt here are folks in areas that are
under-served by root servers or exchange points.
Now with all that said, the main reason we need the /10 is not
critical infrastructure. The real reason is I'm sure somewhere in
this transition mess there will be a new problem. I don't know if
it will be needed for transition technology or some form of critical
infrastructure, or what.
Now, if you want my input, here it is:
- We keep the /10 "off limits" to everyone through "run out", as
is I believe already the current policy.
- We wait a minimum of 6 months, and probably more like 12-18 months to
see what unforseen problems occur around run-out, and if they need
- If after that time there are critical issues, we write policy to
allocate the space in ways that solve the problems.
If thre are no unforseen problems we set aside a very small amount
for critical infrastructure (on the order of a /18-/20 would be my
guess) and hand out the rest via normal mechanisms.
I will admit I'm biased here, but I'm trying hard to be objective and
work in the best interest of the community, and I feel pretty strongly
the plan I've outlined above does that.
Leo Bicknell - bicknell at ufp.org - CCIE 3440
PGP keys at http://www.ufp.org/~bicknell/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Size: 826 bytes
Desc: not available