[arin-ppml] Update on 2009-3: Global Policy for the Allocationof IPv4 Blocks to RIRs
tedm at ipinc.net
Fri Aug 21 15:35:20 EDT 2009
Martin Hannigan wrote:
> On Thu, Aug 20, 2009 at 6:10 PM, Milton L Mueller<mueller at syr.edu> wrote:
>>> by another RIR (or that should be managed by another RIR) well why
>>> can't they just horse trade among each other to say well I have this
>>> block that you could aggregate with some of your stuff and
>>> I'll give it to you, with the expectation that you will give me some blocks you
>>> have that are best managed by me.
>> This strikes me as a pretty good idea. (Details need to be worked out, obviously)
>> While one might want to keep voluntary return to IANA as an option, why not also authorize direct trades among RIRs (just in-kind trades, not purchases).
>> Why centralize a function and introduce an intermediary if you don't have to?
> It is a decent idea. I'm skeptical that the other RIR's would sign up
> for such a system
Right there I think we already have the beginnings of a problem, with
your term of "system"
This implies a formal system, I was speaking of an INFORMAL "system".
Ideally the hostmaster of ARIN should be able to send an e-mail to the
hostmaster of, for example, RIPE and say "I noticed that we got block
W.X.Y.Z/24 given back to us from Wonkulating Gronkulators, and I
also noticed that your managing the 2 adjacent /24's that surround it
and it seems to make more sense for you to take W.X.Y.Z/24 so why
don't you take it over?"
In short they should be able to just do these things without having to
go through all the rackafratz of handing it back to IANA for IANA to
hand back out to RIPE.
Or am I assuming too much about the hostmasters of the RIR's? I had
always thought the ARIN hostmaster we worked with seemed to be a
pretty competent fellow, but maybe I'm wrong and we need Grandma Iana
to supervise this?
since a) they may want it to apply to legacy space
> only since none are likely to hand over their inventories at the end
> (at least greater than 1 x /8) and b) that means that ARIN would be
> the heaviest outflow by default.
So what? The point here is to get the IPv4 in service, right, isn't
that the whole point of reclamation? Who cares if the people that
use it are in North America or elsewhere? The sooner we get the
bits and pieces assigned out the sooner all the scrap IPv4 will
be put in service and then people won't have an excuse any longer
to think that there really is some sort of future with IPv4 and
will have to get serious about IPv6.
How about if AfriNIC and LACNIC stop
> taking /8's and start taking fragments from APNIC, RIPE, and ARIN
> instead? That might promote better utilization. (Details..routing
> table slots, etc...but they are going to get used one way or
> FWIW, voluntary return of address space to ICANN (IANA) has always
> been an option regardless of there being a policy in existence to
> specifically state it. It's not like ICANN would say "no", I think we
> just can't be sure what they'd do with it if we didn't have it
> codified. Modifying the existing ratified policy with a sentence like
> "All returned IPv4 space regardless of it's class will be remanded to
> the free pool and allocated on a needs basis". That change would
> likely fly through the RIR PDP's and accomplish exactly what we are
> "trying" to accomplish here without all of the politics and
> appearances of land-grabs.
"land-grabs" the very word assumes that IPv4 property will have
indefinite future value.
Anyone taking IPv4 right now is a speculator, pure and simple. When
an ISP gets IPv4 they are doing so because they are gambling that
they will need it for the future. Years ago that was a very safe bet,
today with IPv4 running out, it is a question of how many more years
are you going to need it.
If I could tell you that in exactly 2 years, IPv4 will be regarded as
worthless because everyone will be on IPv6, would you spend the time now
to apply for IPv4? What if I told you that in 6 months it will be
worthless? We all agree it's going to be worthless someday, the
question is when.
The situation is akin to the community that has a old bridge in the
middle of it that's falling apart. The community knows that sooner or
later the bridge will fall, but they spend lots of time debating how
much more time they have left before the bridge collapses. Then one
day the bridge unexpectedly does collapse, and everyone then says
"gee, I was sure it was going to collapse but I didn't think it would
More information about the ARIN-PPML