[arin-ppml] Policy Proposal 122: Reserved Pool for FuturePolicy Development

[George Bonser]

> 
> >> The last /10 isn't about making a business with no IPv4 able to
> > compete
> >> against someone who still has half a /8 left.  That is frankly,
> >> laughable.  What it can be about though is help for those who
> planned
> >> well and ran into unexpected consequences.
> >
> It was not my intention to say that it was. However, there is a
> difference
> between the way this happens as a natural result of runout and having
> ARIN actually participate in creating or exacerbating the situation.

I believe what Leo was getting at was that we could set the space aside
yet not designate yet exactly what it is to be used for.  That allows us
some leeway once we see how the chips actually fall.  If we attempt to
designate a specific purpose for those addresses now and it turns out
the addresses need to be used for something quite unforeseen, then we
are stuck with more addresses reserved for something then we need and
not enough for something else.  We don't want to paint ourselves into a
box and being vague now gives options later.  Once some time passes and
we see where the critical needs actually are, we can then assign those
resources where they can be most efficiently utilized.  Attempting to
look into a crystal ball now and tell where the pain is going to be a
year from now is likely impossible.  

In fact, there are going to be so many networks that have waited until
the last possible moment that connectivity problems are going to be
difficult to sort out because they won't know if the problem is in their
network or at the other end because both ends are attempting to migrate
at the same time.  THEY aren't even going to really know what they need
and will be in a mode of just taking everything they can get because
they don't know if they are going to need it or not.  If we wait a
little while, networks will have a better idea of exactly where there
needs are.

If we open up "transitional space" then everyone on the planet is going
to want some "just in case".  If we don't open it up, they will allocate
some of their own space, get things going, and then if they need more,
they will have a better idea of exactly what they need.  I think Leo was
saying not to put too fine a point on things yet.


> >> We need to get out of the mindset that if someone has totally
> ignored
> >> IPv6 transition we have something good waiting for them in the
> wings.
> >> We don't.
> >
> > Absolutely agree.
> 
> I couldn't agree more. If one assumes I have such a mindset, one has
> not been paying attention for some time.

Oh, I don't think anyone was accusing you of having such a mindset.  It
was just that I believe Leo was saying that laying out too much now
might give the impression to some operators out there "we don't have to
worry about that right now, we can just get some 'transitional space'
when the time comes." And it enables additional procrastination by some
operators who may read the thing the wrong way.  Reserve the space but
be vague on exactly how much is to be used for what.  In fact, I would
merge CI and IPv6 transition into one larger block that either use could
be drawn from.  We don't know how much will be needed for what function
but man, I hope there isn't a lot of CI currently being rolled out on
v4.

And I think I miscommunicated the tunneling thing.  I wasn't talking
about 6to4 per the way people have been trying to standardize it.  I was
thinking more like:  "Here is your connection to us.  It is v6.  You can
build a GRE tunnel on your router to send v4 traffic to our v4 router
because our edge units are now v6 only, not dual-stacked.  Oh, and set
your MTU on that link to 9000 so you don't introduce PMTUD issues on
that tunnel".