[ppml] Motivating migration to IPv6

[Ted Mittelstaedt]

>-----Original Message-----
>From: ppml-bounces at arin.net [mailto:ppml-bounces at arin.net]On Behalf Of
>Robert Bonomi
>Sent: Tuesday, July 31, 2007 10:49 AM
>To: ppml at arin.net
>Subject: [ppml] Motivating migration to IPv6
>
>
>
>I'm sure the following idea has to have occured to better minds than mine,
>but I _cannot_ see what the downside to it is --
>
>Given that:
>  1) it is policy to 'encourage' migration to IPv6
>  2) there is a looming shortage of IPv4 addresses available for assignment
>  3) _At_present_ IPv4 address-space *is* viewed by requestors as
>'preferable'
>     to IPv6 space.
>  4) more than 95% of address-space assignments are to entities
>for which there
>     is a reasonable expectation they will be making _additional_ address-
>     space requests in the 'not too distant' future.
>
>Proposed:
>  A) every IPv4 block assignment includes the assignment of an 'equivalent-
>     size'  IPv6 address block ( e.g. assuming '1 IPv4 /32' == '1 IPv6 /64)

Robert, I suggested this about 4 months ago in a discussion and was shouted
down.  You can read the list archives to see all of the objections.  In
summary,
it boiled down to if you give an entity IPv6 space automatically and they
don't use it, you are creating a hazard for a spammer to go hijack the space
and use it - then joe-user on the Internet cannot look at an incoming attack
or spam originating from IPv6 and know if it's legitimate.

One of the big problems with IPv4 right now is all of these legacy holders
who aren't paying a dime every year but are also not advertising the IPv4
space allocated to them.  Since they aren't paying, ARIN has no way of
revoking
the assignment for non-payment, and the legacy holder has no incentive to
stop paying for IPv4 addressing they are assigned since they aren't paying
anything anyway in the first place.  Since there is no advertisement we
don't
know if the space is abandonded or not.  The feeling on the list is by
giving
people resources they aren't asking for you are just creating the same
problem
all over again.

>  B) _subsequent_ v4 requests must show the required utilization levels of
>     *both* the allocated IPv4 *and* IPv6 space.  With
>"utilization" of IPv6
>     space requiring the actual deployment of functional machines in that
>     address-space.

What is the goal here?  To restrict IPv4 handouts?  For what reason?

Look at it this way.  An organization that is going full-bore into IPv6
is likely to be LESS interested in obtaining additional IPv4.  So
restricting
IPv4 handouts to people who are going full bore is a bit like telling poor
people that the only ones who can get welfare cards are the people who are
already not poor.

Your really a lot better off giving away the IPv4 as quick as you can to
organizations that have a real need for it.  The reason is that they will
use
it and then post-IPv4 runout you won't have a lot of unused IPv4 sitting in
the hands of speculators who have no use for it other than selling or
renting it to other people.

>  C) As the pool of available IPv4 addresses gets smaller, the
>ratio of  the
>     relative size of the IPv6 allocation vs the IPv4 allocation
>_increases_.
>
>For 'revenue' purposes, the 'paired' IPv4 and IPv6 allocations are counted
>as single block, as long as both are allocated.  IF the requestor _returns_
>the IPv4 block, they get a significant discount on the IPv6 space for some
>period of time. (50% off for 5 years, maybe?)
>

Once more, what is the goal here?  To get a lot of IPv4 returned?  Which
will
then create clamoring for the returned IPv4 to be handed out again.

What your doing is creating a scenario where shortly after IPv4 runout, orgs
that choose to go full bore into IPv6 will end up dumping all their IPv4
and orgs that choose to NOT get in to IPv6 will just get their IPv4 needs
fulfilled
from all the returned IPv4.  Your creating an inducement to IPv4 orgs to not
update to IPv6.

Obviously you will never get 100% returned IPv4 from IPv6 orgs since they
will
need some IPv4 to dual-stack all the devices they have that offer services
to
the Internet - (ie: webservers and the like) because they will not want to
cut off customers that are still on IPv4-only networks. Orgs that are mainly
IPv4
will do the same thing - for servers of theirs that offer services to the
Internet, they will dual-stack those with numbers out of the minimum IPv6
assignments, then never bother updating the rest of their internal network.

The only way to force IPv6 updating is to make IPv4 unavailable by having
it tied up in allocations.  Because then what will happen is orgs that
need more numbering will have no choice but to use IPv6 internally and
the IPv4 that they currently have will be squeezed into dual-stacked
servers that will fulfill web serving and other service serving and be
used as gateway servers for their internal IPv6 clients that need to get
to IPv4-only resources on the Internet.

>
>An _absolutely_ effective way of driving migration to IPv6 would be to
>condition additional IPv4 address-space allocations on the percentage
>of IPv6 traffic that transits the boundaries of the requestor's network.
>That requires that not only does the requestor deploy IPv6 internally,
>but that they _use_ it with external parties as well.  Nobody can argue
>the efectiveness of such an approach; however I suspect there are a number
>of significant obstacles to actual implementation.
>
>
>As I said at the top of things, I'm sure things like this have already
>occured to far brighter people than me -- I await, with some trepidation,
>being shown  'the **** obvious facts' that I have overlooked, that kill
>such an approach. :)
>

Let me speak as an IPv4-only org for a moment.  What is going to motivate us
to upgrade to IPv6?  I know how to do it and I have plans to do it, I even
have a notion of the cost to do it.  I will tell you.  The day that we see
THE POTENTIAL for SIGNIFICANT customer loss to our competitors by NOT
updating
to IPv6 is when we will update.

Who are our biggest competitors?  Verizon, Qwest, Comcast.  None of whom are
telling customers that the customers HAVE TO RUN IPv6.  Qwest is still
delivering
ActionTec DSL modems that CANNOT run IPv6.  And I would reckon only about
10%
of our customers are running Windows Vista, the remaining 90% are majority
XP but
probably 40% are MacOS Panther, Win2K and so on where IPv6 isn't available.

As for enticing us to update to IPv6 by making new requests for IPv4
unfulfilled,
well the problem there is I've got a billion tricks up my sleeve to extract
IPv4
from our network.  Such as putting in IP-unnumbered instead of /30's on
links -
makes it harder to troubleshoot, but hey.  Such as replacing /29's at
customers
with /30's if they aren't using the additional numbers.  Such as giving a
small
discount to our DSL customers willing to to to RFC1918 numbers instead of
public numbers.  I could also put or modem banks behind translators - it
might
even reduce the number of customers who get broken into.

Until my big competitors start telling customers they HAVE to update to
Vista
to get DSL or pay more money, or they HAVE to pay more money if they want
their
websites to be dual-homed, there is not a snowballs chance in HELL that I
could
pass along ANY update costs to our customers, and even less that I could
require
them to go IPv6.  And furthermore, the day that my big competitors run out
of IPv4
and start telling customers that they have to be IPv6-compliant if they want
a public IP address, I'd be a fool if I didn't try extending IPv4 even
futher.
Hell, it would be a golden opportunity since some of those Verizon, Comcast
and
Qwest customers wouldn't want to switch to Vista or update in any way, and I
would
like to obtain those customers.  God knows the biggies have done their share
of stealing customers from us over the years.

And if you think the biggies don't know all this your nuts.  They do.

IPv6 is basically a game of chicken between me and my competitors.  Each of
us is
involved in a race to be the LAST isp to start forcing customers to update
to
IPv6.  The ONLY way this would EVER change is if a must-have application
came
along that was IPv6 ONLY.

People have joked here that if you put free illegal downloads of movies,
music,
and free porno, on an IPv6-only network, that the question of IPv6 updating
would be a moot issue.  I am sorry to have to say but this is really a lot
closer to the truth, and a lot closer to what is really going to have to
happen.

Think up a killer-app to attract end-users to IPv6 and you will solve the
problem.  I would definitely prefer it than the alternative which is to keep
playing the chicken game with my competitors.

Ted