[menog] IPv4 March 2011 depletion

[Owen DeLong]

On Nov 15, 2010, at 4:36 AM, Brian Candler wrote:

> On Mon, Nov 15, 2010 at 03:30:00AM -0800, Owen DeLong wrote:
>> If you are talking about doing this on a per customer basis, it
>> is absurd.
> ...
>> If you have NAT44+IPv6, you don't need NAT64.
> 
> Yes you do, if there are any IPv6-only devices in the customer's network.
> 
But why would you ever do that?

> Presumably you *do* want to see people starting to deploy IPv6-only devices
> in their network, rather than running dual-stack forever?  Because, for as
> long as you admit that some sort of IPv4 is necessary in the customer's LAN,
> you give people a strong disincentive against deploying V6 in the first
> place.
> 
Perhaps eventually, but, I think that will take the later (possibly much later)
form of deprecating IPv4 on a segment-by-segment basis. As long as you
have NAT44 services on a segment, why would you put an IPv6-only
host on the same segment? There's nothing to be gained by such an action
other than pain and confusion.

>> If you have IPv6 with NAT64, you don't need RFC-1918 or NAT44.
> 
> Yes you do, if there are any IPV4-only devices in the customer's network.
> 
> We don't have the luxury of redesigning every customer's network from
> scratch, nor of telling them what devices they may add or not add to their
> network.
> 
Perhaps not, but, at the point where the customer is on IPv6 with NAT64, it
is unlikely that there is an external IPv4 address available for the NAT44
services you propose. As such, I question whether this will be reality.

>>> 2. new V6-only users, which is where you want the growth to be
>> Eventually, growth will be forced in this direction.
> 
> I am old enough to remember when people connected to two networks:
> 
> (1) X25 and The Internet
> or
> (2) CLNP and The Internet
> 
So am I.

> Today we have:
> (3) IPV6 and The Internet
> 
No... Today, we have the IPv4 internet and the dual-stack internet. There are
significant differences between (1)/(2) and what we have today.

> You don't need me to tell you what happened in cases (1) and (2), despite
> strong government and management pressures to "force" people in a particular
> direction.
> 
There's a bit of a difference between government/management pressure and
numeric reality. The numeric reality of what you call the internet (I call the
legacy IPv4 internet) is that it's running out of numbers and none of the
workarounds proposed as alternatives to IPv6 deployment will change that
for any appreciable period of time.

> Compared to what people have and are using today, V6 gives exactly *zero*
> benefit to the vast majority of users.  Cost without benefit = no business
> case.
> 
Sure... Until we run out of numbers... 

> AFAICS, the types of "forcing" required would be:
> a. ISPs cease to offer V4 services at all (not even NAT44)

Which will happen when ISPs don't have any more external IPv4 numbers to
use for NAT44 services.

> b. ISPs increase their charges to V4 users, or offer discounts to V6-only
>   users

Which will happen after ISPs deploy IPv6 because at that point, continuing
to maintain IPv4 will be all cost and decreasing benefit.
> 
> Maybe there is some mileage in (b). However ISPs work on such small margins
> and in such a competitive market that it may be very hard in practice.
> 
Yes... (a) will be the major driver as a result of address shortage and the rising
costs of addresses (while they are still available at any price).

>> Equipment which does not know about IPv6
>> should safely ignore it like any other unknown protocol.
> 
> I don't think "should" and hope is good enough. Any across-the-board change
> to the service you provide has potential impact, and there are hundreds of
> different types of no-name routers with buggy firmware, partial or untested
> V6 implementations.  If and when those calls hit the callcentre, it would be
> wise to be able to offer a quick solution to the affected customers, rather
> than "sorry, your equipment doesn't work with our service, please buy a new
> one or go somewhere else"
> 
Again, can you cite any specific example of equipment which breaks if IPv6
is turned on, or, is this pure speculation on your part?

> This will of course vary between markets. It won't be as much of a problem
> where the ISP owns and manages the CPE - although the user may still be
> affected if you turn on V6 on the LAN side, it's relatively easy to turn it
> back off again.
> 

It's only easy to turn it off if you have a way to provide IPv4 services. That is
a temporary situation at this point.

Owen