Why do we still need IPv4 when we are migrating to IPv6...

Anfinsen, Ragnar Ragnar.Anfinsen at altibox.no
Sat Feb 14 16:05:41 CET 2015

On 13.02.15, 18.33, "Lorenzo Colitti" <lorenzo at google.com<mailto:lorenzo at google.com>> wrote:

On Fri, Feb 13, 2015 at 5:43 AM, Anfinsen, Ragnar <Ragnar.Anfinsen at altibox.no<mailto:Ragnar.Anfinsen at altibox.no>> wrote:
Reducing the price of the service is not an option for the sales people,
unless there are other benefits, and right now there are none. Spending
for example $650K on IP addresses is far cheaper than reducing the price
by 20% in addition to investing in the technology to enable MAP, lw4o6 or
CGN. So unfortunately, we can put the ideology aside and concentrate on
deploying IPv6 while keeping IPv4 as good as possible. When we finally
meet the magic threshold, we can start discussing which technology is best
for keeping the legacy IPv4 available.

But: at some point, you'll *have* to invest in the technology to enable IPv4-over-IPv6. Not deploying it today means you save X kroner in interest payments on the debt you incur to pay for it today (and possibly save Y kroner if that technology is cheaper in the future).

On the other hand, IPv4 is a sunk cost that is never coming back - because when you are in a position to sell those IPv4 addresses that you bought, it means that you don't need IPv4 any more, and probably other networks don't either. So the price is going to be lower, and you'll lose a lot of the money that you spend on IPv4.

So: compare the interest payments with the write-off with IPv4. What's the outcome?

A few things, 1) interest payments presupposes that one loans money to buy addresses, 2) as long as 40% of all traffic is still IPv4 for DS enabled customer, we need a fairly sizable CGN/AFTR setup.

From our perspective, doing investments on CGN/AFTR technology now can almost be comparable with buying address, as we must consider deprecation on the equipment anyways. If we can wait a bit longer and the IPv4 traffic lowers to for example 10% and then do the CGN /AFTER investment, it would possibly be cheaper and possibly be done with equipment we already have. I guess seen from a pure economics perspective it does not make much difference, but at least we can uphold the native IPv4 until the majority of ISP's and content providers are fully Dual-Stacked.

So it is not just money that drives our service, quality and availability is also important factors.


More information about the ipv6-ops mailing list