0

Given that there is some ridiculous number of IPv6 addresses per every single person on the planet, why do ISPs still charge for IPv4 addresses? Why not just give the user an IPv6 address?

Is it purely on the basis that a layman is not going understand this IP?

I read in my CCNA that IPv6 adoption in the mainstream is slow and usually limited to mobile devices. In my own personal opinion, progress seems throttled because companies still want to charge for IPs and don't want to address that there is no IP starvation in IPv6 (yet).

8

IPv4 and IPv6 are two different protocols. They are not interchangeable without a lot of work.

So if you want to talk to v4 devices you run v4. If you want to talk to v6 devices you run v6. Most of the time, you run both.

4

IPv4 and IPv6 are separate protocols. They can't inter work without some form of translation mechanism.

The original idea of the IPv6 proponents was that we would all move to dual stack running IPv4 and IPv6 in parallel. Then once everything was dual stack, IPv4 could be phased out.

The problem is that plan just isn't attractive from an economic perspective. The costs of going dual stack come immediately, but most of the benefits of IPv6 don't come until some unknown time in the future when IPv4 can be turned off.

The result is that IPv6 remained mostly a toy for techies for over a decade.

In the past few years as the availability of IPv4 worsens we have started to see a bunch more movement towards IPv6, usually in combination with some form of transition mechanism to allow access to resources on the IPv4 internet. Some large ISPs have gone down this route.

Right now though, there are still many v4 only client networks. So if you want to host services that are accessible from networks outside your control (either because they are accessible to the general public or because you have users "on the road" who need to access them) you need public v4 addresses.

It will be interesting to see how far IPv4 prices have to climb before IPv6 becomes ubiquitous.

3

An IPv6 address is useless when there's an IPv4-only partner that needs to connect - they are separate protocols and there's no connectivity between them.

Still, IPv4 is the only protocol allowing a server to offer services to everyone - even users with IPv6 primarily still have some IPv4 connectivity (e.g. DS-Lite), but not vice versa. That is only going to change once everyone has IPv6 connectivity but there are still a lot of white spots on the map.

IPv6 adoption in the mainstream is slow and usually limited to mobile devices is pretty outdated though. Mobile devices were early in IPv6 adoption since the demand for 'some kind of connectivity' was extremely high and the networks were newly built anyway. In the last couple of years, IPv6 has significantly gained ground, so we can likely expect a beginning, slow death of IPv4 in ten years or so.

1

progress seems throttled because companies still want to charge for IPs and don't want to address that there is no IP starvation in IPv6 (yet).

TL/DR: ISPs have a valid reason to charge from IP addressing. The reason behind slow adaptation is money, but profiting of the IP address charges is a tiny minuscule factor in the equation.


As long as there's no direct access to the internet from each and every building, we will need a middleman to arrange for it. Consider that every building in developed world has access to the national power grid, but even that's managed by a middleman. The middleman needs to maintain the equipment etc., pay for employees and create profit for owners. It's only natural that the middleman requests a compensation for services rendered.

Although IPv6 address space is ginormous, it still has to be regulated if only to prevent addressing conflicts. We have already an infrastructure and processes in place that allow for regulation. Again middlemen who need funding to maintain the infrastructure, pay their personnel etc. - compensations of services rendered.

Adopting IPv6 is easy for a home user or a middle-sized company. Not so for large multinationals, governments, military organizations, financial operations, healthcare... or even ISPs. It's not a one-night or even one-year operation. It requires careful planning and step-by-step implementation with monitoring periods and very fast fallback capabilities.

Network equipment doesn't get thrown out and replaced just because there's a new model in the market, it's used as long as it works and serves the purpose. Outdated equipment that doesn't support IPv6 must be replaced. Mid-nineties 80% of the commercial infrastructure was running on custom applications written on Cobol - an insurance company database server doesn't get updated either as long as it serves its purpose, so a great many lines of code needs to be written or systems replaced, which requires again testing and monitoring periods. Personnel needs to be trained, maybe new personnel hired. And so on.

The public sector funded by tax money has on top of all that its very own, separate stumbling blocks on the way with politics entering the picture.

This just from the top of my head for some perspective. We're talking about huge financial investments, and that is, as usual, the real bottleneck.

2
  • "As long as there's no direct access to the internet from each and every building, we will need a middleman to arrange for it." I'm not really sure what that means. The Internet is just companies connecting to each other. There is no Internet as a separate network, so every company on the Internet is a middleman.
    – Ron Maupin
    Oct 1 at 20:06
  • Yes, I'm trying to figure out how to rephrase it better :-) Oct 1 at 20:08
0

The full cutover to IPv6 is still far out. Organizations entire infrastructure will need to change to address this. Most of us aren't running IPv6 at all - period - let alone as the main protocol. This will be big $$$.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.