In reply to a post by RobertoS:

In reply to a post by billford:

(I agree with Ignitionnet- the current system is the least worst way. Run the systems in parallel, when IPv4 addresses run out, site operators have a choice- implement IPv6 or see their hit rate inexorably decline.

But your conclusion is false.

As the plan seems to be that all end users will have an IPv4 address, be it dynamic, static or CG NAT, site operators have no incentive at all to move to IPv6. Only new sites that cannot get an IPv4 addresses will take IPv6, which the end users will be able to access because they are dual-stacked. It's going to be a long time before there are anything like as many IPv6 sites with no IPv4 access as there are IPv4 ones with no IPv6 access.

I've already said I'm arguing from little knowledge of the technology, but logic, human nature, and life experience point towards a right mess with very little chance of resolution for decades.

All of you that argue against that belief seem to be starting from a position of "received wisdom" that the way BT et al seem to be going is the best, without any discussion of alternatives such as the one I suggest. I'm sure there must be others as well.

I already have 100s of sites running dual stack, my own personal sites will be dual stack this year.

Some datacentres are now deliberatly pushing ipv6, sites moving to ipv6 is happening, its just not an overnight change. The UK is slow to adopt tho, doing a bit of research shows that we are behind on ipv6.

Also as ignition said, dual stack is far less complex and messy than cgnat. Modern operating systems support dual stacking natively, modern net applications do. Many routers do as well although you think they dont as the feature gets hidden. Fritzbox doesnt have it hidden, and that has dual stacking capabilities.

All I am seeing is uk isp's making excuse after excuse as to why they havent done any ipv6 yet. There was even a uk body setup which then dismantled itself as they got frustrated with lack of support from the gov.

I will be joining aaisp soon after BT release my MAC so will be dual stack at home also soon. Which will then allow me to be more effective at deploying ipv6, as currently its hard for me to test ipv6 deployments due to lack of native ipv6 at home (currently have to use a tunnel).

The subject says "Daft question" . I still don't get some of this.

Accepting that dual stack does not require CG NAT, which I don't dispute, and given that CG NAT is being tested by BT and has been by Plusnet, it looks to me as if CG NAT is being brought into play because better options have been ignored for years.

Dual stack does not require CG NAT, but the need for an IPv4 address is why CG NAT is coming, to support dual stack. There aren't enough IPv4 addresses available to dual stack without CG NAT.

Is CG NAT with dual-stacking the international norm? You say we are well behind. Or did other countries adopt a different approach?

In reply to a post by RobertoS:

... but the need for an IPv4 address is why CG NAT is coming ...

Afaics, the main reason there's a need for an IPv4 address is because the major players have been so slow in implementing IPv6.

It's more or less a panic measure to cover up for lack of previous effort.

Exactly Bill .

That doesn't make it the best conversion strategy - but no-one has replied yet to how the conversion was handled in the other countries I'm told are ahead of us.

I wonder if the American region may have a bigger problem, as their RIR is not running out quite as fast.

I read discussions from some of the USA WISPs saying they don't need IPv6 yet (or even at all) and neither do their customers. You see some worrying levels of ignorance, such as
- that it is some government conspiracy,
- that a little redistribution and renumbering could extend availability of IPv4 indefinitely,
- "I'm not giving up my IPv4s" as if they have switch off IPv4 and return their allocations in order to switch on IPv6,
- that it's only happening in Europe at the moment, so doesn't affect the internet for USA users
- comments along the lines of "2^128 - it's only 4 times bigger" or "how much longer will the expanded range like 123.456.789.123.456.789 last anyway" I really hope were deliberately trolling but I can't be sure!

Maybe they will be lucky and all networking equipment including home routers will already be dual stack (both in CLI and GUI) by the time they realise they have to move, due to the manufacturers having to release models with updated firmware for the other regions.

As regards the UK, maybe central government should issue a policy that all government departments and agencies must only buy router hardware and ISP / network services that are dual stack capable.
I believe it is currently recommended within procurement guidance but not mandated, which unfortunately in the current climate strict published policies can be spun either way as bad - "red tape to be abolished" or good - "transparency and accountability".

More likely, ISPs who offer no IPv6 will eventually lose customers or be forced to roll it out quickly or without time for thorough testing. But this will take time as it also depends if enough customers ever perceive they are no longer being provided full internet access.

Edited by prlzx (Thu 09-May-13 23:28:38)

Thanks for that.

It may partly explain the UK ISPs' tardiness as well, given that most sites I expect are US based and will therefore be on IPv4. Maybe they also generally use dynamic IP addresses for SOHO?

Back to the UK, I remember a few weeks ago it was reported as having been suggested in government/security circles that there should be legislation to make fixed IP addresses mandatory for all internet-connected devices.

No mention of IPv6 LOL. Shock horror when they discover that's the only way it's feasible, and all the ISPs say it can't be done within the next 20 years .

In reply to a post by RobertoS:

Back to the UK, I remember a few weeks ago it was reported as having been suggested in government/security circles that there should be legislation to make fixed IP addresses mandatory for all internet-connected devices.

No mention of IPv6 LOL...

Oh yes the announcements are full of oversimplifications and misunderstandings in the interests of the soundbites.

I don't think official spokespersons grasp what they mean when talking about logging headers or envelopes. They seem to think that all UK users must send email using addresses issued by their internet provider, and/or only from a computer at a fixed location (home network), and/or that the envelope is always in plain view (like with postal envelopes).

It's like no-one sat down with them with a smartphone and walked them through:

"What do you think this does when it sends and receives emails?
What internet connection(s) does this take place on?
Who (which service providers) sees the envelopes?
Probably not the UK internet ISP if you use GMail, Yahoo, MS (Outlook.com) or Apple (me.com)
Or one based overseas. Or an independently run mail server.
Did you know many email providers allow / encourage you to use SSL/TLS by default?
How do you know which internet connection providers to approach if you want a complete set of logs for a person for a time period (unless they are under house arrest)"

Edited by prlzx (Fri 10-May-13 00:10:59)

yeah cgnat is been used a ipv4 preserving measure, it actually bears no relation to ipv6.

the reason I mention it in ipv6 conversations is because if ipv6 was adopted years ago then the ipv4 situation wouldnt have got so bad. So dual stack will not in the short term relieve ipv4 suply issues as ipv4's will still be assigned to customers, but it does make a ipv6 internet closer.

My view is any isp with ipv4 supply issues should first reduce ip's to customers with more than 1 ipv4 before considering cgnat customers with a single ipv4. I believe this is the policy aaisp said they will do. It is also the policy I am seeing some datacentres do eg. hetzner in germany supply now just 1 ipv4 per server by default.

dual stacking ipv6 with ipv4 is basically having both turned on at once, in such a scenario if you do things like web browsing everything should "just work". if it doesnt then its possible to adjust things like temporary turn of ipv6 or fiddle with priority order. But things are already fairly advanced to the point it shouldnt cause huge issues having a dual stack setup.

double natting however can cause issues, there is the obvious issues with port forwarding, dont forget its not just people running servers, but all those millions who play on games consoles on the internet and need open ports, it can also cause issues doing basic tasks as web browsing can get stuck when traversing 2 nat's. A natted ip may have all its ports tied up because of the shared user's with too many active connections. A shared ip can also have extra issues if a user on it gets banned somewhere and another user on the same ip then cannot access that service, and of course some services check ip for eligibility of use and may flip out if they see multiple users on that one ip at the same time. There is more as well this is just some examples, double natting will need more processing resources as well on the isp side, I cant see how its less complex than supplying ipv4 alongside ipv6.

an example is germany.

one reason the fritzbox has ipv6 support is isp's there are utilising dual stack 'without' cgnat.

a reason why the uk body for ipv6 dismantled itself is the uk gov wouldnt even put its own sites on ipv6. That is one area we are behind other countries.