IPv6 traffic still very low - carriers hanging back according to APNIC researcher -
http://www.networkworld.com/news/2012/0 ... 2012-06-13

Aubrey

_________________
The illiterate of the 21st century will not be those who cannot read and write, but those who cannot learn, unlearn, and relearn. Alvin Toffler, "Future Shock" 1970

Interesting read. Although I do think that more will have to be done to ease the transition pain before IPV6 becomes a reality.
We are still a way off implementing it in the corporate infrastructure and until that happens engineers will be reluctant to go down the IPV6 road.
However I do think that if not more is done to ease this transition a lot of companies will be caught with their trousers down.
And IMO it is another peice of over engineering, I mean do we really need 128 bits, did we have to make the address space that long and that difficult to remember and implement. It seems to me that if more thought could have went in to the design of IPV6 we could have had a faster and simpler transition.
However that said we as network engineers will have to be the ones on the frontline getting this thing to work smoothly.

One point in the interview did strike me as particularly relevant, if private customers are going to be expected to recite IPV6 addresses over the phone then that will be a disaster and also vice versa when a tech rings someone with an IPV6 address............... big trouble.

_________________
Good Luck,

David

re: difficult IPv6 addressing

I have been to a few lectures/training for IPv6, and they keep reinforcing that idea that DNS is to be the solution to overcome memorization of IPv6 addressing. If DNS wasn't already critical enough, it is now even moreso.

I was going to say the same thing.

I've been reading about how it's all going to be auto this and auto that but as we very well know that will never be the case.

It's Y2K all over again, panic panic

_________________
Good Luck,

David

Remember what Jeremy says, if it's auto, you auto not do it...

_________________
"If you're good at anticipating the human mind. It leaves nothing to chance."
-Jigsaw

Actually I am pretty sure a lot of thoughts were put into the design IPv6, as IPv4 as actually meant to be an "experimental" protocol at the time and was never meant or envisioned to be used worldwide (the RFC came out back in 1981, a very different times from now: there is no Internet back then, memory was much, much smaller and was very expensive back then ... BGP hasn't even be invented for almost a decade).

It is also definitely not a piece of over-engineering, as due to the fact that IPv6 is meant to incorporate end-to-end connectivity between hosts (going back to the internet before NATing). Making the address space smaller like 64bit wouldn't make much sense either as we will just end up facing another shortage 30 years down the line.

The one problem in making the IPv6 transition is that there still isn't a "killer app feature" that would make a business jump at going to IPv6, and until that happens I think we will just end up with Service Provider using IPv6 backbone and most Enterprises staying with IPv4, at least for the next few years.

Lastly, I do see some progress made in the IPv6 transition in technology like LISP, which makes IPv6 transition easier.

_________________
"It can also be argued that DNA is nothing more than a program designed to preserve itself. Life has become more complex in the overwhelming sea of information. And life, when organized into species, relies upon genes to be its memory system."

So what you are saying is that 18446744073709551616 addresses wouldn't have been enough to get us by for a few years until it all changes again anyway. It's because the decision was made to make the address space so long is the decision that's holding up the transition.

If more, and by more I mean smarter, decisions were made in the first place, IPV6 would be up and running.

If they would have taken better care with the management of IPV4 addresses in the first place we could have happily be sailing away for a good few years yet and NAT being NAT and router/switch hardware getting better by the day most companies will simply not bother again, more thought.

_________________
Good Luck,

David

I skim read and stopped at this.



This is a technology that the home user just isn't going to start using because it's there. There's no additional features or benefits for mum and dad yet so what are people expecting? It will ideally be a seamless/invisible transition. The biggest growth will be from ISP provided routers being remotely updated to support IPv6.

How do you market a product that doesn't make money, is only meant to replace something that is currently working but will need to be replaced to support growth in the near future? Not to mention what else is going to substitute IPv6 as the IPv4 replacement?

I'm not the biggest fan of network worlds articles and I stopped reading it a while ago. This one is just another nail in the coffin for me.

_________________
- Pete

It's a perfect opportunity for engineers to step up and show their worth. Why should the designers or service providers make it cushy for everyone? Are you an engineer or not? You need to make the case for at least dual-stacking to your superiors. You need to figure out how it works. You need to configure it and support it.

I've already gone down this road with my own management. Like end users they don't give a damn about it as long as everything works with minimal capex now.

So who needs to do more? We as engineers do. The protocol has been in existence for many years now

_________________
www.mellowd.co.uk/ccie/

I certainly agree that as engineers we do need to step up to the mark and we do need to be the ones who understand this technology. My one and only point is that as is usual with internet engineering this solution was invented not to solve a problem but to make sure no one could ever say that we are running out of addresses, again.

A knee jerk reaction to a problem that could have been dealt with in a smarter more intelligent way, simples.

_________________
Good Luck,

David

It was definitely invented to solve a problem, namely IPv4 exhaustion, have your organization tried to get a new public IP block recently? If you have, you'd know it is a real problem. Making the address space smaller than 128 bit won't magically make the transition easier, since it still doesn't give businesses a specific reason to switch.

_________________
"It can also be argued that DNA is nothing more than a program designed to preserve itself. Life has become more complex in the overwhelming sea of information. And life, when organized into species, relies upon genes to be its memory system."

IPV4 exhaustion was accelerated by the ineptitude of the powers that be in controlling the issuance of addresses, as I said if the original problem was tackled with the same gusto as was put in to IPV6 then the problem would not be as serious as it is now.

Why have so many US institutes been issued with full class B address spaces?
Why has a US university been issued with a class A address space?
Does this sound like a good plan, again a smarter more inteligent move would have been to police IPV4 more tightly.

128 bit certainly won't make the transiton harder or easier but, 1 bit, just 1 bit would have doubled or address space and alleviated the problem for quite some time.

_________________
Good Luck,

David

1 bit would still be a work-around. No use keep doing work arounds and work arounds. Let's just get it done. Like I said before, IPv6 has been around for AGES and hence if you're only dipping your toes in now who's to blame?

_________________
www.mellowd.co.uk/ccie/

I'm not really trying to blame anyone, just making a point and yes we all have to just get in to it and pronto. It just annoys me that with all the smart people that are involved in networking and IT in general, that this is the best we can come up with.

_________________
Good Luck,

David

I see nothing wrong with it. It's an expanded address space. It's just an address. We still have IGPs, EGPs, Multicast etc. Don't be scared, embrace it I'm so used to seeing IPv6 addresses now it's become 'meh'

_________________
www.mellowd.co.uk/ccie/

Yeah I suppose although scared would be a little bit strong, I was thinking more along the lines of FFS.
But yeah I suppose you're right get stuck in and the transition will be a distant memory soemtime.

_________________
Good Luck,

David

Working with IPv6 in a production environment I can only say the same. I really don't get all the fuss. IPv6 works. And yes it's incompatible with IPv4, but so would a 33-bit address field be (and given about 10 billion devices connect to the internet, 2^33=8 billion would have given us nothing). And even if you would find a compatible protocol, for example using extention headers, routing headers, whatever, you'd still be trying to fix a car with duct tape.
I'd rather have the redesigned, new car with the new features then.

_________________
http://reggle.wordpress.com

Not sure if I'm a fan of /64s though. I'd prefer a VLSM approach.

@Darren have you guys started giving customers V6 yet? Just wondering if you are using 6rd with /64s or native dual stack with /56.

_________________
- Pete

Not yet. But customers will get /48 - Native dual stack

_________________
www.mellowd.co.uk/ccie/