In reply to a post by Ignitionnet:

In reply to a post by AndrewW:

Trust me there is a lot of forward planning going on. Where i work now ( very large global company) all IT equipment has to support IPV6, but its all turned off. We are waiting for that trigger, and its a massive change as well. not just a home router upgrade, These roll out project costs many millions. Until there is a business case there will be no migration to IPV6 by the vast majority of companies.

Also a quick read of the article you posted shows just what i have said, its not happening....... be prepared but don't enable is a general concept being followed by a large number of companies

This. Most kit in the large enterprise networks I work within supports IPv6 but there's no drive to pull the trigger. IPv4 and RFC1918 are working okay for now. It'll be something done because it has to be as there's so much potential for disturbance.

Is ipv6 that hard to add to networking configurations?

I recognise staff training issues. I recognise old equipment may not suppport ipv6. However adding ipv6 routing to an existing network is one of the easier things I have done, and for safety ipv4 can be set as preference.

The problem of waiting for "when it has to be done" means it will never get done. its one of those thngs where everyone is saying "you first".

For this reason its not impossible ipv6 gets abandoned as already mentioned, it may be the case companies instead just use things like cg-nat instead as its more conveniant, although I think cg-nat is way more complex than adding ipv6 dual stacking.

I think also is a big difference between a company that uses connectvity for its business and one that sells connectivity for its business rolling out ipv6. You guys may consider configuring but not activting ipv6 as the norm for isp's yet I am seeing differently, I use many datacentres and most provide ipv6 connectivity.

Edited by Chrysalis (Mon 09-Dec-13 01:51:23)

In reply to a post by Pipexer:

In reply to a post by jchamier:

Yes, I see this as well. The other issue is training staff who are very familiar with IPv4 in the much more complex addressing structure.

Untrained/unfamiliar staff with IPv6 is probably the biggest risk to some enterprises implementing IPv6. Something, say a whole office or site, inadvertently exposed to the internet via IPv6!

if you talking about generic office staff, why would it affect them? the www still uses domain names on ipv6. dual stack doesnt disable ipv4 either.

also are you the guy who upgraded his office to windows 8? that would be a far bigger impact than adding ipv6 to office staff.

I am not reccomending not just because of ipv6 there is other factors although zen are in the shortlist, the isp is on the list I am passing on to them, however since I take into account principles as well the principles of zen regarding ipv6 had some affect of the one isp I will suggest to them is the best choice. To make it clear I am not telling anyone to avoid zen.

Also a factor is that these companies dont like changing isp's at all, the BE change is sort of enforced on them, an isp that is trying to be proactively future proof for that reason will obviously gain points.

ISP's used to live at the cutting edge, collaborating with each other to build a better internet.

These days ISP's are run by accountants, and It's sad that ISPs dont feel the need to upgrade to IPV6. Who knows what killer apps might come out of it. Multicast TV for example. Better peer to peer technologies, especially with mobiles.

ISP's like Zen are holding back innovation. I could understand if there was a big cost implication, but there isn't. I bet all their hardware has supported IPV6 for years. You just have to spend a bit of time training up staff, and testing it. It's just lazy.

Edited by stuorguk (Mon 09-Dec-13 02:38:11)

In reply to a post by stuorguk:

ISP's used to live at the cutting edge, collaborating with each other to build a better internet.

I wonder if we'll see it coming from the mobile phone world instead? Smart phones seem to pretty much all support IPv6 these days, so this may be where a lot of the IPv6 support ends up coming from. Mobile operators like Verizon wireless mandate that LTE phones must support IPv6, but IPv4 support is optional (!). T-Mobile US also has IPv6 running on their mobile network.

Given that RIPE seems to have about 14million IPv4 addresses left at the moment, it'll be interesting to see what happens when Europe actually runs out.

For better or worse ISPs for now don't need a reason not to deploy IPv6, that's sadly how it is.

Discounting an ISP due to their lack of IPv6 support barring a few extremely limited ranges of customers would be pretty odd.

I apologise for my harshness, your post did give the impression that you were directing people based on your own prejudices rather than towards whatever was best for them, and you have to admit given your posts here that your own choice of ISP is a compromise between price, service, and in no way IPv6 support.

Either way, sorry about that. Could've been phrased far better.

In reply to a post by Chrysalis:

Is ipv6 that hard to add to networking configurations?

If only it were as simple as that!

In reply to a post by Chrysalis:

I recognise staff training issues. I recognise old equipment may not suppport ipv6. However adding ipv6 routing to an existing network is one of the easier things I have done, and for safety ipv4 can be set as preference.

Painless enough to implement, sure, but there's a ton more to it than just implementing it. More to the point if you aren't actually going to use it, keeping IPv4 as the default, why bother? In a mission critical network implementing things for the sake of it is extremely bad practise.

In reply to a post by Chrysalis:

The problem of waiting for "when it has to be done" means it will never get done. its one of those thngs where everyone is saying "you first".

No, it means it gets done when it needs to be done. Corporate networks aren't always, indeed are rarely, running the latest and greatest versions of Windows, Office and Exchange for a good reason. It has nothing to add that will improve productivity.

In reply to a post by Chrysalis:

For this reason its not impossible ipv6 gets abandoned as already mentioned, it may be the case companies instead just use things like cg-nat instead as its more conveniant, although I think cg-nat is way more complex than adding ipv6 dual stacking.

CG-NAT is trivial and uses well understood concepts.

In reply to a post by Chrysalis:

I think also is a big difference between a company that uses connectvity for its business and one that sells connectivity for its business rolling out ipv6. You guys may consider configuring but not activting ipv6 as the norm for isp's yet I am seeing differently, I use many datacentres and most provide ipv6 connectivity.

Indeed they do, because they have a good reason to do so. It can be used as a sales pitch and is fairly painless to implement and maintain as datacentre networks are pretty simple compared to corporate networks.

If you look at the 'equivalents' of these datacentre networks, ISP core and transit / peering networks, you'll note that the overwhelming majority run IPv6 - go here and check the IPv6 address box. BT, Sky and TalkTalk all run IPv6 peering, O2/Be apparently don't, neither do Plusnet or VM/LGI.

ISPs on the whole are not scared of running IPv6 on their own core network, it's their access network, BRAS / CMTS / CPE that scare them more. Comcast had no choice in the matter as they ran out of RFC1918 space to manage their CPE.

In reply to a post by stuorguk:

ISP's used to live at the cutting edge, collaborating with each other to build a better internet

They still do collaborate.

What they don't do is live at the cutting edge. Many people and businesses rely on these ISPs to give them stable connectivity. Running a network on the 'cutting edge' as a giant lab for new technologies is a bad idea with that in mind.

I understand that, but IPV6 is not cutting edge. Maybe 10 years ago.

Nor is it essential just yet. Most of these guys' cost is CPE and that's the rub for right now. Look at Sky's new hub and think about the miniscule cost saving they'd have made going for 100Mb switch ports over 1Gb yet there they are.