No Business Case For IPv6, Survey Finds

alphadogg writes "Business incentives are completely lacking today for upgrading to IPv6, the next generation Internet protocol, according to a survey of network operators conducted by the Internet Society (ISOC). In a new report, ISOC says that ISPs, enterprises and network equipment vendors report that there are 'no concrete business drivers for IPv6.' However, survey respondents said customer demand for IPv6 is on the rise and that they are planning or deploying IPv6 because they feel it is the next major development in the evolution of the Internet."

Comments

Ever?

[WillKemp]

I'm beginning to find it hard to believe that IPv6 will ever be implemented. It seems to have been on the verge of it for close to a decade now.

Re:Ever?

[unlametheweak]

The problem is that the guys that were working on the big IPv6 transition quit there jobs to work on the Duke Nukem Forever project.

Re:Ever?

[bytesex]

As a person who's involved in an implementation of IPv6, let me say that it's difficult to see it implemented without ubiquitous gigabit networks all around, as well as network equipment (routers) that run on the kind of CPUs we don't nowadays expect such hardware to run on. On the one hand, they've made stuff easier (no more checksums on IP level, addresses that tell you something about themselves); on the other they've made it more difficult (potentially quite a lot of headers before you get to ICMP for example, as well as up to seven addresses that any device must listen to, address sizes that don't fit a natural integer), but the network is also busier: network meta-messages fly around all the time - much more so than with IPv4, its ICMP, IGMP and ARP (ARP times out in 20 minutes; link-layer address mapping in IPv6 expires in less than a minute), and don't forget multicast: it's obligatory and used a lot on IPv6, meaning that routers will be so much more busy synchronizing.

Then again; the time that hardware and linespeed catches up, *will* come. It's just not now, and nobody is in a hurry either. But running IPv6 over lines that do 1 Mbps in practice, however doable; it wouldn't make anyone happy.

Re:Ever?

[mellon]

Um, what the heck are you talking about? The ARP timeout is two minutes, not twenty. Speaking as someone who's also implemented IPv6 and used it pretty extensively, it sounds like you really don't know what you're talking about.

There is a known failure mode with ICMPv6 if you have a 127-bit prefix, but this is well-known, there's a fix for it in the standards, and the workaround is that you just don't ever use 127-bit prefixes. There's no particular benefit to using 127-bit prefixes, so this is kind of a no-brainer.

As for CPU consumption, again, what are you talking about? On the backbone, the proliferation of micro-routes for IPv4 is a *huge* problem. IPv6 route aggregation makes things *faster*, not slower, and consumes less CPU time as well.

If you are working over low bandwidth links, you might want to take a look at 6lowpan, which allows you to statelessly compress headers down to under twelve bytes.

Bottom line, the conclusions you've drawn are, as far as I am aware, complete nonsense. I'm sure you believe what you've said, and it's the result of real things that you saw, but without a bit more back story, I don't think it contributes any useful knowledge to the discussion.

Re:Ever?

[QuoteMstr]

Me, I would have preferred to extend the dotted-quad notation over using the colon-separated hex format usually used for IPv6. Dotted quads look more familiar for network administrators, software developers, and so on. As you noted, IPv6 addresses look strange and scare people. This fear of the unknown is a barrier to adoption. Any unnecessary break with IPv4 hurts IPv6 adoption, and we can't afford that; IPv6 with dotted quads is better than IPv4.

Re:Ever?

[QuoteMstr]

Your post demonstrates my point perfectly: the colon-separated hex notion screws up URL parsing, requiring algorithm changes for everyone, and as you see, lots of people still haven't gotten it right. Dotted-quad notation wouldn't have required nearly as much effort. The new notation was an unnecessary barrier to adoption.

We're talking about Joe Sysop and Joe Programmer, whose opinions regarding IPv6 are far more important than Joe Plumber's. These people see IPv6 as something exotic and frightening, and try to avoid it as long as they can. IPv6 should have been made as similar to IPv4 as possible; instead, the IETF tried to do too much too fast, and now we're paying the price.

Re:Ever?

[mellon]

Ohforgod'ssake. You're going to *type in* raw IPv6 addresses in a URL? I don't *think* so. I do it for debugging, but there's no way I'd ever ask an end user to type one in, and if I did there's no way the end user would do it. Which makes it a non-problem.

Decimal dotted quads are too big, and they wouldn't look like IPv4 dotted quads anyway. For instance, my IP address as a dotted quad is:

32.1.31.56.2.6.0.0.2.23.191.255.254.133.196.90

In hex, it's:

2001:1938:206: :223:dfff:fe85:c45a

You really prefer hex? You really think that's going to look familiar and comfy to a person who can't handle the hex format? Naw, dude - this is really a great way to weed out people who shouldn't be on staff - if they can't handle the hex, there are a lot of other much more important things they also can't handle, in IPv4-land as well as IPv6.

Admittedly, there's always resistance to new stuff by a certain number of people, and that's perfectly understandable and not grounds for firing. But those people will get over it after a bit of hands-on.

Well,

[TinBromide]

In a world without sharp objects, knives, or sidewalks, there would be no business case for bandaids. IPV6 is a solution to a problem that hasn't asserted itself. How often do you buy cough medicine when you haven't been sick in a while? This goes the same for ipv6. Until ISP's start charging more for ipv4 addresses due to scarcity, nobody is going to switch beyond digital survivalists and people who like to tinker with new technology.

Re:Well,

[mellon]

I guess you don't care about end-to-end connectivity. P2P, VoIP, skype, stuff like that? Obviously not something you want.

As we run out of IP addresses, we will have more NATting of IPv4 networks. This will mean that instead of having a single global IP address with your ISP, you will have an RFC1918 address. The people who have global addresses will be fewer, and so Skype's nat traversal will depend more heavily on them, which they will notice and which will decrease Skype's popularity. Same with p2p.

Consequently, at some point it will be the case that the only applications that are well-supported on the Internet are walled-garden apps run by commercial sites. Innovation will drop off.

It's not a pretty scenario. To me, the main selling point of IPv6 is *not* that we are running out of IP addresses and need more. It's that end-to-end is getting less and less available as the internet grows. Deploy IPv6, and end-to-end comes back. That's why we need IPv6.

It will happen

[Daimanta]

With the rate IPv4 adressess are running out it is only a matter of time before we will switch to ipv6. It might be 3 years from now or perhaps even more but when ipv4 becomes scarce(and it will), people and (internet)companies will try and make the switch to ipv6.

Don't get started about the turd that is called NAT, that's a problem posing as a solution.

Re:It will happen

[arkhan_jg]

NAT is the only reason we still have ipv4 - if we hadn't had that nasty hack, we'd have had to move to ipv6 out of necessity some time ago. I'm really looking forward to going back to having every PC with a globally routable IP address, it will make application communication work so much easier, and firewalls can stick to being allow/deny/drop firewalls instead of all this stateful masquerade hack-job stuff on top.

The main sticking point for me is all UK ISPs are IPv4 only. There's not much point running IPv6 internally if you're only going to have to tunnel it or 6to4 it once it leaves your network, though I'm thinking of converting a VLAN or two internally to IPv6 for a systems and applications trial.

Re:It will happen

[QuoteMstr]

You've hit the nail on the head. NAT dovetails very nicely with the "castle mentality" many network administrators have: this is mine, and you can't touch it. It's about control, and there are fewer more tangible symbols of control than your own network numbering scheme. Nobody wants to give up that sense of control by moving to IPv6.

But since 2005, you don't have to: IPv6 now has private address ranges just like IPv4's. Also, NAT has always worked with IPv6.

Since 2005, all four combinations of address spaces can work in principle: IPv4 inside, IPv4 outside, IPv6 outside; IPv4 inside; IPv6 outside, IPv4 inside (with DNS proxying), and obviously, IPv6 inside with IPv6 outside.

Whether this "castle mentality" is appropriate is a different debate. Moving to IPv6 for the public internet is too important to get bogged down in talking about NAT.

Aside from the obvious "business driver."

[palegray.net]

As a developer and network security professional, I frankly can't wait until everything under the sun is addressable. I really do want my car to be able to talk to my electric razor.

Customer demand should be the business case.

[mellon]

I don't mean customers should want IPv6. I mean that that's what should drive IPv6 deployment. Address depletion is a problem, but it's a problem that has workarounds, and to the extent that customers aren't bothered by the workarounds, there will be no IPv6 deployment.

The main impact of the workarounds is twofold. First, your outward-facing global IPv4 address will go away. Right now, your ISP has probably assigned you a real IPv4 address, not an RFC1918 address. So people can get packets to your gateway directly. That will go away.

The second impact is that we will have more and more layering of NATs. This will make peer-to-peer applications harder and harder. Also, as more users are piled up on single IP addresses, we will start to see port starvation. What this looks like is that iTunes will start acting funny - displaying some things, showing error messages for others. DNS lookups will fail, and you'll have to retry. Google maps tiles won't show up, so you'll see a partial map, and have to reload (possibly to see different tiles not show up).

So yeah, things will keep chugging along. But it will work less and less well as time goes on.

And I think that is what can, and should, be driving demand. If you don't want that, you might want to start fantasizing about how to get IPv6 into your own home. I have it in mine, it works a treat. I think it's too hard for the average person to do right now if their ISP doesn't support it, but that's a problem that we ought to try to solve if we want the internet to keep being a place where peer-to-peer is possible, and where innovation is possible.

Running out of address won't kill the internet. But it will suck the life out of it.

Cell phones

[FranTaylor]

If cell phones turn into real computers, which has probably already happened, then we will need IPv6 if all those phone users want to surf.

Let's flip the question....

[mark-t]

People ask what can IPv6 offer that NAT cannot. Try running multiple servers on multiple machines behind the same NAT, where one would like them to be accessible to the outside world via default port numbers. No amount of NAT configuration can get around this limitation, so saying NAT solves all the problems that IPv6 is supposed to answer is nothing more than self-delusional. Let's flip the question now.... what can NAT do that IPv6 cannot? Especially considering the fact that even *IF* for some reason that didn't involve how many IP's you actually have available, you still wanted to utilize NAT for some reason, you still could do that with ipv6... no problem at all. So what does NAT do that IPv6 can't? The only answer that might actually exist to this is that it arguably costs less to implement. So in reality, it's not that there's no business case of IPv6, it's really the case that these businesses are just cheap.

Self-defeat.

[numbski]

I tell this story all the time, and I'll tell it again.

I *tried* to build up a new fiber network in downtown St. Louis using IPv6. I couldn't get the address space!

It's insane - I could get 3x/24 blocks (non-sequential) assigned to my ASN, but in order to get an IPv6 allotment, I had to show proof that I *already* had utilized a full /24 of IPv6 addresses (which is NOT 256. It's 256*256*256!) They said to get it from my upstream provider - they said they don't do that, get it from ARIN. I go back to ARIN, ARIN says "They're full of it, get it from your upstream provider."

Even more insane? IPv6 allotments are FREE! I had to pay per year for an IPv4 allotment, but the free stuff? Pfft...we have it, we'll never run out of it within your lifetime, but you can't have it.

WTF?

Re:Self-defeat.

[mellon]

Instead of getting upset, get smart. ARIN is correct - you're supposed to get your allotment from your upstream provider, unless you're peering on the backbone (which it seems you aren't, since you have a provider). Your provider is probably used to the IPv4 way of doing things; the problem with that is that it produces fragmentation, which produces huge routing tables. In order to keep the routing tables small, the IPv6 allocation policy is to allocate hierarchically, so that you would get your addresses out of your provider's space.

When your provider runs out of space, you either renumber or fragment; renumbering is obviously preferred, and in v6 it's also easy, because you can do a soft transition - deprecate the old addresses, but keep using them for a month; by that time, all existing connections will be using the new addresses, and in the meantime all the connections that used the old addresses have faded away.

This is sufficiently different than the way things are done in IPv6 that it's not surprising that your provider doesn't understand it yet. So you need to help educate them - this isn't a situation where people are deliberately fingerpointing, but rather an opportunity for some education.

Minor nit - ARP cache timeout

[karl.auerbach]

This is a minor nit - ARP cache timeouts are normally on the order of 300 seconds, not two minutes.

A less minor nit is this: IPv6 does not help decrease the size of routing tables as seen by major providers. Nor does IPv6 reduce the burden of sending routing updates so that routing updates are propagated faster than the underlying rate of change of usable net paths. (Enterprise subnets, whether IPv4 or IPv6, don't generally propagate into the routing announcements as seen by the big carriers.)

The compelling argument, for me at least, is that IPv6 is really a new internet that runs along side of the existing IPv4 net - there is no direct interoperability. This means that pretty much any new expansion of the net is going to require IPv4 connectivity, and IPv4 addresses, to reach the legacy net. And that makes IPv6 redundant from the user's point of view. That sort of drains the oil out of the IPv6 crankcase.

Of course the biggest argument of all is that IPv6 does not solve the hard issues of propagating routing information and finding usable paths across the net, particularly as the demands of human-conversational traffic and the political acts of nations are (unfortunately) driving routing to become increasingly aware of the types of traffic being routed.

I'm waiting to be shown that I'm wrong - I helped do the very first calculation of IPv4 address consumption back in the mid 1980's. And I was in the group at Sun back in the very early 1990's where IPv6 took form. I spent time at Cisco wrestling with questions like how to efficiently mechanize 128-bit longest-prefix matching on 32 and 64 bit hardware. And my company currently has IPv6 testing products. So I've been watching IPv6 for what will soon be two decades.

To me one of the tilt-points of IPv6 will be when I can go into Frys Electronics and find IPv6 capable print servers and other widgets of that ilk on the shelves.

I saw ISO/OSI come and go (I was rather a fan of TUBA - which included the use of ISO/OSI CLNP for the new IP layer - when the various IPv4 alternatives were being considered in the early 1990's.) It would not surprise me to see IPv6 go the way of ISO/OSI.

Re:Minor nit - ARP cache timeout

[Anonymous Coward]

To me one of the tilt-points of IPv6 will be when I can go into Frys Electronics and find IPv6 capable print servers and other widgets of that ilk on the shelves.

We're starting to see this already. The Apple Airport Express/Base Station products are IPv6 capable and do 6to4 tunnelling when used as gateway devices, out of the box.
The HP CPxxxx series network printers are also IPv6 capable.
Now we just need the other tilt-point of broadband providers handing out IPv6 allotments, and we'd be set.