IPv4, IPv6, and a sudden change in attitude

https://apenwarr.ca/log/20200708

by LaSombra • 6 years ago

111 182 6 years ago

quaintdev• 6 years ago

One terrible way to switch to Ipv6 faster is to make existing services available only on ipv6.

Or create some cutting edge application that will work only on ipv6

I know ipv4 users like author mentioned will still be able to access them because someone else will plug in. We are stuck with ipv4 for decades, aren't we?

toast04• 6 years ago

We are stuck with ipv4 for decades, aren't we?

Probably. All the IPv4 only devices being made now aren't going to get IPv6 and they don't have the capacitor plague or junky lead free solder of the early 2000s to kill them. They'll probably live a long life, or at least until 2038.

DanAtC• 6 years ago

Most IPv4-only things I see are cheap IoT junk whose requisite cloud services will shut down well before that.

zokier2• 6 years ago

Ipv4 will not be going away because the long tail is very long, but it'll become irrelevant at some point for day-to-day stuff

Polylactic_acid• 6 years ago

A vps host I use offered v6 only servers for $2 cheaper but they removed that plan probably because people kept getting confused.

dtech2• 6 years ago

We are stuck with ipv4 for decades, aren't we?

This was always going to be the case, even with more speedy adoption op IPv6. It's also not really a problem.

Making services just available for IPv6 is going to be a recipe for disaster of the service, you've now made the service unavailable for 80-90% of your audience, and they can't do a thing to fix it (their ISP has to). You need to be better than start-up-era Google, Facebook, YouTube or Netflix to push through that and force ISPs to adopt IPv6 to support you. Basically impossible.

Forced adoption is a much better model. Apple forcing iOS apps to work on only IPv6 connections if they want to get into the app store probably is one of the largest drivers of adoption by businesses small and large, and IaaS providers especially.

AndrewDucker• 6 years ago

You're locking out 67% according to Google's latest figures.

51% in the USA.

Still too many to make it practical, but getting there.

bzb3• 6 years ago

Given that ipv6 makes it much easier to track individual devices, I'm sure ad companies would be happy to have it deployed. I think Google could strongarm isps into supporting ipv6.

alexarnesen• 6 years ago

The article was worth it just for this gem:

'If we were feeling snarky, we could perhaps describe IPv6 as "the String Theory of networking": a decades-long boondoggle that attracts True Believers, gets you flamed intensely if you question the doctrine, and which is notable mainly for how much progress it has held back.'

Sean-Der• 6 years ago

This trend has been bothering me a lot lately.

In standards bodies you will have a 'True Believer' about a particular topic. They will push it for years and years, and eventually get their way. The idea isn't bad, and would be great if included in the spec from day one. Unfortunately adding it to what we have today causes massive breakage/incompatibility.

Maybe I am just being resistant to good change. It just is frustrating because in most cases the 'True Believer' isn't going to go worry about the real-world impact of the change.

This phenomena seems to be hurting a lot of programming languages as well. So much harder to say no, and idealistic people are always going to find a way :)

woodson• 6 years ago

I like to see the other side (that of the true believer) as well. If all (or most) can agree on that it is the right thing to do and the current way of doing it is wrong, why oppose the change for 20 years saying it will cause breakage? If the change had been made at the beginning, any breakage would have been dealt with long ago. (I’m specifically not talking about change for the sake of making changes.)

m4633• 6 years ago

I think, like python 2->3, the chaos and confusion of trying to pull off a "clean slate reboot" is very rarely successful or worth it.

walshemj• 6 years ago

Back in late 90's back when I was working in OSI networking I realised that IPv6 was a bad standard.

hkt6• 6 years ago

Honestly, having worked for a hosting company in the past that was fastidious about IPv6 support, I can say a couple of things:

* It is almost definitely not worth it commercially, unless you've carved out your space in the community of people who specifically want IPv6

* It is much, much harder to work with than IPv4 and I don't believe this is only a lack of exposure

* Dual stacking is expensive and requires staff to pursue training with very uncertain levels of reward

IPv4 address space shortages.. could have been addressed by doubling the number of bits in an IPv4 address, rather than throwing out the many tools that already worked.

The last point in particular is heresy to network engineers, but perfect sense to commercial types. Adoption should be cheap, if not free. The huge up front human cost of training people to operate IPv6 is uneconomical. A 64bit IPv4.1 would be fine for decades.

throwaway2048• 6 years ago

There is a weird persistent idea that all anyone really needed to do was "extend ipv4 addresses" and everything ipv4 related would have remained compatible.

No, it still absolutely would have completely broken everything and anything that used ipv4, all the tools would still need to be thrown out.

There is basically no way such a proposal could work and maintain any sane level of compatibility.

Its evident right on its face, how exactly would an ipv4 only tool connect to a 64 bit "ipv4.1 address" ?

AmericanChopper• 6 years ago

Maintaining compatibility was never an option. But the most sensible solution to running out of address space is simply to extend the address space. Instead IPv6 decides to change the format into something which is not really human readable, and decides to kill NAT as well. Which was really a terrible decision imo. I like NAT, I like having addresses that I can remember, and I hate the idea of having a unique globally routable address for every device.

The primary argument for adopting IPv6 is that IPv4 will be exhausted. Not that there’s something good about IPv6 that I would want to have. Personally I hope it never succeeds in getting sufficient adoption, so that eventually we can have a good IPv7 that’s just a bigger version of IPv4.

privateSFacct2• 6 years ago

Do you admin a medium or larger network? Despite the strength of your comment it also illustrates some ignorance.

"No way", "weird persistent idea".

This despite many reasonable people suggesting it.

Deploying IPv6 at scale is deploying a totally different protocal. What is irritating is that it's not just a larger set of bits, everything changed making adoption and tooling MUCH much harder.

"All the tools would need to be thrown out"

Totally and absolutely false. Because an extended Ipv4 would have the same underlying concepts you could modify the tools and continue to use them.

From address assignment (3 ways now) to the dynamic address privacy extensons (don't actually play well with IPSEC configs) to doing renumberings on prefix changes (100% nightmare) to all the training / learning new things (costs money in bigger orgs) they seem to have purposely made this change extremely hard.

Good news, I'm on board more or less with the migration at this point, and if I am a good marker of average reasonable interested in new things but not wasting tons of time then this is a good sign.

But boy they could have made this whole thing easier

mprovost2• 6 years ago

There were proposals for backwards compatible addressing schemes. But they were rejected for a "clean slate" approach. Almost 30 years later, we can see how successful that was...

EIP (Extended Internet Protocol) [0] was proposed in 1992 as a replacement for IPv4:

"EIP achieves maximum backward compatibility with IP by making the extended space appear to be an IP option to the IP hosts and routers.

When an IP host receives an EIP packets, the EIP Extension field is safely ignored as it appears to the IP hosts as an new, therefore an unknown, IP option. As a result, there is no need for translation for in-coming EIP packets destined to IP hosts and there is also no need for subnet routers to be upgraded during the transition period."

[0] https://www.rfc-editor.org/rfc/rfc1385.html

p_l4• 6 years ago

A huge part of that is due to BSD Sockets, which require in-depth application support for any difference in connection parameters.

Thiez• 6 years ago

Where would these extra 32 bits come from? This simple IPv4.1 would still require all networking hardware and software to be updated. The protocol would also not be backwards compatible and people would still want dual stack. I imagine adoption would not be that much easier than IPv6 adoption, while missing out on the additional improvements that come with IPv6.

oarsinsync2• 6 years ago

while missing out on the additional improvements that come with IPv6.

Can you give some examples of what these additional improvements are?

hkt6• 6 years ago

Dual stacking with 64 (or more) bits and 32 bits would otherwise use the same software. It would be cheaper, require less training, and require less development. It would therefore not face the old, established IPv6 problems covered in the article, and could even treat the old IPv4 ranges as a /32 within the 64 bit address space. Rather than being more than twice as expensive in terms of expertise and development, it would be nearly as cheap as doing nothing.

dylz• 6 years ago

I'm sorry, what? Extending the bits in IPv4 will break compatibility completely. Everything will need upgrades just the same as to support IPv6, so you might as well go all the way at that point.

How is it much harder to work with?

galvin3• 6 years ago

It's a minor detail, but an IPv4 address is easier for humans to use.

It's easier to read, type, remember, share over the phone or even just plan with in a spreadsheet.

toast04• 6 years ago

It is almost definitely not worth it commercially, unless you've carved out your space in the community of people who specifically want IPv6

I mean if you're addressing mobile, IPv6 hosting means skipping carrier NAT64 for at least t-mobile and jio. Probably other carriers too.

p_l4• 6 years ago

Generally high chances you'll get better connection with any LTE network, as having only v6 to the mobile terminal (phone/modem) is much cheaper - which drove v6 adoption heavily.

mprovost2• 6 years ago

The extension to a 64 bit address with few other changes was called SIP and was proposed in 1992 (!).

https://tools.ietf.org/html/draft-historic-simple-ip-03

p_l4• 6 years ago

TUBA, which used existing technology and tools to support 18-20 byte addresses (yes, byte, not bit) was ready and implemented on two different platforms by ~1991.

prussian• 6 years ago

Funny enough, to the last point, someone drafted an IPv4+ spec that does something like that.

see https://tools.ietf.org/html/draft-tang-ipv4plus-03

Basically adds an optional area code to encode more addressable bits.

gorgoiler3• 6 years ago

IPv6 is great if you control lots of components that need to talk to one another. Life is much simpler when you can trivially have thousands of globally routable addresses. It provides great value without having to be 100% deployed and for that, I am grateful.

I liken it to a commercial kitchen. Large numbers of identical steel pans, everything measured in grams, Cambros, walk-in refrigeration. Ideas and tech that are unlikely to catch on in the home kitchen, but when you operate at a larger scale, invaluable.

tomalpha2• 6 years ago

...everything measured in grams...
Ideas and tech that are unlikely to catch on in the home kitchen

Given the posting time I’d guess you were likely not US based so I found this surprising.

Even if that is so, then I continue to be surprised every time I rediscover the fact that the US hasn’t moved towards metrification.

Edit: rereading this the wording is harsher than intended. So to add: Either way the wording tickled me for some reason. Whether intentional or not, thank you (genuinely) for the smile this morning.

nikanj• 6 years ago

Metric units are a mouthful. Centiliter <-> ounce. Kilometer <-> mile. Centimeter <-> inch.

I think having easy-to-pronounce names for the metric units would help. Kliks are much better than kilometers.

aequitas• 6 years ago

Aren't all imperial units standardized on metric values nowadays? So technically Americans are using metric, they just use a different notation?

rascul• 6 years ago

the US hasn’t moved towards metrification.

I couldn't care less about which system is used, i just want there to be only one. I'm not a fan of owning two sets of tools, and I'm not a fan of working on a car and coming across something I need my second tool set for because people can't figure out what system to use.

guenthert3• 6 years ago

IPv6 is great if you control lots of components that need to talk to one another.

And that's the fallacy. Those things don't need to talk to each other (my toaster doesn't need to talk to your fridge), they do talk to a well defined peer (gateway, server, whatever). There might be someday more than a few billion networked things out there, but there won't be even remotely a billion gateways on the Internet in the foreseeable future.

Enterprises generally don't like outside elements access elements within their network at will. That's not going to change, it will always be through some form of gateway.

Dagger28• 6 years ago

There are still benefits to using unique address space on all networks though. The fact that my fridge doesn't need to talk to your toaster isn't a reason for us to not do v6.

If you've never had to deal with clashing RFC1918 blocks when doing a VPN or network merger, consider yourself lucky. It's a giant headache that can never be properly fixed, yet v6 completely avoids it.

bsder2• 6 years ago

IPv4 continues to exist because consumers are subject to ISP monopolies.

"Symmetric bandwidth" or "Running my own server" are not criteria that can be used to choose an ISP because there is NO ISP that offers either one in most of the country (US).

Since nobody can run their own servers, there is no pull for a larger chunk of addresses that would drive IPv6 adoption.

shmerl7• 6 years ago

Then relatively easy solution is to not give ISPs licensees to operate if they don't provide IPv6. Sounds more like an administrative problem for the government than a technology issue. Same applies to many other problems caused by monopolies. That's why there is anti-trust, or at least should be.

logicchains• 6 years ago

The ISPs only have a monopoly because the government literally makes it illegal to start a competing ISP in many areas. Even local governments were forbidden from competing with Comcast et al. in some cities.

dx0343• 6 years ago

Not offering symmetric bandwidth has technical reasons. You only have a limited amount of bandwidth with most technologies and the vast majority of consumers is much better off having higher download than upload speeds.

I'm actually glad that ISPs default to only giving one IP address per household. Not having most devices directly reachable from the internet is an extra layer of security. It should never be the only one but can be an extra step to make it harder for atttackers to introduce malware.

shmerl7• 6 years ago

No technical reason today, except for ISPs not upgrading to modern technology (fiber optics) because of being short term profit cheapskates.

Jonnax3• 6 years ago

Your broadband argument does not work in 2020. The technology exists for symmetric.

NAT is not security.

Here's an article from F5:

https://www.f5.com/services/resources/white-papers/the-myth-...

And why would the devices be reachable over the internet?

Your router should firewall by default its LAN side subnets.

galvin3• 6 years ago

The vast majority of consumers don't know or care about IPv6.

There seems to be more competition here in France than in the US but while I've previously used ISPs with IPv6 support my current one does not. Even as someone who prefers to have IPv6 it's not something I considered when signing up and I probably wouldn't pay much extra for it, as a consumer.

Symmetric bandwidth and connections for anything more than a personal server are available here but only on business plans. Those are easily more than 10x consumer prices so I can't imagine going for them unless you need the SLA's.

sandstrom2• 6 years ago

Still some distance to go, but Google see 30% of traffic from IPv6 now. It’s slowly getting there.

https://www.google.com/intl/en/ipv6/statistics.html

Also, with most cloud hosting (AWS et. al.) it’s fairly trivial to enable dual stack support.

teraflop2• 6 years ago

Google Compute Engine still does not support IPv6 for internal networking, only for external-facing load balancers, despite having made vague noises about it being "in progress" for the better part of a decade.

dsl• 6 years ago

Are there any specific features of IPv6 you need for inner-cloud connectivity?

If you are spinning up enough resources to exhaust even a /16 of IPv6, you can probably use your monthly bill as leverage to force the issue.

zokier2• 6 years ago

Also, with most cloud hosting (AWS et. al.) it’s fairly trivial to enable dual stack support.

I say AWS IPv6 support is pretty atrocious. You still can not set up pure IPv6 VPC, even for internal use, afaik most AWS services are still accessible only through IPv4, IPv6 VPC has lots of weird limitations, etc etc. Somewhat sad, considering that one would think AWS would be one to benefit from v6

sandstrom2• 6 years ago

Agree, it's sad and strange that they can't provide that.

Still, hosting an IPv6 service on AWS is fairly simple (at least in my experience).

LinuxBender• 6 years ago

What percentage of that is from cellular networks?

ac292• 6 years ago

Interestingly, you can clearly see the effect of the pandemic if you zoom into this year. The swing between access from IPv4-leaning business networks during the week to more IPv6-leaning home/mobile networks on the weekends has greatly reduced.

ohazi• 6 years ago

This motivated me to bite the bullet and jump through the hoops to set up an IPv6 address for my personal server. I got all the way to the end and then just when everything looked like it was working, I tried to load the page and... nothing. Turns out my home network doesn't currently have an IPv6 address. Comcast in the SFBA seems to alternate giving out IPv6 addresses and then taking them away again every few months or so.

Works on my phone, though, so ¯\_(ツ)_/¯...

divbzero2• 6 years ago

My ISP doesn’t support IPv6 either. I get 0/10 on WiFi vs. 10/10 on mobile using Test Your IPv6. [1]

[1] https://test-ipv6.com/

I’ve considered emailing to ask them to support it but haven’t come up with a persuasive reason for why they should.

Jonnax3• 6 years ago

Find out what your ASN is (e.g. https://www.ultratools.com/tools/asnInfo) and then look it up on peering dB:

https://www.peeringdb.com/

It'll give you an idea if your ISP is advertising IPv6 routing on their network if they have some prefixes. So whether it's something they're working on or not.

executesorder66• 6 years ago

but haven’t come up with a persuasive reason for why they should.

I don't know about you, but whenever I see an ISP, hosting provider, or website that supports IPv6 I think: These guys know what they are doing, they care about quality, and they actually plan for the future.

I'd much rather do business with someone that supports IPv6 because of the above impression.

If I was running my own business that provided network related products or services I'd make IPv6 support required because I don't want to seem like I'm incompetent/lazy/ignorant.

So in short the persuasive reason is that it would improve their reputation. Technical people notice these things and base their recommendations on these sorts of impressions.

lizknope• 6 years ago

I have gigabit fiber through AT&T. So do my parents 30 miles away.

I get a 0/10 from the test-ipv6.com site

I logged into my parents network and they get a 10/10

As I said we live 30 miles away and we both got fiber 2 years ago.

apearson• 6 years ago

Look into setting up a ULA IPv6 network

tenebrisalietum• 6 years ago

Hurricane Electric was giving out free tunnels for a while, they might still be.

chousuke• 6 years ago

They still do that; I have one set up on my router.

The main problem with it is that the tunnel server is hosted in Sweden, and while sometimes it's actually lower latency than my IPv4 connection (due to better routing, I guess), every geolocating website (eg. YouTube) thinks I'm Swedish :/

angrygoat2• 6 years ago

The 'noise' on this plot of Google IPv6 adoption is interesting: https://www.google.com/intl/en/ipv6/statistics.html

It looks quite periodic when you zoom in: at first I thought it might be different adoption rates in different timezones, but there's only one sample available per day so that doesn't explain it. Anyone know what might be going on?

teraflop2• 6 years ago

It's higher on weekends and holidays, which is presumably connected to mobile devices having a significantly higher rate of IPv6 adoption.

why_only_152• 6 years ago

You can see on the graph between March 10 and March 24 the significant increase in the number of people working from home by the proportion of ipv6 usage

angrygoat2• 6 years ago

That makes sense - thanks for that!

bluGill3• 6 years ago

Not just mobile, the big US cable ISPs have been doing IPv6 for years. People go home and everything is IPv6 and they don't even know.

pferde• 6 years ago

That is, the ability for your connections to keep going even if you hop between IP addresses. If you had IP mobility, then you could migrate connections between your two internets in real time, based on live quality feedback.

This bit is solved pretty well by https://en.wikipedia.org/wiki/Multipath_TCP, but of course, both ends of a TCP connection have to support it.

PaulAJ• 6 years ago

The people who make the decisions about IPv6 adoption are not users, they are ISPs. You get an IPv4 address on your router WAN port because that is what your ISP gives to you.

From an ISP's point of view IPv6 is a lot of No Business Case. Those three words are the death knell for any proposal to do anything in any business that expects to be here next year. It has exactly nothing to do with the geeky I-could-have-designed-it-better arguments about the technology.

If you are an ISP that is a going concern, with a bunch of customers sitting on IPv4 addresses, then handing out IPv6 addresses makes no difference, except when it breaks something. You still have to give your customers just as many IPv4 addresses. So why bother?

If you plan to migrate your customers to IPv6 then you are a lunatic. Its going to break stuff. Lots of websites don't exist on IPv6, and customers are going to notice. Also your customers have spent the last 20 years slowly picking up bits of IPv4 lore, like the vital importance of 192.168.0.1, and are going to be puzzled when this doesn't work any more. All of this translates into higher support costs and more customer churn.

Also, your allocated block of IPv4 addresses is a valuable asset in its own right; not only does it have real financial value (around $20 per address at present), but it also acts as a barrier to entry for competitors; if you want to set up in business as an ISP you are going to have to acquire some IPv4 address blocks from somewhere, and they aren't making any more of them. Managers are trained to look for barriers to entry to their industry, and the current IPv4 situation is exactly that. No sane manager is going to do something to make it easier for new competitors.

Eventually, of course, the dam is going to break and IPv6 will become ubiquitous. ISPs will decide that buying blocks of IPv4 addresses costs more than providing new customers with IPv6 plus some kind of carrier-grade NAT for legacy IPv4 addresses. More website hosting companies will support IPv6 in response, and suddenly IPv4 will be so last-decade.

antientropic• 6 years ago

If you plan to migrate your customers to IPv6 then you are a lunatic.

According to https://www.google.com/intl/en/ipv6/statistics.html#tab=per-..., IPv6 adoption in the US is above 42%. Was that all done by lunatics?

goalieca2• 6 years ago

I would not be surprised if the breakdown of those users was near 100% for mobile users and less for broadband.

barbegal• 6 years ago

By migrate I think the OP meant take away their IPv4 address. IPv6 adoption may be at 42% but unless it is at 99.999% you will have a lot of issues when not allocating an IPv4 address.

glogla• 6 years ago

Yeah, we live in a culture where thinking about anything but next quarter revenue is discouraged in companies. In such a culture, anything but a status quo is not going to happen easily.

IPv4, IPv6, and a sudden change in attitude

Related Stories