There are unallocated /8's that are privately routed. A lot are controlled by the U.S. DoD. But every time they come up, the response is always "so what? use IPv6!"
I generally agree. Putting those in circulation would actually be a bad thing. It would allow IPv4 to continue to zombie-walk for another couple years, maybe another decade, and in so doing to become more deeply entrenched, harming us in the long-run. IPv4 is technical debt.
It's also important to consider why DoD (and some large corps too) likes to sit on these address spaces. The problem with IPv4 isn't just that there isn't a big enough address space for every device on the Internet. IPv4 also lacks enough address space to permit the easy allocation of non-conflicting private address spaces. Ask any large-scale enterprise network engineer what happens when two companies merge or want to interconnect corporate networks and both use 10.0.0.0/16 as their address space. The answer is a visit to the bottom-most circles of NAT hell where sinners are boiled in firewall port remapping rules for all eternity. Either that or one of the merging/linking entities must renumber their entire network, which is often even more painful.
It really, really pains me that so many cloud providers are taking so damn long to support IPv6. Does Amazon even support it yet? I heard Digital Ocean was finally doing a limited beta. Jeez.
The cloud provider problem is definitely real. I have an idea for a service that requires both IPv4 and IPv6, but Linode and Rackspace were the two I found that both support native IPv6. Otherwise, it's something involving a HE.net tunnel, which I'm not comfortable with for more than experimental use.
A note on Rackspace: we see about 2–3% of the VMs we allocate come with only an IPv4 address. We have no idea why — we suspect it's a bug on Rackspace's side (we allocate everything automatically, through the same code path, so it isn't us!).
We currently only use the IPv4 address (I know, I know!), but it does present a bit of a problem when/if we decide to move.
I work at Rackspace and I'd love to get our product engineers involved and review this issue for you. Would it be possible for you to email your account number and any pertinent server information (hostname/IP address/etc) to help@rackspace.com so that we can investigate?
Thanks!
Andy Pape
Social Media Support
@Racker_Andy
> It really, really pains me that so many cloud providers are taking so damn long to support IPv6. Does Amazon even support it yet?
I don't work at Amazon, but I spoke with someone about this recently, and I believe the issue is that Amazon provides virtual hardware and a virtual networking stack, and the tooling required to virtualize the networking stack in IPv6 isn't mature enough yet.
For providers that offer dedicated hardware, this isn't an issue, but Amazon's services abstract over a very complicated and intertwined stack, so I can understand why it would take them longer.
KimeraLive.com supports IPv6 since 2012, we have more than a few private data-base servers that have IPv6 only access.
Cloud providers are not the problem, Internet Service Providers are, at home I'm stuck with IPv4 still.
This is the recovered pool that's finally running out, so short of reclaiming something drastic like the class A block 10 (you know, a really really bad idea), dang.
Anyone know what IPv6 adoption is looking like these days? Google is showing like 3% of their connections are native to it [1], which does not look promising.
Great strides could be made if Google, Wikipedia, and other major destinations had a day where they displayed a box to non-IPv6-enabled visitors telling them to contact their ISPs and ask for IPv6 and explaining why.
Great strides will be made when the cost of each IPv4 address rises significantly. That's when you'll see real progress, though not necessarily in the right direction.
If only there were sites / protocols that everyone wanted, which could only be reachable by IPv6. Maybe a new file sharing protocol, which uses the vast IPv6 address space to hide better, or something like that.
LTE is in large part about IPv6 (with the aim of using the PS IPv6 net to move all audio/video longer-term), often with carrier-level NAT for IPv4, simply because unlike the incumbent telecos, many mobile operators simply don't have many IPv4 addresses and there isn't the pushback against carrier-level NAT on mobile.
The World IPv6 Launch site has monthly measurements available that show the amount of IPv6 being seen from network operators participating in the measurement effort:
Even if a block is not publicly advertised, it is still entirely possible that it is used internally. Using the 10/8 range was not common place, and certain organisations have IP's that are technically publicly routable in their internal infrastructure.
As for 44/8, there are still those that use it, taking it away would mean having to renumber all of the equipment on that network...
No, trying to reclaim IPv4 addresses by making people go through herculean efforts to renumber their networks is not the way forward and simply delays the inevitable. IPv6 is the way forward, start pushing that, and get more people using it.
Ford has an entire /8, a significant chunk of the entire address space. Ford does not have as much reason to need that much address space compared to, say, major ISPs.
In practice, I imagine they'd outsource the internet connectivity bit to cellular carriers (as is the case for OnStar and similar at the moment) and those companies would be the ones to worry about provisioning IP addresses.
I would agree, and say that universities with an entire /8 probably shouldn't have them anymore either. That's a case where it definitely made sense early on, but stopped making sense quite a while ago.
You could never recover 10/8. Simply impossible. Due to the fact that IP address space is running out, 10/8 what everyone uses for NAT; how are you going to solve the address exhaustion problem by removing the main tool people are using to fight it?
Well before that, you could recover 18/8, MIT's prefix; how many addresses does a single university actually need, even MIT? Stanford already returned their /8 prefix. Or you could potentially recover a variety of DoD /8s, or start assigning some of the ones that are "reserved for future use" like 240/4.
Of course, an even better use of 240/4 would have been to use it as a backwards-compatible way to move to a larger address space. IPv6 should have been specified with IPv4 compatibility via NAT from the beginning, so that it could be a gradual migration. I mean, we're moving that way anyhow, but with a much more painful period in the middle in which NAT is necessary but IPv6 isn't ready yet for people who want to end the pain of NAT.
> start assigning some of the ones that are "reserved for future use" like 240/4.
The 240 block is problematic because a huge amount of legacy routing equipment is hard-coded to blackhole it. The best thing for it is probably to reserve it for unspecified local use, which would allow you to use it to NAT IPv6 addresses into locally, or use as additional RFC1918-style address space, or as ORCHID-style address space for IPv4-only applications using the likes of cjdns, etc.
Last time I tried to acquire a new IPv4, the distributor (my VPS host) required a technical reason for the request. If the request wasn't up to par, you would be denied an IPv4 + given an IPv6.
It's trivial to get new IP4 addresses on AWS and Rackspace. I guess it depends on the company. Microsoft's Azure does a weird thing where instances all share the same IP but are accessible via different ports when in a group.
Depends on the RIR that assigned your block. RIPE For example requires you provide a technical reason, AWS gets around this by saying that technically every instance could have an SSL certificate and thus should have it own IP for compatibility reasons. RIPE will and has reclaimed /24's and /16's from LIRs who didn't follow their guidelines (tl;dr of which is "must be a technical limitation of IP sharing")
I set up a dedicated server a few months ago. I was given, without even asking for it, a /29 (sort of 8 IPs), and could pay more for additional ones with justification like wanting one per virtualhost, etc.
This was not at a particularly expensive provider.
Hasn't that been the case for a long time? I'm pretty sure I remember seeing stuff along the lines of, "If you'd like an additional IP address for your server, contact us and tell us why you need it" a decade or more ago.
15 years ago I had to fill out an IP justification form - the justification was I had multiple SSL sites that all needed separate IP addresses. Wasn't a big deal, but was told it was standard practice (it was a smaller server center). And then I've had other dedicated server companies give me a block of 8 (5 usable) as just part of the monthly price.
What seems so wasteful is the "3 unusable" out of the 8 that are allocated. I have a vague understanding that those 3 are used for some routing/addressing stuff internally, but it's always felt incredibly wasteful.
You can actually use two of these addresses (network & broadcast) - but you cannot really call it a proper /X block anymore. You will also have to make sure to disable the broadcast address you've used up on all of your hosts.
This is all very fuzzy and some admins will look at you weird for doing this - but sometimes you've gotta do what you've gotta do.
The third address (router) is actually used by your ISP's router's to reach your subnet.
Probably the same day that the Voyager spacecraft finally leaves the Solar system. What I mean by this, is that the line is really fuzzy -- you can always buy a block of IPv4 addresses, if you pay enough.
When I changed my home provider 6 months ago I didn't get a real static IP by default. I had to call them and explain why I would need one to get it. That may not sound like much but here in Bulgaria just a few years ago we used to get 2-5 IPs with faster home plans and 1 with even the cheapest. The pressure is slowly hitting the end users so it can't be that much longer until it becomes cheaper to switch.
So what exactly is the "IANA IPv4 Recovered Address Space"? Today's news sounds almost exactly like February 2011 when the IANA announced that it finally allocated the last remaining blocks: http://www.nro.net/news/ipv4-free-pool-depleted
The announcement in Feb 2011 was that IANA had allocated the last IPv4 address blocks from the planned allocations. IANA (ICANN) and the RIRs then worked to recover some blocks of unused IPv4 addresses. These became the "recovered IPv4 pool" and the agreement was that allocations from this pool would begin once one of the RIRs reached the last /9 of their available IPv4 address space.
LACNIC hit that mark and triggered ICANN's policy to start allocating out of this recovered address space pool. The NRO has an announcement up about this, too:
It would be nice if certain websites we all use every day supported v6 better...
firstmillion:~▻ host -t AAAA news.ycombinator.com
news.ycombinator.com has no AAAA record
firstmillion:~▻ host -t AAAA twitter.com
twitter.com has no AAAA record
firstmillion:~▻ host -t AAAA cloud-images.ubuntu.com
cloud-images.ubuntu.com has no AAAA record
firstmillion:~▻ host -t AAAA ubuntu.com
ubuntu.com has no AAAA record
Easy problem to solve, so many entire /8 blocks are owned by single corporations, which should either be forced to sell or give up this outdated privilege. Does Xerox really need all of 13.x.x.x? Or Ford Motor Company all of 19.x.x.x?
And how long will that last? And there's a presumption here that businesses with lots of IPs that got in early aren't already selling them for fun and profit. They, such as Nortel, already are.
Between 2008 and 2010, IANA was allocating 10 /8s per year. In 2010, they allocated 20 /8s. There are only 256 total in the system, and not all of those are even usable.
Rapidly-developing Asia already has about as many people as there are IP addresses by itself.
Where are you going to find enough /8s, how are you going to recover them fast enough to do any good, and who's going to pay the massive legal bills from the ensuing litigation?
They probably assign static IPs in their range all the time. They just aren't publicly routable. Taking away their /8 would require a huge amount of reconfiguration work, so not sure what the motive would be, until the prices get really high.
So there is a lot of sentiment that I've been hearing over time about when exactly will IPv6 become a thing. Here's how I see it:
Step -1: We started running out of IPv4 addresses. The internet is IPv4-only for all practical purposes. Nobody is doing IPv6 because it's still experimental. Your weird neck-bearded friends are talking about it, that's about it.
Step 0: Inception! IPv6 is now a standard and you can route IPv6 over the tubes. Two networks have connected!
Step 1: Dual stack is a thing. Now you can set up both IPv4 and IPv6 on the same network and they co-exist. Some hosts (goole.com, facebook.com) try this and finally turn it on for good in 2012.
Step 2: Dual stack intensifies. This is where we are now. You cannot yet run a full blown IPv6-only network that you want to talk to the Internet and do anything practical with it, but all your cleanly shaven friends are now telling you that you should support IPv6 first, then add IPv4 for compatibility. Google sees as much as 3.5% of their traffic over IPv6.
Step 3: The cost of any single IPv4 address rises dramatically. For an end user it's already roughly $1/month. Buying in bulk is much cheaper, but not free like it used to be. ISP's and other service providers (Amazon, Rackspace, Digital Ocean) are going to start noticing the costs on their bottom lines. Severe restrictions are imposed by IPS's and service providers on anything but the bare minimum of IPv4 usage.
Step 4: The future. The first IPv6-only networks start to pop up. They use special network translation to talk to IPv4-only internet. (You can do this at home now! It's called NAT64: http://www.litech.org/tayga/).
Step 5. The number of IPv6-only networks is growing rapidly. IPv4 is now a legacy protocol. All your friends are telling you that they cannot believe we are still using IPv4 and how much of a pain it is to support it. Sales of T-shirts with "There is no place like 127.0.0.1" drop dramatically.
Step 6. The same people that put together IPv6 World Launch Day put together an IPv4 funeral. Major companies turn off IPv4 presence.
The crucial point here is not that someone demands IPv6 by calling their ISP, etc. It is fun to play with and you should at least get a tunnel set up through Hurricane Electric if you are reading this, but the demand here does not matter. What matters is supply: supply of IPv4 that is. Once that goes to nil, the cost rises and some executive at every ISP and service provider has a bright idea to save the company millions by switching everyone to IPv6. That's it. We, the people in the trenches, cannot affect this. We cannot speed it up and we cannot slow it down. The invisible hand of the market will make IPv6 happen all on its own.
Yeah, but thats only for those who meet the requirements.
The Real world scenario is that people pay about $500/mo for a /24 I heard - kind of a black market thing.
Hmm. I would argue that we're still in step -1, though the clean shaven folks have finally started listening to the neckbeards in a few circumstances and are slowly moving around towards step 0.
Most people still can't actually route IPv6 over their tubes without going through a tunnel broker. At least for home, small office, and companies that aren't tech heavy, it's still firmly in the neckbeard camp.
Well, you definitely can route IPv6 over the public backbone. If you get an IPv6 address in SF and I get one in NYC, we can already talk to each other. That is how we got past step -1. Dual stack is now a pretty standard setup on a non-trivial, though still small number of networks. For example, Linode, Comcast, T-Mobile all support dual stack on at least some of their networks.
What we have not quite reached yet is the point where IPv4 addresses are really expensive. You still get one with every VPS you sign up for and with every home connection. Once that can no longer be done, IPv6 will become standard pretty much across the board, starting in Step 3 and 4.
As another example, when I am using LTE on Verizon Wireless, I am connecting over IPv6 to sites running IPv6 (ex. Google, Facebook, Wikipedia, etc.). I can go to sites like http://www.test-ipv6.com/ and see that I have full IPv6 connectivity.
Hey, so I had posted the above primarily because I've been trying to get a routable IPv6 address for a while, and haven't been successful; I've repeatedly asked Comcast when my CMTS would be IPv6 enabled, but their techs have just said "try enabling IPv6 on your computer and see if it works", they haven't been able to give me a schedule or even tell me if it was supposed to work yet.
However, just checked last night and discovered that they'd added a new feature that would tell you when your CMTS was IPv6 ready, and it turns out that they've finally upgraded me. I could finally switch off my tunnel and use a real, routable address. Woo!
Aside from ::/128 (undefined address) and ::1/128 (local loopback), the entirety of ::/96 was reserved for ipv4-compatible addresses and had been deprecated for a while [1]
It is safe to assume that they won't be allocated in the future.
You can easily assign them to your loopback interface. e.g. on linux:
$ ip addr add ::2/128 dev lo
You will then be able to use them over loopback just like ::1, and can populate your /etc/hosts file accordingly.
$ ping6 -c 1 ::2
PING ::2(::2) 56 data bytes
64 bytes from ::2: icmp_seq=1 ttl=64 time=0.078 ms
1 packets transmitted, 1 received, 0% packet loss, time 0ms
rtt min/avg/max/mdev = 0.078/0.078/0.078/0.000 ms
We've been saying goodbye to IPv4 for years and years. No one will care unless money is lost by not running IPv6. That is when adoption rates will go up.
My provider (Suddenlink) support rep responded to my question about IPv6 support with "No, we do not support it since almost no one uses it currently." S
Perhaps you can point them to sites like http://www.worldipv6launch.org/measurements/ and show them some of the stats on networks that are using it. You can also point them to the fact that Facebook, Netflix, Wikipedia and all of Google's sites are available over IPv6. So the "almost no one uses it" doesn't quite fly.
Isn't IPv4 good for privacy? My IP address doesn't uniquely identify me unless you get records from my ISP. But wouldn't I be more likely to keep the same IPv6 address consistently?
It also makes it hard for websites to block users based on IP without also blocking lots of legitimate users sharing that IP. I think this is a good thing to keep the internet more equal and anonymous.
Time to start auctioning IPv4 ranges. Is there anything in the terms that prevents such an auction? Putting a price on ranges should redistribute IPv4 ranges to those who need them. It's certainly simpler than coming up with a set of rules for usage.
I have the feeling that fighting this is about as important as protecting net neutrality. I wouldn't like a world where only those who can afford to buy expensive IPv4 addresses can serve their content without any obstacles, especially when the issue would be completely solved by wider IPv6 adoption.
Steve Reeding, one of the creators of IPv6, shares his crucial insight on how some networks adapt, survive, and impose themselves [1]:
"I have a different take on that. Yes, a lot of people were dissing tunnels yesterday. I'm actually a big fan of tunnels. I think the way the IPv4 Internet was built originally was basically by tunneling over the phone system. Ignore, you know, getting leased lines, putting computers on the end, building a network with no cooperation at all from the phone company. In fact, I once saw a phone company memo that referred to the Internet as a 'hostile overlay'. You know, and we basically, you know, demolished their billing plans and everything, and built this network in spite of the desires of the carriers. And, as I see the IPv4 Internet getting more and more like the phone company, I thought one way to deploy IPv6 is basically tunnel over all that junk. Iterate the same thing again. And I gather, you know, some people, you know, who are concerned about this are looking at other ways of sort evolving the Internet, where you, you know, tunnel over HTTP or whatever you have to do to basically get over all that cruft and build a new network on top of it. So that, to me, was one possible way, you know, as the IPv4 Internet rotted the underneath. We would tunnel over it and then we would throw away the rotted bits and replace them with wires."
I also wondered about this and any throttling-based protection / rate limiting based on IP. With IPv6 the address space is huge and makes blacklists far less effective (or much more memory consuming and slow, but that's something that can be resolved with things like Bloom Filters I imagine).
It would take some time to adapt implementations. I speculate this gives attackers some opportunities to use old-style hacks that were largely mitigated, but suddenly re-appear with IPv6.
The quickest way to gain mass adoption to IPv6 is if facebook, twitter and google+ collectively decide to disable their IPv4 and leave only IPv6 as a connection option.
I figure the subsequent avalanche of user pressure on isps will solve the problem in a week or so. Added bonus: during that week productivity will be at an all time high.
Ive been given a static ipv4 and a big ipv6 range from my isp, problem is despite getting the latest and best router
http://www.asus.com/Networking/RTN66U/
I cant get ipv6 to work with this router, tho if i connect laptop directly to the wan cable i get an ipv6 address without issues and can use google via ipv6 etc
If a slightly unshaven geek like me cant get ipv6 to work easily what hope is there for rest of the world :(
It feels like I've been reading this exact headline once a week for a couple of years.. I started to believe that there is an infinite reserve of ipv4 blocks and we'll be just fine.
Meanwhile, most ISPs have done sod all about IPv6 and will probably only start when they start getting complaints from users that foo site doesn't work.
110 comments
[ 0.18 ms ] story [ 185 ms ] threadI generally agree. Putting those in circulation would actually be a bad thing. It would allow IPv4 to continue to zombie-walk for another couple years, maybe another decade, and in so doing to become more deeply entrenched, harming us in the long-run. IPv4 is technical debt.
It's also important to consider why DoD (and some large corps too) likes to sit on these address spaces. The problem with IPv4 isn't just that there isn't a big enough address space for every device on the Internet. IPv4 also lacks enough address space to permit the easy allocation of non-conflicting private address spaces. Ask any large-scale enterprise network engineer what happens when two companies merge or want to interconnect corporate networks and both use 10.0.0.0/16 as their address space. The answer is a visit to the bottom-most circles of NAT hell where sinners are boiled in firewall port remapping rules for all eternity. Either that or one of the merging/linking entities must renumber their entire network, which is often even more painful.
It really, really pains me that so many cloud providers are taking so damn long to support IPv6. Does Amazon even support it yet? I heard Digital Ocean was finally doing a limited beta. Jeez.
We currently only use the IPv4 address (I know, I know!), but it does present a bit of a problem when/if we decide to move.
I work at Rackspace and I'd love to get our product engineers involved and review this issue for you. Would it be possible for you to email your account number and any pertinent server information (hostname/IP address/etc) to help@rackspace.com so that we can investigate?
Thanks! Andy Pape Social Media Support @Racker_Andy
I don't work at Amazon, but I spoke with someone about this recently, and I believe the issue is that Amazon provides virtual hardware and a virtual networking stack, and the tooling required to virtualize the networking stack in IPv6 isn't mature enough yet.
For providers that offer dedicated hardware, this isn't an issue, but Amazon's services abstract over a very complicated and intertwined stack, so I can understand why it would take them longer.
Not sure about cloud providers, but several VPS companies already offer discounts for you to go IPv6 only.
Anyone know what IPv6 adoption is looking like these days? Google is showing like 3% of their connections are native to it [1], which does not look promising.
1: https://www.google.com/intl/en/ipv6/statistics.html#tab=ipv6...
http://www.worldipv6launch.org/measurements/
http://www.internetsociety.org/deploy360/ipv6/statistics/
Eric Vyncke and Lars Eggert have two good sites at:
https://www.vyncke.org/ipv6status/
https://eggert.org/meter/ipv6.html
Cisco 6Lab has a nice map at: http://6lab.cisco.com/stats/
10/8 is off limits, but there are a bunch of /8 blocks which could likely be reclaimed if we really needed to, including:
- About half a dozen which were allocated to the US military in the early 1990s, most of which aren't even publicly advertised at all
- 44/8, reserved for amateur radio
- A number of other /8 blocks owned by corporations, which could potentially sell or return chunks of them if properly convinced
Edit: I wish there was an updated version.
That's enough to find out that this particular one is called a Hilbert curve.
https://en.wikipedia.org/wiki/Hilbert_curve
This page has similar images from 2012
As for 44/8, there are still those that use it, taking it away would mean having to renumber all of the equipment on that network...
No, trying to reclaim IPv4 addresses by making people go through herculean efforts to renumber their networks is not the way forward and simply delays the inevitable. IPv6 is the way forward, start pushing that, and get more people using it.
Consider subcontractors needing to connect to servers in a particular department.
Ford has 181,000 employees as of 2013 (http://en.wikipedia.org/wiki/Ford_Motor_Company).
That's very roughly 92 publicly routable IP addresses per employee.
I highly doubt they need that many.
It would take years to arrange the return of 100 million addresses compared to the demand for a few billion addresses with even conservative growth.
Well before that, you could recover 18/8, MIT's prefix; how many addresses does a single university actually need, even MIT? Stanford already returned their /8 prefix. Or you could potentially recover a variety of DoD /8s, or start assigning some of the ones that are "reserved for future use" like 240/4.
Of course, an even better use of 240/4 would have been to use it as a backwards-compatible way to move to a larger address space. IPv6 should have been specified with IPv4 compatibility via NAT from the beginning, so that it could be a gradual migration. I mean, we're moving that way anyhow, but with a much more painful period in the middle in which NAT is necessary but IPv6 isn't ready yet for people who want to end the pain of NAT.
The 240 block is problematic because a huge amount of legacy routing equipment is hard-coded to blackhole it. The best thing for it is probably to reserve it for unspecified local use, which would allow you to use it to NAT IPv6 addresses into locally, or use as additional RFC1918-style address space, or as ORCHID-style address space for IPv4-only applications using the likes of cjdns, etc.
When will it finally be impossible to buy an IPv4 address?
Last time I tried to acquire a new IPv4, the distributor (my VPS host) required a technical reason for the request. If the request wasn't up to par, you would be denied an IPv4 + given an IPv6.
This was not at a particularly expensive provider.
What seems so wasteful is the "3 unusable" out of the 8 that are allocated. I have a vague understanding that those 3 are used for some routing/addressing stuff internally, but it's always felt incredibly wasteful.
This is all very fuzzy and some admins will look at you weird for doing this - but sometimes you've gotta do what you've gotta do.
The third address (router) is actually used by your ISP's router's to reach your subnet.
It is space that has been surrendered to IANA for re-allocation.
LACNIC hit that mark and triggered ICANN's policy to start allocating out of this recovered address space pool. The NRO has an announcement up about this, too:
http://www.nro.net/news/iana-allocates-recovered-ipv4-addres...
Basically, after this recovered pool of IPv4 addresses is allocated by IANA, there aren't any more IPv4 addresses to give out.
Am on phone so not read 100%
BTW, there is a nifty Chrome extension that tells you whether the site is doing IPv4 or IPv6. Always nice to see a green 6 in the address bar.
[0] https://chrome.google.com/webstore/detail/ipvfoo/ecanpcehffn... [1] https://addons.mozilla.org/de/firefox/addon/ipvfox/
http://en.wikipedia.org/wiki/List_of_assigned_/8_IPv4_addres...
Rapidly-developing Asia already has about as many people as there are IP addresses by itself.
Where are you going to find enough /8s, how are you going to recover them fast enough to do any good, and who's going to pay the massive legal bills from the ensuing litigation?
Step -1: We started running out of IPv4 addresses. The internet is IPv4-only for all practical purposes. Nobody is doing IPv6 because it's still experimental. Your weird neck-bearded friends are talking about it, that's about it.
Step 0: Inception! IPv6 is now a standard and you can route IPv6 over the tubes. Two networks have connected!
Step 1: Dual stack is a thing. Now you can set up both IPv4 and IPv6 on the same network and they co-exist. Some hosts (goole.com, facebook.com) try this and finally turn it on for good in 2012.
Step 2: Dual stack intensifies. This is where we are now. You cannot yet run a full blown IPv6-only network that you want to talk to the Internet and do anything practical with it, but all your cleanly shaven friends are now telling you that you should support IPv6 first, then add IPv4 for compatibility. Google sees as much as 3.5% of their traffic over IPv6.
Step 3: The cost of any single IPv4 address rises dramatically. For an end user it's already roughly $1/month. Buying in bulk is much cheaper, but not free like it used to be. ISP's and other service providers (Amazon, Rackspace, Digital Ocean) are going to start noticing the costs on their bottom lines. Severe restrictions are imposed by IPS's and service providers on anything but the bare minimum of IPv4 usage.
Step 4: The future. The first IPv6-only networks start to pop up. They use special network translation to talk to IPv4-only internet. (You can do this at home now! It's called NAT64: http://www.litech.org/tayga/).
Step 5. The number of IPv6-only networks is growing rapidly. IPv4 is now a legacy protocol. All your friends are telling you that they cannot believe we are still using IPv4 and how much of a pain it is to support it. Sales of T-shirts with "There is no place like 127.0.0.1" drop dramatically.
Step 6. The same people that put together IPv6 World Launch Day put together an IPv4 funeral. Major companies turn off IPv4 presence.
The crucial point here is not that someone demands IPv6 by calling their ISP, etc. It is fun to play with and you should at least get a tunnel set up through Hurricane Electric if you are reading this, but the demand here does not matter. What matters is supply: supply of IPv4 that is. Once that goes to nil, the cost rises and some executive at every ISP and service provider has a bright idea to save the company millions by switching everyone to IPv6. That's it. We, the people in the trenches, cannot affect this. We cannot speed it up and we cannot slow it down. The invisible hand of the market will make IPv6 happen all on its own.
Most people still can't actually route IPv6 over their tubes without going through a tunnel broker. At least for home, small office, and companies that aren't tech heavy, it's still firmly in the neckbeard camp.
What we have not quite reached yet is the point where IPv4 addresses are really expensive. You still get one with every VPS you sign up for and with every home connection. Once that can no longer be done, IPv6 will become standard pretty much across the board, starting in Step 3 and 4.
However, just checked last night and discovered that they'd added a new feature that would tell you when your CMTS was IPv6 ready, and it turns out that they've finally upgraded me. I could finally switch off my tunnel and use a real, routable address. Woo!
The nerdy t-shirt industry experiences a boost in margins, as they can now save ink by printing the much shorter "::1".
In IPv4, you have an entire /8 at your disposal, just for talking to yourself!
It is safe to assume that they won't be allocated in the future. You can easily assign them to your loopback interface. e.g. on linux:
You will then be able to use them over loopback just like ::1, and can populate your /etc/hosts file accordingly. [1] http://tools.ietf.org/html/rfc4291#section-4EDIT: formatting.
FTFY.
It also makes it hard for websites to block users based on IP without also blocking lots of legitimate users sharing that IP. I think this is a good thing to keep the internet more equal and anonymous.
"I have a different take on that. Yes, a lot of people were dissing tunnels yesterday. I'm actually a big fan of tunnels. I think the way the IPv4 Internet was built originally was basically by tunneling over the phone system. Ignore, you know, getting leased lines, putting computers on the end, building a network with no cooperation at all from the phone company. In fact, I once saw a phone company memo that referred to the Internet as a 'hostile overlay'. You know, and we basically, you know, demolished their billing plans and everything, and built this network in spite of the desires of the carriers. And, as I see the IPv4 Internet getting more and more like the phone company, I thought one way to deploy IPv6 is basically tunnel over all that junk. Iterate the same thing again. And I gather, you know, some people, you know, who are concerned about this are looking at other ways of sort evolving the Internet, where you, you know, tunnel over HTTP or whatever you have to do to basically get over all that cruft and build a new network on top of it. So that, to me, was one possible way, you know, as the IPv4 Internet rotted the underneath. We would tunnel over it and then we would throw away the rotted bits and replace them with wires."
[1] "Internet as a 'Hostile Overlay'" -> http://www.youtube.com/watch?v=mwRVNwa6nJc#t=11m07s
I really like that.
How are RBLs and other useful internet institutions that work on the basis of IPv4 addresses along the way with their IPv6 adoption?
It would take some time to adapt implementations. I speculate this gives attackers some opportunities to use old-style hacks that were largely mitigated, but suddenly re-appear with IPv6.
I figure the subsequent avalanche of user pressure on isps will solve the problem in a week or so. Added bonus: during that week productivity will be at an all time high.
I cant get ipv6 to work with this router, tho if i connect laptop directly to the wan cable i get an ipv6 address without issues and can use google via ipv6 etc
If a slightly unshaven geek like me cant get ipv6 to work easily what hope is there for rest of the world :(