IPv6 Achieves 50% Reach On Major US Carriers (worldipv6launch.org) 150
Long-time Slashdot reader dyork brings new from The Internet Society: IPv6 deployment hit a milestone this month related to the four major US providers (Verizon Wireless, T-Mobile USA, Sprint, AT&T): "IPv6 is the dominant protocol for traffic from those mobile networks to major IPv6-capable content providers."
A graph on their "World IPv6 Launch" site shows those carriers are now delivering close to 55% of their traffic over IPv6 to major IPv6-capable content providers -- up from just 37.59% in December. "This is really remarkable progress in the four years since World IPv6 Launch in 2012, and the growth of IPv6 deployment in 2016 is showing no signs of abating." In fact, the NTIA is now requesting feedback from organizations that have already implemented IPv6, noting that while we've used up all the 4.3 billion IPv4 addresses, IPv6 offers 340 undecillion IP addresses -- that is, 340 followed by 36 digits.
A graph on their "World IPv6 Launch" site shows those carriers are now delivering close to 55% of their traffic over IPv6 to major IPv6-capable content providers -- up from just 37.59% in December. "This is really remarkable progress in the four years since World IPv6 Launch in 2012, and the growth of IPv6 deployment in 2016 is showing no signs of abating." In fact, the NTIA is now requesting feedback from organizations that have already implemented IPv6, noting that while we've used up all the 4.3 billion IPv4 addresses, IPv6 offers 340 undecillion IP addresses -- that is, 340 followed by 36 digits.
In before... (Score:5, Funny)
Here, let's get the resistance out of the way:
"But, but, if we can't have NAT then we'll be h4xx0r3d! And I can't remember all those hex digits LOL."
Re: (Score:2)
ip6tables -t nat -- this NAT? There are good uses of NAT, although not what most people are thinking of.
Re: (Score:2)
You can NAT IPv6. Works just like NAT in IPv4.
As for address length, my public IPv4 network number is 15 characters long, whereas my IPv6 network number is only 13 characters long.
Re: (Score:2)
Re: (Score:2)
Actually RFC 6269 dis-endorses NAT. RFC 6269 provides the least worst form for those that irrationally just have to have NAT.
For reasons discussed in [RFC2993] and Section 5, the IETF does not
recommend the use of Network Address Translation technology for IPv6.
Where translation is implemented, however, this specification
provides a mechanism that has fewer architectural problems than
merely implementing a traditional stateful Network Address Translator
in an IPv6 environment. It also provides a useful alternative to the
complexities and costs imposed by multihoming using provider-
independent addressing and the routing and network management issues
of overlaid ISP address space. Some problems remain, however. The
reader should consider the alternatives suggested in [RFC4864] and
the considerations of [RFC5902] for improved approaches.
Re:In before... (Score:4, Informative)
No. PAT involves using port numbers to supplement IPv4 addresses, so that a hybrid static-dynamic NAT in IPv4 can get a 1:1 mapping b/w the local IP addresses and the external IP address coupled w/ the port number. That's one of the things that IPv6 eliminates, but it also removes things like load balancing or address isolation.
NPT - Network Prefix Translation - is different. It keeps the Interface ID unchanged - the part of the address that's not part of the network address - is not touched. Instead, a public unicast address is converted into a site local address, w/o touching port numbers. What you have is a 1:1 relationship b/w local and routable addresses. In PAT, what you have is a 1:1 relationship b/w a routable address coupled w/ one of 65536 port numbers to a local address. Which is a mess, b'cos if one needs ports for anything else (like map segments in a mapping application), one has to keep tabs on the ports used for PAT vs the ones used for the applications that need them
Re: (Score:2)
Re: (Score:2)
And I can't remember all those hex digits LOL
And THIS is the best thing about IPv6: it might finally stop enterprise IT teams and programmers from using IP addresses to access everything, rather than using their names. Because IPv4 numbers are easy to remember, it's tempting to use them in config files, command lines, and code. But this is a dangerous practice, considering that many IP addresses change assignments regularly, even if they are "fixed" addresses. I've seen entire VM clusters inadvertently wiped out by IT staff because they mis-typed a
Re: (Score:3)
It doesn't help when enterprise IT teams come up with DNS naming conventions that cryptically encode all the info about an asset into the name, and then apply that naming policy not only to desktops and laptops, but the servers that everyone needs to access, and steadfastly refuse to acknowledge the existence of CNAME records. The IP address is the easiest thing to remember where I work (there are only two
Re: (Score:2)
Also, as a server admin, having IPv6 open increases your traffic, not because more people are visiting but because a lot of bot nets are scanning IPv6 looking for vulnerabilities.
I'm very skeptical of this. What's the Venn diagram of "people who know what IPv6 is" and "people who think you can scan IPv6 space before the heat death of the universe"?
IPv6 early vulnerabilities proliferate (Score:2)
Re: (Score:2)
Many of the routers with ipv6 support are linux based, the linux ipv6 stack is quite mature already... /64 ipv6 space is impractical.
V6 also comes with some security improvements that v4 never had, like temporary privacy address and a huge address space - scanning an ipv4 range for targets is commonplace but scanning someone's
Also although v6 typically has fully routable addresses, all the consumer oriented routers i've seen block inbound connections by default so it's no worse than the default ipv4 setup w
Re: (Score:2)
Re: (Score:2)
Routers? The only routers here are high end IPv6 routers used in mobile networks. The majority of cheap home routers are still very much on IPv4 and those who aren't (like myself) have incredibly shit IPv6 support.
Re: (Score:2)
Re: (Score:2)
No I didn't make my point clear. IPv6 support in shitty routers is not attacked because it's not used. No one is running around actively scanning for open IPv6 connections, and by far the default configuration even if IPv6 is available is to use the IPv4 (carrier grade NATed) connection first. I have an IPv6 connection, but good luck actually talking to it. In the mean time a browser exploit through an ad network will give you a few millions of hits or so.
Malware is now a industry and follows the rules of c
Re: (Score:2)
Re: (Score:3)
You'll still see intrinsic difficulties that aren't there for V4. For example, if I set my AP wide open, you'll have all kinds of fun finding the 5 out of 4 billion addresses in my prefix that have anything on them.
I suspect malware will continue more or less as is in the form of drive bys and trojans. v4 or v6 won't matter much. The router won't matter much.
Re: (Score:2)
You'll still see intrinsic difficulties that aren't there for V4. For example, if I set my AP wide open, you'll have all kinds of fun finding the 5 out of 4 billion addresses in my prefix that have anything on them.
There are some new problems that didn't exist before too. Using the example above one of them is now external actors spamming a /64 results in ND broadcast transmissions of router asking network if anyone matching spammers request is home. Given /64 is essentially infinite for purposes of response caching this can negatively affect available bandwidth between systems on switched networks and eat away at batteries of mobile devices connected via wireless Ethernet.
Re: (Score:2)
True, but easily defeated at the firewall.
Re: (Score:2)
And that's why I'm suggesting that the proliferation of it will make it a sweeter pot
That was the second part of my point. The proliferation that we're seeing now is not new home networks but rather carrier grade routers in mobile towers used to cope with 1.5 billion smart phones that have been added.
One would hope that someone with a brain programmed the IPv6 implementation on those rather than the lowest cost H1B import from India that seems to be in charge of home routers.
Re: (Score:2)
What do you mean? (Score:2)
What kind of vulnerabilities do you think would exist in IPv6, but not IPv4?
Re: (Score:2)
Not progress (Score:2)
This isn't progress at all. We've done little to nothing to move people to IPv6. The only problem is that we've run out of addresses and the easy solution to adding millions of smartphones was IPv6. The majority of home connections are still IPv4 and the majority of ISPs still only offer this.
As is true with all human nature where a profit centre is involved, we won't make "progress" until we're absolutely forced to.
Re: (Score:2)
There has been quite a lot of progress in residential broadband too. The "Networks" tab of Akamai's IP adoption visualization page [akamai.com] shows Comcast at 44%, TWC at 22%, and Sky Broadband at 53.5%, alongside the mobile carriers moving to IPv6.
The smartphone migration is also progress as it has helped to remove the old chicken-and-egg problem for IPv6. Why should websites take the effort to support IPv6 when the eyeballs aren't there? Well now the IPv6 eyeballs are there, and there's a lot of content for them:
Re: (Score:2)
It is a good first step, however. Everybody was never gonna move to IPv6 at the same time, so it's good that the carriers - the main area where the growth has been - have adapted them in such a big way.
As far as the broadband providers go, they do need to get moving. At Comcast, I have IPv6 at work - the Comcast Business (from my look at it, it seems to be dual stack lite or maybe dual stack - when I run IPconfig, I don't get a public IPv4 address) but at home, there is no IPv6. The default settings on
IPv6 deployment is not a switchover (Score:3)
What you say is not wrong, but many people will interpret it incorrectly as suggesting that there is a "switchover" from IPv4 involved. That's not how IPv6 was designed and planned at all. IPv6 was designed right from the start to run alongside IPv4, and "migration" or "transition" are poor words for what will mainly be an expansion of IPv6 use, and it
Re: (Score:2)
I understand that the changeover from 4 to 6 has to be gradual, and I suppose the fact that all the new cellphones are using IPv6 is significant. Still, I wonder if we will ever be able to shut off IPv4 in home installations -- or on phones. Realistically, we can't do it until every server out there supports IPv6.
With Comcast service, I am now fully dual stack, and it's nice to see more of my traffic using IPv6. But there have to be extra overhead and security issues when running two IP systems compared
Re: (Score:2)
You can mostly run a NAT64+DNS64 network with no native v4 right now -- the only problem with it is v4-only client software (not v4-only servers). And even that could be fixed by client OS support for 464XLAT or some sort of automatic mapping of v4 sockets into a v6 prefix (which is something that everything should've supported years ago but unfortunately doesn't look like it's ever going to happen).
Re: (Score:2)
That's not how IPv6 was designed and planned at all. IPv6 was designed right from the start to run alongside IPv4, and "migration" or "transition" are poor words for what will mainly be an expansion of IPv6 use, and it may have very little early effect on IPv4.
Indeed. My point is that IPv6 is now legally old enough to vote in the USA and yet the most recent router I received from my ISP still doesn't have support. Based on the useful life of even industrial grade gear the entire world should be at IPv6 by now.
Instead we've done little. A crumb or two here and there, and a growth so slow and painful that it makes you wonder if it's actually moving at all. In the mean time all those lovely adoption figures are for new technologies like the mobile market where someo
What the fuck are you whining about? (Score:2)
What do you mean we've done nothing to move people to IPv6? Do you think it is magic? Do you think we just wave a wand and people are on v6? No, what it takes is rolling out support on the OS, router, ISP, and so on. That has been happening, lots. Have a look at Google's IPv6 chart: https://www.google.com/intl/en... [google.com] what you see is exponential growth happening. This is actual IPv6 connections as well, Google is counting the percentage of people hitting their site with v6, which means an end-to-end connectio
Re: (Score:2)
What do you mean we've done nothing to move people to IPv6? Do you think it is magic?
Yes. It should have been magic. IPv6 is now 18 years old. Think about that for a moment. You could have had a child and raised him to an eligible voter in the time IPv6 has been around. How many routers did you replace in that time? 3? 5? I probably would have gone through around 4 with my jumping between ISPs. The most recent of which was last year. Guess what my router does NOT support.
That has been happening, lots. Have a look at Google's IPv6 chart: https://www.google.com/intl/en [google.com]... [google.com] what you see is exponential growth happening.
And thus you missed my point. People haven't been moved to IPv6. People have been given new devices on new networks which
Re: (Score:2)
No, that's not the approach you take. If you think it is, well you need to grow up. You don't cause massive compatibility problems and huge disruptions just for the fun of it. Instead, you do things as smoothly as possible. There is no need to rush out IPv6, it isn't like the world will blow up. IPv4 works, and will continue to work.
You thinking that implementing something like this on a worldwide scale being cheap, easy or quick just shows a massive lack of experience and perspective.
Re: (Score:2)
What the fuck are you talking about. I was agreeing with your last point, just pointing out that it isn't actually happening.
Grey Goo Limit (Score:2)
I recall a joke scenario from a couple years ago:
Earth is in the throws of a Nanotech Grey Goo scenario. The microscopic self-replicating robots have converted about half the planet to more of themselves. And then they stop. The few surviving humans, observing from space, are puzzled.
Zoom in. Thought balloon from the mass of Grey Goo: "Damn! We shouldn't have stuck with IPV6. We've run out of addresses!"
Re: (Score:3, Informative)
You recall xkcd 865 [xkcd.com].
Re: (Score:2)
You recall xkcd 865.
Yep, that's it. Thanks.
In the meantime Canada ISPs are behind (Score:4, Interesting)
Still frustrated that the ISPs in Canada are still lagging on getting IPv6. The biggest failing ISP is Bell, with no publicly announced plans.
There has been the "Call Your ISP for IPv6" campaign by the guys over at Sixxs:
https://www.sixxs.net/wiki/Cal... [sixxs.net]
Re: (Score:2)
Telus is offering native ipv6 as well. Teksavvy, by extension, is also offering native IPv6 when using Telus as the carrier. Shaw is still stuck in ipv4 land though, which prevents Teksavvy from offering ipv6 on those links.
Re: (Score:2)
Which US ISPs? (Score:2)
Cox is dual-stack on their entire network. Comcast is likewise. Time Warner is about 90% done with IPv6 on their network. That most of the US's cable providers right there, with Charter being the only major that doesn't have IPv6 yet and they are working on it actively.
Not every ISP has it, of course, when you count DSL CLECs, dial up, and so on there are literally thousands of ISPs in the US. However it seems that most of the major cable providers do, and combined those guys serve a massive part of the US
Re: (Score:2)
I have Comcast. Like I said above, at work, we have a Dual-Stack Lite or a Dual-Stack setup from Comcast Business. But at home, I don't have IPv6. I'm talking about the defaults Comcast gave, w/o me saying a word.
I had Charter in Atlanta a year ago, and TWC in Charlotte a year before that. Both of them had pages that described IPv6, but vaguely spoke about their plans. But in both these cases, I tested IPv6, and got it on neither. If TWC has it, it has to be more recent: it certainly wasn't there in
Re: (Score:2)
I can't speak authoritatively to Comcast, not having it, but everything I see says they have dual-stack on their entire residential network. Have you tried it? You have to set up DHCP-PD on your router (that is how most ISPs are doing it) and they should give you a prefix that your devices can use.
Re: (Score:2)
Which isn't surprising, actually, because I believe LTE, besides eliminating pure voice support (LTE is data-only), LTE also has NO support for IPv4. That's right, LTE is forward-facing and IPv6 only. Of course, most people want to hit IPv4 sites, so there are mechanisms that get you over - like IPv5 to IPv4
Re: (Score:2)
We have more land per person in the world, why shouldn't we also have more IP addresses per person?
Internet royalty walks among us (Score:5, Interesting)
Human (Score:2)
Unfortunately, and as far as I can tell, I am either a human or a holographic projection with limited storage capacity. I need IPv4 cause I can't memorize an IPv6 address. Seriously, who can remember an address like 2001:0db8:0a0b:12f0:0000:0000:0000:0001 .. you have got to be kidding me
Re: (Score:2)
Have you actually used v6? It's not really that hard. For starters, that address is 2001:db8:a0b:12f0::1. (Why did you write it with all the extra zeros?) Secondly, let's compare the v6 case with the inevitably-NATed v4 case:
2001:db8:a0b:12f0::1
vs
192.0.2.215+192.168.189.1
So, v6 is shorter. If you have trouble memorizing v6, then you should be having even more trouble with v4.
I'd also like to introduce to this wonderful thing called DNS [wikipedia.org] that eliminates the need to remember most addresses. It's a pretty matur
Re: (Score:2)
No idea what you're trying to achieve with your IPv4 plus sign there. 2 separate addresses? You remember them as 2 separate addresses, depending on which side of the NAT you're on.
One on either side of the NAT, yes. And v6 is basically the same: it's the prefix (2001:db8:a0b:) plus the subnet and host (12f0::1).
Try memorising a SLAAC address or any other autoconf.
Yeah, obviously you don't do that. This is what DNS is for.
Try building out an infrastructure provider PD hierarchy [...]
Ultimately, it doesn't matter what you want here. v4 is too small for the internet, v6 is the replacement. As an ISP it's your job to deal with it. And it won't be as bad as you think it is.
Re: (Score:2)
Re: (Score:2)
Check `ipconfig /all`. Or I can tell you it's 2001:db8:420::53 because you deliberately picked a short address for it, because why would you pick something long and unrememberable like 2001:db8:420:f4ca:c6fb:d174:620e:37f9 for the one specific machine that you have to remember the IP for?
Re: (Score:2)
Check `ipconfig /all`.
Your servers run windows?
Your servers use DHCP? I mean, some of them, yes. But some of them... no
Re: (Score:2)
No, I just figured that GP probably did, since any Linux sysadmin should already know how to look up which resolvers their system is using.
DHCP seems somewhat orthogonal here.
Wastage (Score:2)
What's clear is that huge swaths of the address space will be wasted by being bought up, monopolized, misallocated, and overused. I expect us to functionally exhaust the IPv6 space within a decade or two.
Re: (Score:2)
Really? it's _already_ happening.
For example sprint owns 2600:: - 2600:7:ffff:ffff:ffff:ffff:ffff:ffff ..which means they own billions of trillions of addresses .. to be clear .. sprint owns 633,825,300,114,114,700,748,351,602,688 ipv6 addresses. They don't have that many customers. Sprint is just one example. Similarly there will be a few hundred ISPs that will grab vast amounts of the address space an sub-allocate the addresses over-generously. It's not exactly easy to take it back especially if the numb
Re: (Score:2)
1. Sprint, a major ISP, has 2600::/29 - two billionths of the possible IPv6 addresses
2. ????
3. We're doomed! Somehow.
You should show your math for running out in 20 years. That takes a lot of /29s (five hundred million). Also remember that end users can get a /48, which is 524,288 times smaller than a /29, or /56, 256 times smaller again.
Re: (Score:2)
v6 has a lot of addresses. There's no point counting the IP addresses somebody has, because the answer is always "too many". And surely that's a good thing? Would you rather people had too few addresses instead?
Sprint's block is not overly generous -- in terms of overall consumption of the v6 space it's like allocating them 8 IPs in v4 space, to cover 60 million customers. We'll be fine, and even if we aren't and we somehow manage to run out of space in 2000::/3, we can break into the 5 other reserved /3 bl
Re: (Score:2)
The range of what's been allocated has been cleanly shown on IANA's site [iana.com]. 2400 is assigned to APNIC, 2600 to ARIN, 2800 to LACNIC, 2a00 to RIPE and 2c00 to ARFINIC.
So it's ARIN that has 2600::/12, and they allocate downstream. From that box, Sprint has been given something lesser i.e. the number you see after the / for Sprint would have to be something above 12.
So whatever number you see after the / - let's say it was /24, subtract that number from 64, since the bottom 64 addresses are the Interface I
Get with it cloud providers. And network providers (Score:2)
Every time I see a "new big features" announcement from the big 3-5 cloud vendors (AWS, Google, Azure, etc). I keep hoping that one or the other is going to really buy in to IPv6. And I keep being disappointed.
There are some ways to get them playing moderately nicely with IPv6 (especially if you're buying load-balancing services from them), but most of their networks are IPv4 internal-routing subnets.
Meanwhile, the middle range VM places (Linode, DigitalOcean, etc) are far more IPv6 friendly. My understand
Re: (Score:2)
> 's amazing IPv6 has as much traffic as it does.
It's really not been necessary. I've not seen a single business or service provider failing to find, or provide for its customers, some IPv4 space to host their services, even if it's a name based proxy. Can you think of or find a single commercial service whose IP addresses are only IPv6, without any accompanying IPv4?
Re: (Score:2)
That's because those businesses are paying extra money to continue to support v4 -- which is of course being passed straight on to their customers.
Would you rather have waited until companies were being bankrupted by the need for v4 support until we did anything about it? (Because it sure seems like a lot of people would...)
Re: (Score:2)
insanity like the Cogent-v-Hurricane split of the IPv6 internet (holy crud... it's SEVEN years now since Hurricane baked Cogent that cake begging them to peer with the world's largest IPv6 network... and it's still broken),
It's irritating that those companies care more about interconnection politics than about serving their customers but I don't think it's that important in the grand scheme of things. Decent hosting providers are usually multihomed and thus reachable from both HE and Cogent.
equipment... (Score:2)
IPv6 often is faster to address and has been better monitored however
end user equipment that route's is lacking for example google OnHub is not IPv6 compliant
( https://on.google.com/hub/ )
whats the process for certification ?
thanks
John Jones
I had to switch IPv6 off (Score:2)
I had to switch it off. All of a sudden Netflix decided that my registered tunnel with my own IPv6 subnet was an indication of me not being in the place I was supposed to. So netflix just stopped working. (I'd cut them off by that point, but the rest of the family didn't see it that way...)
So the final and workable fix was to switch off IPv6 on my internal network. Now it's only my gateway that is v6 routable.
Talk about "giant leap for mankind" backwards. Thanks Netflix. (Or rather "MPAA" I guess.)
Re: (Score:2)
I've got IPv6 Dual stack here and it works fine with Netflix. The netflix traffic is using IPv6.
Re: (Score:2)
I've had country blocks (most of Asia, Africa and Russia) for some time now and have seen a dramatic drop in crap that's trying to crawl up the tubes from those countries. Mainly automated out-of-the-box spam bots and hacking scripts (I was shocked to discover e.g. that w00tw00t is still a thing). This takes off some of the unnecessary load in the backend, too, and these countries are the ones that usually spam the Internet with fresh exploits, so that's a better-than-nothing first line of defence also.
Re: (Score:2)
Re:is that math correct? (Score:5, Informative)
IPv6 = 256^6 = ... 340x10^36 ???
Not sure if bad attempt at joke or not, but in case it isn't: the 6 in IPv6 isn't the number of octets used in the addresses, it's a version number. IPv6 uses 128 bit addresses, and 2^128 = 3.4 * 10^38.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
An individual IP should be used by one machine. It'd be more useful to track the number of networks/subnets (where each subnet gets a /64). But houses can have more than one network, so they should be getting more like a /56 (with companies getting /48 or so for their bigger networks).
The numbers are still crazy though. 10 million houses per person? We should be able to keep under that.
Re: (Score:2)
The correct calculation is 2^(2*2^4) vs 2^(2*2^6))
There are 5 trillion /56 blocks (Score:5, Interesting)
IPv6 has five TRILLION /56 blocks.
There are enough /64 to give every person on earth 2,635,249,153 of them.
128 bits allows for HUGE numbers.
Long ago, when we were developing IPv6, I was part of the group who argued for 128 bit addresses rather than 64 bit. I've decided I was wrong. 64 bits would have been more than enough, and could be processed on 64-bit processors, in standard databases, without hassle. Since my side won the argument, we have 128-bit addresses, which are so big they are a pain in the ass in Microsoft SQL Server and elsewhere.
Re:There are 5 trillion /56 blocks (Score:5, Insightful)
Re: (Score:2)
and could be processed on 64-bit processors, in standard databases
It may surprise you that 64bit processors don't limit your ability to work with numbers higher than that.
Okay you knew that already but the world is built around needs and use cases. The use case for 64bit processors came out of memory limits just like the use case for increasing the IP addresses came out of its limits too. If you need a 128bit processing ASIC to get the performance you require then they will appear on the market, just like bitcoin number crunchers did.
The world is full of applications inhe
Re: (Score:2)
> It may surprise you that 64bit processors don't limit your ability to work with numbers higher than that.
Larger numbers create an additional storage, memory, and data access cost at some very deep layers of the stack. That cost is, in fact, a profound limit on the ability of network software, and hardware, to operate under load.
Re: (Score:2)
Larger numbers create an additional storage, memory, and data access cost at some very deep layers of the stack.
All of which is offset by being able to intelligently route by using larger blocks. One of the fundamental problems right now is that routing tables are huge and growing exponentially not due to an increase in the number of devices, but a fragmentation of address space.
Re: (Score:2)
So, IPv6 does not support private ASs and (more importantly) addresses owned by private companies?
Currently, the company I work for has its own AS and its own IPv4 subnet. We can use whatever ISP we want and still keep our IPs (we use two ISPs for redundancy and are able to change the ISPs to new ones if we need to). Would this be not possible with IPv6?
Re: (Score:2)
And the company continues to use that AS with IPv6. AS's are independent of IPv4 and IPv6.
Your company should just request a /48 per site from the RIR's. You already qualify for IPv4 so you qualify for IPv6. The cost is max(IPv4 cost, IPv6 cost).
You can do it the same, or 1:1 nat (not PAT) (Score:2)
You can get an IPv6 assignment:
https://www.ripe.net/publicati... [ripe.net]
You also use the opportunity to no longer need to work with the next ISP to have your addresses routed by using one-to-one NAT (not the far more commom port address translation, which is yucky). With one-to-one NAT, each machine still has a seperate IP, you can just map the network prefix from FF08:x to BEEF:x or whatever at the router. You can change ISPs instantly in an emergency.
Re: (Score:2)
We can change the ISPs pretty much instantly now. We just change the priorities and prepends on our BGP router and traffic now goes through the new ISP.
Re: (Score:2)
Re: (Score:2)
You mean a non-routable address space for internal use only, becuase your IP addresses are really no one else's business? See http://www.networkworld.com/ar... [networkworld.com] and a dozen other articles like it about private IPv6 address spaces.
1000% performance penalty on Ivy Bridge (Score:2)
64-bit CPUs *can* process 128-bit numbers, or anyway they can run code that emulates it. And it takes ten times as long compared to using native 64-bit types. Your mileage may vary, of course, but that's one benchmark on an Ivy Bridge - a 1000% performance penalty.
Actually try working with 128-bit numbers, IPv6, in common software like SQL Server. There IS no 128-bit unsigned number in SQL Server. You *can* jack around binary types, I have. It's a time-consuming pain in the ass. Speaking of databases, you
Re: (Score:2)
And it takes ten times as long compared to using native 64-bit types.
Depending on operation it should take twice as many calls.
in common software like SQL Server
Border gateway routers do not run SQL Server.
Actually they don't run Ivy Bridge platforms either. None of what you say and your talk about CPUs is at all relevant to moving our data around on the internet except for maybe a microscopic penalty at the end point which now in addition to having to serve the content to the client needs to add a few bits to the packet.
We wish (Score:2)
>> And it takes ten times as long compared to using native 64-bit types.
> Depending on operation it should take twice as many calls
Figure out how to manage that and I'll make us both billionaires. Maybe you'd care to demonstrate by showing us how you can two add 4-bit numbers using 2-bit operations.
Are you under the impression that border routers are the only thing that ever sees IP addresses?
Re: (Score:2)
Already, IPv6 consists of 2 parts - the global prefix, which is assigned from IANA right down to the subscriber, and then the interface ID, that is either autoconfigured or can be assigned using DHCPv6 or manually. Treat those 2 things separately, but when forming the address, couple/concatenate the global prefix w/ your interface ID. Job done.
Unless you are busy playing w/ things like loopback address ::1/128, or home ::/128
Might save me a lot of time, except SQL is signed (Score:2)
You got me thinking. You're right,
If SQL had a 64-bit unsigned int, I'd use a pair of them. Alas, it doesn't. Postgres has an IP type which works, but my design has to work for SQL server. On the other hand, Microsoft SQL server does have decimal type, numeric. Hmmm ..
On the third hand, the idiot before me decided to store 32-bit integers (ip addresses) as four seperate bytes, in four separate columns (in some tables). That's pretty silly. So when rewriting it to handle IPv6, my first step would be to b
Re: (Score:2)
IPv6 has five TRILLION /56 blocks.
There are enough /64 to give every person on earth 2,635,249,153 of them.
128 bits allows for HUGE numbers.
Long ago, when we were developing IPv6, I was part of the group who argued for 128 bit addresses rather than 64 bit. I've decided I was wrong. 64 bits would have been more than enough, and could be processed on 64-bit processors, in standard databases, without hassle. Since my side won the argument, we have 128-bit addresses, which are so big they are a pain in the ass in Microsoft SQL Server and elsewhere.
You made the right call!
64-bit addresses would have made sense only if the idea had been to just physically extend the addresses, as opposed to starting from scratch given everything that had been learned about networking in the previous decades. 128 bits make sense not b'cos of any physical limitation, but rather, numeric structural ones. As I discussed elsewhere, once one tries packing hierarchical meanings into the addresses, they either have to grow, or be grotesquely complicated.
The 64:64 split
64 allows 2 billion IPs per person. 2GB limits (Score:2)
> Is the Microsoft SQL Server thing the only reason why you think 64-bit would have been better?
SQL server is one example that 64-bit software, on 64-bit computers, natively handles 64-bit numbers, while 128 bit requires gymnastics.
Generally, I think 64 bits would have been more than enough. It would have allowed us to assign 2 billion addresses to each person. :) Not that we'd actually do that, obviously. We would have done perhaps 256 addresses (8 bits) for most end users, while reserving 80%-90% of t
Re: (Score:2)
Or make custom CPUs just for routers. They don't need to use Atoms: something like a customized 128-bit MIPS would do the trick. And such a CPU would just need those instructions needed by a router, and could get rid of everything else.
In normal computing, I doubt that we'll ever even get to surpass 64-bit computing, where we need more than 2^64 bytes of memory that need to be addressed. If you addressed all the RAM, Flash and Hard disk storage, I still doubt you'll exceed 64 bits.
Re: (Score:2)
I think overkill was the right call. I'm not convinced that 64-bits would be sufficient for everybody to get away with NAT indefinitely. I think it might be, but even if so I think realistically ISPs would've given allocations that were too small.
Case in point: ISPs giving /60s or even /64s in 128-bit v6, even though they easily have enough space to do /48s. In a 64-bit v6 world, that would probably translate into people getting 256 individual address or so, which technically is enough for "most" people tod
Re: (Score:2)
Either way, I'm waiting to hear about impending IPv6 exhaustion.
Your going to be waiting a while as just 1/8th of the total address space is currently in play. If things unexpectedly go off the rails there is opportunity for IANA to reign it in with policy changes for allocation from remaining 7/8's.
decide that giving out /56's to everyone calling themselves an ISP wasn't such a good idea
It's more like /32 or more... We pull a /56 from our ISP. A so-so rule of thumb for understanding allocation difference between IPv4 and IPv6 is every "ISP" is allocated IPv4 address space equivalent of a single IP address. On order of a billion ISP like allocations and y
Re: (Score:2)
The GP was dismissive about address exhaustion, but there are not just possible, but plausible reasons why that would happen.
Currently IANA has already allocated a number of /16 blocks to the RIRs, and it's up to them to allocate it as they wish. While ARIN has been assigning blocks like birdseed (the way Jon Postel did in the early days of IPv4) downstream in /48s, RIPE and APNIC have been more conservative, and assigning them in /56 blocks.
The way address exhaustion is likely to occur is not distribu
Re: (Score:3)
You're right that hierarchy in address allocation increases the address space requirements (you could say "wastes" addresses, but they're not wasted, they're being used to reduce routing table fragmentation). But it doesn't follow that v6 exhaustion is actually plausible as a result. v6 is really damn big (for precisely this reason!) and we're only allocating from one /3; over 60% of the total space is outright reserved at the moment.
Your example allocations are all shifted to the right a bit. ISPs are gene
Re:is that math correct? (Score:4, Informative)
While I do think IPv6 addresses are wasted, I don't think the wastage happens in the global prefix. Rather, it happens in the Interface ID area. 64 bits is way overkill for an interface ID, since no subnet will have anywhere close to even 4 billion nodes, much less 2^64. In the meantime, in the global prefix side, things are squished w/o getting into a hierarchical routing, which would have been a real godsend, had it been implemented.
Yeah, my example probably shifted things, due to my use of documentation's 2001:db8::/32, and your scenario would be the more likely one. Still, I believe that the upper 4 words should have been strictly global prefixes, not including subneting: it should have been used to hierarchically route from IANA to RIR to country to ISP to organization/families/individuals. At a gateway, allow for either a /64 or a /96 (w/ 4 billion subnets), depending on the need. So if a home router needs 2 network addresses for 2 SSIDs, either get 2 /64s or subnet 1 /64 into /96s.
The reason that is given for the 64 bit interface ID is auto-configuration, but that's a lame excuse. First of all, even w/ 64 bits, it's unlikely, but not guaranteed that there won't be an address conflict, and at any rate, there is DAD to resolve that in IPv6. Then, using things like MAC addresses or IMEA numbers to obtain these addresses creates a potential for spoofing agents to deduce a target address, assuming that a network wants those things either hidden, or difficult to find w/o initiating from the user's end. 32 bits would easily have been enough for any subnet - even in the most crowded spot in Guangzhou, I doubt that there would be anything close to 4 billion devices that would be under a subnet, and even if there was, that network would grind to a halt w/o redundant APs, repeaters and other signal enhancing agents. So the IETF could have designed IPv6 to have the first 64 bits strictly the global prefix, then either have a 16:48 or 32:32 split b/w subnets and nodes.
Re: (Score:2)
Currently IANA has already allocated a number of /16 blocks to the RIRs
Actually they allocated them /12s . There are also some smaller older allocations. So currently less than 6 /12s of global unicast space have been allocated.
Currently IANA has already allocated a number of /16 blocks to the RIRs, and it's up to them to allocate it as they wish. While ARIN has been assigning blocks like birdseed (the way Jon Postel did in the early days of IPv4) downstream in /48s, RIPE and APNIC have been more conservative, and assigning them in /56 blocks.
The standard allocation for an ISP is generally a /32, they then suballocate to customers in smaller chunks (/56 is currently considered best practice as a default allocation for small customers).
The way address exhaustion is likely to occur is not distribution (for obvious reasons) but rather, lending structure to those addresses. While route optimization seems to have been abandoned for now
Mainly because the Internet is NOT a network with a strict and static heirachy, it's a network of private companies involved in constantly shifting relationshi
Re: (Score:2)
But it is a two-way street. it isn't just about the phone being able to route v6, it is also about destinations being v6 as well. So in reality, this is more about destinations than the carriers themselves (which at least tmo has had v6 enabled this entire time for years now)
Re: (Score:2)
I have Cox HSI in Las Vegas. Prior to Cox turning their native ipv6 on a few months ago, I had a Hurricane Electric 6to4 tunnel. When Cox enabled ipv6 here, I tried 5 different times to switch to their DHCP6-PD configuration. The connection would work fine for a few days, then I'd lose v6 connectivity. Usually a reboot of the router would bring it back for a while, it would drop out again.. I tried pinging their tech-support but their first-line support is totally useless for advanced issues like this, and