Black Market May Develop For IPv4 Addresses

GMGruman writes "Everyone knows that we're running out of traditional IPv4 Internet addresses and that switching to IPv6 is the answer — yet foot-dragging by IT departments and vendors means the problem is still on the back burner. IPv4/IPv6 coexistence is now expected to last for 5 years. In this article, Mel Beckman explains how this is all leading to a black market in traditional IPv4 addresses that will catch many people off-guard, and boost Internet access prices sky-high."

Comments

Public IPs at premium prices

[DigiShaman]

You mean they will start NATing more often for residential customers. Long gone will be the default of having a dynamic Public IP address. Want one of those? That will cost extra.

Re:Public IPs at premium prices

[wish bot]

Double NAT, double good!

P.S. 10.1.x.x going cheap. Mail me!

Re:Public IPs at premium prices

[Anonymous Coward]

Except places like Slashdot, they won't let a sibling NAT address do anything useful.

Re:Public IPs at premium prices

[Mordok-DestroyerOfWo]

"What would you do if you had a million dollars?"

"I'll tell you what I'd do, man: two IPv4 addresses at the same time, man."

Re:Public IPs at premium prices

[youn]

my provider uses 10.* addresses there therefore I had to change the addressing scheme on my LAN because I didn't want stuff routed accidentally... pisses me off... I'm sure more and more providers will do that... and that's accidents waiting to happen.

Ahhh, I long for the days when a private address was garanteed to be private. why don't they switch already to ipv6... it's been 15 years. I know it will irk some people but it's stable enough and it's about time... and as time passes it's going to get harder because people will be more dependent on the internet.

Most OSs & routers are compatible... I say it's time to require the change... give people a 6 month warning and switch... should be plenty of time to address most issues

Re:Public IPs at premium prices

[pspahn]

We seemed to make a relatively painless procedure of the digital TV switch. Or maybe I never noticed since I have been using internet for TV for so long. I don't even know how to set up a regular TV anymore.

Re:Public IPs at premium prices

[optikos]

Not quite. Numerically, you will still have the same "public" IPv4 address that you have today (either dynamic or static). It is just that it will be like that _Star Trek NG_ episode where, upon hearing something munching on the Enterprise's hull, Dr. Crusher asks the ship's computer "What is the nature of the universe?" to which the answer comes back "The universe is an oblate spheroid one kilometer in diameter." In the IPv4-lives-on-forever world, "public" will be redefined to "among all of the subscribers of the same ISP" (not "worldwide" anymore). Then *all* IPv4 addresses (other than loopback and test ranges) will be NATed between ISPs/carriers. In other words, there will not be one Internet address-space anymore, but rather one IPv4-sized address space per ISP/carrier/telco. The goal is to carve the single Internet up into multiple per-telco Internets with interworking at the telco-to-telco or ISP-to-ISP boundary. There will be the AT&T Internet and separately the Verizon Internet and separately the Deutsche Telekom Internet and so forth.

Re:Public IPs at premium prices

[feepness]

It is just that it will be like that _Star Trek NG_ episode where, upon hearing something munching on the Enterprise's hull, Dr. Crusher asks the ship's computer "What is the nature of the universe?" to which the answer comes back "The universe is an oblate spheroid one kilometer in diameter."

Dude, there was nothing munching on the hull, people were disappearing because she had dropped into a collapsing warp bubble of her own private universe!

CRUSHER: Computer, what's happening?
COMPUTER: Explosive decompression decks five through fourteen. Sealing off forward sections.
CRUSHER: Cause?
COMPUTER: A flaw in the ship's design.
CRUSHER: Show me. Analysis.
COMPUTER: No ship's structures exist forward of bulkhead three four two.


God!

Re:Public IPs at premium prices

[ls671]

> my provider uses 10.* addresses there therefore I had to change the
> addressing scheme on my LAN because

Why did you chose 10.x.x.x for your LAN in the first place ? I doubt that you are planning to connect 16,777,216 machines to your LAN ;-)

Guide lines are to use:

192.168.X.X if you need 65,536 IP addresses or less

172.16.X.X-172.31.X.X if you need between 65,536 and 1,048,576 IP addresses

10.X.X.X if you need between 1,048,576 and 16,777,216 IP addresses

Routing is slightly faster with more bits in your netmask. Although I do not think that you will notice a difference especially nowadays. I think this was one of the reasons for these guidelines. Following the guide lines also ease connectivity to bigger nated networks, your provider in your case.

Re:Public IPs at premium prices

[DavidRawling]

Because every damn organisation I VPN to uses 192.168.x.x addresses, or 172.[16-32].x.x addresses. By using a 10.x with a 24 bit mask I can use space that doesn't route, doesn't conflict with the orgs I VPN to, and that doesn't require me to reconfigure.

Re:Public IPs at premium prices

[plan10]

Why not use 10 if it's a home network? The point of private reserved is that it should be, you know, private.

Re:Public IPs at premium prices

[JWSmythe]

    How are they doing it? Can you hear the entire 10.0.0.0/8 network, or are they simply providing a single NAT address for internal use (i.e., on your final point to point link).

    I'm actually surprised that more residential providers haven't put the majority of their customers on NAT's, and providing public IP's as the exception, not the rule.

    I'm annoyed when some SEO "expert" wants a dozen /24's to stick on a handful of boxes because "the search engines will know I'm link spamming", but it's even more annoying to know millions of IP's are wasted because the providers don't want to conserve address space.

Re:Public IPs at premium prices

[Anonymous Coward]

"I'd buy a complete /28 subnet for my children!"

Re:Public IPs at premium prices

[KiloByte]

10.X.X.X if you need between 1,048,576 and 16,777,216 IP addresses

Or if you want to have any sanity in your VPNs.

Like, 10.X.X.*/24 is a physical network, 10.X.*.*/16 are networks at one localisation, 10.*.*.*/8 are all connected private networks together.

Re:Public IPs at premium prices

[nmg196]

10.* addresses are not really for home use - so I'm not really surprised you ran into conflicts. Home routers tend to use 192.168.x addresses for that reason. 10.x addresses are really for where you conceive you might need hundreds of thousands of IP addresses on your own private network (eg, you're probably an ISP or a university campus).

Re:Public IPs at premium prices

[hankwang]

"What would you do if you had a million dollars?" -- "I'll tell you what I'd do, man: two IPv4 addresses at the same time, man."

Not even so unrealistic if the US dollar starts to hyperinflate a due to the national debt. ;-)

Re:Public IPs at premium prices

[T_Tauri]

I use 10.a.b.c because almost nobody else uses it! When trying to create VPN tunnels to one of the several other companies I have to connect to I got fed up with them saying "Oh but we use 192.168.1.something for our [servers | LAN | DMZ ] etc. Can't you use something else?" By choosing an obscure 10. subnet for my network I have avoided these problems and can setup firewall to firewall VPN tunnels to the people supporting our database, phone system, network support, etc without worrying about routing.

Re:Public IPs at premium prices

[dotgain]

The big telco in the country I live (NZ) shipped routers with a default IP of 10.1.1.1/8

What a nightmare!

Re:Public IPs at premium prices

[dotgain]

That's what I use 172.16-31.x.y for - some people don't even know about that range!

Re:Public IPs at premium prices

[Rogerborg]

Calm down, dear. It's TNG we're talking about - a fictional show. Now if he'd misquoted the Documentary Series, we'd have a problem.

Re:Public IPs at premium prices

[Anonymous Coward]

You do understand that you can't have private addresses in IPv6?

Re:Public IPs at premium prices

[Anonymous Coward]

"What would I do ?"

"Besides, two IPv4 address at the same time ?"

"I'll keep those two IPv4 adresses unused."

"You dont need a mil $ for unused address man".

Re:Public IPs at premium prices

[Anonymous Coward]

using a 24 bit netmask in the 10.* range isn't logically any different than using the a 24 bit net in 192.168.*. CIDR allows you to up the network bits even further.

Re:Public IPs at premium prices

[wiz_80]

My ISP has done that since forever. At least they set up the routing so that I can get to the IPs of other subscribers, but I can't SSH into my machine from work, at least not without a bit of work up front.

Re:Public IPs at premium prices

[vtcodger]

I wouldn't worry too much about high inflation due to central government spending until housing demand exceeds housing supply and the economy starts to heat. Which will likely be a number of years.

Now moderate inflation due to paying to import ten million barrels of expensive crude oil a day ... that might be a problem. Could do stuff like driving smaller cars and driving them slower of course, but that's unamerican. More fun to whine.

If government spending bothers you, you might want to look into downsizing the US's preposterously large military. It's possible that the best place to cut expenses is the place where the money is being squandered.

Re:Public IPs at premium prices

[Anonymous Coward]

sorry but I want to be connected to internet, not to a wan that nats my packets to internet servers.

that, or make it work in a way that online games are not affected by it.

Re:Public IPs at premium prices

[vtcodger]

***Ahhh, I long for the days when a private address was garanteed to be private. why don't they switch already to ipv6...***

Probably because deep down inside, a whole lot of people think ipv6 is going to be a massive bundle of fascinating bugs and general grief. They wish to let somebody else clean it up to the extent that is possible. It's a matter of avoidance of pain.

So, feel free to switch over to ipv6. But don't be too surprised if you look back and don't find a crowd following you. Be sure and write and tell us how it works.

Re:Public IPs at premium prices

[Hognoxious]

give people a 6 month warning and switch... should be plenty of time to *address* most issues

I see what you did there.

Re:Public IPs at premium prices

[lindoran]

I agree do it now before the government becomes involved and forces ISP's to adopt ipv6... we need another debacle like the DTV switch as much as we need a hole in the head.

Re:Public IPs at premium prices

[CAPSLOCK2000]

There is a neat little trick that's only usefull with 10.0.0.x: leave out the inner zero's.
You may know this from IPv6, but you are allowed to leave out leading zero's and field that are entirely zero.
Instead of 10.0.0.1 you can write 10.1 . It's not 100% foolproof as not all software knows about this
possibility, but ping & ssh understand it just fine.

Re:Public IPs at premium prices

[ls671]

I know this. I use one networks with 26 bits and another one with 28 bits in the netmask.

Note this wasn't true in the very beginning of time although. 10.X.X.X was considered as one unique physical network and a netmask of 8 was implied back then.

See:
http://en.wikipedia.org/wiki/Classless_Inter-Domain_Routing

http://en.wikipedia.org/wiki/IP_address

http://en.wikipedia.org/wiki/Classful_network

Re:Public IPs at premium prices

[Anonymous Coward]

Just wanted to point out that your guidelines are written with one major and incorrect assumption, which is that you only need ONE subnet worth of addresses.

192.168.x.* is a single /24 of 256 IPs (-3 ofc), and the entire 192.168. space is 256 subnets of 256 IPs.

Most people have never heard of the /12 private space under 172.16 so that is fully expected of the public.

10.x.x.x is the only block left that can be subnetted more flexibly than only 256 IPs per subnet.

The main problem is, even though these IPs are not supposed to be used for routing, they definitely cause many problems with routing when two private networks on the same subnet boundaries want to talk, say over a VPN.

For my home network, i 'stole' a reserved /8 from the IANA blocks, which is a HUGE no-no and will screw me a little bit once that network is allocated, but for the past 6 years it has kept me from dealing with matching subnets for VPN tunnels, which at one point I was the VPN concentrator for around 10 end points so something was needed.

For the work network I maintain, that is not really an actual option, so subnetting 10. into two layers was the best answer on this end.
10.x.y.z is the format, x being which division, y being a department or logical subnet assignment within a division, and z being the host.

Re:Public IPs at premium prices

[xQx]

Well I guess it's a good thing the IPv4 overlords are still withholding Class-E addresses for "future use".

Of course, we can't hand them out to hosts, no matter how low we get on the rest of the CIDR space. They are for the future, not now!

Re:Public IPs at premium prices

[TheCycoONE]

(You do understand that) you're spreading FUD:
http://en.wikipedia.org/wiki/Unique_Local_Address

Re:Public IPs at premium prices

[cartzworth]

Actually, I'd start at the horrendous entitlement commitments we have. Like, the ones with outstanding liabilities that equal all the private wealth in the country. Then, maybe the military. Or maybe fending off more entitlements, like the Federal gov't being the only originator of student loans. And maybe cut off Fannie and Freddie from the taxpayers pockets. Just some thoughts.

Re:Public IPs at premium prices

[comm3c]

> my provider uses 10.* addresses there therefore I had to change the
> addressing scheme on my LAN because

Why did you chose 10.x.x.x for your LAN in the first place ? I doubt that you are planning to connect 16,777,216 machines to your LAN ;-)

Guide lines are to use:

192.168.X.X if you need 65,536 IP addresses or less

172.16.X.X-172.31.X.X if you need between 65,536 and 1,048,576 IP addresses

10.X.X.X if you need between 1,048,576 and 16,777,216 IP addresses

Routing is slightly faster with more bits in your netmask. Although I do not think that you will notice a difference especially nowadays. I think this was one of the reasons for these guidelines. Following the guide lines also ease connectivity to bigger nated networks, your provider in your case.

We're in the era of VLSM. Please stop spreading this nonsense of classful addressing.

Re:Public IPs at premium prices

[ls671]

>Just wanted to point out that your guidelines are written with
> one major and incorrect assumption, which is that you only
> need ONE subnet worth of addresses.

This isn't true anymore, at least since 1993, so you can do things like 192.168.128.0/17, 192.168.128.0/28, 192.168.0.0/16 or 10.3.4.0/24.

http://slashdot.org/comments.pl?sid=1639534&cid=32083348

Re:Public IPs at premium prices

[idontgno]

http://xkcd.com/386/

Re:Public IPs at premium prices

[Anonymous Coward]

There is a neat little trick that's only usefull with 10.0.0.x: leave out the inner zero's. You may know this from IPv6, but you are allowed to leave out leading zero's and field that are entirely zero.

The correct number of zeros in the word zeros is zero.

Re:Public IPs at premium prices

[hitmark]

must be a rule for linking xkcd by now...

Re:Public IPs at premium prices

[ls671]

> We're in the era of VLSM. Please stop spreading this nonsense of
> classful addressing.

WTF ? What I wrote has got nothing to do with classful addressing.

Here is an example of somebody who thinks in a classful way. You can also read my reply to him.

http://slashdot.org/comments.pl?sid=1639534&cid=32083386

Re:Public IPs at premium prices

[squallbsr]

I use 172.16.99.x and 172.16.42.x and I happened to have a client that I often had to clean-up their messes that also used BOTH 172.16.99.x and 172.16.42.x.

That was annoying...

Re:Public IPs at premium prices

[Amouth]

i always wondered about that

- How . .if the edge of the universe was collapsing causing bulkheads to be incomplete via walls no longer existing.. How exactly was there Explosive decompression? is the edge of the universe a vacuum? For it to explode and decompress then where did it explode too? you would have thought it would have just been something then nothing/ and no one would have been the wiser

Re:Public IPs at premium prices

[fast turtle]

Once this happens I hope my ISP finally gives me proper IPv6 support so I can surf all the free pron out there.

Personally, I'd love to have a dedicated IPv6 address since it would then make it easy for me to access my systems using the Win7-Homegroup properties when ever I'm on the road.

Re:Public IPs at premium prices

[Compaqt]

Well, if they're using 192.168.1.x, can't you use 192.168.2.x ?

Also, if you have two ethernet interfaces, one of which is 192.168.1.x, and a Netgear WGR640 is on the other trying to also use that block, the router will autoconfigure to use 10.0.x.x . Does anyone know offhand if you can use other private (or not) netblocks manually?

Re:Public IPs at premium prices

[Compaqt]

>Instead of 10.0.0.1 you can write 10.1 .

It works!

Wow, the stuff you learn on Slashdot comments.

Just confirmed it works with ping, ssh, and telnet on Ubuntu.

Re:Public IPs at premium prices

[skids]

Because we'd be all so proud to live in a country that kicked it's elderly to the curb after they had "served their purpose"

Re:Public IPs at premium prices

[Anonymous Coward]

Classful netmasks can screw up even globally routeable netblocks that have been since CIDRed, but it's becoming less and less of a problem. Last time I had issues was dealing with split tunneling on crusty old VPN clients.

More here on that.

Re:Public IPs at premium prices

[CrimsonAvenger]

If government spending bothers you, you might want to look into downsizing the US's preposterously large military. It's possible that the best place to cut expenses is the place where the money is being squandered.

Just fyi, the US military budget is just about 1/2 the DEFICIT for 2010.

In other words, reduce the military budget to zero, and you still have an annual deficit of over $500 billion

Note that mandatory spending plus the State Department, Health and Human Services, and Transportation Department budgets just about use up the entire tax revenue of this fiscal year.

Note further that an across the board doubling of income taxes (assuming that such a thing didn't have any real effect on people's behaviour) would not quite do away with our current deficit.

Re:Public IPs at premium prices

[Lars T.]

Because every damn organisation I VPN to uses 192.168.x.x addresses, or 172.[16-32].x.x addresses. By using a 10.x with a 24 bit mask I can use space that doesn't route, doesn't conflict with the orgs I VPN to, and that doesn't require me to reconfigure.

Unless somebody comes up with the same clever idea - see the OP.

Re:Public IPs at premium prices

[Abcd1234]

And *that* is one reason why publicly-routable IPs on private networks is a good thing. It makes things like VPNs *far* easier to manage, as you don't have to worry about conflicting private networks.

Re:Public IPs at premium prices

[Blakey Rat]

They probably wanted to keep it a light science fiction action series and not a dense philosophical dirge by completely over-thinking every possible detail.

Just a guess there, though.

Re:Public IPs at premium prices

[spazdor]

Yeah, we should consider the possibility that anonymous posting may suffer. As more and more AC's end up accessing Slashdot from behind the same gateways, they'll get that "Wait 10 minutes, asshole" message more and more often.

Re:Public IPs at premium prices

[JWSmythe]

    It should be perfectly doable. Where I am, from the desktop to the Internet there's a wireless bridge, then a NAT, and then I'm on. I haven't had any problems playing games online. Most new installs from what I've seen give a router that gives the LAN a NAT address anyways.

Re:Public IPs at premium prices

[quanticle]

If government spending bothers you, you might want to look into downsizing the US's preposterously large military.

Its always easy and fun to bash the military, but the fact remains that military and VA spending accounts for only about 23% of the overall federal budget. The lion's share goes to Social Security and Medicare. Between the two of them, they're taking about 40% of the budget.

It's possible that the best place to cut expenses is the place where the money is being squandered.

By all means. The places where the money is being squandered are Medicare and Social Security.

Re:Public IPs at premium prices

[quanticle]

I'm not a fan of kicking the elderly into the streets either. That said, the US spends the most and gets the least out its health care system, as compared with other industrialized countries. Our per-capita costs are the highest in the world, while our outcomes are amongst the worst in the industrialized world.

Re:Public IPs at premium prices

[optikos]

Hey, troll-fisher! You are quoting extraneous lines that precede the "nature of the universe" part to which I am referring.

Re:Public IPs at premium prices

[feepness]

I was actually being silly. I was more referencing both the "things munching on the hull" (explosive decompressions) rather than the nature of the universe.

My apologies.

Re:Public IPs at premium prices

[ls671]

> Because every damn organisation I VPN to uses 192.168.x.x
> addresses, or 172.[16-32].x.x

it is 172.[16-31].x.x. At 32, you are in no longer in the 172.16.0.0/12 reserved for private use IP space.

Re:Public IPs at premium prices

[ls671]

> That was annoying...

Yep, I estimate that by year 2115, these types of annoyances will be resolved, thanks to IPv6.

Re:Public IPs at premium prices

[ls671]

Classful netmasks can screw up even globally routeable netblocks that have been since CIDRed, but it's becoming less and less of a problem. Last time I had issues was dealing with split tunneling on crusty old VPN clients.

More here on that.

I suspected this could still occur although I have never encountered it.

Thanks for confirming ;-))

Re:Public IPs at premium prices

[T_Tauri]

Oh yes, I can change my IP ranges to 192.168.2.x. Only two problems with this:

Firstly I manage about 15 servers with static IP's - changing all those IP address is not easy especially when you add in the firewall rules etc.

Secondly there is nothing stopping the next company I connect to using 192.168.2.x themselves. Or 192.168.1.x/24. By using an obscure 10.a.b.c subnet this is still potentially a problem but far less likely.

Re:Public IPs at premium prices

[Anonymous Coward]

This is what differentiates engineers from computerists: the latter too often think they can get away by doing the minimum and always patching later. Changing is hard and expensive.

Truth is

[Anonymous Coward]

IPv4 is like oil. It'll never go away.

Luckily, IPv4 isn't a bad technology.

Re:Truth is

[Anonymous Coward]

IPv4 might be, but IPv4 addresses certainly are not like oil. The remaining addresses are not harder to find or costlier to acquire. The rate at which these addresses are assigned will increase right up to the very end, when suddenly there won't be new any new allocations, first by the IANA to the RIRs and then by the regional internet registries to ISPs. The supply of IPv4 addresses is finite. We know that we need more addresses than there are, the vast majority of addresses are already assigned and the rate of assignment is increasing. Right now everybody's hoarding IPv4 addresses: They're used as leniently as possible. Got a server? Get three addresses automatically. Why? Because that's a good enough excuse to get an allocation from the RIR. When there are no more allocations available, then the big redistribution of IPv4 addresses begins. Customers will have to start paying for addresses that were formerly included for free. This way there will still be addresses available for new applications, but they'll be taken from other applications (ones where they're not really needed at first, but eventually it will be a matter of who pays more.)

Re:Truth is

[mattventura]

Luckily, IPv4 isn't a bad technology.

On paper, it looked like a great technology when it was created, but nobody ever expected the internet to spread like wildfire. Not we're faced with exhaustion of it's (comparatively) tiny address space. In addition to that, ISP customers get a single IP and have to use NAT, which as we all know, causes lots of problems.

I guess an addressing crisis has it's ups and downs. The good part of this is that it will light a fire under ISPs and get them to seriously think about IPv6. The bad part is that IPv4 will become an even worse situation, with not only customer equipment NAT, but ISP-wide NAT, which causes routing messes, and will give ISPs yet another thing to gouge you over. Want a real IP? Oh, that's $20 a month extra.

Real pushes towards IPv6 deployment would make this entire point moot. However, half the big-name ISPs refuse to do anything about it, and until we see some real support from the home/small business ISPs, IPv6 will pretty much go nowhere, and we'll be stuck in the IPv4 hellhole.

Re:Truth is

[Architect_sasyr]

If you had to argue the point out with most router admin teams, I would guess that your main argument comes down to IPv6 is so hard to remember. Ok, so let's make it easier and build on IPv4 - rather than 255.255.255.255 let's add another octet: 255.255.255.255.255 - it's not going to be that much harder to remember, and it increases the address space quite a bit, hell go wild and add another two rather than just one. It has been an eternal hatred of mine that representation of IPv6 looks more like a MAC than anything else - I don't know why they didn't just fix it properly in the first place to make it more human readable. I mean, if you're going to talk about design flaws at conception...

Of course this is all just IMHO and may be significantly flawed. Just sayin' is all.

Re:Truth is

[Anonymous Coward]

Or they could, you know, write them down or something. Hell, give them all iPads to sweeten the deal if that's the only hold up on IPv6 - with an App to let them assign friendly names to their favourite addresses in a lookup table, so instead of remembering some obscure string of data they can just remember the finance router is called "bob".

Re:Truth is

[gmack]

Ipv6 was designed for sparse allocations to allow for easier routing that and if they figured if they need to painfully break something make it large enough so that it only has to be done once.

Ipv4 32 bit.. ipv6 128 bit so really if you wanted to reference the number the same way it would be 255.255.255.255.255.255.255.255.255.255.255.255.255.255.255.255.

Re:Truth is

[vtcodger]

***I don't know why they didn't just fix it [address nomenclature] properly in the first place to make it more human readable. I mean, if you're going to talk about design flaws at conception...***

Indeed. The failure to make ipv6 addressing readable; the failure to make ipv4 a proper subset of ipv6; and the failure to support existing ipv4 NAT all make me wonder if the people who designed ipv6 knew what they were doing. I'm really inclined to let the early adopters find out. I think I have a lot of company on that.

Re:Truth is

[Miamicanes]

Forget what's officially on paper regarding the lower 64 bits of an IPv6 address. As a practical matter, the bits are yours to use, or yours to intentionally discard and ignore. Nobody can force you to use an unholy address with 128 pseudo-meaningful bits. In reality, a typical home cable internet or DSL customer should get a /48 address. That means the upper 48 bits are assigned by your ISP, and the lower 80 bits are yours to use or ignore as you please.

48 bits translates into three chunks of the address. Initially, they'll look like this:

2001:xxxx:xxxx

where xxxx:xxxx is unique for your global service provider (read: AT&T, Global Crossing, Telia, etc), and the top few bits are unique to your country. Eventually, the top 16 bits (first chunk) will be unique to a country. So unless you're in Europe, 99% of the addresses you deal with daily will start with the same chunk. A little more annoying than a.b.c.d, but not quite insurmountable.

Now, the lower 80 bits. Officially, there's some nonsense somewhere about using the MAC address to derive it. It's voluntary, and lots of people have said they won't be following it because it turns your ethernet card into a globally-unique cookie that can be shared via out of band means. It's led to no small number of fights between the autoconfiguration faction and the DHCP6 faction. In the long run, the DHCP6 faction will win. So, getting back to those 80 bits...

Right now, you're dealing with 64 bits whether you know it or not: the 32 bits in your public IP address, and the 32 bits in your NAT'ed private addressses. Your home PC's IP address isn't 64.82.19.34, it's 64.82.19.34 + 192.168.100.69, with a port-forwarding router in the middle to glue them together. There's nothing to stop you from treating your IPv6 address the same way, even if it slightly stomps on the original official intent of how the bits should be used. If you want to, your home PC's 80-bit IPv6 address can be ::1. More likely than not, your router (still needed for 6-4 translation) will be ::1 though, so let's make your home pc ::2. So, here's your home PC's new IPv6 address: 2001:aaaa:bbbb:cccc::2 (aaaa, bbbb, and cccc are 16-bit hex values). Or maybe uuuu.aaaa.bbbb.cccc (where "uuuu" is the same for most addresses you encounter at the raw address level daily).

Let's suppose you want to be cute, and have a home network whose 4 and 6 aspects cleanly map to each other, and you don't mind doing some ugly binary math behind the scenes for the sake of human-readability. Your internal network's ipv4 addresses are in the 192.168.100.x block. Your 4-6 router is 192.168.100.1, and your desktop PC is 192.168.100.2. So, you configure your IPv6 addresses to be:

uuuu:aaaa:bbbb:cccc::192:168:100:1 and uuuu:aaaa:bbbb:cccc::192:168:100:2 for the router and PC. Yes, in this case, you're abusing bits and pretending that 0x0192 means something special (the way 192 meant 0xC0, which had a special meaning to the router), but it's mainly to illustrate that IPv6 addresses don't HAVE to be horrific trainwrecks of unmemorable bits. If you're determined, you can make them look quite a bit like ipv4 addresses within the realm of your own network if you really want to. At the end of the day, is

24:3cf2:8d99::2 (your home PC), or even 24:3cf2:8d99::192:168:100:2

really any worse than

64.87.142.98 -> 192.168.100.2 ??

The truth is, the IPv6 crowd has been its own worst enemy by trying to force needless address complexity. IPv6 has lots of bits, and if you ignore the majority that you're entitled to ignore, it doesn't have to be a lot worse than ipv4. It can even be better.

Re:Truth is

[petermgreen]

but eventually it will be a matter of who pays more.
And in turn it will also be a matter of cost of continuing to use those public V4 addresses VS cost of other options. Other options include IPV6 and private IPV4.

Re:Truth is

[Anonymous Coward]

Just wait until a few Class A's get dumped into Gulf of Mexico. They'll be IP addresses washing up on shore for decades to come.

Re:Truth is

[Ares]

or better yet we could create a hierarchical distributed database that maps names, divided by dots, to ip addresses and give the "owner" of each level of their particular branch of the hierarchy control over how they map names to those addresses.

wait a minute.

nevermind.

Re:Asleep at the Switch

[gyrogeerloose]

That shit was funny for a while but now it's just annoying. Give it up, dude.

I fail to see the black market part

[Sycraft-fu]

So if I have IPs, and someone else needs IPs, I sell them some of my IPs... What's the problem here? For that matter, that is how it works for anyone who's not a big provider. When I wanted static IPs for my cable connection I asked my cable ISP. They said sure, $5/month/each.

I'm just not sure I see a problem. Goes double since higher IPv4 prices may encourage IPv6. Consider:

Say I'm an ISP, we have all old v4 hardware. To the extent our routers support v6, it is all in software meaning that any significant amount of IPv6 will overload them. They only have IPv4 ASICs. I don't wanna upgrade because it is expensive. So I keep getting more and more customers that want IPs. However, I run out, my allocation is gone. ARIN says "Sorry, all space is allocated." So I go looking around. Turns out I can buy a /24... But for 500x what I used to. Ouch. Well then, maybe time to get some IPv6 hardware.

Likewise it could encourage customers to want IPv6. A company buys a net connection and says "We need 32 IPs." ISP says "Well you can have 32 v4 IPs for $3200/month, or you can have as many IPv6 IPs as you want, and 1 IPv4 IP for 6-to-4." Company says "Oh ok, v6 may be more of a pain, but it is worth it to save the money."

What it comes down to is we need to migrate away from IPv4. That'll be a long process, but one thing that'll help it along is if there's economic incentive to move to IPv6. Right now, the situation is generally that there is an economic DISincentive to move to v6. You need new hardware, sometimes new software, etc. It costs money and IPv4 works fine. However, if v4 starts costing more, that makes v6 more attractive.

So I don't see this as a "black market" nor do I see it as a big problem.

Re:I fail to see the black market part

[Tuzanor]

Any market that forms that people don't want to form is a black market. They'd prefer some "benevolent" agency to dole out the limited amount, nevermind that a few organizations are holding massive amounts of unused IP ranges. Making them worth money will encourage them to release them, but these people are afraid of markets.

Re:I fail to see the black market part

[Anonymous Coward]

What I see will end up happening is that the V4 prices will go up, and eventually for a static IP (v4 or v6), the price will be insane ($50-$100 a month, citing how few IPs are available.) Of course, when someone mentions that IPv6 doesn't have the number exhaustion issue that V4 does, the sales person will look at you and start drooling, just like you were trying to tell a chimp about vector calculus.

So even if IPv6 becomes common, eventually expect to either be forced to a dyndns-like service or be paying big bucks for even a single static IP, be it a V4 or a V6.

Re:I fail to see the black market part

[Anonymous Coward]

What I see will end up happening is that the V4 prices will go up, and eventually for a static IP (v4 or v6), the price will be insane ($50-$100 a month, citing how few IPs are available.) Of course, when someone mentions that IPv6 doesn't have the number exhaustion issue that V4 does, the sales person will look at you and start drooling, just like you were trying to tell a chimp about vector calculus.

So even if IPv6 becomes common, eventually expect to either be forced to a dyndns-like service or be paying big bucks for even a single static IP, be it a V4 or a V6.

with IPv6 it would be a static /64. It is hard to give out anything smaller than a /64.

Re:I fail to see the black market part

[optikos]

black market = a market of illegal goods, such as narcotics or other against-the-law contraban, such as weapons of mass destruction
gray market = an unauthorized after-market of commercial goods
This IPv4 /8 market would be a gray market, because IPv4 addresses are not narcotics or weapons of mass destruction. Original poster is engaging in hyperbole for a troll-fishing effect.

Re:I fail to see the black market part

[TubeSteak]

Any market that forms that people don't want to form is a black market. They'd prefer some "benevolent" agency to dole out the limited amount, nevermind that a few organizations are holding massive amounts of unused IP ranges. Making them worth money will encourage them to release them, but these people are afraid of markets.

I think you're confusing "afraid of markets" with "afraid of unregulated markets".

If you don't understand why unregulated markets are bad, feel free to pickup a history book and look at the American business landscape in the 100 years preceding the 1930s.

Re:I fail to see the black market part

[optikos]

Any market that forms that people don't want to form is a black market.

No, you have effectively defined "gray market" instead: an unauthorized market in commercial goods. Now if we were to pass a law that makes possession of an IPv4 address (or /8 IPv4 address) a crime (especially a felony instead of misdemeanor), then it becomes a black market. black = crime in criminal courts. gray = unauthorized breech of contract in civil courts.

Re:I fail to see the black market part

[timeOday]

They'd prefer some "benevolent" agency to dole out the limited amount, nevermind that a few organizations are holding massive amounts of unused IP ranges.

Not "nevermind" - that's exactly why the IP addresses should be doled out by a regulator and NOT resold. Just because some chain of company buyouts leads back from you to somebody that requested a /8 when they were given away for free, does NOT mean you have somehow "earned" millions of dollars in any meaningful sense. There is NO reason to financially reward such behavior.

The free-market-true-believers-under-all-circumstances crowd is correct that markets always find some solution, but why can't they see that sometimes it's a bad one?

Re:I fail to see the black market part

[shadowbearer]

  Except where the corporations can buy whatever regulated market they wish to, and individuals have no choice in the matter. I agree with you, but the last decade, at least, has shown that government regulation in this country is for sale to the highest bidder.

SB

Re:I fail to see the black market part

[shadowbearer]

Any market that forms that corporate/government interests don't want to form is a black market.

  Fixed, with respect.

SB

Re:I fail to see the black market part

[Lorens]

Right. Add in the fact that black or gray market implies not going through the regulator, which means that the Whois will not be updated by said regulator. You'll also have bigger IPv4 routing tables and even more complicated and fragile BGP configurations since you'll have a financial incentive to let your IPs be used by people you do not have a direct interconnection with. For the routing tables it probably doesn't matter since routers should handle even more for IPv6, but the instabilities and uncertainties are not good.

Re:I fail to see the black market part

[Rakishi]

So if I found a now rare coin in my attic that my great granddad got from a bank than you say it's bad for me to sell it? If I bought a stock that then went up is it wrong for me to sell it at a profit? Or is it only evil when it's a method in which YOU can't make money yourself?

What's a better solution? IP4s are a limited resource, a market would allocate them in the most efficient manner possible. They're a valuable resource as well which means people are both willing and will pay for it. You can try to regulate that away but it won't work, my parents bought many things in the soviet union by trading in their alcohol allotments on the black market.

Re:I fail to see the black market part

[takev]

IPv6 will use less space in the router tables as the address space is split into a geographical tree. Or at least it used to be when I looked at it the last time, which was a long time ago. Anyway on that scheme the first few bits show the continent, then the country, then the ISP.

Re:I fail to see the black market part

[timeOday]

A market based solution would be good if IP addresses were leased instead of (permanently) given away in the first place. Curently, even a charge of e.g. $1/address/year would free up millions of addresses given away in huge blocks in the early days to a small number of businesses and universities.

Re:I fail to see the black market part

[JumpDrive]

You have just pointed out the reason why a number of providers are dragging there feet about switching
They don't have to train anyone, hire anyone else, update any hardware, just sit and wait and in a couple of years everybody will be screaming.
"WE'RE OUT OF IP's , WE'RE OUT OF IP's"
It will make it to the business section 2 columns half a page, one paragraph explaining solution. But most of the article will talk about the shortage and the burden it has placed on ISP's who now have to manage and ration IP addresses. So that $5 IP address is now $12 dollars/month. They have now increased their cash intake 140% doing nothing.
So the only way this is going to happen is if someone like Google,Motorola,Nokia, Verizon, AT&T, Sprint, becomes truly hampered by this and see a loss of product growth do to it. Then the money will flow to Congress and they will have a hearing on how the industry can't manage itself and eventually pass regulations requiring usage of IPv6.

I can't wait for the analogy of the big pipe having many little pipes with addresses (I may start taking book on which Congress man does it first).
This may also be a wonderful time for the ISP's to explain why they must control the flow of traffic with certain cost penalties. Partially so they can recoop the 140% gain for doing nothing they will loose if they cooperate.


I can't tell whether I'm being insightful or funny, please tell me.

Re:I fail to see the black market part

[Patch86]

So if I found a now rare coin in my attic that my great granddad got from a bank than you say it's bad for me to sell it?

No. But if you're using "found in attic" as the basis for a whole economic industry worth millions, then I'd say it was a little far from ideal.

We allow people to play the lottery, but no-one is suggesting that a lottery would be the best way to, for example, allocate corporate support contracts, or rail franchises. What is OK on a small scale for individuals isn't necessarily OK for the world as a whole.

Re:I fail to see the black market part

[houghi]

When I wanted static IPs for my cable connection I asked my cable ISP. They said sure, $5/month/each.

If that is for ADDITIONAL IPs, that would be somewhat logical. If it is for the firs IP, then it is not. Cable modems connections are NOT the same as PSTN connections. Cable (and ADSL) will be likely to be connected all the time.

This means that the provider needs at least the amount of IPs as it has customers. This means to have a fixed IP or a Dynamic IP does not make any difference in the number of IPs needed. Providers however use the excuse that IP adresses are rare as an excuse to charge more money from it. There are even providers that force a different IP on you after a certain time.

But even then there is the question of charging for something that they themselves do not need to pay for. In Belgium the average price for a fixed IP is about 50EUR. The provider pays a fixed price for all his IPs. http://www.ripe.net/info/faq/membership/newlir-billing.html and http://www.ripe.net/membership/billing/procedure.html

In that last URL you see that a provider will pay 5.000EUR for its IPs. And as you can see almost all are in that group. I would say that 5.000EUR is more operational cost then a real cost. It also is a fixed cost. So charging 50EUR for an IP address is a bit of black market to me.

Re:I fail to see the black market part

[Anonymous Coward]

Does anyone else think that IPv6 may be overdesigned? You can have a huge number of addresses, but you can't handle the routing table? Why didn't they do a 64 bit address, or even just 40 or 48 bits?

Nobody needs a /64, really. And it's difficult for me to imagine it ever being necessary.

Re:I fail to see the black market part

[Rakishi]

I don't think you understand. The stuff has already been allocated, that's the whole problem.

The black market auctions are to fix the allocation mess that was already created. Get a time machine if you want to fix that mess, the rest of us have already shrugged and moved on.

We allow people to play the lottery, but no-one is suggesting that a lottery would be the best way to, for example, allocate corporate support contracts, or rail franchises. What is OK on a small scale for individuals isn't necessarily OK for the world as a whole.

Well congratulations, what do you think is the point of having black market auctions? The lottery has already happened. A landlocked farmer in Wisconsin just won an oil tanker and it's rusting in a harbor. Now we're onto the point of him selling that piece of floating iron to someone who can profit from it.

Not that it may even matter, it's unlikely the cost of selling ips will be worth the cost of changing ip ranges.

Re:I fail to see the black market part

[Rakishi]

I like the idea of leasing in principle (granted, $2+ billion a year seems amusingly high) but practically it may not work. Hell, practically black market auctions probably won't work either.

Changing ip ranges is apparently very expensive to do for large entities, almost as expensive to do so once as moving to ip6. With selling there's a way of offsetting the costs of such a move and no one is forced to change their use.

Re:I fail to see the black market part

[u38cg]

Because people's standard of living was so much better in 1829 than 1929?

Re:I fail to see the black market part

[Anonymous Coward]

The primary reason for not changing the rules so late in the game is that it would be pointless to do so. Sure, it would satisfy a lot of people's sense of fairness, but it would not solve the technical problem, which is that there are not enough IPv4 addresses. No matter how efficiently we assign them, we will run out of addresses. A fee per address would create an incentive to replace addresses with NAT now and then prolong this deplorable state of affairs because it would slow down IPv4 address depletion.

Re:I fail to see the black market part

[Anonymous Coward]

In which country is government regulation not for sale to the highest bidder according to this standard?

Furthermore, given the diversity and number of countries, if the answer is between 0 and 3, isn't it surprising that no country has chosen a different path?

Re:I fail to see the black market part

[Anonymous Coward]

You could argue there is a fundamental fairness and governance issue when something that has been doled out for free and willy-nilly suddenly takes on a huge value.

Imagine if IPv4 was carbon emission permits, and you initially handed out a bunch to whoever asked, then suddenly say "Oh, sorry, there are no more, you have to trade the ones you have". New entrants would be pissed off, whoever got lots to spare will run off with the money. Even the law where I am from recognises the concept of "undue enrichment".

Re:I fail to see the black market part

[Pinky's Brain]

So don't use it ... it really doesn't matter to the upstream router whether you use one or a million IPs in the subnet.

Re:I fail to see the black market part

[petermgreen]

Say I'm an ISP, we have all old v4 hardware. To the extent our routers support v6, it is all in software meaning that any significant amount of IPv6 will overload them. They only have IPv4 ASICs. I don't wanna upgrade because it is expensive. So I keep getting more and more customers that want IPs. However, I run out, my allocation is gone. ARIN says "Sorry, all space is allocated." So I go looking around. Turns out I can buy a /24... But for 500x what I used to. Ouch. Well then, maybe time to get some IPv6 hardware.
One big problem is that IPV4-only hosts and IPV6-only hosts cannot talk to each other. Therefore migrating to V6 only isn't an option for most hosts until large parts of the internet are on dual-stack. Migrating from V4 only to dual-stack does not save you any IPs.

Teredo and 6to4 may help a bit with this but they have issues of their own (teredo is very fragile, 6to4 needs a public v4 IP)

Re:I fail to see the black market part

[Anonymous Coward]

recoop

To put the chicken back inside?

Re:I fail to see the black market part

[petermgreen]

Not that it may even matter, it's unlikely the cost of selling ips will be worth the cost of changing ip ranges.
Depends what the IPs are being used for and what the price per IP gets like.

I'd think the first thing to go will be free public IPs for home users. Home users tend to be on dynamic IPs anyway so the cost of readdressing them to private (or for those who pay extra for a public IP to a new smaller public block) will be relatively low and the addresses will be able to be recovered in relatively big blocks.

Low end hosting/colo/dedicated packages will probablly get less generous in their IP allocations as well but because those IPs tend to be static I'd expect those IPs to be more suited to internal reuse than to resale.

Re:I fail to see the black market part

[Myopic]

Whoa, dude, whoa. Do NOT try to start a fight with a free-market ideologue and try to use facts or historical evidence to support your position. That's even more foolish than trying to convince a creationist about evolution by using the fossil record as evidence. These people will never, ever, ever admit that their simple little ideology doesn't perfectly explain the complexity of the real world.

Just let it go, man. Save yourself the trouble.

Re:I fail to see the black market part

[timeOday]

A fee per address would create an incentive to replace addresses with NAT now and then prolong this deplorable state of affairs because it would slow down IPv4 address depletion.

Why don't you think it would spur migration to IPv6?

Re:I fail to see the black market part

[Anonymous Coward]

-or in the 10 years preceding the 2010s.

Re:I fail to see the black market part

[Lord Ender]

Arin is supposed to be a benevolent communist dictator. Any free market of address space is a black market, just like any free market in the USSR was.

Re:I fail to see the black market part

[Anonymous Coward]

IPv6 addresses already are much simpler to get, but everybody keeps using IPv4, because that's what everybody else uses.

Re:I fail to see the black market part

[Anonymous Coward]

I would much rather be a farmer in 1829 then working in pretty much any factory in 1929

Re:I fail to see the black market part

[amorsen]

IPv6 isn't all that much more geographical than IPv4. Regional registries have assigned blocks, just like IPv4, but they will last a bit longer with IPv6 so the fragmentation will be less. Noone stops you from advertising addresses in a different region than they were acquired though.

Re:I fail to see the black market part

[amorsen]

Dynamic IP's are going away, to be replaced by NAT'ed addresses. Mobile Internet is far along that road already.

Re:I fail to see the black market part

[u38cg]

Maybe you would, but why did scores of rural poor migrate to urban factory work everywhere in the developing world - including in Africa and Asia today?

And most big ISPs won't care...

[Anonymous Coward]

They'll still rake in the profits without upgrading much (beyond some NAT hardware), and the times when you had end-to-end network layer connectivity between hosts across the Internet will be just a memory.

IPv4 Address depletion?

[euyis]

They've been talking about this since maybe 2005 or earlier, and yet I didn't see this happen...

Re:IPv4 Address depletion?

[paul248]

IPv4 depletion was a looming problem in 2005, and today it's an even more closely-looming problem. It's not like we discovered more numbers since then.

Re:IPv4 Address depletion?

[Ihmhi]

It's not like we discovered more numbers since then.

My good friend Sean Connery would disagree with you, sir.

Re:IPv4 Address depletion?

[Anonymous Coward]

IPv4 depletion was a looming "problem" since the late '90s.

It is, in a word, bullshit.

I can go to $RANDOMHOSTINGPROVIDER and buy out an entire class C for $256/month, if not less. Seriously, $12/year for a dedicated IP is so common it hurts.

If we were in any real danger - if IPv4 addresses were anywhere near the level of rarity that keeps being touted by fearmongering fools - they'd be a *hell* of a lot more expensive.

I'm sure we'll get to that point eventually - and that's a good thing. Crying about IPv6 isn't going to get the world to switch. Switching costs money. Switching can end up costing a metric ton of money - consider the number of legacy systems out there that simply cannot handle IPv6 and *cannot* be upgraded. Not to mention switching layers, routing, firewalls. It all adds up, and it all adds up to a hell of a lot more than $1/month IPv4 addresses.

Re:IPv4 Address depletion?

[Anonymous Coward]

Actually, cutting-edge research mathematicians are currently investigating the possibility of a new, heretofore unnoticed integer between 6 and 7. The application possibilities are endless. For example, any meeting room with 10 seats may actually have additional space for another person between seats number 6 and 7, which could result up to 10 % savings in real estate costs at businesses all over the world.

Re:IPv4 Address depletion?

[Pinky's Brain]

There is no mechanism for the price to increase as the pool of remaining adresses decreases, there is in fact a mechanism in place for $RANDOMHOSTINGPROVIDER to rent them out cheaper ... whatever he has when the clock runs out is what he will be stuck with.

Re:IPv4 Address depletion?

[Anonymous Coward]

Why not just use more bandwidth. Double book an IP number and route all packages to both machines.

Can /8 companies resell subnets?

[HockeyPuck]

Could an organization with a /8 resell a block of their IP addresses? I can't imagine how someone like MIT or US Postal Service, could use 16million IP addresses, or HP use 32million (they have their own plus Compaq/DECs).

Re:Can /8 companies resell subnets?

[Anonymous Coward]

The problem with this is that it's likely to be more expensive to reconfigure their networks, internal address allocations, and everybody's routing tables in the world for those cut up /8's than it will be to just upgrade to IPv6. If somebody really needs v4 addresses that badly and want some space from these /8 holders, then they'll need to make it worth their while to start splitting them up.

Re:Can /8 companies resell subnets?

[Anonymous Coward]

The /8 holders will be as willing to give up their allocation as they are their trademarked domains. The older companies who have been around to be able to get such IP ranges know their value, and are going to sit on them, as part of their core corporate identity. The only way their holdings will ever be taken away is not by them selling the ranges out, but if one of the companies goes bankrupt, and someone knowledgeable buys the range just to break it apart to sell. Which likely won't happen because I'm sure there will be a bailout from the US taxpayers before this ever would happen.

Re:Can /8 companies resell subnets?

[jochem_m]

Actually, I know some of the big boys use external IPs internally, because they have them anyway... So it's far more than trivial to just sell parts of them...

Re:Can /8 companies resell subnets?

[Anonymous Coward]

Funny thing - if the government decides they want the land that my house is on, to build a new railway or some such, they can force me to sell it to them at whatever price they decide is fair. I don't see any reason these companies wouldn't eventually be forced to sell their addresses, and I don't see any reason why they should be able to dictate the prices, if enough governments get together and agree to give the necessary powers to the necessary regulators. Of course, and rather depressingly, this would only happen if the cost of buying enough members of said governments to block the legislation is more than the potential loss of revenue from selling the addresses.

Re:Can /8 companies resell subnets?

[jimicus]

Apparently someone's already done the arithmetic - IIRC at the current rate of growth, all you'd do is delay the exhaustion of IPv4 addresses by some stupidly small amount like 6-12 months.

Re:Can /8 companies resell subnets?

[petermgreen]

There is no question at this point that the virtually free and relatively easy IPV4 IPs we enjoy today will stop in the not too distant future and there is NOTHING that can be done about that. Recovering legacy allocations will as you say make no significant difference to when this happens.

The question is what happens after that. It seems certain that IPs will be reassigned from less important/lucrative uses to more important/lucrative ones. The only question is will that only be allowed to happen inside companies/ISPs, will there be a grey market in IPs or will there be some kind of official resale structure.

Re:Can /8 companies resell subnets?

[Anonymous Coward]

Was just thinking the same thing- HP could be rich selling off parts of 15/8 (HP) and 16/8 (DEC). Addresses coming out the ears. But yeah, it would probably be cheaper to change to IPv6 than to reconfigure the entire organization to adjust for new network configs. But once the organization moves to IPv6... possible cash cow? :)

A black market may develop

[jimmyfrank]

Or, a year before IPv4 addresses are estimated to run out, foot dragging IT departments will switch over to IPv6.

Re:A black market may develop

[pipedwho]

According to the article, that time was yesterday.

The authors of TFA estimate that in less than a year ARIN will have no more /8 blocks left to allocate.

Re:A black market may develop

[RobertM1968]

According to the article, that time was yesterday.

The authors of TFA estimate that in less than a year ARIN will have no more /8 blocks left to allocate.

Which has nothing to do with how many are sitting unused by ISPs and large companies sitting on big IP blocks.

Re:A black market may develop

[pipedwho]

The article also estimates that the RAR allocations will deplete with at most a 6 month lag behind ARIN. So, the /8 depletion rate is a good indicator as to the rate that IPv4 addresses are being exhausted.

Also, according to the article, it's not so easy for the average organisation to hoard addresses:

Such a move could mean price increases as depletion nears. Under today's rules, a small organization would pay a minimum of $1,250 annually for a /24 assignment, which represents 256 addresses, the smallest block that can be portably routed on the Internet. Smaller allocations than this must be obtained from an ISP, at the cost of a few dollars per month per IP address. Larger organizations could pay between $4,500 and $18,000 per year, but in all cases address holders must provide justification to their registry to continue using IPv4 allocations.

And the percentage of addresses that are held by the 'old guard' are far too few, or too entrenched to do any more than stay the crunch for a few extra months.

Re:A black market may develop

[RobertM1968]

Also, according to the article, it's not so easy for the average organisation to hoard addresses:

Sure... as long as they use them. Like the 200+ Microsoft bots that hit my forums at the same time indexing the same content? I guess Microsoft is using those IPs... for what I dont know. Does each of their search servers need to do it's own index of my forums? Why will 15 or 20 be in the same thread at the same time?

I know... it doesnt matter...

Re:A black market may develop

[RobertM1968]

Sorry, I'm just kinda cranky because even after adding a robots.txt entry to slow them down, they still send a bunch to the same pages and way more bots than needed - just was noting the ton of their IPs on the server at the same time I was responding.

Re:A black market may develop

[pipedwho]

The current use of IPv4 addresses is far from optimally efficient, but I don't think it's as sparse as many people seem to imply.

For example, those search bots might be on non-co-located machines. Yeah, companies could start making greater use of NAT gateways and better CIDR suballocations, but that isn't going to be easy or convenient. For many of those companies, changing the infrastructure to do that is close to the effort required of just going straight for the better solution of dual stack IPv4/IPv6.

For Sale

[gooman]

Slightly used internet address.
Act now and 127.0.0.0 could be yours today!
Only $5.00!

Re:For Sale

[Anonymous Coward]

Slightly used internet address. Act now and 127.0.0.0 could be yours today! Only $5.00!

no, it is only $5*10^999 (EVERY computer uses it)

Re:For Sale

[daveime]

There's only an estimated 10^82 atoms in the universe. Just to put your 10^999 in perspective.

Re:For Sale

[blueg3]

People don't understand numbers, they just hit the "0" and "9" keys until the number they're typing in "looks big" or "looks small".

Re:For Sale

[bky1701]

Oh yeah? Well, our country's national debt is 10^2000 dollars!

Re:For Sale

[Anonymous Coward]

Fond memories of the salesman who gleefully attempted to fulfill a customer's request for a userid at a local ISP. The requested userid? root.

Re:For Sale

[Anonymous Coward]

I'm angry about numbers. Numbers make me mad.

Re:For Sale

[gmhowell]

People don't understand numbers, they just hit the "0" and "9" keys until the number they're typing in "looks big" or "looks small".

0.00000000000000000000009

I keep putting more zeroes in, and it's just not getting bigger.

Re:For Sale

[Anonymous Coward]

Great seller! Perfect transaction! A+++++++++

Re:For Sale

[kenj0418]

Oh yeah? Well, our country's national debt is 10^2000 dollars!

Really, I didn't realize Zimbabwe had Internet access.

Re:For Sale

[Anonymous Coward]

Funny thing is that there is a concept of "used" IP addresses on the Internet. I'm sure that a customer is going to pay an ISP more for a "fresh" IP address, as opposed to one used by either spammers or dynamic IPs which means the IP likely is in a lot of blackhole lists, some of which have a policy of once on, always on.

Re:For Sale

[plan10]

I think he was talking about Greece

Re:For Sale

[anarche]

Damn, I was hoping that joke wasn't taken...

Re:For Sale

[Anonymous Coward]

The Greek debt is less than the UK debt in dollar terms. Its in % of GDP that it gets big.

Re:For Sale

[KarolisP]

welll... if you keep adding 9's then technically it gets bigger... but not by much :)

Re:For Sale

[Anonymous Coward]

Oh yeah? Well, our country's national debt is 10^2000 dollars!

Let me guess, you are Greek!

Re:For Sale

[confused one]

I have ranges of addresses available. Only $1.00 ea. Beginning with 10.0.0.1 but they won't last, so act now!

Re:For Sale

[Anonymous Coward]

Nobody cares about your 'sperging

Re:For Sale

[Anonymous Coward]

Hey! I bought this address yesterday you thief!

Re:For Sale

[marcosdumay]

Fiat money doesn't care about your physical restrictions.

Re:For Sale

[kenj0418]

Actually I was referring to Zimbabwe because of the crazy inflation they had. Before they pulled their currency entirely, it was taking about 300,000,000,000,000 Zimbabwe dollars to = 1 US$. http://en.wikipedia.org/wiki/Zimbabwean_dollar

Psst! Wanna buy a slash 32? Primo qualtity.

[khasim]

I know a guy who can get you a slash 29, but it'll cost you.

More of a technical issue ... how are the people in this "black market" going to handle the routing?

Re:Psst! Wanna buy a slash 32? Primo qualtity.

[the_other_chewey]

More of a technical issue ... how are the people in this "black market" going to handle the routing?

Tar. Routes don't get much blacker than that.

Black Market

[Arancaytar]

How do you secretly buy something that only works, by definition, if the public routing table knows it belongs to you?

Re:Black Market

[maugle]

How do you secretly buy something that only works, by definition, if the public routing table knows it belongs to you?

Shadowy figure in alley: *Pssst*..... Hey, buddy, wanna buy a billboard ad?

Re:Black Market

[mbeckman]

Actually, the public routing table isn't so neatly managed. It contains whatever ISPs are willing to advertise via BGP and the RADBs. Actual ownership is seldom verified unless someone complains about spam or other abuse. That's why ARIN has a new policy letting IPv4 holders designate a specific recipient when they release IPv4 space, essentially enabling IP address sales of /24 and larger prefixes. It's very important that the address registration records accurately reflect who is actually responsible for a particular IP allocation.

Re:Black Market

[dissy]

How do you secretly buy something that only works, by definition, if the public routing table knows it belongs to you?

A similar way that little chinese ISP ended up routing half of the worlds networks to them for a few hours.

Any core routers willing to advertise the route will get traffic for that IP block.

This only causes a problem when two route entries are 'out there' and the wrong one ends up taking control and everyone notices.
In a case like this, the original owners won't be filing complaints and there shouldn't be two routes, just the one new one.

Why would the ISP of the new-ip-block-owners question it if they have a valid ASN and it is not currently being advertised (routed, or in use) on the internet now?

The only thing that might stop this if ARIN choose to, is their IP contracts specifically ban reselling entire IP blocks (You can only reallocate blocks out of it, not reassign)

Now I've never owned more than a /20 through ARIN, so not exactly a big player here, and I've never heard of or seen that policy being enforced, but that doesn't mean they can't start doing so if this black market thingy starts annoying them.

Re:Black Market

[Rich0]

The problem is that if you try to advertise a route for a /24 all those big routers will just ignore you.

Selling IPv4 addresses only works if you sell them to somebody on your subnet, or if you sell a LOT of these addresses. Your other option is to offer to tunnel traffic.

You can't just take 10 IPs you aren't using in California and ebay them off to some guy in Denmark unless you have some way of actually directing traffic to him.

Re:Black Market

[dissy]

The problem is that if you try to advertise a route for a /24 all those big routers will just ignore you.

Very true.

But it seems the article is referring more to companies that own a /8 (or at least larger than a /16) and selling off large blocks out of it.

The smaller blocks will definitely have the problems you describe. But that might just be a condition of the 'black market' sale, left to the buyer to figure out.
A /24 would most likely be useless as far as core routers go, but a /20 or /21 might work Most places, or at least enough for the intended use of the buyer.

Re:Black Market

[amorsen]

/24's are widely accepted at this point. I wouldn't be surprised if we went to /25 or /26 in the next few years as the limit, but anything much lower would force a lot of upgrades.

So what?

[Anonymous Coward]

Can someone please explain what any of this means to me, as an end user? I have a router (Time Capsule) seems to support IPv6, and my computer does too.

Re:So what?

[Anonymous Coward]

But does your ISP support IP6?

Re:So what?

[petermgreen]

What it means is that assuming you are on a normal home connection you have a very high chance of ending up behind ISP level IPV4 nat (which practically speaking means no incoming connections) as your ISP reclaims your IP for more lucrative customers and/or resale.

If you are lucky you may be offered public IPV6 addressing as well.

Re:another shit kdawson post

[Sanhedran]

i guess.. in the same way the universe will eventually run out of energy.

"It was like a million high school physics teachers crying out in unison, then suddenly silenced."

I'll be rich!

[Illogical Spock]

1) Connect to my dynamic IP address ISP
2) Post ad on eBay for the IP
3) Sell it
4) Disconnect
5) Repeat from 1 to 5
6) Profit!

Re:I'll be rich!

[Anonymous Coward]

You'll never make any profit that way...

Re:I'll be rich!

[Anonymous Coward]

Yeah, where's the question-mark step?!

Re:I'll be rich!

[HyperQuantum]

line 6: warning: unreachable code detected

Re:I'll be rich!

[shallot]

That's why it was funny, you know. :)

Re:I'll be rich!

[shentino]

In this case, wouldn't it be your ISP that owns the IP?

IP address squatting?

[cgenman]

How would this be any different than the current state of domain-name squatting? I'm personally afraid that as soon as the black market finds it, the naming bodies will start auctioning off IPv4 addresses to the highest bidder. As more and more number blocks go, the remaining few become more and more valuable. What would be their incentive to switch everyone over to IPv6, any more than ICANN might want to fix its broken naming system?

Re:Now selling...

[Anonymous Coward]

Too late man, I just bought the whole 127.0.0.0/8 block from some guy a couple of posts above yours.

Bah!

[Greyfox]

Google just needs to start a nationwide ISP with IPv6 on all that dark fiber they're sitting on. They could probably kill off the rest of the Internet inside the USA overnight.

Re:Bah!

[paul248]

Building an ISP with dark fiber is easy! Just set up millions of tiny apartments inside the Internet Exchange buildings, and get some cat5.

Re:Bah!

[Greyfox]

So you're saying they're going to build a nationwide network with dark fiber and solve the homeless problem? Gotta hand it to Google, they're pretty smart!

Google and facebook

[Anonymous Coward]

get google (including youtube) and facebook to switch to IPv6 only sites and see how long it takes ISP to make the move them. They will be fighting for the users. Hay use out ISP we have google and facebook xyz ISP dosn't :)

Re:Google and facebook

[Issarlk]

That would be more like "Wew, no more of those annoying bandwith hungry websites ; we can cancel upgrades to our network."

crap

[gavron]

What a bunch of crap. Congrats for getting /. to repeat it. This is all crap.

Now go worrying about something real...

This isn't it.

E

Skyrocketing prices solves the problem

[Kohath]

If IPv4 addresses become very expensive, people will just ... switch to IPv6.

Yeah. That's how free markets solve problems, be they black, or any other color.

Re:Skyrocketing prices solves the problem

[Sycraft-fu]

Also as I said, I think that's what's really needed. IPv6 adoption will cost money, no question about that. Equipment will have to be replaced, software will have to be upgraded, bugs will have to be fixed, etc, etc. It won't just be "Flip this switch and everything magically works." Because of that it is real easy to have a lot of inertia with IPv4. After all it works, and it works well. All of us chattering here are a testament to that. Also there's been technologies to really easily extend it. NAT is a great example. Homes are getting more computers these days and a single IP would be a problem, but NAT solves it nicely in a way that not only doesn't bother most users, it actually helps them (as NAT implies a default deny on inbound connections, thus increasing security).

So because of all that, we've just stuck with IPv4, long after the major technical issues with v6 have been worked out. We can get the hardware we need these days to route it at high speeds, software support isn't a big deal in most cases, you can do tunneling and so on to support legacy devices and networks and so on. While for awhile it was something that really couldn't be done (the necessary stuff to support it just wasn't there) not it just isn't being done because it costs more.

Thus we likely need a shift in cost to make it attractive. When IPv4 space starts to cost a lot because we've run out, the costs and hassle of IPv6 become worth it.

Re:Skyrocketing prices solves the problem

[marcosdumay]

Only if at least one ISP provides IPv6, of course. And that is how free market players collude to destroy any deregulated market.

ip geolocation databases to suffer. yay.

[Anonymous Coward]

I just hope that the unregulated market for IPv4 means that ip-geolocation databases become invalid and thus irrelevant. ip-geolocation is used for evil.

Re:ip geolocation databases to suffer. yay.

[Anonymous Coward]

Except that IPv6 addresses are essentially globally unique. Geolocation will become much more accurate.

Re:ip geolocation databases to suffer. yay.

[paul248]

IPv4 addresses are globally unique too. That's why we're running out of them.

Re:ip geolocation databases to suffer. yay.

[petermgreen]

nah companies will keep using them because their content suppliers will sue them into the ground if they don't. Their content suppliers will do this because the next level up of content suppliers would sue them into the ground if they didn't.

Broadcasters are generally limited to a particular geographical region and want to be able to advertise an exclusive in that region so content owners sell them an exclusive license for that region.

As long as internet broadcasters are taking reasonable steps to perform geographic restriction they will probablly be ok just as some spill over boarders is tolerated for traditional radio based transmissions.

Not to be too picky ...

[Wrath0fb0b]

But usually when a chart contains a combination of past data and future predictions, it is customary to color the two sections differently or otherwise make a clear distinction between the two. I read that plot and thought (for a second) "Holy shit there are only 2 /8s left!" before realizing that it wasn't December 2010.

Here is a good example: http://media3.washingtonpost.com/wp-dyn/content/graphic/2009/03/21/GR2009032100104.gif

Real problem?

[Anonymous Coward]

(perhaps silly) question:

Isn't it technically trivial to hide 99% of residential customers behind routers which are IPv6 but let the customer think they're IP4?

(with a 5 year timeframe you should have the ability to transition at least as quickly as new legitimate server addresses are needed)

Give the big ISP's a tax break or something if it's really an issue...

Re:Real problem?

[paul248]

While it's technically possible to build a NAT gateway that allows IPv4-only clients to access IPv6 servers, it would be a rather messy blob of state tracking and fake IPs. You're better off deploying IPv4 NAT alongside native IPv6. If a customer doesn't have IPv6-capable equipment, then they'll still be able to access most of the webby stuff for years to come, since anyone running a server of consequence will have IPv4.

IPv4 NAT will impede P2P networking, but if the customer really wants that, they can upgrade to some IPv6-compatible equipment.

Re:Hold on right there!

[Shikaku]

http://en.wikipedia.org/wiki/Internet_Stream_Protocol

hi

[dianearbus]

Except where the corporations can buy whatever regulated market they wish to, and individuals have no choice in the matter. I agree with you, but the last decade, at least, has shown that government regulation in this country is for sale to the highest bidder. Brac Apartments

Haliburton's /8

[soundguy]

ARIN had a booth at Interop last week. I asked them why they don't confiscate the /8 that's assigned to Haliburton, which is mostly wasted. They said they don't want to get shot.

Doubt it

[TheLink]

Seems more likely that what will happen is that all "normal users" get RFC1918 IPv4 addresses.

So you could have say 1000 to 2000 ISP users behind one ISP public IPv4 address, which will be shared to access the Internet. One /8 public IPv4 range can then serve 16 to 32 billion users.

Stuff like WoW, google, facebook, gmail, IM will still work.

But running a public server, Bittorrent and other P2P stuff will be difficult. If you are lucky the ISP might allow you to serve to peers within each RFC1918 "district".

The Media Companies and Powers-That-Be might consider this a feature and not a problem. Since this means locking in to a world of few talkers and many listeners.

Re:Doubt it

[Anonymous Coward]

....or you could use IPv6.

Re:Doubt it

[Amouth]

Bit torrent and P2P would still work - if the client and server/tracker understand how to make connections through double nat..

Re:Doubt it

[TheLink]

You'll need cooperation from the ISPs because you'll have this scenario:

1000 users -> ISP1 NAT device-> Internet <- ISP2 NAT device <- 1000 users

How would one of the users on the left side connect with one of the users on the right side? You can only connect to the public IP of the NAT devices, but to which port?

So the ISPs will have to NAT a port each on their NAT devices back to each user behind the NAT that wants to be a server (do P2P uploads etc). And the user will need to know what that port is.

Re:Doubt it

[TheLink]

Only if all the servers I want to use support IPv6 AND our ISPs support IPv6 and are able to route/tunnel the IPv6 packets.

Otherwise if the servers I want to use only support IPv4, I will still:

a) need an IPv4 address to talk to them (but those are running out)
Or
b) need a proxy or special IPv6 to IPv4 NAT device (which will also need an IPv4 address of its own)

This scenario does NOT work:

IPv6 only client <-> IPv4 only server

No amount of tunneling will help the two talk to each other. You will need to translate or proxy the connections e.g.

IPv6 only client -> IPv6-IPv4 NAT device/proxy -> IPv4 only server.

Or use an IPv4 NAT as mentioned:

Client with RFC1918 IPv4 address -> Conventional IPv4 NAT -> IPv4 only server.

Re:Doubt it

[Amouth]

one of the methods i have seen is using a server to cordinate the transfer (i think AIM does it to set up driect connection between 2 nat connections)

we have 3 computers
A = person behind nat
B = person behind nat
C = publicly addressable server

A and B both talk to C (say C is a tracker)

A wants direct connection to B so it tells C

C then tells B that A wants to talk to it

B then creates a second session to C opening anew reply port in the NAT, C then takes that return contact info and passes it to A

A uses the return info C got from B's second connection to contact B allowing B to now have a conversation with A via A's new connection info.

I know thats a wall of text but it is easier to explain on paper or in person

Re:Doubt it

[TheLink]

Either:
a) all the connections will be via intermediate servers on the Internet - in which case it's no longer P2P. It's Client-Server. Many clients connecting to a few servers.
or
b) The ISP NAT devices have to be configured to map ports back to the P2P clients, as I said before.

Remember, you no longer have a device with a public IPv4 address that's under your control.

Without the cooperation of the ISP's NAT device(s) even if you run your P2P client on 10.1.1.1:1111 and say "connect to 1.2.3.4:5555" (where 1.2.3.4 is the ISP nat device's public IP) to the tracker server, the other side will fail to connect because the ISP's NAT device just isn't translating 1.2.3.4:5555 back to 10.1.1.1:1111.

The ISP can set up static NATs, but someones got to set up some sort of standard to do it:

They could set things up so that:
10.1.1.x will get port 10000+x on 1.2.3.4.
10.1.2.x will get port 10256+x on 1.2.3.4
so
10.1.1.15 will announce on the tracker that it's 1.2.3.4:10015
10.1.1.123 will announce on the tracker that it's 1.2.3.4:10123
10.1.2.5 will announce on the tracker that it's 1.2.3.4:10261

Anyway I think we'll still be using IPv4 for a few more years, the move to IPv6 appears to be rather slow.

ISPs are far more likely to put "normal" customers on RFC1918 IPv4 addresses (with maybe public IPv6 addresses), and make them pay more for public IPv4 addresses.

It'll be crazy to only give their customers IPv6 addresses and no IPv4 addresses at all (because we've run out of them).

Maybe after 10 years of this coexistence we'd move completely to IPv6.

Or, as I said earlier (and got modded troll for it :) ) the various companies might happy with this situation and make efforts to keep things that way.

Re:another shit kdawson post

[bky1701]

Pretty sure he means that entropy is increasing, which in some sense is the loss of energy; at least, useful energy. On the other hand, we're probably just going to be sucked into black holes.

It's not for everyone

[VincenzoRomano]

To really do blackmarket on IP addresses you cannot be anyone.
You need to be an ISP, with at least a LIR to route those IPs almost everywhere or a hosting company so you don't move the IP but host the applications ... that article sounds a little bit non sense.

I'd be using it...

[dandart]

I'd be using IPv6 if only my ISP supported it. I think all ISPs should get on with getting it out there, and then give us one by default. Then no one would have to worry, except the ISPs. Because IPv4's going to run out so soon, I'd recommend a nice round date for the deadline for the Internet switchover - 1/1/2011.

Re:I'd be using it...

[imakemusic]

Yeah, that's a great idea - a massive technical switch over on a day when most people are off work.

Re:I'd be using it...

[dandart]

Woohoo! Massive technical inconvenience! Aww, you saw through my plan... at least it might have been a quiet day... Alright, 2 weeks after everyone gets back?

Re:I'd be using it...

[imakemusic]

Well, I'm not sure that it's appropriate to celebrate a digital landmark on a date with a 2 in it.

Re:I'd be using it...

[dandart]

10[2] * 111[2] binary days? >. Stop punching holes in my logic!

Re:I'd be using it...

[Abcd1234]

I'd be using IPv6 if only my ISP supported it.

Meh, why not use it anyway? Grab yourself a free tunnel from, say, Hurricane Electric, set up your router (well, assuming it supports 6to4... I use m0n0wall, which works brilliantly), and voila, you should be good to go!

Re:I'd be using it...

[dandart]

Tried it for a while. May have to soon again.

Re:I'd be using it...

[Abcd1234]

OOC, is there a reason you went through the trouble to set up IPv6, only to turn it back off again? After setting it up at home, I largely haven't thought about it, save that when I hit google, I notice my browser hits a v6 address (I have an extension that shows the server IP in my status bar).

Re:I'd be using it...

[dandart]

Because I have an OS upgrade by now. That was maybe a year ago.

Re:black market for penis entrance into your mouth

[Chrisq]

I hope you are talking about an exchange for a static IP address, otherwise your post might be seen as offtopic

When IPv6 finally arrives ...

[Taco Cowboy]

... what would happen to the IP Mafia(TM) ?

another selling point

[Chuck Chunder]

Great ping times, test it now!

Re:another selling point

[jaavaaguru]

I just tried it on Windows XP. It's a millisecond or two slower than Ubuntu on the same hardware.

Pinging 127.0.0.1 with 32 bytes of data:

Reply from 127.0.0.1: bytes=32 time=1ms TTL=128
Reply from 127.0.0.1: bytes=32 time=1ms TTL=128
Reply from 127.0.0.1: bytes=32 time<1ms TTL=128
Reply from 127.0.0.1: bytes=32 time=2ms TTL=128

Ping statistics for 127.0.0.1:
        Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
        Minimum = 0ms, Maximum = 2ms, Average = 1ms

Re:black market for penis entrance into your mouth

[upyourshomo]

Can I exchange some of your saliva with my semen?

Only a problem if your ISP doesn't support IPv6

[cheeseandham]

Do something about it, you are a customer after all. (Assuming you have a choice about which ISP you give your business to, and aren't in some horrible monopoly situation)

i) Complain to your ISP, ask them why they don't support IPv6
ii) Threaten to switch to an ISP that does support IPv6
iii) Actually switch to an ISP that supports IPv6, and tell your old ISP why you are moving.

Companies will listen to their wallets, if nothing else.

and yes, my ISP supports IPv6 native & tunneled and has a 6to4 gateway if you don't want to dual-stack

Re:Only a problem if your ISP doesn't support IPv6

[Bert64]

No, they won't unless a large number start demanding it...

Even ISPs like A&A who actually support v6, combined with customers running OS which support it, will probably have routers which don't support it rendering it useless...

What needs to happen, is for government to step in... Require that any provision of internet connectivity supports both v4 and v6 on an equal footing, and that all equipment sold does too. Once every site and every user is running on v6 then v4 can gradually be phased out. Look at TV, every TV sold in the last few years has been capable of receiving both analog and digital signals.

Re:Only a problem if your ISP doesn't support IPv6

[Bert64]

Out of interest, i know someone who runs an ISP which supports IPv6... And just for fun we looked the systems where ADSL connections terminate... Of just under 5000 active connections, 5 of them had IPv6 active - one of them was me, one was his testbox...

It's a sad state of affairs when even geek oriented sites like slashdot don't support IPv6...

Re:Only a problem if your ISP doesn't support IPv6

[cheeseandham]

It's a sad state of affairs when even geek oriented sites like slashdot don't support IPv6...

That's a very good point, and very sad.

Re:Only a problem if your ISP doesn't support IPv6

[Anonymous Coward]

It's a sad state of affairs when even geek oriented sites like slashdot don't support IPv6...

The problem is that the number of people with broken IPv6 configuration is much larger then the number of people with an IPv6-only connection (should be close to zero). So if Slashdot would offer IPv6, it would break the site or slow it down massively for some people, while providing zero benefit to anybody else.

The only way I can see IPv6 starting up is if starts at the client. Use IPv6 for P2P, gaming, file transfer, VoIP and similar things that require client/client communication and always cause trouble with IPv4. For the big servers on the other side IPv4 seems to be here to stay for a long while to come, as even when its NAT'ed or proxy'ed, it doesn't cause much of an issue.

use too much space in 10.x.x.x/8? use 11.x.x.x

[speculatrix]

I worked on contract in IT dept of an international bank based in the UK, actually somewhat in the north, more I dare not say. They used addresses from 10/8 like crazy, and when they ran out, started using 11/8

Some of the skeptics will think I'm making it up. Battle-hardened IT pros will probably facepalm and know it has the ring of truth and that no-one would possibly come up with such a stupid plan and therefore it must be true!

Re:use too much space in 10.x.x.x/8? use 11.x.x.x

[Junta]

I've had the exact same situation.

More commonly, I see people use 192.x where x is not 168 thinking 192 is a class A private network (except they have no idea what those words mean).

Re:use too much space in 10.x.x.x/8? use 11.x.x.x

[AK Marc]

I worked in a place where they used 172.0.0.0/8 for their addresses. It worked fine except all 172.0.0.0 addresses that are public were inaccessible because they'd go into the VPN network and get black holed or go to the private server with the same IP.

Re:use too much space in 10.x.x.x/8? use 11.x.x.x

[speculatrix]

hmm, modded funny. why is there no moderation value "+5 horrific if true"?

Supply vs demand - tail wagging the dog.

[Anonymous Coward]

At the risk of insulting myself by including myself in this group, what's new? The nerds have known about IPv4 depletion now for YEARS. Short supply of a commodity will generally result on a black/secretive market for selling it.
It may reflect poorly on us compared to perhaps the US or Europe, but down here in Australia there really isn't any SUPPLY of IPv6 available regardless of what demand exists in the marketplace.
Telstra, the countrys biggest wholesale/retail telco and ISP *still* doesn't actually have IPv6 available to customers 'in the wild' apart from a few discrete pilots http://www.telstraenterprise.com/SiteCollectionDocuments/Whitepapers/IPv6_whitepaper_TEG_0255.pdf
In fact, there is only 1 ISP touted to be offering v6 (http://ipv6.internode.on.net/) and even then it is without the benefit of some 'features' like 'free unmetered content' that the user gets with a v4 connection.

How can we as users kick up a stick MORE such that the business who generally dole out a IP address (Internet service providers) to us will take note and start doing something? I'm certain the demand is out there for IPv6 but maybe there just isn't the avalanche of users and public opinion that would force the ISPs to get their shit together. Right now it's like the tail trying to wag the dog.
Maybe when users actually start to be either denied a new connection because of no address space - or the ISP service desks are overloaded with issues from people complaining about their RFC "private" address on their supposedly public network interface???

I have a shell company I'll glad sell, which comes

[ajmcello78]

I have a shell company I'll glad sell, which comes with a grandfathered, 15 year old /21 :)

Worse than IE6?

[Anonymous Coward]

I bet that there will be a situation where users will have trouble accessing large parts of the internet in the future due to corporate networks refusing access. This could be worse than the IE6 lock in because even modern browsers would be locked in to the IPv4 space.

We thanks you

[SmallFurryCreature]

Thank you subject #415780. Your cooperation has been noted. We will take immediate action to discipline the member of staff of ARIN you talked to, so that in future his replacement will say the correct thing.

Your commanding master, Haliburton.

Re:IPv4 Address depletion? Not a problem for CSI!

[Anonymous Coward]

Once again, CSI Miami are thinking ahead:-
359.33.9.234
http://adamjh.blogspot.com/2005/10/359339234.html

Re:black market for penis entrance into your mouth

[Anonymous Coward]

You're starting too small. You need to think big if you'll make any money.

Found off...

[spyingwind]

the coast of , in a sunken ship. Get your gold IP for just 10 payments of 29.95!!!1!!!!!one!!!!111

this is a good thing not a bad thing

[Colin Smith]

you just gave gthe economic argument for IPv6.

Nobody gives a shit (about anything) till it starts costing them money. When IPv4 gets expensive, people will move to v6 and sell their v4 addresses.

Don't worry about it.

There'll be a market for them, of course.

[jonadab]

> IPv4/IPv6 coexistence is now expected to last for 5 years.

I'm pretty sure that's a lowball estimate. I do not see IPv6 overtaking IPv4 as the primary protocol of the internet for several times that long at the very least, and quite frankly my money is on never. I see IPv6 being the next MicroChannel Architecture or Itanium: how's it going to ever get businesses (other than the core IT shops who are already behind it) on board, given that it's not even slightly backward compatible? It's a classic prisoner's dilemma, with dozens of participants: if *everybody* does the "right thing", everybody's better off, but if even *one* of the major players doesn't bother, the whole thing is pointless and the practical thing to do is ignore it.

> this is all leading to a black market in traditional IPv4 addresses

I'm not sure I would call it a black market, since there is (to my knowledge) nothing illegal or unethical about reselling such resources. It's not the way things have traditionally been done, but it will be. But yeah, IPv4 addresses are going to be resold (as in, ISPs will probably buy back unused subnet blocks from their customers and then sell them to other customers). Duh.

> that will catch many people off-guard,

Everything that ever happens catches some people off guard, no matter how obvious it is ahead of time that it's going to happen.

> and boost Internet access prices sky-high."

Non sequitur. There are more than 2^31 total public IPv4 addresses, which is in practice (with NAT for systems that don't need to provide public services) more than enough to meet the actual need. Consequently, they don't need to be very expensive, and they won't be. Anybody who tries to charge completely outlandish prices will be underbid.

IPv4 addresses *will* have to cost money, though, going forward, because when they're basically free, as now, organizations snap up entire Class-A blocks, because they can, even if they're only planning three servers and a couple of NAT gateways for workstations. That's why we're running low: it's human nature to take more than you need as long as it's free or nearly free. So the price of IPv4 addresses will go up enough to convince some people to sell off some of the extra ones they're holding that they don't need. An equilibrium will be reached, and that will be that.

My guess is home users who want a public IPv4 address (e.g., so they can remote-desktop into their home PC from work or whatever) will end up paying an extra dollar or two a month for a public IPv4 address (either as a line-item on their bill, or by using a slightly more expensive ISP that bundles it). This is a very rough estimate, but I think the price point that's needed is somewhere around that level. Much more, and they'll be a glut on the market; much less, and nobody will be willing to give them up for sale.

Re:There'll be a market for them, of course.

[Anonymous Coward]

You really don't have a clue when it comes to little insignificant things like routing, do you?

Nortel IP address range

[Mainframes ROCK!]

I thought Nortel Networks, which is bankrupt and almost gone now, had a class A address range (47.x.x.x). Will that not be freed up?

Re:Nortel IP address range

[Anonymous Coward]

And why does the US dept of defense need 184 million ips ? or HP 32 millions. or ford, apple, halliburton, xerox, merit, lilly, merck 16 million each. it boggles the mind. but the stupidest has to be assinging a whole class A to Interop. really. wtf where they thinking back then?

Re:Nortel IP address range

[Abcd1234]

Woohoo! Congrats! You've just delayed the problem by, hmm, let's see... 2 weeks? Maybe a month? Well done!

Re:Nortel IP address range

[petermgreen]

wtf where they thinking back then?
Before the advent of classless routing the only options were a class A, a class B or a class C. Therefore companies/institutions who needed more than a class B got a class A.

T-Mobile USA uses 25.0.0.0/8

[eladts]

T-Mobile USA uses 25.0.0.0/8 for their NAT endpoint addresses. This is not some IT department in a bank, this is a telecommunication company. They should know better.

I'll say it once and I'll say it again...

[Anonymous Coward]

Disconnect China. They don't want to be a part of the public internet anyways. IPv4 problem solved for another 50 years.

Trading "IP options" anyone?

[FenwayFrank]

eom

DNS

[damuhatori]

Roll out IPv6. Anybody that cares enough about having a large address should buck-up and run their own DNS. Problem solved.

Of course, IPv6 still lacking

[Junta]

IPv6 has some immature aspects to it. IPv4 has had decades of hardening in the practical space. As the only viable protocol, all the issues encountered had to be solved and people couldn't bail on IPv4 if it were not fixed. Much of the experience has been translated over, but owners of IPv6 have been keen to try to rethink every aspect of how things are done, frequently missing corner cases or demanding a workflow change for net admins accustomed to IPv4. With no hard requirement driving these admins to adopt IPv6, rejected admins just report back to their organization that IPv6 is infeasible, and walk away from the table since IPv4 works today.

ISC's DHCP server still doesn't have full function in IPv6 compared to IPv4, and resorts to violating RFC 3315 to try to hack in a feature from IPv4.

There are also many standards that are still IPv6-incapable. Examples include IPMI (net configuration directives have 4 bytes for addresses and such) and PXE.

Of course, this is all part of the chicken and the egg, standards improving requires demand in the industry that lacks in an IPv4 dominated world, and IPv6 won't displace IPv4 so long as the standards aren't improved. There are enough people chipping away at it to make some progress, but the rate is significantly slower than it would be if people didn't just go back to IPv4 and take a wait and see position.

Re:Asleep at the Switch

[couchslug]

Can they take her sooner, or is there a waiting list?

Uh, depletion of /8 netblocks? Go figure.

[geekmux]

OK, you had me right up until the point where I noticed on the "shock and awe" graph in TFA that showed the "rapid decline, oh noes!!" of /8 netblocks.

Gee, go figure. We're running out of blocks of IPv4 address space that are 16 million addresses wide. Shocking, I know.

what happened to

[kpjlfm]

What happened to IPv5?? Did it go the way of the Oral A toothbrush?

Re:what happened to

[Darinbob]

The public has long since mentally blocked out the debacle that was Preparation G.

Re:what happened to

[marcosdumay]

IPv5 was an experimental version that never left the lab. (And I'm serious here. Really, no sarcasm.)

How long...

[Restil]

How long would it actually take to switch over? All of the IPv4 addresses are already available as IPv6 addresses. If everyone just magically started using IPv6 tomorrow, the currently allocated addresses would all still work, as would the routing.

I see a few problems. First off, everyone who is currently familiar with how to configure and manage IPv4 would be expected to know how to do all of the same things with IPv6. While I'm certain the learning curve isn't TOO high, there are certainly many people who simply haven't bothered because they haven't had to yet.

Secondly, what legacy operating systems, hardware, and application software still exists that has no support for IPv6? This is no small problem, as I'm certain there's a large number of 20+ year old software still in service that may not be feasible to upgrade. It might have survived Y2K, but this one could finally kill it off. Of course, if this software need not connect to the internet at large, it could live on in an internal network forever.

Of course, no matter how much preparation you do, no matter how much you advertise the change. On D-Day, ISP's are going to have a busy day on the phones, convincing their customers that "no, in fact, the web ISN'T broken. No, we don't still support 16 bit netscape on windows 3.1."

Since, in theory anyway, IPv4 and IPv6 can co-exist right now, it does make sense to demand.. or at least strongly encourage all hardware and software developers to ensure full IPv6 capability along with IPv4 in all of their products from this point forward, assuming they haven't been doing so already, and also make the configuration seamless. If you configure IPv4, IPv6 is already configured using those addresses and works as-is, so on the day that IPv4 goes dead, the software still works without a hitch.

The biggest issue of all, of course, is who decides when to pull the plug on IPv4? While nobody will really just turn off IPv4, there WILL come a day when someone starts using IPv6 addresses that don't correspond to IPv4 addresses and from that point on, it will be necessary for the entire Internet to be in compliance. Who makes that call? And do we listen to them? This isn't the same scale of a problem as converting the broadcast TV to digital was. They couldn't even keep a schedule on that, and it was only one country. How do you ensure that the entire world, including countries we don't all entirely get along with, to agree to a mass-scale, instant upgrade on a specific date? This is likely to be more than just a mild inconvenience. The Internet is used for far more critical applications than twitter and porn distribution. At least, it'll be an interesting news day... assuming I can get online. :)

-Restil

IF we run out

[nurb432]

1 - More NAT for those that don't need public IP ( most consumer devices don't really *need* it . Most business don't need 1000's either )
2 - More people move to v6 and the problem goes away.

Tho i do wish i had bought a small block long ago, but i didn't NEED it so i was responsible and passed.

n00b here with a question

[BenEnglishAtHome]

Well, n00b at some things, not at others.

I pay my ISP a few extra bucks for some static IPs. I fool around with server builds. I build 'em, play with 'em, trash 'em. Wash, rinse, repeat. I never bother to register anything with anybody to associate my IPs with any names. I just expect that if I'm sitting anywhere in the world and I want to look at that page I put on my playtoy of a web server at home, I only need to type in http://xxx.xxx.xxx.xxx./ If I need to fetch the file that I left on that miserable excuse for an ftp server that I have at home, I expect to be able to plop down in any internet cafe in any place in the world, fire up their locked-down, browser-only interface, and type ftp://xxx.xxx.xxx.xxx and grab said file with a couple of mouse clicks.

That's why I pay my ISP for static IPs.

Is there anything about all of this that will break that functionality?

Re:n00b here with a question

[cowboy76Spain]

Yes and no.

In the first part, the address space is not yet over so your ISP can get IPs that it needs in the usual conditions.

If it comes that IP space is finished, or that measures to avoid that happening makes getting more IPs antieconomical/unpractical for your ISP AND your ISP needs more address, it may be tempted to end this option or to translate the costs to you (of course it would normally warn you with enough time to adapt your network, it is not something that happens from a day to another after all).

If you lose your fixed IP (not available or too expensive) then there are two possibilites:

• Your ISP gives you a variable IP. Maybe in your case (connection on all day) it would make no real difference for IPs availability, but for most users (on/off) it means that they can server 50 users with (maybe) 20 IPs.
• Your ISP gives you a private IP that you cannot reach from outside your ISP network. Then you should get a deal from them so they redirect access to a fixed port (say 8080, or 18080) to your servers (10.0.0.1:80, for example). Only specific ports (but that is good since acts like an extra firewall). In this case you could set up a VPN into your network and get a redirection from the ISP public IP to it; and then after connecting you would be in your local network so you only need one NAT redirection, but it means trouble if you try to get in from a ciber where the VPN software is not enabled/you don't have rights to configure it.

Anyway, if it comes to that, I would expect they to chose the latter option because now most of the "casual" users must be using a variable IP already. Of course, it only matters if your ISP needs more IP (or it thinks that your IP is worth more in the market than what you are paying them).

Re:n00b here with a question

[BenEnglishAtHome]

Thanks. I appreciate the info.

I've seen worse, though not in this century

[FreeUser]

I worked on contract in IT dept of an international bank based in the UK, actually somewhat in the north, more I dare not say. They used addresses from 10/8 like crazy, and when they ran out, started using 11/8

Some of the skeptics will think I'm making it up. Battle-hardened IT pros will probably facepalm and know it has the ring of truth and that no-one would possibly come up with such a stupid plan and therefore it must be true!

When I worked at SG Warburg (15+ years ago) I was shocked that they'd just gone ahead and used all kinds of class A networks ... not just the 10/8 reserved for private networks, but any ol' class A they pleased, whether or not it belonged to MIT, the DoD, or whoever. When I told them they would have to reengineer their entire network before they could hook up to the Internet, the lead network engineer replied breezily: "We won't be letting anyone on the 'net for at least another ten years, and by then we'll all be using IPv6 anyway." I replied that, knowing how long COBOL had (at that time) held on, I'd be surprised if we moved to IPv6 within 20 years, and if we did, it would be because the IPv4 address space was well and truly exhausted, and that, if SG Warburg was still not on the Internet in even half that time, it probably wouldn't exist as an investment bank (or any other entity, for that matter).

The company tanked about a year later (for reasons unrelated to my prediction, bad network practices, or indeed technology at all. It was an interesting lesson in what happens to an organisation rife with nepotisim and led by boarding school buddies when markets get a little tough, but I digress), was ultimately absorbed by Swissbank to become UBC, and as far as I know had their entire network infrastructure chucked in favour of something that was at least moderately compatible with public IP addressing. Not that connecting to the exchanges doesn't remain NATting hell, and probably will for at least another decade (they'll be dragged kicking and screaming into the ipv6 world long after the rest of us have moved, and I don't see too many organisations west of the Atlantic moving anytime soon).

Re:I've seen worse, though not in this century

[FreeUser]

UBS, not UBC. Damn typos.

Re:I've seen worse, though not in this century

[TheTrueScotsman]

Not quite. SG Warburg was bought by SBC (Swiss Bank Corporation) and became SBC Warburg. SBC merged with UBS later in the decade. Cool place SBC Warburg. It had wall-to-wall NeXT machines, the taste for which SBC had inherited from O'Connor.

Re:I've seen worse, though not in this century

[FreeUser]

Not quite. SG Warburg was bought by SBC (Swiss Bank Corporation) and became SBC Warburg.

Yeah, I was there for that bit (we bought our Chicago operation back from S.G. Warburg when SBC took over, and went on as a very successful prop trading company for another decade-plus. I did muddle up the acronyms after that ... UBS is indeed correct. SBC Warburg did indeed inherit its NeXT machines from SBC--Alas, when I was there we had no such cool toys to play with (though sparc stations weren't a terrible consolation prize, and lots of world travel sweetened the deal).

He really didn't understand

[kevmeister]

As one of the primary sources for this article, I'm rather distressed that the author really missed much of the point of a talk I made several months ago. When I spoke with Mr. Beckman, he was not clear on how the Internet numbering system works and, while he was was close in this article, he still does not appear to quite get it.

One thing he got exactly right is that "If people have legitimate rules that permit address transfers, they'll use them instead of a black market." There is now a formal ARIN transfer policy which will allow transfers of address space for payment. This is the critical bit that will probably prevent any significant black market from developing and, more importantly, having any real impact on the Internet at large.

The other thing that is absolutely right was his calling me an "pseudo economist". I am an engineer, not an economist, even if I do play one from time to time.

the one things I must say is that the IPv4 address space is near exhaustion and things will change. The adoption of IPv6, it undertaken soon and in a competent manner, looks to be far the most likely way to the future. Not the only way, but the only way I see to continue the growth of the Internet as we know it today. It does not mean that massive NAT implementation, which will eventually re-shape the Internet into a very different thing, won't be what happens.

Then again, I am only a "pseudo Economist" and even the real economists don't agree very often.

Re:He really didn't understand

[mbeckman]

When I spoke with Mr. Beckman, he was not clear on how the Internet numbering system works and, while he was was close in this article, he still does not appear to quite get it.

Kev, Thanks for participating in the article. In our interview it was never my intention to try and convey to you my depth of knowledge, but rather to elicit from you a clear explanation of your NANOG paper. Which I did, thank you very much! I assure you, however, that I'm reasonably competent at IP addressing concepts, since I run BGP myself and also deploy BGP and IPv6 implementations for enterprises. And I teach IPv6 courses around the world (I was in Toronto doing ao last week, which is why I had a deadline crunch on your interview).

And in other news...

[Anonymous Coward]

The terrorists may win, film at 11.

How am I an ideologue?

[Sycraft-fu]

Because I suggested that this wasn't a black market? It's not, regardless of what you think about regulation on any level. A black market is, by definition, selling illegal goods. Selling IP space wouldn't be a black market at all it would be, well, just a normal market.

Or am I an ideologue because I dared suggest that it might help solve a problem? Here's news for you: Free markets can solve problems. Not all problems, but then nothing solves all problems (if you think heavy regulation solves everything the former USSR would like a word with you). However there are certain problems they are well equipped to solve. One is dealing with demand for a scarce resource, as IPv4 IPs now are.

I'm not saying that "The free market will fix everything," I'm just saying that a rising cost in IPv4 space may help lead to an economic incentive to go to IPv6. If you've followed the current situation at all, you cannot argue that the current setup has done a good job promoting IPv6. The tech is there, the will to move isn't. Ok, well make IPv4 more expensive and perhaps that'll change. If the situation is as it was, a couple million (or more) to upgrade to IPv6 equipment or a couple thousand (or less) to get more IPv4 addresses, well it is easy to see why people choose more IPv4. However if the cost of the IPv4 addresses shoots up, maybe then it becomes worth changing.

Also there have been continual calls for companies with /8s to give some back. However, for now, there's no reason. Renumbering your IP space is a bitch and takes a lot of time and thus money. However if those IPs are now worth something, then maybe they'll take the time to do so to be able to sell some.

So sorry, but I don't see the "ideologue" thing here. I'm not saying this is a a sure fire solution, I'm saying maybe it'll help break the status quo we have now. If you can't see that, perhaps you are an anti-free market ideologue who simply refuses to acknowledge that a free or semi-free market can ever accomplish anything.

Re:How am I an ideologue?

[Myopic]

Sorry my man. I was taking a potshot at free-market ideologues in general, not at anyone specifically -- but if it were directed at anybody it would be Tuzanor, not you. You took offense where there was none directed at you, so we need not argue, we can be friends.

Re:another shit kdawson post

[robot256]

If the universe is expanding, but energy is conserved, then the energy density of the universe is approaching zero and there will be no perceivable energy! But that theory is as full of holes as yours is.

Re:black market for penis entrance into your mouth

[upyourshomo]

bummer your right, my baby-carrot sized wiener is wayyy too small to make any money. can I borrow someone else's (plz)?

I could tell you...

[Anonymous Coward]

... but you might not swallow that.

Nortel's Class A

[Anonymous Coward]

Has anyone factored in that Nortel is in bankruptcy and will be liquidated within the next couple of years? Nortel has a full Class A address space (47.x.x.x) and a few Class B's. Whether these addresses are reclaimed or sold in bankruptcy, it should serve to stabilize the price.

Re:Stop trying to play expert gmhowell, you poseur

[Anonymous Coward]

Oh look, it's the troll that's been stalking Clone53421, Squiggleslash, Tom Hudson, and now GM Howell too.

For those new to this, the troll is someone claiming to be a defender of a certain Alexander Peter Kowalski, the author of a tool, apkapp2backgrounddaemonprocessengine, generally considered malware by a large number of anti-malware companies and organizations.

CA

PestPatrol

SpywareDB ("Dangerous!")

Freedom Anti-Spyware

Spycheck (Spanish-language) - "Recomendacion: DESACTIVAR Y ELIMINAR"

Spyware No-More ("Threat risk: High risk", "Advice: Remove This is a very high risk threat and should be removed immediately as to prevent harm to your computer and / or to protect your privacy")

Mr Kowalski, or his admirer, got upset because someone had the audacity to link to a threat describing Kowalski's attempts to remove some embarrassing comments posted under his name. Rather than deal with it maturely, this person has been attempting to stalk said poster and those who pointed out Kowalski wasn't doing himself any favors.

So if you see these comments posted as replies to clone, squiggleslash, gmhowell, or Tom, now you know why they're appearing. And if you feel like joining in, making it clear to Mr Kowalski that spamming, sliming, and stalking are unacceptable, well, come on in, the water's lovely!

gmhowell now posting as AC to avoid answering?

[Anonymous Coward]

Do you have a degree in CSC or CIS gmhowell? No?? We thought not. You're by no means an expert to comment on anything in the art and sciences of computing then. The same things you note have happened to others that write good softwares, such as Dr. Mark Russinovich of Microsoft even and also Nir Sofer of Nirsoft, to name only a couple. Your lack of expertise is showing itself gmhowell. Go get a degree and then get back to us, because until then? Your credibility as an expert is non-existent and your avoidance of answering a simple question just says it all.