Slashdot Mirror
IPv4 Unallocated Addresses Exhausted by 2010
An anonymous reader writes "Ars Technica is reporting on how the unallocated IPv4 address pool could run out as soon as 2010. The IPv4 Address Report gives details on just how fast the available pool of IPv4 addresses is diminishing. Will ISPs be moving towards IPv6 any time soon? Or will IPv4 exhaustion become the next Y2K?"
Despite the best efforts of organizations like ARIN, the simple fact is that, compared to IPv4, IPv6 gives you access to very little content and very few users. So far, nobody has been able to get past this chicken-and-egg issue, although a The Great IPv6 Experiment proposes to change this by giving away free access to "10 gigabytes of the most popular 'adult entertainment,'" but only over IPv6.
Is IPv6 so unappealing that they've gotta bribe people with pr0n to use it?
They could delay the inevitable by reallocating existing IPv4 space more efficiently. Many old/historical allocations are inefficient. Apple Computer, for example, has all of the 17.x.x.x space, comprising 256^3 = more than 16 million addresses, which is just plain absurd in this day and age.
predictions
every year there is a new nutter predicting the end of the world. Havent we heard of this argument before? Would it be a good idea to take ownership of those class A spaces that quite a few companies are hoarding??
That would require a change to all TCP/IP stacks, and replacement of core routers.
Why not just fix the problem outright if you are going to do that?
Ease adoption how, exactly? You still need to update the protocol, then update all the software, and all the hardware, and all the documentation and training... you can't just tack that on to existing implementations of software.
If you're going to force all that change, then change to something that isn't a silly half-arsed hackjob.
That's really just not true. With IPv6, you can get a lot more anonymity than you have now with IPv4. v6 has all sorts of special provisions for randomly assigning addresses, letting you reset them when you want, so that you can appear to be a new user in the middle of a browsing session. That's tough to do with IPv4; even if you try a DHCP release-and-renew from your ISP, generally they won't issue you a new address until the other one has expired.
IPv6 doesn't force you to give up any privacy, and there's no 'user serialization' unless you buy into it voluntarily.
"Ladies and gentlemen, my killbot features Lotus Notes and a machine gun. It is the finest available."
I'm continually amazed at the number of people in the IT and Net industry who keep "wondering" when IPv6 will arrive. Its been here for a long time. I'm running a series of web servers for internal company use that have native IPv6 addresses. For public consumption, we have an IPv4 reverse proxy that allows us to run our entire web services behind one IPv4 address. Any customer who has an IPv6 address gets to talk to the individual servers.
/64 v6 address for a cheap price. You'll design your websites to be usable on v4, but for management tools, etc, you'll need to install a v6 tunnel.
The advantage comes when you consider management. In order to have 20 SSH/FTP/etc accessible Internet servers, I'd either need 20 separate IPv4 addresses (getting a decent segment of a class C here is expensive), or I'd have to play fun games with ports. All our technicians have IPv6 on their laptops, and use tunnel brokers for access to the v6 network.
Most of our clients have IPv6 connectivity, though they don't notice it. When we put in a firewall, IPv6 comes default setup with tunnel brokers.
People keep asking, when's there gonna be v6 content? There is no v6 content (ok, their is full colour ascii starwars). Any content provider would be nuts to say "you have to have v6 to see our content" at this point (with the exception of mobile phones). IT Techs brought v4 to the public, we'll bring v6 to the public. Its technicians like myself who appreciate having an Internet accessible toaster (ok, so its not yet accessible) that have already started the ball rolling.
Before long you'll see hosting providers saying, you can have one web gateway shared v4 address and a
I use to have a funny sig, but slash cut it off, and I forgot what the punchline was.
Just move slashdot to an IPv6 only address; voilla by monday every corporate will have a functioning IPv6 setup... ;-)
Oil out of supply in 1999, Global warming killing everyone in 2005, P2P piracy ends with Napster, Limewire, Kazaa.
Seriously it's all just FUD, There's an expiration date, but 2010? What happens when we make a few Class As into Class Bs? oh that's right, more time. I think the key is to figure out how to make the best "IPv6" and a way to make it so my old commodore 64 is willing to work with it (whether that be ISP level conversion or a inexpensive hub, note INEXPENSIVE)
Do I have a commodore 64? Not any more but the point remains there's literally a million devices out there only able to communicate with IPv4. There's actually a million people out there not willing to go through the hassle of going to IPv6 (and probably about that many who are unwilling to change) and if the way they are pushing to get people to switch with FUD like this, I'm guessing it's more than a couple million who don't want IPv6, so it's time to ask ourselves, how can we make IPv6 more attractive than staying with IPv4, and implement these ideas. IPv6 will likely overtake v4 one day, but come on, let's find a way to make people switch rather then just wait for it to happen.
They don't allocate IP addresses, they allocate routes entry and with route entries, you get way more addresses than most need. The solution for this is to start allocating non-contigious /24... Force everyone to fix their routing and treat the wold as a 2^24 /24 ranges and get over it. To do this right requires less than 8mb of cache tag ram in most routers that want full feeds and enough ram to process the bgp routing updates.
Going to IPv6 doesn't fix the fact that routers are running out of routes. This problem will get plenty of attention in about 2 months when the big Cisco routers start to dump routes because they are too big and adding IPv6 only makes the problem much worse.
But if you talked to @Home's people as individuals rather than Corporate Employees, almost all of them would say "Well, Duh! Napster is the reason that people are *buying* broadband internet connections, of *course* we like it."
And, ok, the paranoia about servers on home cable modems was partly because their early trial equipment didn't work very well and they had no way to regulate individual upstream bandwidth usage, and PacBell's dishonest "Cable Modem Web Hog" ads made them really worried about perceptions of slow performance, but they were worried that somebody would run a pr0n webserver from home, become Cool Site of the Day because doing that on cable modem would be cool, and trash their neighborhood's network performance while causing a lot of publicity. And unfortunately most of the cable companies have not only not recovered from that attitude, they've been propagating it to the DSL providers, and they've been learning other cluelessly paranoid attitudes from the Australian ex-monopoly who thinks you should cap the total monthly download of their users (since that used to be expensive in Oz), and cap it to a ridiculously low level like 1GB/month, which is like 1.5 days of continuous 56kbps usage.
But when I had my corporate hat on, especially if I was talking to non-California customers, it was certainly much more proper to talk about the big internet usage being for music piracy than for pr0n
Bill Stewart
New Fast-Compression-only CPR http://preview.tinyurl.com/dy575ks
Many years ago, I worked at Network Appliance. For their webcache servers (which ISP's could use to cache popular pages for their users), every so often we could see what people were looking at. This was because once in a while the kernel would crash, and so we'd get a core dump back. Complete with images of what was in RAM.
For many (if not every) coredump, 90% of the web pages were porn. I kid you not. If you dont believe me, just go ask Guy Harris (yes, of Ethereal/Wireshark and other fame) as he was one of the top kernel debuggers back then.
This was truly depressing for someone who has spent much of his working life building up the Internet, from the protocols, to various UNIX OS's as well as other stuff. Yes folks, the major use of the Net is Porn. NetApp confirms it.
Sigh.
Claiming to be pedantic on Slashdot is asking for trouble
I made the assumption that each state/lata/switch/tower was doing their own nat.
09f911029d74e35bd84156c5635688c0
Your scheme would only permit a _single_ tcp connection between any pair of hosts attached in the fashion you describe, since a TCP connection is identified by the tuple (src ip, src port, dest ip, dest port). So you'd wind up inventing a whole load of connection multiplexing to go with that NAT.
Frankly, that sounds like more engineering work than switching to IPv6.
So right now we have a flat address space of 32bits or so. Why not
create multiple internets, one per country lets say. Everyone
gets to keep their existing internet address. Its just encapsulated
within a country network.
In order to get to country A address B.B.B.B you have to use
a route. Each ISP would have a special router address that would
send packets to that country accross a "dedidcated" connection. Your
computer would know that when DNS assigns a "zip" for a particular
connection, it locks the routing for those packets to go out via
the local ISP dedicated router address.
Your computer knows what router to use because it got the "zip code"
for that route when it did the DNS lookup.
Yes, I realize there would be problems. But perhaps less problems then
with IPv6 adoption?
This is moving to a hierarchial model. And the extra address space
comes from the routing tables.
Its just an idea. Please be kind.