What Happens When IPv4 Address Space Is Gone 520
darthcamaro writes 'We all know that IPv4 address space is almost all gone — but how will we know when the exact date is? And what will happen that day? In a new report, ARIN's CIO explains exactly what will happen on that last day of IPv4 address availability: '"We will run out of IPv4 address space and the real difficult part is that there is no flag date. It's a real moving date based on demand and the amount of address space we can reclaim from organizations," Jimmerson told InternetNews.com. "If things continue they way they have, ARIN will for the very first time, sometime between the middle and end of next year, receive a request for IPv4 address space that is justified and meets the policy. However, ARIN won't have the address space. So we'll have to say no for the very first time."'
Hmmm (Score:4, Insightful)
However, ARIN won't have the address space. So we'll have to say no for the very first time.
Hmmm, maybe that's part of the problem? They never say no to anyone. Do all those companies really need all those IP blocks? Maybe if they had said "no" once in a while we'd have another year or so to work out how we'll get everyone over to IPv6.
Re:Hmmm (Score:1, Insightful)
Much like how if we had conserved our petroleum resources in the beginning, we wouldn't be freaking over the potential for shortage in this age...
Re:Hmmm (Score:5, Insightful)
Whatever. The world has had how long now to move to IPv6? If we had two additional years, we'd be talking about this two years from now instead of right now. I've been using it for nearly 10 years now. I just hope that this threat is finally becoming significant enough to get ISPs and other organizations moving faster in the right direction.
Re:Hmmm (Score:5, Insightful)
To be fair, we've had almost 10 years. Strike that, 12 years.
We've even had all OS and router support for 5 years.
Fact of the matter is, nobody's moving to IPv6 until they *have* to. We can cry doom and gloom all we want (we have been, after all), and nobody cares. When Comcast can't address new customers, they'll get off their ass.
Though that's a bit of a gamble. The right answer is moving to IPv6, the best answer is doing that in advance, but they'll definitely consider just NATting new customers. Hopefully they'll do things properly, but this is ISPs we're talking about.
Re:Why run IPV6? (Score:5, Insightful)
The Internet was designed so that any computer could connect to any other computer. This is evident in the design of things like FTP, etc.
Every phone, watch, fridge, TiVo, computer, and printer should have a public IP address. Imagine if you didn't need to port forward for Bittorrent, if Skype could connect right to your friend's computer, or you could print to your home printer by just entering its address. That's how the internet was/is supposed to work.
NAT breaks this. Behind a NAT box, nobody can address a specific computer - only the NAT itself. This happens to lend some security, but is essentially accidental. With IPv6, your home router will instead be a firewall. Each computer will be addressable, but will still need to pass through.
Plus, there's enough address to give each subscriber many thousand. And they don't need to change. No more charging for a static IP...
Also, routing is more efficient since it can be done properly by hierarchy.
So there's a bunch of reasons. Pick some.
Re:Hmmm (Score:3, Insightful)
but they'll definitely consider just NATting new customers.
Trouble is, 99% of users won't even notice. If they profile the users to figure out which ones won't notice beforehand, even more.
in the short term... (Score:3, Insightful)
in the short term it will add value to IPv4 addresses, and organizations not using them might *gasp* make money getting rid of ones it doesn't need. That's not a bad thing. We have this problem with spectrum too, there's no particular cost in having a huge chunk idling away once you've got it. Anything which motivated more efficient utilization is good, and money creates a motivation.
A short term will drive up the cost of IPv4 addresses will, in turn, make IPv6 look much more economically viable to people who actually pay for things. As with everything else in the real wold: money makes things happen. IPv6 isn't magically cheaper than IPv4, so no one has been all that bothered about it, so either you lower the cost of IPv6 or raise the cost of IPv4, and running out of IPv4 addresses manages the latter nicely.
Re:So now the question is... (Score:3, Insightful)
They could reclaim blocks from companies and then hand out 1 IP for them to run behind a NAT firewall
I believe that's already being done. Though I believe the biggest single owner is DoD.
they could start to charge for IPv4 addresses on a yearly basis
Good idea. Never happen.
I've advocated charging a higher fee for second level domain names for a long time. After all, if you really need one, paying $30/year or even a lot more, is a minor expense compared to your hosting costs. It would put an end to cybersquatting. But every time I suggest it, I get flamed half to death. People won't pay a penny more than they have to for something, and never mind the consequences. Call it the WalMart effect.
The only solution is to move to IPv6. But, as you point out, people won't do this until they have to.
No, worse, they won't even begin preparations. Not a big deal for most of us, but the changeover is going to be non-trivial for ISPs, manufacturers, and a lot of other people who do Internet infrastructure.
When I was at Sun, I was on a product team for a new product with an embedded Service Processor (for remote control, diagnostics, lights-out management, etc.). Whenever I suggested that the new SP have IPv6 support, I was told "none of our customers is asking for this feature."
Re:Auction? (Score:3, Insightful)
Yeah, there are calculations. They all come to the same conclusion: The effort needed to get those addresses back in to use is enormous and the benefit would be that the final deadline moved 12-18 months forward...
In other words, it's not even close to being worth it.
Re:The Internet is Full (Score:5, Insightful)
That's not necessary, IPv6 already has the IPv4 address space blocked off and reserved for IPv4 addresses, so all you need is protocol translation for the systems that can't understand IPv6. It's not a hard problem. Yeah it will cost a little money, but really it's a drop in the bucket compared to everything else a business needs to deal with.
You band-aid it until you can justify the necessary overhaul. Eventually everyone will be on IPv6.
In other words, the reason nobody is rushing to fix it is because it's not that big of a deal. The problem is small enough that you won't really need to worry about it until it actually comes up.
Re:Hmmm (Score:3, Insightful)
The reason nobody is rushing to fix it is because it isn't a big problem.
It's not like the Y2K bug, where stuff could blow up if it wasn't fixed before the clock struck midnight.
You know what is going to happen the first time ARIN says no? The organization will go "Oh, ok.Can I get a nice block of IPv6 instead?" and add some protocol translation to their network to deal with anything that can't handle IPv6. Done. Problem solved.
In other words, there is nothing to freak out about at all.
Seriously people, get a grip! We've known the solution to the problem since the early 90's, at least, and implementing it is trivial.
Re:Why run IPV6? (Score:3, Insightful)
Well, personally I'm not into BSDM. NAT is an unnecessary pain and a ugly hack that raises complexity and breaks stuff.
Contact tne Class B holders (Score:3, Insightful)
...and offer them some serious wonga to switch to IPv6 and/or make more use of DHCP/NAT etc.
A lot of Universities have class B blocks (and a lot of those addresses are assigned to Ethernet cards now sitting in dusty cupboards and landfills). Still a non-trivial job, but probably easier for universities than big business.
Universities are gagging for cash at the moment - and even if all the cash is spent on the switch
Or the gub'ment can make them do it. Here in the UK, back in the 80s, the powers that be were forcing universities to use the ISO networking protocols: forcing them to switch to IPv6 is far less silly than that (e.g. unlike the ISO stack the IPv6 protocol actually exists and has been implemented by people).
Re:Hmmm (Score:5, Insightful)
However, ARIN won't have the address space. So we'll have to say no for the very first time.
Hmmm, maybe that's part of the problem? They never say no to anyone. Do all those companies really need all those IP blocks? Maybe if they had said "no" once in a while we'd have another year or so to work out how we'll get everyone over to IPv6.
Too late. Hindsight is 20/20, etc. Does MIT really need a /8? No. Does HP need two? No. But as with any scarce resource when no more IPv4 addresses are available they will rise in value and people will auction off their space. The price will have an upper bound at the cost of deploying IPv6. That'll buy us another few years. And then people will NAT even more. That'll buy us a few more. And by that time most people will be ready to move to v6. There really is no need to panic here. I'm not sure where all of the anxiety stems from. The people that understand the issue and care about it are aware of it and on top of it. I suspect an ulterior motive.
Hosts that dont need visibility will be NATd (Score:3, Insightful)
Phones, TVs, and millions of other devices that will never need to act as servers will be forced behind NAT walls.
There will be two price structures, client access and server addresses.
Client, will be NAT only. Server will have a real address whether it be fixed or variable.
Maybe they will even charge by DHCP lease time statistics.
Eventually, the entire IPv4 address range will be relegated to servers. And all the clients will be IPv6. They will be told that the "tunneling" is just temporary, but it will in fact be permanent.
In related news, Pacific ocean found (Score:4, Insightful)
Similar to the expansion of the US "wild west", we're due for years of backfilling and territory arguments. Look ahead to the owners of /8 address ranges having them confiscated. (MIT, for example, hardly needs it: they should be NAT'ing all their internal traffic anyway to prevent "computer science majors" from pulling stupid stunts like the David LaMacchia case (http://www.informit.com/articles/article.aspx?p=169520).
NAT is notoriously lighter weight to support than IPv6, and helps provide some border control of undesirable services from inside your network. Replacing the router infrastructure and the configuration tools for stable, legacy systems to support IPv6 is expensive and the benefits of IPv6 are frankly underwhelming. It's exciting "auto-configuration" is, in most cases, a horrendously bad idea for public facing systems, and private systems don't need it. Useful security features, such as IPsec, were backported to IPv4. And the robust technical features of IPsec seem to be overwhelmed by the far easier to use client behavior of PPTP.
Multicast? Oh, dear. Do _not_ get me started on the flaws of multicast programmers decided that the lack of information about missed packets in multicast forcing them to rewrite TCP, badly, as an unstable software layer on top of multicast.
Promote IPV6-only free porn, games and warez (Score:3, Insightful)
Re:Hmmm (Score:1, Insightful)
I think the receivers will be more than willing to sell that class A...
Re:The Internet is Full (Score:5, Insightful)
No, it's artificial scarcity because the demand only exceeds the supply because those who control the demand (e.g. ISPs) choose to limit the supply by not upgrading their networks to use IPv6.
Re:Over twenty years ago (Score:2, Insightful)
.. and I've worked in telecommunications for over twenty years and suffered under OSI during that period, and can assure readers that OSI sank under its own weight and richly deserved the end to which it came, like similar "committee-generated" works.
The author might be familiar enough with the current activities of the ITU-T (the drivers of OSI - still responsible for international telecommunication standards) to see that it's the bodies like the IEEE and IETF that are making the running in modern comms standards and have been for some time.
I say all this as a present ITU-T member, by the way.
Re:The Internet is Full (Score:2, Insightful)
it's not artificial, because the choice to upgrade is not cost-free for the ISP.
There is nothing artificial about the costs of V6, implementing is very expensive and difficult to justify, when there are no major content providers using IPv6 address space -- so providing IPv6 connectivity gives basically no gain and no immediate competitive advantage.
There would be large hardware costs in terms of network equipment.
And large software licensing costs for updates.
And large administrative costs in the form of evaluating all services for IPv6 compatibility and rebuilding systems that are not, using V6 compatible software (e.g. re-doing DNS systems using V6-compatible DNS server software, which may increase hardware requirements).
Implementing end-to-end IPv6 is expensive, so is the re-training of all network operations.
Re:The Internet is Full (Score:5, Insightful)
Uhuh.
Let's pretend, just for the moment, that this idea isn't ridiculous (it'd be simpler to deploy v6 than to get all those operators to re-number their networks). The current projected timeline for the remaining 20 /8s to run out is September, 2011, which is 17 months away. You propose to return 26 /8s to the pool. So, assuming the rate remains constant (which it won't), that gives us, what, 24 more months? Maybe?
Wow, way to go big guy! Instead of 2011 for IANA exhaustion, it'll now be 2013! Problem solved.
Re:The Internet is Full (Score:3, Insightful)