Slashdot Mirror
IPv4 Will Not Die In 2010
darthcamaro writes "A couple of years ago, the big shots at IANA (that's the people that handle internet addressing) issued a release stating that the IPv4 address space was likely to be gone by 2010. Here we are in 2010 and guess what, IPv4 with its 4.3 billion addresses will NOT be all used up this year. In fact there could be another two years worth of addresses still left at this point. 'We're at about 10.2 percent (IPv4 address space) remaining globally,' John Curran, president and CEO of ARIN said. 'At our current trend rate we've got about 625 days before we will not have new IPv4 addresses available. We're still handling IPv4 requests from ISPs, hosting companies and large users for IPv4 address space, but that's a very short time period.'"
IPv4 doesn't die - it just runs out of available addresses.
IPv4 not dying... Enterprise Networking Planet confirms it!
That just doesn't have the same ring to it. ):
THL phish sticks
Another two years? Good, now we can all can put off panicking for another two years and not do anything to resolve this in the meantime.
Does it make you happy you're so strange?
"...At our current trend rate we've got about 625 days before we will not have new IPv4 addresses available..."
I think this:http://www.xkcd.com/605/ sums it up
Domain squatters and the like use one IP (and one server) for thousands and thousands of domains. They're parasites but they're not using anything like a significant fraction of the available IP space.
Thanks to Apache and the miracle of Virtual Servers, one can use one IPv4 address to host thousands of domains! This depends on HTTP1.1, though, and old browsers can't handle it, but nobody cares about them.
See: http://en.wikipedia.org/wiki/Domain_name#Use_in_web_site_hosting
In conclusion, your argument is invalid.
For context, click Parent.
We'll never be able to justify the cost of implementing IPv6 properly until it becomes something customers are demanding, and that won't happen until there is stuff on the Internet people want that to reach couldn't get hold of an IPv4 address.
Still, I suppose I just have to be patient.
Big F...ing deal! How many predictions are accurate for three or more years? The original prediction was made in May, 2007 and current prediction has slipped the date from December 2, 2010 to November 18, 2012 not quite a 2 years. I challenge anyone to find accurate predictions that are 3 1/2 years old.
We need to be moving to IPV6 as quickly as possible. We may have a bit longer than was predicted 3 1/2 years ago. The thing that is scary is have we made much progress in moving to IPV6 in the last 3 1/2 year? I think not much. So, whatever the actual exhaustion dates are for IPV4 address. We can be certain that we are 3 1/2 years closer than we were and we have done almost nothing to prepare.
...another financial crisis. Because that's the reason there was a slump in allocation rates. The current best projection for IANA pool exhaustion is Sep/Oct 2011. Without the financial crisis that would have been end of 2010. The IANA guys would have been dead on, if not for a once in a 100 years financial event.
/8 is theoretically usable, even before subtracting the legacy allocations. The summary makes it sound like it was a doom-and-gloom prediction that didn't happen to be true, but that's not the case.
The tone of the submission is really silly. There wasn't 4.3B allocatable addresses in the first place. Out of the 256 "/8s" only 219.914
Also, it's "not the next 2 or 3 years", based on the available number of addresses 1.5 years for the IANA pool and 2,5 years are hard bars until RIRs (regional internet registries) run out.
It takes a man to suffer ignorance and smile
Be yourself no matter what they say
Nothing new there. The university I work at have a /16 network. Everything has its own ip, even projectors. And by God thats how its supposed to work
Having live IP addresses is the way it should be done. NAT offers no more security than a simple firewall in this case.
I predict that 2012 we will still have available IPv4 addresses.
This will happen because some IPv4 addresses will be reallocated as client-side doesn't need IPv4 addresses in IPv6 to access IPv4 resources. So IPv6 adaptation it self will slow the need to migrate to IPv6 as singular Internet Protocol.
Whether or not the issue will be forced, the problem is that for most of the developing world they already are either running out or pretty damn close. Because of this, if the US doesn't jump on the band wagon we will continue to be outpaced by countries like China that are already neck deep in rolling out IPv6. This isn't a matter of when, just if, and really ought to be done gradually, but quickly, rather than wait till a moment to be forced. I encourage anyone that can to move as quick as they can towards this rather than sit and wait and watch the world pass them by.
The musings of just another geek and his junk.
...if we actually went after those who currently hold "monster" /8 and even /16 blocks that aren't doing squat (pun intended) with them.
When the IPv4 addresses run out, those "monster" holders will be doing something with them. Selling them.
The "monster" holders are big IT players, and they would never give away something that they see could be a valuable asset in the future.
Go knock at HP's door, with a bowl in your hand, and say: "Please, Sir, can I have some more IPv4 addresses?"
"More? You want MORE!"
Schroedinger's Brexit: The UK is both in and out of the EU at the same time!
No. Repeat after me, there is no market in IPv4 addresses. The current rule is that when a RIR requests a block from IANA that would bring the IANA pool below 5 /8s, then every RIR gets one last /8 from the "final five". Then IANA is done and the RIRs have whatever addresses they have left in their unused pool. For AfrNIC it'll last decades, for APNIC/ARIN it's curtains in about a year.
There is no market in IPv4. There never will be, because reclaiming addresses is too hard and routing can't handle it atm (routing too small blocks). Let's switch to IPv6 already, for fuck's sake, we'll have to do that anyway even if a miracle happens, technical problems get worked out and someone sets up an IPv4 market, about 6 months after.
It takes a man to suffer ignorance and smile
Be yourself no matter what they say
You're correct... I'm careful to point out the uncertainty when doing the interviews, but reporters tend to lock onto the IPv4 depletion countdown number regardless...
IPX won't die in 2010, either.
But, in all seriousness, there's a few things to remember here.
1. The v4 address space will be exhausted in the foreseeable future.
2. Reclaiming large blocks only delays that inevitability by a few months.
3. With a few exceptions, modern, supported OSes (Windows [2003, 2008, Vista, 7], GNU/Linux, all of the BSDs, OS X) support IPv6 perfectly.
4. Most of the critical applications support IPv6 perfectly.
5. The big holdup on the consumer side has been with the ISPs. The DOCSIS 3.0 roll-out is ongoing in many places.
6. The US government has mandated it. The compliance date was in 2008 for all of the Federal agencies on their backbones. It's just a matter now of getting ISP access to those sites, and configuring lower-level systems.
The luddite attitude here about this is amazing. If you're really all that concerned about it, and don't want to focus too much on the nuts-and-bolts, here's some advice: Learn BIND. Setting up your resolvers properly will spare you headaches.
I use IPv6 every day. I get lots of e-mail over IPv6 (netbsd and freebsd mailing lists, to name just a couple). I enjoy being able to ssh to all of my machines at home directly. It's here. Evaluate your crap, and see what's not going to work. Plan to replace that stuff. Most of it probably will need replacing by the time you get assigned a /64 or /48 by your ISP, anyway. This isn't rocket science. /rant
Reclaiming all the legacy IP addresses would buy us 6 months tops. So we delayed the problem by 6 months, during which we would be fucking up the DoD, IBM and a handful of other companies, I'm sure it'd be worth it, it's not like the military would be fighting somewhere and they could pull off a massive networking restructure in less than a year for the 8 /8s they're holding on to.
It takes a man to suffer ignorance and smile
Be yourself no matter what they say
Maybe THIS is the end of the world everyone's talking about...
Will his website be at address 255.255.255.254?
If something is so important that you feel the need to post it on the internet... It probably isn't that important.
Give them 5 class B's of their choice from their current class A, and reclaim the rest, then this issue will go away for a long time.
No, it won't. The equivalent of one class A net of IP addresses gets assigned every 28 days. It would buy a few months and the resources are better invested in the transition to IPv6.
why can't all little-visited domains be on virtual hosts and share their IP address with many others?
Many can but as of right now if you want to use ssl/tls you pretty much need your own IP.
with ssl/tls the server does not have the http request at the time the connection is negotiated and certificates checked so it can't use the name from it to decide what certificate to present.
You can in principle have multiple domains on one certificate but it makes the certificate management far more of an administrative PITA (essentially the host would have to apply for a certificate on behalf of all the domains they host on an ip and get a new one every time a domain needed to be moved between machines)
There is a ssl/tls extention which tells the server which domain is being requested during the ssl handshake so it can send out different certificates for different domains. Unfortunately the built in ssl support in xp doesn't support it (both IE and chrome use the windows built in ssl support, firefox doesn't).
note: i'm known as plugwash most places but i screwd up registering that here somehow in the past and now can't register
Having live IP addresses is the way it should be done. NAT offers no more security than a simple firewall in this case.
Ah, no, having live IP addresses is the way it was done, back before viruses, trojans, 10,000-node botnets, and Microsoft got involved. That "simple" firewall you speak of is now absolutely mandatory for damn near any business or home today.
Trying to run the Internet "traditionally" on a 30-year old protocol is like trying to drive a Model-T on the freeway. Neither of the original designers ever envisioned what the future would bring.
And yes, I realize that IPv6 design will help eradicate NAT, and get it back to the way it "should" be from an addressing standpoint, but one can also see that just from a Security view, IPv6 is a far cry from v4.
The projector is mounted on the ceiling, and your laptop is on your desk. You wirelessly hook up the projector as an external monitor.
You're in a lab, and your media server is in your room. You hook up and start streaming music. The possibilities are endless.
All sorts of things Just Work so much better with ipv6.
Yeah if only there was a way to assign names to these hard to remember numbers, universal names that anyone can use to get to that exact ip address that they wanted. Dumb people will use the name system all the time!
OH I KNOW! We'll name it DUMB NAME SYSTEM! DNS for short!
Yes there are a lot of ipv6 addresses, however looking at the total size of a 128 bit address space is very misleading as ipv6 addresses are designed to be allocated in a more heirachical manner and designed to support stateless autoconfiguration for clients. Originally end sites were meant to be allocated a /48 though ripe now seems to be pushing for smaller allocations to smaller end sites.
Plus only 2::/3 is assigned to the ipv6 internet with other address space being reserved for other purposes.
Those figures would give us 2^45 end sites, this should be enough that we don't run out of addresses any time soon but it's a lot less than the ammount people assume from just looking at the number of bits in the IP.
note: i'm known as plugwash most places but i screwd up registering that here somehow in the past and now can't register
They can't sell em, at best they can go into the ISP business with them.
There are going to be lots of little solutions that will stretch out IPv4 for a while.
One that I came up would be to offer a financial incentive to reclaim unused blocks of addresses. Ten or fifteen years ago, IP address space was handed out like candy. You could get a class C block readily, and class B blocks just needed a little justification. I did some contract work in the late 90s for a company that I still keep up with, and they have a few entire C nets in their possession and not in use. Now how do we get these back? There is going to be demand for IP addresses, and as the supply becomes more and more limited, that demand will make people desperate.
So why not let people who already have address space sell what they have? It does reward unrightfully holding onto stuff, but if these addresses are needed, then hey!
then every RIR gets one last /8 from the "final five"
And if BSG taught us anything, it's that the final five won't be who you expect!
"Destroy science and religion. Science would re-emerge exactly the same; but not religion." - Penn Jillette, paraphrased
Once again, I'll ask a simple question:
How long until it is possible to pull up the main page on Slashdot, using nothing but IPv6 packets?
IMHO, every time one of these "OMFG IPv4 gonna run out RSN!!!1!11!" stores hits the front page, the Slashcrew should have to state where THEY are in becoming IPv6, and what is preventing them from doing so already.
www.eFax.com are spammers
126 * 28 days = 3,528 days, or just over 9 years, and that's only by re-allocating wasteful Class A's currently in use.
9 years is a lot more than a few months.
That's also the problem with these extrapolations, because as the addresses become more scarce the assignment rate slows down. When IPv4 addresses were first being handed out, they were given at a rate of 5-10 Class A's a week. We're down to one a month now, in a year it will be one Class A every two months, if that.
It would buy a few months and the resources are better invested in the transition to IPv6.
You're obviously not very good at business, because with any change the conversion to a new system is never worth it until the cost to maintain the current system is as expensive or at least very close to the same cost. Right now for the Tier 1 ISPs, the backbone of the internet, it is still significantly cheaper to extend IPv4 than to make the switch to IPv6. Doing so could cost them hundreds of millions, if not billions in manpower costs (the hardware should be pretty much there already).
Security is mostly a superstition... Avoiding danger is no safer in the long run than outright exposure. - Helen Keller
You could easily use 64 bit addresses in IPV4 by sticking the rest of the source and destination addresses in the options field of the packet. But why? We would still need to upgrade all the routers and software to work with the new system, except we would be upgrading to a hack not a solution, and we would be ignoring the solution that has already been implemented in many computers around the world. As likely as not, IPv6 already works with the computer you are holding, it just needs to be turned on.
IPv6 is the way to go, and everyone is already heading that way. By the time IPv4 addresses run out, the biggest difficulty may be explaining to your friends how to fix their internet that is no longer working and they don't know why.
Qxe4
DJB is correct, in that the IETF considered it outside their scope to do a "transition plan for the Internet"... This means that instead of having one standard model for how to get to IPv6, we've seen a veritable parade of transition and coexistence technologies. The combination of no clear transition plan plus no new end-user features makes deployment of IPv6 challenging, and I noted the same thing 15 years ago in RFC 1669. Despite all of the above, IPv6 remains the only viable answer if we want to keep growing the Internet.
Storm
Um, this is basic math. If we would have been giving out 5-10 /8s per week, we would have exhausted the complete range of addresses in less than a year. I don't think that happened back in the 90s.
It takes a man to suffer ignorance and smile
Be yourself no matter what they say
I think the position will change, but only after some kicking and screaming. As for unused allocations, even if all these were freed up, it would only provide a few more months of capacity, so the debate over what to do with them is really moot.
[FUCK BETA]
The IPv6 designers made a terrible mistake by not including backward compatibility with IPv4.
How, praytell, would they have gone about doing that?
IPv6 is a lot like Intel's Itanium processor. It's unclear right now whether the the anointed successor will gain ground or whether some IPv4 extension hack will come along and make fools of the IPv6 crowd. (Wait, what's the opposite of "crowd"?) BTW, I'm using a x86_64 processor right now, like most people.
Sorry. Try again. The first Itaniums had IA-32 compatibility. Later ones do not. And AMD did some incredibly stupid things with x64, which are becoming rapidly apparent as time goes along. I'm guessing you haven't noticed. But the reason Microsoft has an emulated 32-bit XP VirtualPC instance running for compatibility is that amd64 can't do vmm86 when running in 64-bit mode. Consequently, sixteen-bit applications can't run at all, natively. Sure, they run fine if you run the processor in 32-bit mode, but then you still are stuck with PAE for >4GB memory, and there's no way to directly access 64-bit registers (which you can do on IA-64, Sparc64, and POWER). I'm not saying amd64 isn't without its merits, but, backwards compatibility sure ain't one of them.
Yeah, that's why this whole "peak oil" thing is bogus - because of course as we run out of oil, our rate of using it will go down!
...
You see the problem, right? At some point we're going to start feeling some pain - we'll be foregoing the use of an IP address that we could really put to productive use - but we can't because getting one is too difficult/expensive. The point is that your problems start long before you use up the absolute last bit of a valuable, scarce resource./p.