Slashdot Mirror
ARIN Is Down To the Last /8 of IPv4 Addresses
An anonymous reader writes "On 3 February 2011, the Internet Assigned Numbers Authority (IANA) issued the remaining five /8 address blocks, each containing 16.7 million addresses, in the global free pool equally to the five RIRs, and as such ARIN is no longer able to receive additional IPv4 resources from the IANA. After yesterday's large allocation (104.64.0.0/10) to Akamai, the address pool remaining to be assigned by ARIN is now down to the last /8. This triggers stricter allocation rules and marks the end of general availability of new IPv4 addresses in North America. ARIN thus follows the RIRs of Asia, Europe and South America into the final phase of IPv4 depletion."
They've been talking about this day for what seems like an eternity... Finally, we can start complaining about something else!
Pretty outrageous that the whole of North America has to go on a diet earlier because Akamai somehow needs a whole fucking /10.
ARIN's behavior has made it clear: you can get all the IPs you want as long as you're a big guy paying big fees. But a small company asking for a /22? Go away, small businesses don't deserve to be able to do business.
Now capitalism kicks in and people start buying and selling spare IP4 addresses.
Kinda like that other thing they ain't making any more of....land.
It's going to be interesting to see this year.
There's no place like ::1
"If any question why we died, Tell them because our fathers lied."
Thanks, Obama!
Oh wait.... never mind.
Years back, my boss got a whole class C for a company with ~5 employees and network footprint nothing more than one website. Maybe they can get some of the corporations with class As to give some back? (yeah yeah I know)
I swear to God...I swear to God! That is NOT how you treat your human!
Phase 1 Recieve last IPv4 /8 Adresses /8 Adresses /8 Adresses /8 Adresses
Phase 2 Giveaway down to last 3x IPv4
Phase 3 Giveaway down to last 2x IPv4
Phase 4 Giveaway down to last 1x IPv4
Phase 5 ???
Phase 6 Profit
What comes first, a widespread NATed internet or IPv6?
285 million addresses reserved for no compelling reason. sure, let's push onwards to ipv6, but saying "our hands are tied" when over 1/16th of the entire space is still available is a bit irritating.
The IPv4 address exhaustion is a useful case study in human behavior in response to resource exhaustion.
http://www.albartlett.org/presentations/arithmetic_population_energy_transcript_english.html
Relevant quote: "Remember our conclusion from the cartoon of one person per square meter; we concluded that zero population growth is going to happen. Let’s state that conclusion in other terms and say it’s obvious nature is going to choose from the right hand list and we don't have to do anything—except be prepared to live with whatever nature chooses from that right hand list. Or we can exercise the one option that’s open to us, and that option is to choose first from the right hand list. We gotta find something here we can go out and campaign for. Anyone here for promoting disease? (audience laughter)"
In this case, fortunately, it's extremely unlikely that violence and death will occur as a result of this specific resource exhaustion, but the study of human behavior in response to the resource shortage is telling.
We've been aware for years that zero IPv4 address availability is going to happen. It's absolutely certain. The only way to make it not happen, or not *care* that it happens, is to do something about the problem. But of course, even for such a technically manageable problem, humanity on the whole chooses to do nothing. The exact same thing will happen for fossil fuel exhaustion, arable land exhaustion, etc.
And now nature will choose for us from the right-hand list of IPv4 exhaustion: here comes corporate greed, lawsuits, slow and inconvenient CGNs (one bad actor in your ISP's network causes you to be banned from the services you use), etc.
Humans are hard-wired to be reactionary, not proactive -- and at that, only reactionary to immediate problems. "Oh, I can't get a new IPv4 address. What do I do?" or "Oh, I can get a new IPv4 address, but it's too expensive. What do I do?" -- These are the kinds of things we will start thinking about, and making people start to care. NOT "Oh, we better deal with this problem that is likely to happen in 5 years."
As flawed as we are, it's probably a good thing that we won't survive long enough to leave our solar system and populate the cosmos. We don't deserve it. We're just too *dumb* as a species.
It didn't matter whether it was last year or next...IP usage was accelerating into the wall anyway. The GOOD part about this is that now the US is out of addresses certain parts of the Internet industry are more likely to take IPv6 seriously.
Sadly, ISPs in other parts of the world have proven adept at further avoiding the problem by downgrading consumer connections to carrier-grade NAT, so we have another 5 years of eking out of old order before people REALLY have to take notice.
NAT works great, no issues here [Ducks!]
Kidding guys :)
"It's the end of the world as we know it, and I feel fine"
--- REM
http://www.youtube.com/watch?v...
We're running out of free ones. And like any freely available resource, they've been squandered. Once the free supply is exhausted, they'll simply no longer be free - meaning that actual incentive will exist to conserve them and organizations will have incentive to sell unneeded blocks. Economics 101, people.
Help save the critically endangered Blue Iguana
Now that addresses have run out, they have become a valuable resource for the ISPs that own them. If those ISPs implement IPv6 then there will be no shortage of addresses, and they will lose all their value.
So the monopolist ISPs will now do everything in their power to prevent IPv6 from being adopted.
You assume wrongly that most of the world behaves like the british and people from english-speaking countries, who still haven't made the move to the universal measurement system, and neither have reformed their language to make grammar and pronounciation rules consistent.
I doubt the organizations with those large blocks will sell them unless they become very expensive (which I don't think will happen for a long time). The costs of restructuring the network for a lot of these companies would far outweigh the gains.
What I see as far more likely is ISPs implementing carrier grade NAT as the default, and potentially charging a small fee for those who need a unique IP. The vast majority of users won't care, and as long as getting an IP if you run a game server or use skype or whatever is an easy process, it's actually not a bad solution. I figure we've got 10 years or so before we actually see IPv6 really take off.
We're running out of free ones. And like any freely available resource, they've been squandered. Once the free supply is exhausted, they'll simply no longer be free - meaning that actual incentive will exist to conserve them and organizations will have incentive to sell unneeded blocks. Economics 101, people.
Why would you choose that option when we have a way of bypassing it? Isn't progress generally about creating plenty? We have the ability to create plenty, and not have to deal with buying and selling IP addresses. Just because you can create a market doesn't mean you should.
Clearly we should have invested years ago in finding renewable sources of IP addresses...
There are no languages where grammar and pronounciation rules are completely consistent. Spanish and Indonesian come close with regards to grammar, and Dutch and Czech, when it comes to spelling vs. pronounciation, but there just is no language which is completely consistent. Even artificial languages like Esperanto have their inconsistencies.
Because there is a very high one-time-only cost involved in switching to ipv6, compared to a small running continuous cost of continuing in ipv4, and for now, it is advantageous to become in ipv4. No one wants to be the one to switch first.
Just think of all sort of problems large ISPs will have to deal in terms of support if they switch to ipv6, in terms of phone service, visits, substitution of cable modems, support for old machines running none/bogus ipv6 implementation.
Just think of all the programs coded years ago, with ipv4 hardwired in (I know 4to6, but your client does not).
Not easy as flick a switch.
I'm glad to know I'm not the only one using the universal Klapagorg measurement system!
Get free satoshi (Bitcoin) and Dogecoins
As flawed as we are, it's probably a good thing that we won't survive long enough to leave our solar system and populate the cosmos. We don't deserve it. We're just too *dumb* as a species.
How is anyone supposed to take a person like this seriously?
Space game using normal deck of cards: http://BattleCards.org
That was over three years ago, why is this news?
So let's finally move on to IPv6. ISPs, I'm looking at you.
I refuse to sign
I run several Xen virtual servers rented from a VPS vendor.. They come with two ip addresses, one of which I don't need. I asked them about returning the extra address to them, as I didn't need it.. They said "no, we have plenty" ... ????? Hey! I thought the ipv4 address space was running out... hmmm.. guess not.. At least THIS vendor does ipv6.. each vps comes with 3 (count 'em THREE) ipv6 addresses.. WAAAY cool!!
Most of the ISPs I've dealt with here in Canada do not offer routable IPv6 allocations to users. They certainly don't readily offer static ones for business use like they do with IPv4.
- Michael T. Babcock (Yes, I blog)
$ host -t aaaa slashdot.org
slashdot.org has no AAAA record
If you go by http://ipv4auctions.bstocksolutions.com/ a /8 is worth roughly USD 100M. There are a few unused ones by companies out there...
You brightened my day.
You are wrong. Bulgarian is a language with 100% persistent pronunciation.
Except this still won't fix the fact that v4 is simply too small.
so we have another 5 years of eking out of old order before people REALLY have to take notice.
Possiblly much more than that.
XP and andriod 2.x are dying. They aren't dead yet but in a few years time their relavence will likely have declined to the level where website operators think it reasonable to stop supporting their default browsers. Once that happens we will be able to use SNI (and tell the holdouts still on XP to "use firefox or chrome damnit")
Once that happens it will be possible to put multiple SSL websites behind one IP reducing the IP demand on the hosting side. With end lusers put behind CGN, SSL web hosting running multiple sites per IP and basic VM hosting using front end load balancers to let them share IPv4 IPs it should be possible to keep IPv4 going for a long time.
One interesting question is what price will IPv4 addresses reach, currently it seems to vary from about $7-$25 per address depending on block size (http://ipv4auctions.bstocksolutions.com/)
note: i'm known as plugwash most places but i screwd up registering that here somehow in the past and now can't register
Here at DHCP, we're committed to providing only renewable and conflict-free IPs.
How can I believe you when you tell me what I don't want to hear?
Why would you choose that option when we have a way of bypassing it?
Because people will do what is individally best for them, not what is best for the community as a whole.
If I want to run a server for the general public to access over the internet it needs to have an IPv4 address until such time as the vast majority of clients can reliablly access IPv6 servers (I would not consider teredo to be "reliable", it's overcomplicated and fights against NAT rather than working with it).
Similarly if I want my users to be able to access resources on the public internet I need IPv4 addresses for the intent side of my nat boxes until such time as the vast majority of servers are available on IPv6.
If I deploy IPv6 it will not change whether those systems need IPv4 addresses. To do that requires OTHER PEOPLE to deploy IPv6 which they are often unintertested in doing.
note: i'm known as plugwash most places but i screwd up registering that here somehow in the past and now can't register
So now the exact same thing that happened with land will happen with "land".
You clearly have absolutely no clue of how linguistics work.
First of all, your Bulgarian claim is not true: https://upload.wikimedia.org/wikipedia/commons/1/12/Bulgarian_dialects_by_Todor_Bozhinov.png
Secondly, if it were, you'd just have to wait for around 25 years to start noticing things changing.
You know which languages *could* (but don't) have 100% persistent pronunciation? Dead ones.
A large number of companies from all over the world set up shell companies in Africa.
Newer mobile phones should have been IPv6 from the beginning. China mandated that years ago. T-Mobile is IPv6. (You can supposedly open up an end to end IPv6 connection between two T-Mobile phones). It's suprising that the cellular phone companies didn't fix this, since they have control of both network and handset.
Bulgarian doesn't even note emphasis in the writing, and also not the length of the vowels, thus it doesn't have a fully written pronounciation. It is still possible to follow the full set of pronounciation rules in Bulgarian and still read a text completely non-understandable for a native speaker.
Now:
Static IP Address: Requires a business account, plus $20/month.
Future:
Public IP Address: Requires a business account, plus $20/month.
Static Public IP Address: : Requires a business account, plus $100/month.
(Note: Numbers are completely made up. But in my experience, that's is how most residential ISPs structure it.)
That is wrong, first thing we learnt in class was (and from my textbook):
Slashdot doesn't support unicode, so, I'll be using romanization
@ = er golyam
j = i kratko
"In both e- and i- verbs, there is a lack of correspondance between pronunciation and spelling in the "I" and "they" forms:
cheta', cheta't are pronounced as if written chet@', chet@'t; /subota/, is often pronounced /supta/ (u in these examples are IPA (ram-horn)
pi'ya, pi'yat are pronounced as if written pi'jo, pi'jot
"
In addition to other irregularities: s@bota, nominally
the way to get people to respect scarcity is to give them price signals. I guess the concern was that your favorite actor would buy a /8 for no reason, but lots of idiots have blocks for no reason now.
We're running out of free ones. And like any freely available resource, they've been squandered. Once the free supply is exhausted, they'll simply no longer be free - meaning that actual incentive will exist to conserve them and organizations will have incentive to sell unneeded blocks. Economics 101, people.
There has been pressure for near two decades now in the form of allocation policy and documentation requirements where lack of plentiful IP resources has lead directly to proliferation of 1:Many NAT.
If you go by http://ipv4auctions.bstocksolu... a /8 is worth roughly USD 100M.
This assumes that either the seller is allowed to split the block or the price per IP for a /8 is comparable to the price per IP for a /8 block is comparable to the price per IP for the much smaller blocks you see sold on that site.
note: i'm known as plugwash most places but i screwd up registering that here somehow in the past and now can't register
remember, there were only 256 /8 nets. So a /8 is a lot of adresspace.
There is lesson to be learned here:
Every fix length field should have a reserved value for an extension. .
Back in the early days of the internet each header bit was precious, so it was important to have packet headers as small as possible. However even then we could easily have reserved say 255:255:255:8 as the extensible value of the IP address.
By the time they are needed, namely in present day, it is a rather trivial mod to make network gear which reads another 32 bits past the end of the standard TCP/IP headers to collect the extended IP address, and presto IPv4 address shortage is gone.
You're not supposed to. This is called pandering to your base, however, as most people on Slashdot are anti-human trendy-wannabe psuedoliberals who really, honestly believe that the human race is a cancer upon the planet. They are the cancer in the human race.
Because there is a very high one-time-only cost involved in switching to ipv6, compared to a small running continuous cost of continuing in ipv4, and for now, it is advantageous to become in ipv4. No one wants to be the one to switch first.
Nobody is switching to IPv6 they are *adding* IPv6. IPv4 is not being turned off by anyone well into the foreseeable future.
Most large content providers are already offering service via IPv6 and millions already have IPv6 access via their ISPs.
Just think of all sort of problems large ISPs will have to deal in terms of support if they switch to ipv6, in terms of phone service, visits, substitution of cable modems, support for old machines running none/bogus ipv6 implementation
The migration to IPv6 takes a while and does not involve turning off IPv4 anytime soon. There is no need to rush to replace gear. It will eventually break or become obsolete in the next few years anyway.
Not easy as flick a switch.
For most consumers it will be easier than a flick of a switch. They get it without having to expend any effort at all or ever even knowing they have it. This happens either immediately or after their old router or CE has broken and gets replaced.
Think of it. Here is this scare resource, IPv4 addresses, and no more are going to be allocated in North America. I see great potential in profit, online exchanges opening up allowing the trading of IP addresses, etc. etc. To quote the Ferengi, my lobes are tingling.
It's really quite a simple choice: Life, Death, or Los Angeles.
We've been using IPv6 for about a decade now.
Didn't you get the memo?
-- Tigger warning: This post may contain tiggers! --
Human beings are terrible at planning ahead. Just look at the financial condition of just about any modern nation. Knee deep in debt with only plans to spend more money without thinking about how to pay it off. It takes bankruptcy for them to change their ways.
Same goes for renewable energy - yeah there is some token adoption as long as it is heavily subsidized. It will take actual depletion of the current resource to drive full-scale adoption.
IPv4 is just another example of human-driven resource exhaustion with immense resistance to the future plan until the current resource is actually completely exhausted.
The same could have been said about the buggy whip manufacturers. Everything eventually needs to move on and improve. That's just how it goes. If you're on XP and you're on the internet, you shouldn't be and we should make it hard on you, not the other way around.
Well, he's thought about it a bit more than you have, obviously.
Tell us: what precisely do you find in his statement hard to take seriously, and what are your thoughts that you can present as an alternative viewpoint?
I started to follow this on 2000. "The end is here sooon". Now it starts to feel like generation after generation of new forget their history. Almost like some great war in the past. Kids just don't learn.
Sure do that if you want to break the whole fucking internet overnight.
---Saying gnome 3 is better than windows 8 not so much a compliment as it is damning with light praise.
I prefer the slightly cheaper black market blood IPs.
Add to it millions of mobile devices that can't deal with TLS (forces you to use SSLv3), or SHA-2 (forces certificates to use SHA1), or large RSA (force certificates to a maximum of 2048 bits)... and cannot deal with SLI at all.
Universal SLI support is still at least 5 years away.
Not to mention NATed IPv4 is the bane of P2P...
What! No way... Drill Baby Drill should be the new manta.
Meh, I know it is asking too much of /. editors, but LACNIC (which covers latin america, which includes South America entirely) is still in phase 0.
ARIN entered phase 4 with one /8 left. The other registries have different phasing. LACNIC has three more phases to go, it is still in "if you ask, we have it" mode (phase 0), and will remain in phase 0 until there is a /9 left (we're currently at 0,66 of the last /8, so we have a few weeks left in phase 0). It will enter phase 1 when there is one /9 left.
I expect lots of shaddy business in the LACNIC region when phase 2 begins, which might well require that LACNIC deploy a rule that all resources received under phase 2 and 3 must never be sold or transferred.
ARIN has always been weird (mostly because much of the pre-ASO allocations are in ARIN), I doubt they can do anything about people gaming their final phases to get IPs to sell. Other regions are a lot less tolerant of monetization of the IP address space and have increased control, so they can reclaim such IP and ASN resources easier than ARIN.
The future is now
I think we can pipe some down from Canada.
While technical innovations could theoretically make IPv4 last a while longer, the reality in Asia and Europe is that end users are seeing severely degraded service due to CGN, for example with VoIP services. New ISPs, on the hosting and the access side, can not get IPv4 addresses for all their needs. Making IPv6 work is basically their only option.
IPv6 will be used, and when it can do the job, then there will be no point in sticking with IPv4 and all its complicated multiplexing methods.
This is a false comparison. People are doing NAT to deal with IPv4 exhaustion. You can say NAT is bad, but so far it's worked out ok. People are not head in sands, they are simply postponing the cost until they are convinced it's needed.
Progress is about creating plenty, yes. However, the granparent specifically cited economics, which is typically about creating an artificial shortage and then profiteering.
Forget magic. Any technology distinguishable from divine power is insufficiently advanced.
IPV6 was built to solve a host of problems we don't have anymore, and introduces a bunch of new of problems we don't need.
It's time to simply add an octet to IPv4. Balancing scalability with ease of use is important, and IPv6 failed at the second part.
I don't know about grammar, but for pronunciation, reading Hiragana and Katakana (Japanese) seem to be 100% consistent. There is only one way in which I can read those individual characters. Also Korean seems to be 100% consistent.
There are no languages where grammar and pronounciation rules are completely consistent.
Have you really studied all 4000 or so written languages in the world to make that claim?
when we run out of IPv6 addresses? It's a serious and pressing concern, like what happens when Intel runs out of nanometres to cut off their processors. Fools, the lot of them!
I think he's talking about a virtual network adapter in the VM with a NAT subnet created specifically for those thse VMs. Just another way of natting a VM.
Making IPv6 work is basically their only option.
Trouble is "Making IPv6 work" requires other people to deploy it. I dunno what things are like where you live but here in the UK the only access providers that offer IPv6 connectivity are the boutique ones.
note: i'm known as plugwash most places but i screwd up registering that here somehow in the past and now can't register
without net neutrality will just be multiple pingable anal probes up you ass anyway. Who cares about IPv7 without decent standards, not me. I'd prefer the dark ages I think...
Fuck you. If you have or intend to have any children you'll change your tune very quickly. You MUST remain positive about the future of humanity or end up like one of those morons who thinks he knows better than everyone else but appears on the outside as a miserable cunt, someone no-one wants to spend any time with because they see the future as hopeless. Don't be one of those miserable cunts, please.
As flawed as we are, it's probably a good thing that we won't survive long enough to leave our solar system and populate the cosmos. We don't deserve it. We're just too *dumb* as a species.
As flawed as single celled organisms are, it's probably good that they will not survive long enough to leave our solar system and populate the cosmos...
Evolution is still happening. The "we" that you are postulating will not exist in a thousand years. If we were to separate 2 groups of humans for a thousand years, it is possible that they would be totally separate species if there were sufficient evolutionary pressure.
In other words, kill yourself* to improve the species if you think the current state of the human species is so terrible.
*not saying that to be a dick, just pointing out that is how evolution happens: death of the organism, growth of the new slightly different organism.
"Someone needs to talk to the tree of liberty about its ghoulish drinking problem." by ohnocitizen
Like I said elsewhere in this page, even if ISPs implement IPv6, they'll still need their IPv4 addresses for dual-stack connections, which will be needed as long as a majority of the customers are IPv4.
Most of the ISPs I've dealt with here in Canada do not offer routable IPv6 allocations to users. They certainly don't readily offer static ones for business use like they do with IPv4.
You can get IPv6 connections here in the UK so long as you pick your ISP carefully. Unfortunately, the last time one of my customers shopped for a new internet connection (i.e. expensive leased line), they directly asked the ISP if they did IPv6 and got a "yes" reply, and only after the connection was installed and paid for did it become apparent that the ISP lied. (unfortunately it wasn't important enough to the customer for them to kick up a proper fuss about the ISP mis-selling).
http://blog.nexusuk.org
If ipv4 address are rare, then other ISP's will offer ipv6 as a cheaper option.
All we need is some ISP competition...
All we need is some ISP competition...
And I want a pink unicorn.
Since when is a link local ip in ipv6 *not* localhost? Note, link-local, not site-local, the latter of which would correspond to a subnet.
What you are thinking about is 'Node-local', which is the loopback address - ::0 in IPv6. Note that this falls within the IPv4-compatible addresses, which is currently deprecated by the IETF. As Bengie pointed out, link-local is your subnet, where addresses are automatically assigned, and is fe80::/10. Unique-local addresses are fc00::/7, and a site-unique address fd00::/7 is also there: fc00::/7 is meant to be universally unique, but that would require the IANA or the RIRs assigning them, while fd00::/7 can be unique just to a site, or domain.
In IPv4, all of the above tend to get conflated under 'private addresses'. Here, link-local and Site-Unique addresses are clearly delineated. The link-local addresses are what are randomly assigned by stack, and visible throughout the subnet.
There are plenty of other dependencies (OpenSSL versions, Apache versions, Java versions), so it's even worse than that.
PS: I can't see what the hell you responded to -- it's not in the parent comment!