Slashdot Mirror
America Runs Out of IPv4 Internet Addresses
FireFury03 writes: The BBC is reporting that the American Registry for Internet Numbers (ARIN) ran out of spare IP addresses yesterday. "Companies in North America should now accelerate their move to the latest version of the net's addressing system. Now Africa is the only region with any significant blocks of the older version 4 internet addresses available." A British networking company that supplies schools has done an analysis on how concerned IT managers should be. This comes almost exactly 3 years after Europe ran out.
Out of IP addresses? Sounds like a good time to invade somewhere where they mine them!
At this point, ISPs need to mandate that customers use SNI where possible; too many IP addresses are allocated just for an SSL certificate. I think we'll start seeing more Let's Encrypt-type Subject Alternate Name management tools, too.
As most people do not type these number and do not need to remember these numbers, I do not see any problem with longer numbers. Especially when there are methods to write them shorter than that: 0000::0000::0000::0000::0000::0000::0000::0000
For example zeros ca be omitted. see also https://en.wikipedia.org/wiki/...
If you are typing or using IP addresses for ANYTHING other than you primary DNS servers, you're doing something wrong.
Seriously - set statics on your DNS servers (which can even be IPv4!), plug that into your DHCP etc. servers. Done.
This is the problem with IPv6 - those people whining about it aren't in charge of networks where it could be an issue anyway.
P.S. likely your mobile phone and maybe even your cable setup has been using IPv6 addresses for a few years now. They are specified and necessary in related standards. Did you notice? No. Because nobody types in IP addresses any more, not even on their home networks, work networks, thousands of servers, etc.
To be honest, MAC addresses are much more problematic to me, but I barely ever have to type those either.
I just checked my IP address and it's 192.168.1.102. Whew, I'm glad I got one before they ran out. No one else can have my IP address!
No, that's just an artifact of the different policies for assigning the last addresses. RIPE (the European registry) throttled assignments by making the requirements much more strict. That change of policy was considered the point when RIPE ran out of IPv4 addresses, because the remaining addresses are not given out just for asking. Unlike the other registries, ARIN did not institute a policy to extend the availability of IPv4 addresses for transitioning purposes, so they burned through the last 16 million addresses like no tomorrow and are now truly out of IPv4 addresses to assign. They are in fact the first registry without IPv4 addresses in stock. RIPE still has almost a full /8, APNIC has two thirds of an /8, LACNIC has one seventh of an /8, and AFRINIC still has 2.3 /8 blocks.
According to google's ipv6 stats, about 21% of its American visitors access the site via ipv6.
https://www.google.com/intl/en/ipv6/statistics.html#tab=per-country-ipv6-adoption&tab=per-country-ipv6-adoption
That is not as high as Belgium (almost 36%), but it is a start.
The problem with switching is IMHO three fold, 1.- It is gonna cost probably a couple hundred million in routers and modems that cannot support IPV6, in fact if you look at places like Amazon and Newegg there are more routers being sold that doesn't support IPV6 even today than not, 2.- Years of treating IT workers as disposable means we simply do not have enough workers that can support all the headaches that are gonna happen with the switch, I know in my area most of the greybeards simply went into other fields because they were tired of being fucked by the MBAs, and my own personal beef 3.- Assigning everyone a unique IP means it will be trivial to track everyone, its gonna be meat on the table for your *.A.As and copyright trolls.
So you can see why switching hasn't bee a priority for most, its gonna cost a mint, shit is gonna break everywhere, and I wouldn't be surprised if it will end up with a shitload of requests from the *.A.As spamming the ISPs as they will be able to argue that "IP address does not equal individual" no longer applies.
ACs don't waste your time replying, your posts are never seen by me.
The real WTF is that Slashdot has been running IPv6 articles for years...and *still* doesn't support IPv6.
Facebook on the other hand - not a tech site, but a site for angsty teenagers, baby pics, cat memes and partisan squabbling - has supported IPv6 fully for years.
It's embarrassing that a tech site can't do what a non-tech site has been doing for years.
Oolite: Elite-like game. For Mac, Linux and Windows
exactly as prophicised. I knew this was coming when Gene Ray went into hiding.
the preceding comment is my own and in no way reflects the opinion of the Joint Chiefs of Staff
No thanks. IPv6 addresses are a mouthful, typically 3x as long when printed. We should move to a version that makes them 1 byte longer.
IPv6 was a poor decision. It's like someone who ran out of toilet paper once so they went and filled their entire basement full so they won't accidentally run out again. 192.168.23.17 compared to AB34:34ED:AB34:34ED:AB34:34ED:AB34:34ED
As we're now pretty much stuck with ipv6, they would be better off locking out all the later bits until the transition is complete and make the ipv4 directly translatable. I.e. 192.168.25.25 becomes just FFFF:C0A8:1919 and all other ipv6 numbers are off limits until the transition is complete.
FFFF:C0A8:1919 isn't much more difficult than 192.168.25.25 and would make the transition much simpler than giving everyone a ipv4 number and a completely different ipv6 address.
Doing it this way, everyone could still access the websites via either their ipv4 or ipv6, it would only be the higher order ones that you would need to upgrade in order to access. Similar things have happened with phones and websites. When new area codes were introduced or new top level domains, a few people had problems accessing the new areas with older equipment if the older equipment was hardcoded somehow.
- What are we running out again? I thought we ran out last month! They are crying wolf!
- IP addresses are assigned by region we only just ran out.
- NAT makes this a non issue. Just use NAT!
- NAT is a broken concept that breaks end-to-end connectivity!
- I won't move to IPv6 they are too hard to type.
- Why are you typing IPv6?
- I can't NAT on IPv6 so it breaks my firewall and its insecure.
- NAT is not a firewall, you can firewall IPv6
- Why don't we just steal some of HP's IP addresses? They have some spare.
- Break the internet by splitting up routing tables even further.
- But NAT has protected us for many years everything works on NAT.
- Everything now needs to connect to a command server. No end-to-end connectivity and nasty workarounds in routers to make applications work.
- But DHCP doesn't work for IPv6!
- DHCP isn't needed, and if it is needed yes it does.
- But we can NAT the NATTING NAT NAT!
- Go fuck your NAT.
If we just shut down all the porn sites on the Internet, I'm sure we'd get back a good 98% of those IP addresses...
Incipiamus, fratres, servire Domino Deo, quia hucusque vix vel parum in nullo profecimus.
Copy/paste them. Or use DNS, it's hardly a new technology.
And if you really can't do either, then pick your addresses better. If you pick addresses like 2001:db8:42:a57e:a92f:2c3d:30c5:7562 rather than 2001:db8:42:1::2 and refuse to use DNS for them, then you can't complain about how hard they are to remember.
The problem with this is that some of the original recipients of those really big blocks like GM and HP were given those addresses, not leased them. They, for all practical purposes, own that address space.
I know the organization I work for is a part of the problem. Before ARIN existed, a group of three schools (I work for one of them) were granted a /8 as a part of our research status. We have no relation with ARIN, and there isn't even a way to really give back 100 of the /16's we don't use.
No, not again.
Last year, ARIN hit one /8 left (that's the second article you linked). Back in July of this year, ARIN had to make their first ever refusal for an allocation on the basis of not having the IP space for it (that's the first article). They still had some space remaining for small allocations. Now, as of yesterday, they have to refuse all allocations on that basis, because they ran out of space altogether. That's this article.
Apparently, the idea that reaching 0% involves going through 10% and 1% first is hard to grasp...
How can you be so ignorant of how IPv6 works and still have the hubris to propose a modification that supposedly fixes it?
Oh silly me, this is Slashdot.
IPv6 was a poor decision. It's like someone who ran out of toilet paper once so they went and filled their entire basement full so they won't accidentally run out again. 192.168.23.17 compared to AB34:34ED:AB34:34ED:AB34:34ED:AB34:34ED As we're now pretty much stuck with ipv6, they would be better off locking out all the later bits until the transition is complete and make the ipv4 directly translatable. I.e. 192.168.25.25 becomes just FFFF:C0A8:1919 and all other ipv6 numbers are off limits until the transition is complete. FFFF:C0A8:1919 isn't much more difficult than 192.168.25.25 and would make the transition much simpler than giving everyone a ipv4 number and a completely different ipv6 address. Doing it this way, everyone could still access the websites via either their ipv4 or ipv6, it would only be the higher order ones that you would need to upgrade in order to access. Similar things have happened with phones and websites. When new area codes were introduced or new top level domains, a few people had problems accessing the new areas with older equipment if the older equipment was hardcoded somehow.
The stuff you are describing was initially contemplated, which is why we had IPv4 compatible addresses (::192.168.2.5) and IPv4 mapped addresses (::ffff:192.168.2.5). Problem was that that wasn't a simple way to resolve the addresses due to NAT in IPv4 among other things, which is why you have different transition mechanisms. Some of them have been used, like 6rd, Dual-Stack lite, Teredo, et al.
The toilet paper analogy is not quite correct. Rather, it's more like a case of discovering a new fuel that's a million times cheaper than gasoline, doesn't emit greenhouse gases, but which would require all engines worldwide to be changed. Since that would be an expensive process, the guys who design the replacement engines are working w/ the fuel engineers to ensure that the engines would never need to be redesigned again. In the case of IPv4, even making it 33 or 40 or 64 bits would have required an overhaul of all the world's networking gear, which is why the jump was made to 128 bits.