Posted by
ryuzaki0
on from the ip-address-in-every-pot dept.
miladus writes "According to a story at Zdnet,
Asian countries are running out of IP addresses. China, for example,
was assigned 22 million IP addresses (for a population of 1.3 billion)
under IPv4. The US owns 70 percent of current IP addresses. Perhaps IPv6 will solve the problem."
Corporations are at fault?
by
sinergy
·
· Score: 5, Informative
I personally know of many large corporations that have several Class-B networks that they use for non-accessible internal routing. I'm sure their numbers are much higher than just the one's I've come across.
Couldn't somebody review who has all of those assigned addresses and help(force) them to migrate to private ranges?
-- ...
Re:Corporations are at fault?
by
tomhudson
·
· Score: 4, Informative
... and some of them have class A addresses that they cannot possibly fill.
IANA Address assignments
003/8 May 94 General Electric Company
004/8 Dec 92 Bolt Beranek and Newman Inc.
005/8 Jul 95 IANA - Reserved
006/8 Feb 94 Army Information Systems Center
007/8 Apr 95 IANA - Reserved
008/8 Dec 92 Bolt Beranek and Newman Inc.
009/8 Aug 92 IBM
010/8 Jun 95 IANA - Private Use
See [RFC1918]
011/8 May 93 DoD Intel Information Systems
012/8 Jun 95 AT&T Bell Laboratories
013/8 Sep 91 Xerox Corporation
014/8 Jun 91 IANA - Public Data Network
015/8 Jul 94 Hewlett-Packard Company
016/8 Nov 94 Digital Equipment Corporation
017/8 Jul 92 Apple Computer Inc.
018/8 Jan 94 MIT
019/8 May 95 Ford Motor Company
Re:Corporations are at fault?
by
LinuxHam
·
· Score: 2, Informative
Think about manufacturing.. how many devices are IP-enabled nowadays.. now go through your list and think about companies that produce no less than millions of parts per year, and therefore have tremendous manufacturing facilities that have ip-enabled sh*t all over the place..
General Electric Company - Massive production lines Bolt Beranek and Newman Inc. - They (not Gore) invented the 'Net Army Information Systems Center - um, the **ARMY** Bolt Beranek and Newman Inc. - again IBM - (my employer) HUGE MANUFACTURER and 300k+ (technical) employees DoD Intel Information Systems - The Dept. of Defense AT&T Bell Laboratories - AT&T fer chrissakes, IP *everywhere* Xerox Corporation - Another big mfgr and computing co. Hewlett-Packard Company - Even bigger now with Compaq Digital Equipment Corporation - Also HPaq, ok 32 million IPs is a bit much Apple Computer Inc. - They'll never need 16m addys:) MIT - your point.. a bit much Ford Motor Company - Back to massive manufacturing facilities worldwide
Re:Corporations are at fault?
by
mattsouthworth
·
· Score: 2, Informative
Who is this 'someone'? And how can they take something that another entity owns? These class A assignments came well before IANA and whoever doled out/24s to whoever could figure out a SWIP.
Class A and B owners shouldn't have to move to 'private' (RFC1918) address space. 1918 space used in a one-to-many NAT is a hack that breaks end-to-end. IPv6 maintains e2e and is preferable. I'm sitting on a huge network numbered out of RFC1918 right now, which is a pain in the balls.
And while I'm soapboxing, although 'security' (broken e2e) is a side-effect of NAT it's not a reason for NAT. One could be just as secure with a properly configured firewall, and (all together now) not break e2e.
Re:Corporations are at fault?
by
spif
·
· Score: 3, Informative
Re:They should really swap to IPV6 then..
by
Sexy+Commando
·
· Score: 5, Informative
China and Japan will invest millions to develop IPv6. For example, June last year, both governments pledged US$32 million into network construction and testing, system development, application technology development and standardization, she said
NAT is simply not the long term solution, and it is going to cause headaches when dealing with wireless devices.
Re:IPv6? Yes because NAT is too limited
by
jcdr
·
· Score: 5, Informative
NAT is pefect to extend the network of one single entity, but is a very limited solution to extend the network to several entity.
If you have only one public adresse you have a single port for each services. Despite the fact that most services can extended by virtual one this is not the case for all of them (think SSH, or IPSec for example) and this require a high degre of coordination between the entity.
So IPv6 could be the cheapest way to solve the problem. And this could provids a good boost to the network market...
Nah, NAT will solve the problem - about a zillion times less expensive to implement.
Nope, absolutely wrong.
While all computers on the same NAT can directly connect to others, it cannot do so easily to others on another NAT, or other 'real' IP addresses. This effectively prevents anyone from running any server that can serve to networks outside the NAT, unless some ports are designated at the NAT router level specificly for that particular server. I don't see ISP's or network admins designating specific port ranges for every computer, as it takes work, and it could conflict with applications that uses specific port ranges (such as file transfers on MSN used by illiterate users who can't use ftp).
I would say using NAT to solve this problem is all but a cheap bandage that will cost more in the long run. IPv6 must be implemented soon to ensure the continue growth of the Internet.
-- Please direct all bug reports to/dev/null
Re:whats the ratio?
by
EdgeShadow
·
· Score: 2, Informative
With IPv4, China has only 22 million IP addresses for its population of 1.3 billion people. Last year, it had about 17 million Internet subscribers, and the figure will hit 62.5 million in 2007. Japan and Korea will also run out of addresses soon, she said.
Regardless of ratios, the fact is that China, Japan, and Korea are simply running out of addresses. Though costly, implementing IPv6 is a necessity and will take care of the address shortage. Besides, most of the newer OS's (XP, Linux 2.2 or higher, OS/X) already have IPv6 built in.
IPV6 won't happen...
by
scovetta
·
· Score: 2, Informative
until a business need exists for it. Since US companies (ISPs, schools, etc) will have to back it first, and most won't make any extra money from Lao Po in Beijing having a class D instead of a NAT'ed single IP, I don't see this happening. And since US ISPs are now NAT'ing and giving people gateways like 255.255.255.254, they've delayed our running out of IPs for quite a while.
And to the guy who mentioned that MIT probably isn't using their entire class A, I think they've given a unique IP to each student, book, pencil, and brick on campus.
-- Wer mit Ungeheuern kämpft, mag zusehn, dass er nicht dabei zum Ungeheuer wird. --Nietzsche
Asia doesn't get numbering from the American Registry for Internet Numbers; they come from APNIC (Asia Pacific Network Information Center).
Otherwise, a good point.
Re:IPv6 adoption
by
caluml
·
· Score: 3, Informative
I agree with you. I work pretty much purely over IPv6 now. I can administer our entire network with IPv6.
I keep banging the IPv6 drum, but people are naturally lazy, and don't want to change unless they have to. It explains the Microsoft/Linux thing too - people can't be bothered to try it, as MS works, to a fashion.
Unfortunately, this lack of IPv6 adoption is due to Microsoft. As 90% of the online-population can't use it, the people running the services can't be bothered to support it. And while there aren't any decent services on IPv6, the impetus to upgrade it is low.
I think it can be all summed up by asking: Why don't you make all the sites you administer IPv6 only? Because then most of your audience wouldn't be able to see it.
I love IPv6. I've played with it in the lab, and it's nifty! I'm in charge of restructuring my company's IP layout, guess what I suggested. Interestingly enough, when I proposed my plan on #ipv6 on freenode, the answer was a resounding DON'T DO IT. I have too much legacy stuff laying around that just won't support IPv6. Funny thing is, we are doing well on technology. I think of all the other businesses in worse shape than us, and I start to think. There is no way in hell IPv6 migration will happen any time soon. It's sometimes hard for us to see, especially when we do transparent stuff at home. What we forget is all the weird hardware that companies still depend on. There is some stuff that just won't go. We bought a Cisco router 3 years ago, its IOS won't support IPv6. That's only 3 years ago! Think of the legacy crap that was installed 10 years ago that still runs! NT servers that no one upgrades because they still work. We still have a Windows 3.1 machine that does its job, and in fact we broke trying to upgrade! Still works, it's easier to leave it alone. This kind of stuff happens everywhere, I've seen plenty of businesses with old hardware that's costly to upgrade and not broken.
IPv6 is great in the lab, and with brand new networks it's wonderful. Too much legacy hardware is going to keep it from being adopted on a large scale, and it won't happen anytime soon.
--
There is no reasonable defense against an idiot with an agenda :wq
Re:IPv6 adoption
by
rplacd
·
· Score: 3, Informative
If the world switched tomorrow, linux users would probably be the first ones up and running.
Wrong. Linux is nowhere near as IPv6-friendly as the *BSDs. To enable IPv6 in FreeBSD, for example, put 'ipv6_enable="YES"' in/etc/rc.conf and reboot. It'll autoconfig based on router advertisements, etc. You also have the option of enabling it at install time, so you can install over IPv6.
Each FreeBSD CD comes with a bunch prebuilt IPv6-ready apps, like apache, wget, etc -- apps that don't have native IPv6 support. Linux distributions are way behind when it comes to IPv6 adoption.
AEven Microsoft is on the bandwagon here. XP shipped with a "dev release" of their IPv6 code, and service pack 1 upgraded that to a production-ready release. To enable it, type "ipv6 install" at a command prompt, and you're set (no need to reboot!). The new 2003 server release comes with production quality IPv6 code as well.
Re:IPv6 adoption
by
Jordy
·
· Score: 3, Informative
Let's take Redhat 9; to enable IPv6 you have to go into/etc/sysconfig/network and stick the line 'NETWORKING_IPV6="yes"' in, then restart networking with 'service network restart.'
This same config file also will set auto tunneling 6to4, forwarding, router setup, etc. It is about as easy as you can get.
The Redhat CDs have IPv6 enabled applications and many patched apps as well. It even installs ping6, traceroute6, etc. by default for goodness sakes.
There are some pieces of IPv6 Linux is missing, but don't make it seem like there isn't any support. Linux currently is missing 6over4 (different from 6to4), proper TOS bit handling, IPsec ESP transport and AH tunneling modes (AH transport works), full mobility support (supposedly almost there) and a couple other minor things.
-- The world is neither black nor white nor good nor evil, only many shades of CowboyNeal.
Certified FUD
by
FunkyMarcus
·
· Score: 2, Informative
Fear. Uncertainty. Doubt.
The reality is that there's plenty of IPv4 space available and there will be for years to come. APNIC may be nearing the exhaustion of the address space currently assigned to that region of the world, but that doesn't mean that more space won't be assigned.
The Asia-Pacific region got additional IP address space in April 2003. And February 2003. And July 2002. And December 2001, September 2001, and December 2000. And so on.
And there's more space where that came from. Plenty of it. In fact, nearly half of IPv4's 32-bit address space is still up for grabs.
For more, see this and various other pages reachable by cutting-edge clickable links!
Slashdot Mirror
I personally know of many large corporations that have several Class-B networks that they use for non-accessible internal routing. I'm sure their numbers are much higher than just the one's I've come across. Couldn't somebody review who has all of those assigned addresses and help(force) them to migrate to private ranges?
...
RTFA
NAT is simply not the long term solution, and it is going to cause headaches when dealing with wireless devices.
NAT is pefect to extend the network of one single entity, but is a very limited solution to extend the network to several entity.
If you have only one public adresse you have a single port for each services. Despite the fact that most services can extended by virtual one this is not the case for all of them (think SSH, or IPSec for example) and this require a high degre of coordination between the entity.
So IPv6 could be the cheapest way to solve the problem. And this could provids a good boost to the network market...
Nope, absolutely wrong.
While all computers on the same NAT can directly connect to others, it cannot do so easily to others on another NAT, or other 'real' IP addresses. This effectively prevents anyone from running any server that can serve to networks outside the NAT, unless some ports are designated at the NAT router level specificly for that particular server. I don't see ISP's or network admins designating specific port ranges for every computer, as it takes work, and it could conflict with applications that uses specific port ranges (such as file transfers on MSN used by illiterate users who can't use ftp).
I would say using NAT to solve this problem is all but a cheap bandage that will cost more in the long run. IPv6 must be implemented soon to ensure the continue growth of the Internet.
Please direct all bug reports to
With IPv4, China has only 22 million IP addresses for its population of 1.3 billion people. Last year, it had about 17 million Internet subscribers, and the figure will hit 62.5 million in 2007. Japan and Korea will also run out of addresses soon, she said.
Regardless of ratios, the fact is that China, Japan, and Korea are simply running out of addresses. Though costly, implementing IPv6 is a necessity and will take care of the address shortage. Besides, most of the newer OS's (XP, Linux 2.2 or higher, OS/X) already have IPv6 built in.
until a business need exists for it. Since US companies (ISPs, schools, etc) will have to back it first, and most won't make any extra money from Lao Po in Beijing having a class D instead of a NAT'ed single IP, I don't see this happening. And since US ISPs are now NAT'ing and giving people gateways like 255.255.255.254, they've delayed our running out of IPs for quite a while. And to the guy who mentioned that MIT probably isn't using their entire class A, I think they've given a unique IP to each student, book, pencil, and brick on campus.
Wer mit Ungeheuern kämpft, mag zusehn, dass er nicht dabei zum Ungeheuer wird. --Nietzsche
One quibble-
Asia doesn't get numbering from the American Registry for Internet Numbers; they come from APNIC (Asia Pacific Network Information Center).
Otherwise, a good point.
I keep banging the IPv6 drum, but people are naturally lazy, and don't want to change unless they have to. It explains the Microsoft/Linux thing too - people can't be bothered to try it, as MS works, to a fashion.
Unfortunately, this lack of IPv6 adoption is due to Microsoft. As 90% of the online-population can't use it, the people running the services can't be bothered to support it. And while there aren't any decent services on IPv6, the impetus to upgrade it is low.
Windows XP users: ipv6 install
RedHat: http://gk.umtstrial.co.uk/~calum/ipv6-intro/
I think it can be all summed up by asking: Why don't you make all the sites you administer IPv6 only? Because then most of your audience wouldn't be able to see it.
Get your own free personal location tracker
But it does work with IANAPDN -Public Data Network, etc. Hadn't noticed that.
When is /. going to support IPv6?
I love IPv6. I've played with it in the lab, and it's nifty! I'm in charge of restructuring my company's IP layout, guess what I suggested. Interestingly enough, when I proposed my plan on #ipv6 on freenode, the answer was a resounding DON'T DO IT. I have too much legacy stuff laying around that just won't support IPv6. Funny thing is, we are doing well on technology. I think of all the other businesses in worse shape than us, and I start to think. There is no way in hell IPv6 migration will happen any time soon. It's sometimes hard for us to see, especially when we do transparent stuff at home. What we forget is all the weird hardware that companies still depend on. There is some stuff that just won't go. We bought a Cisco router 3 years ago, its IOS won't support IPv6. That's only 3 years ago! Think of the legacy crap that was installed 10 years ago that still runs! NT servers that no one upgrades because they still work. We still have a Windows 3.1 machine that does its job, and in fact we broke trying to upgrade! Still works, it's easier to leave it alone. This kind of stuff happens everywhere, I've seen plenty of businesses with old hardware that's costly to upgrade and not broken.
IPv6 is great in the lab, and with brand new networks it's wonderful. Too much legacy hardware is going to keep it from being adopted on a large scale, and it won't happen anytime soon.
There is no reasonable defense against an idiot with an agenda
:wq
If the world switched tomorrow, linux users would probably be the first ones up and running.
Wrong. Linux is nowhere near as IPv6-friendly as the *BSDs. To enable IPv6 in FreeBSD, for example, put 'ipv6_enable="YES"' in /etc/rc.conf and reboot. It'll autoconfig based on router advertisements, etc. You also have the option of enabling it at install time, so you can install over IPv6.
Each FreeBSD CD comes with a bunch prebuilt IPv6-ready apps, like apache, wget, etc -- apps that don't have native IPv6 support. Linux distributions are way behind when it comes to IPv6 adoption.
AEven Microsoft is on the bandwagon here. XP shipped with a "dev release" of their IPv6 code, and service pack 1 upgraded that to a production-ready release. To enable it, type "ipv6 install" at a command prompt, and you're set (no need to reboot!). The new 2003 server release comes with production quality IPv6 code as well.
Let's take Redhat 9; to enable IPv6 you have to go into /etc/sysconfig/network and stick the line 'NETWORKING_IPV6="yes"' in, then restart networking with 'service network restart.'
This same config file also will set auto tunneling 6to4, forwarding, router setup, etc. It is about as easy as you can get.
The Redhat CDs have IPv6 enabled applications and many patched apps as well. It even installs ping6, traceroute6, etc. by default for goodness sakes.
There are some pieces of IPv6 Linux is missing, but don't make it seem like there isn't any support. Linux currently is missing 6over4 (different from 6to4), proper TOS bit handling, IPsec ESP transport and AH tunneling modes (AH transport works), full mobility support (supposedly almost there) and a couple other minor things.
The world is neither black nor white nor good nor evil, only many shades of CowboyNeal.
Fear.
Uncertainty.
Doubt.
The reality is that there's plenty of IPv4 space available and there will be for years to come. APNIC may be nearing the exhaustion of the address space currently assigned to that region of the world, but that doesn't mean that more space won't be assigned.
The Asia-Pacific region got additional IP address space in April 2003.
And February 2003.
And July 2002.
And December 2001, September 2001, and December 2000.
And so on.
And there's more space where that came from. Plenty of it. In fact, nearly half of IPv4's 32-bit address space is still up for grabs.
For more, see this and various other pages reachable by cutting-edge clickable links!
Mark