Level of IPv6 Usage Is Vanishingly Small 626
An anonymous reader writes "The impending IPv4 address allocation shortage has led to a lot of speculation on the future of IPv6 (including here). A new study says that Internet IPv6 migration is not just going slowly — it has basically not even begun. After spending a year measuring IPv6 traffic across 87 ISPs around the world, the study concludes 'less than one hundredth of 1% of Internet traffic is IPv6... equivalent to the allowed parts of contaminants in drinking water.'"
Re:The end is nigh? (Score:3, Informative)
No, but you won't be able to make a site with a new ip-address, which is highly annoying. New people are not able to "join the internet" when the ISP runs out of IP-addresses. It's basically nasty.
That's why I hope they will be prepared when the time comes.
Re:You know what would help? (Score:5, Informative)
Re:Not needed. (Score:2, Informative)
There should be a karma hit for not using the preview button. It should be -1, Dumbass.
That second line should read "With more intelligent allocation of IPV4 address we wouldn't be needing IPv6 anytime soon
Re:Stupid arbitrary units of measurements (Score:3, Informative)
You didn't read the article.. Only 3 voices cried out in terror!
Re:Solution looking for a problem (Score:5, Informative)
Also, most of the world is using Windows XP. Can you show me where in my TCP/IP settings panel I am supposed to enter my IPv6 information? Exactly.
You don't. As is the benefit of IPv6, if it's installed it should be automagically configured. It shouldn't require manual configuration.
France is ready, except for Windows (Score:1, Informative)
In France, all the major ISP (Free, Orange, Neuf, etc.) and several small ones (like Nerim) provide a /64 segment as part of the usual "triple play" package for 30€.
The 2 minors problems are :
- the user has to activate the IPv6 for the gateway, in a web interface of the ISP (easy, just a checkbox)
- the user must have to have an IPv6-ready OS : nothing to do for GNU/Linux, a choice in a menu for MacOS, but a pain in Windows
As for the servers, IPv6 support is mandatory in all administrations (all equipment must be able to route et handle IPv6, but the application can still be in IPv4) and IPv6 has been declared "strategic mission" in CNRS (National French Research Center; that is all public research).
Telecom companies are also beginning to use IPv6 in the mobile television on cellphone.
So we can say that in France the situation is OK, except that IPv6 must be manually enabled by the user.
Re:Makes me happy (Score:5, Informative)
It may be just me, but I always felt IPv6 is a solution looking for the problem. [..] And lots and lots of NAT or proxying.
And NAT is a problem masquerading as a solution.
Anyway, I am ready to bet some cash that IPv6 will never become a major transport protocol.
I know I will do whatever I can to keep it far far away.
And I'll keep on enjoying all the free services people provide for IPv6 enabled hosts.
Re:Stupid arbitrary units of measurements (Score:2, Informative)
HALF a page of one book in the library of congress is IPv6. Everything else (except one stupid book in the back room) is IPv4.
Re:Makes me happy (Score:3, Informative)
"[IPv6 addresses do] not fit into any native data type (and won't until we move to 128 bit architectures - which does not seem to be very soon)"
Wow are you serious? Never heard of structs? And we all know NAT is a very annoying 'solution'. I think the real problem with IPv6 is that is isn't sufficiently backwards compatible with IPv4 (hence all that 6-over-4 and 4-over-6 nonsense.
That and it isn't really needed yet.
Re:How to really accelerate the migration... (Score:2, Informative)
http://www.ipv6porn.co.nz/ [ipv6porn.co.nz]
Re:The end is nigh? (Score:3, Informative)
Caveat - only 1 HTTPS per IP. But that really isn't that big a deal either
Maybe a few of the Class A holders like Apple or IBM should give up some of their blocks. Take IBM as an example - they subclass internal networks so they have very very few 'real IP's routable.
Or maybe if they use the evil bit [wikipedia.org] within packets we could double our existing IP4 range!
Re:Makes me happy (Score:5, Informative)
One of the key features of ipv6 is simplified routing (it was pretty much the #1 design improvement), so the amount of processing routers have to do goes way down, in spite of the higher bit count.
Please read the first page of this:
http://en.wikipedia.org/wiki/IPv6 [wikipedia.org]
and of course more if you are seriously interested.
Re:Reasons. (Score:5, Informative)
There is a killer app, It's called
news.ipv6.eweka.nl
It has 120 (!) days retention, and comes to you at gigabit speed.
All for FREE if you use ipv6.
Re:Should have gone to A.B.C.D.E.F.G format. (Score:5, Informative)
IPv4 addresses can be represented in IPv6 as 0::10.10.1.12 (Or as 0::FFFF:10.10.1.12 in some cases.)
I don't see that using dots instead of colons makes a transition any easier.
Re:Solution looking for a problem (Score:1, Informative)
*bluuaaaargh*Reverse VNC [plenz.com]*pfffft*
Re:It is obvious (Score:2, Informative)
Got something more recent to back up that 99% claim?
Re:Should have gone to A.B.C.D.E.F.G format. (Score:5, Informative)
Really? The dots vs colons thing is the single most problematic thing I've encountered. No seriously - network level is easy, just upgrade firmware or hardware. It when working with configuration files and addresses that IPv6 sucks. Firstly, : was already very widely used used, for separating IPv4 address from port number.
Just using abcd.abcd.abcd.abcd.abcd.abcd.abcd.abcd would have meant that abcd.abcd.abcd.abcd.abcd.abcd.abcd.abcd:443
would have worked much like 123.123.123.123:443, though obviously distinguishably - hex and more sections.
People seem to have settled on enclosing the IPv6 address in square brackets to make it work reasonably parseably (given abbreviation, see below) into config files and urls and stuff, at least that seems to be the most widely used convention. i.e. [abcd:abcd:abcd:abcd:abcd:abcd:abcd:abcd]:443
It works okay, but it could have been simply avoided, damnit.
Secondly, the :0000:0000:000: to :: abbreviation rule was actually a terrible mistake. It makes parsers somewhat harder to write, and means that IPv6 addresses can't be munged with regexes nearly as handily as IPv4 addresses, which seriously inconveniences time-pressed sysadmins. Yes, Ipv6 address are long if unabbreviated. But without the abbreviation they would have been REGULAR.
Re:Why it doesn't matter (Score:5, Informative)
Affect/effect are one of those amusingly nasty little hand grenades in English. Handy crib sheet:
Affect, n: emotional response. "The Minister for Granola appeared to be displaying flattened affect during his speech, leading to suspicions that he was abusing his own product."
Effect, n: causal result. "The effect of the proposed granola reform would be catastrophic."
Affect, v: alter. "The proposed reforms will affect the granola industry greatly."
Effect, v: put into immediate action. "If elected, I will effect sweeping reforms of the granola trade."
Re:I existed before NAT (Score:4, Informative)
And yet they're more secure than NAT, which you do trust?
Ever wonder how you're able to receive calls on Skype through NAT? I'll give you a hint: your network is not terribly private behind NAT ;). Private from TCP packets, sure, but NAT has to be incredibly stupid when it comes to UDP.
If you want to keep your network private, you should get a firewall that keeps your network private. NAT does not do that, but there are a lot of firewall implementations that will.
In short, when it comes to security, public IP + firewall > NAT.
Re:Should have gone to A.B.C.D.E.F.G format. (Score:2, Informative)
ipv6calc supports the rfc1924 format, even if few apps do...
apt-get install ipv6calc
ipv6calc -Ibase85 -Oipv6 '4)+k&C#VzJ4br<0wv%Yp'
1080::8:750:8052:72a8
What regex problem? (Score:3, Informative)
Looking at an app that uses regex to match both IP4 and IP6 precisely (as opposed to numbers and dots or hexchars and colons), the IP4 pattern is:
PAT_IP4 = r'\.'.join([r'(?:\d|[1-9]\d|1\d\d|2[0-4]\d|25[0-5])']*4)
RE_IP4 = re.compile(PAT_IP4+'$')
and the IP6 pattern is:
RE_IP6 = re.compile( '(?:%(hex4)s:){6}%(ls32)s$'
'|::(?:%(hex4)s:){5}%(ls32)s$'
'|(?:%(hex4)s)?::(?:%(hex4)s:){4}%(ls32)s$'
'|(?:(?:%(hex4)s:){0,1}%(hex4)s)?::(?:%(hex4)s:){3}%(ls32)s$'
'|(?:(?:%(hex4)s:){0,2}%(hex4)s)?::(?:%(hex4)s:){2}%(ls32)s$'
'|(?:(?:%(hex4)s:){0,3}%(hex4)s)?::%(hex4)s:%(ls32)s$'
'|(?:(?:%(hex4)s:){0,4}%(hex4)s)?::%(ls32)s$'
'|(?:(?:%(hex4)s:){0,5}%(hex4)s)?::%(hex4)s$'
'|(?:(?:%(hex4)s:){0,6}%(hex4)s)?::$'
% {
'ls32': r'(?:[0-9a-f]{1,4}:[0-9a-f]{1,4}|%s)'%PAT_IP4,
'hex4': r'[0-9a-f]{1,4}'
}, re.IGNORECASE)
Longer, but not any less handy. I mean, what do you care care once the
expression is compiled?
Re:Should have gone to A.B.C.D.E.F.G format. (Score:3, Informative)
And Slashdot chewed my url.
It should have been [::1]:8080
Re:The end is nigh? (Score:2, Informative)
In fact I have this vision of everyone in the world getting one routable IPv4 address
One small problem: we already have over six billion people in the world, and 32 bits provides only about four billion values. Thanks for playing.
Re:Not needed. (Score:4, Informative)
IPv6 has a feature that allows an admin to renumber an entire network quickly an easily.
See RFC2894
Re:The end is nigh? (Score:2, Informative)
Not anymore. Modern SSL versions provide a hostname hint in the (unencrypted) clienthello so single IP ssl virutal hosting is possible.
Re:congested? really? (Score:2, Informative)
Re:Should have gone to A.B.C.D.E.F.G format. (Score:2, Informative)
Close, but no cigar... think up not across (Score:2, Informative)
People have no problems remembering up to four three digit groups. So why not, expand the address space to support 0-999 values instead of just 0-255. Sure, 999 isn't a byte, but it's close enough to 2^10. Sacrificing the remaining 25 values won't hurt much. But more importantly, it would increase the address pool from 4.2 billion (minus invalid values) to 1,000,000,000,000 (a trillion) which still allows something like 200 IP addresses for every person on the planet. And with technology like NAT which should be employed for security purposes should be more than we could ever use.
Not we just need some genious to figure out how best to map that mechanism to the base-2 or IPv6 world
Re:Should have gone to A.B.C.D.E.F.G format. (Score:3, Informative)