"The next-generation protocol, Internet Protocol version 6, is the future of the Internet. Learn how IPv6 compares to Internet Protocol version 4, understand the version 6 address formats, discover the benefits of IPv6, and learn which IT products comply with this new standard."
Post a Comment
Member since:2005-07-06
Member since:
2005-07-06
Member since:2005-11-11
IP6 is supposed to save more money than it costs by far and wide. The costs of the ever expanding routing tables in ipv4 are large and growing. Each country gets its own address space by design in ipv6.
Key advantages to using IPv6
IPv6 not only solves the shortage problem for address space, but it solves a number of other key problems as well. IPv6 includes these important benefits:
* Efficient management of address space
* Enhanced security support
* Easy maintenance of administration TCP/IP
* Elimination of the network address translation (NAT) role
* Better mobility support
Member since:
2006-04-12
Member since:
2006-01-07
Member since:
2005-11-15
IANA allocates addresses to ARIN, LACNIC, RIPE, AfriNIC, and APNIC, who allocate them to ISPs. Since IPv6 is so huge, everyone will get enough.
http://www.iana.org/ipaddress/ip-addresses.htm
Who will foot the bill is a good question.
Member since:2006-01-24
Member since:
2006-02-15
Member since:2006-01-03
I don't see how IPv6 is going to replace IPv4. As it stands now, it must be a coordinated effort. Yes, you can tunnel v6 through v4, but I bet most of the 2nd and 3rd tier will be waiting for backbone to convert first, but it doesn't look that 1st tier has any particular reasons to convert either. Next, the cost of conversion. Try to convince me that IPv6 solves all the problems we have with IPv4. Yes, NAT is a cludge, but it works, and it is firmly established and also provides the security. Huge routing tables? Well, there have been some work done optimizing the searching those tables, so besides having some extra memory, it's a problem you can deal with. In addition, you can do 'traffic engineering' (MPLA). The bottom line: there is no *really urgent* need to convert to IPv6 right now, and the longer we wait the larger the established base of IPv4 becomes, and so the more difficult is to convert from IPv4 to IPv6. Frankly, I don't believe conversion to IPv6 will ever happen. Except if somehow the current infrastructure gets completely wiped out, and you have to start from the scratch.
Member since:2005-11-11
Nice input on the subject. Considering how fast things change on the Internet how can you be sure that ipv6 will never happen? IMHO it just needs something to trigger it. Once a large % of computers support it, the move en' masse can (and I think will) happen pretty quickly (3-6 months). Seems to me most of the routers support it. No?
edit: by 'trigger' I mean something about ipv4 that really causes problems that ipv6 addresses (better). It may be a few years yet before it happens, of course.
Edited 2006-06-11 20:34
Member since:2006-01-03
> Considering how fast things change on the Internet ...
on application level Internet is highly dynamic. But on a lower level (addressing scheme and routing protocols) it's pretty static. Seems like NAT and MPLA are two major enhancements (for better or for worse) since ipv4 was introduced.
the major problem with ipv6 is that it solves primarily the problem with addressing space but comes with major problem of its own: conversion. Oh yeah, cost of retraining everyone dealing with IP, from traffic engineering (folks from NANOG) to troubleshooting to developers.
As I've said, we need nothing short of a major cataclysm, something that brings the current infrastructure to complete halt, and only v6 can solve that problem. I can't think of anything of that kind.
Incidently, since NAT, the rate of growth of IP address space dramatically slowed down. Nothing like predictions made a decade ago. I've worked for a couple of companies which --- after implementing NAT internally --- returned large blocks of unused addresses (several class B's) to IANA.
Member since:
2005-10-22
Member since:
2005-07-06
Member since:
2006-06-11
Member since:
2006-05-14
....well, almost 
There's a good chance that the core internet router serving your ISP is already using IPv6 in some fashion or another. It may be configured incorrectly, but it's already being used throughout the internet core.
http://www.sixxs.net/tools/grh/dfp/all/
I believe that the use of NAT is a major contribution to the slow adoption of IPv6 at the user level. It (NAT) has helped extend the life of IPv4 for a decade or maybe two (as many have already pointed out). No pending crisis, so why change....the "it's good enough for now" attitude. But I think that could be a dangerous attitude to take. The roll out of IPv6 may be an extended one, but one that shouldn't be ignored. We had decades to prepare for Y2K (non-event as it was for many), but look at all the time and expense caused by that event and the procrastination surrounding the planning for its passing.
Why not at least prepare for IPv6's future implementation now? As we look at evaluating, purchasing and deploying new hardware and software, I think it wise to consider the IPv6 "readiness". It appears some are already preparing for the next greatest software experience out of Redmond by beefing up current computers or new purchases. That's probably a wise choice, as would investing in IPv6 compliant solutions. If not, well there may be some more lucrative consulting contracts for some of us in the future
Member since:
2006-01-16
I already have a native IPv6 connection to the internet, although there are not many sites available through it though ;-).
IPv6 has many huge advantages over v4, not only there won't anytime be a shortage of addresses anymore - which is a huge problem in europe and even more in asia/africa where just too less address spaces were given to. Also IPv6 has very good auto-configuration features, routing/gateway features and so on. For example:
To be in IPv4 I need to either setup an IP adress or DHCP, which works, or not. I can't work as server on the net, as I either haven't got a static IP address - or even worse - am behind a proxy server.
To be in IPv6, I just have to connect the cable. An address is automatically assigned to my machine according to my mac address (in my particular example) and it will never change.
Everytime something new and fancy comes along, there are always people growning and telling "I don't need that, it's not good for anything" and so on, without even knowing a shit about this piece of innovation. Come on, guys! There are people who need the bigger address space. There are people who know that DHCP is a pain in the a**. There are people suffering from problems with NAT and video streaming (VoIP,..) protocols. There are features you likely would benefit from, but just don't know about yet. Get informed!
AND BTW: Even Internet Explorer supports IPv6, even on Win2k, you just need to install some drivers form Microsoft for free. Not to mention that nearly every decent and important piece of open source software supports v6 already, where it makes sense. For example, if I connect to any host in university per SSH, it will connect to the IPv6 address, as this is preferred by default (the IPv6 address is given in the AAAA-Record on the nameserver, alike the A-Record for v4). Even mplayer always tries (and mostly fails) to connect to v6 first. Did I mention that multicasting is builtin in v6?
Edited 2006-06-12 13:58
Member since:
2005-08-18
"As it stands now, it must be a coordinated effort"
Eh, no it doesnt. There's already plenty of IPv6 deployment in Europe and Asia, especially Japan. The U.S is the only ones slacking. IPv6 can easily co-exist with IPv4. I really dont get what you think needs to be "coordinated".
"Yes, NAT is a cludge, but it works, and it is firmly established and also provides the security."
There's no security in NAT.
"Huge routing tables? Well, there have been some work done optimizing the searching those tables, so besides having some extra memory, it's a problem you can deal with. In addition, you can do 'traffic engineering' (MPLA)."
MPLA? Maybe you mean MPLS but that has nothing to do with routing table sizes and neither has "traffic engineering" in general.
Edited 2006-06-13 03:33
