IPv4 Addresses are like Bitcoins for next 10 years
Checking the current market /23 block of IPv4 addresses cost as much as an Bitcoin. Finding three re-allocatable /8 blocks means someone is getting a big check.
APNIC, the regional internet address registry for India, China, and 54 other Asia-Pacific nations, has found about fifty million unused IPv4 addresses under the couch. The organisation recently shared an analysis of a 2018 survey of its stakeholders and members, which concluded reclamation of unused IPv4 space should be a high …
If carrier-grade network address translation is applied to them all, it could permit hundreds of millions of devices to hook up to the internet using IPv4 rather than IPv6
Carrier-grade NAT is in place all over Asia, which is why they know all about it. China and India coudl each easily use the entire IPv4 address space. Seeing as the growth of devices is driven by mobile devices, which generally spend most of them their time on carrier networks, which already have to support IPv6, carriers probably aren't that interested in downgrading to NAT. These addresses will make more sense for legacy systems where 6-to-4 gateways aren't practicable. But I reckon "cloud" providers will be the biggest users.
On a related note: seeing an IPv4 address block that you used to manage for many, many years suddenly in-use with a new, unrelated owner is a bit like spotting your old car being driven by someone else after you flogged it to a second-hand car dealer: you're all nostalgic and teary-eyed for a moment, then you remember why you moved on. :-)
A condition of 4G/5G and DOCSIS cable standards are IPv6 support.
Google gets a ton of IPv6 queries - something like 25+% of their searches are on IPv6.
People are just too lazy to actually deploy it.
(Hey, Reg, any AAAA records yet? About the 6th/7th year of "we're working on it?"
Yes. In many ways they are like petroleum. There is a limited amount. Nobody is really sure when we will hit that limit but we have gotten far enough that there can be problems with the supply. Some groups control a massive amount for no good reason. Large parts of the world have next to none compared to their populations. There are replacements that might be useful if more people were to use them.
IPV6 has many problems, and making any change is difficult, but it is already chaotic to try to find and keep IPV4 addresses. This block may have reduced that pressure for a little bit, but in only one region of the world and only for so long. Given the aggressive CGNAT used in some parts of Asia, I imagine demand for these addresses will be fierce.
if you already have an IPv6, maybe to get the IPv4 you set up a PPPoE (or other) kind of solution, then tunnel directly into the correct router via IPv6, Voila! (that way you won't need /30 netblocks just to get one usable IPv4 address, as I've seen happen in the past).
i looked at ipv6 a long long long time ago.
i take a look every now and again too.
they put too much unnecessary nonsense ion ipv6 and tried to persuade people it was a good thing.
no nat by design
every thing publicly addressable by design
the ip address being derived from the MAC address permitting device tracking by design (ok mainly addressed now)
etc etc etc
the simplicity of ipv4 has ensured it has survived longer unexpected as we've learnt to bolt things around the standard rather than in the standard.
> they put too much unnecessary nonsense ion ipv6 and tried to persuade people it was a good thing.
That can be said of IPv4.
However, the truth is the "they" you are referring to are not IPv6 people. "They" are a host of IPv4 advocates trying to make IPv6 behave like IPv4 when such behaviour is counter productive and inefficient. As a result there have been a huge number of extension, compatibility, address translation, address mapping, address re-assignment, and tunneling specifications defined. Each used briefly then thrown away when the "IPv4 way" proved, time and again to be unnecessary or worse than the originally defined "IPv6 way".
> no nat by design
That can be said of IPv4.
NAT is an extension in both IPv4 and IPv6. It is largely unused because once admin take the step to migrate they find out how useless the IPv4 types of NAT really are in IPv6 networks.
> every thing publicly addressable by design
The IPv4 core specification defines all IPs as globally routable.
IPv6 core specification defines several ranges which must never be permitted through routers.
> the ip address being derived from the MAC address permitting device tracking by design (ok mainly addressed now)
Link-local address allocation is such a popular feature it got back-ported to IPv4 despite major performance loss on allocation, issues with IPv4 not coping with multiple IPs per machine interface, and the 169/8 range being globally routable adding security issues the IPv6 dedicated private range does not have.
You also aware of a little old protocol called ARP? The one which puts the word 'address' in the term "MAC address".
IPv4 has the same feature. Its called DHCP static address assignment, and is surprisingly popular with IPV4 admin.
IPv6 just removes several layers of complexity and dependency on DHCP servers.
> extensible headers
The great failing of IPv4. Fixed at last. No more need to, how did you put it, "bolt things around the standard".
> etc etc etc
Do continue please. So far all we can see is a list of common myths.
> the simplicity of ipv4 has ensured it has survived longer unexpected as we've learnt to bolt things around the standard rather than in the standard.
IPV6 is much, much simpler than IPv4. It is the great expanse of legacy IPv4-only software and hardware around the world combined with misinformation from IPv4-focussed people like yourself which is keeping IPV4 alive. IPv6 is a mature protocol, in so far as any network protocol ever is.
so much nonsense in this its hard to know where to start
"However, the truth is the "they" you are referring to are not IPv6 people. "They" are a host of IPv4 advocates trying to make IPv6 behave like IPv4 when such behaviour is counter productive and inefficient."
--this is so not true, ipv6 advocates are constantly telling us ipv6 is as easy and simple to use as ipv4 when its not. It turns out its advantageous for many many reasons to change the apparent source/destination or both IP addresses & even use equipment to break the end to end sessions which is handily easily achievable in IPv4 but IPv6 tries to prevent that in its very design and further anti nat stance.
"The IPv4 core specification defines all IPs as globally routable."
--rfc 1918 provides for a /8 /20 & /24 non internet routable private ip addresses
"IPv4 has the same feature. Its called DHCP static address assignment, and is surprisingly popular with IPV4 admin."
--DHCP is not built into IPv4 its a bolt on. The initial SLAAC is a privacy failure, current versions will now mean domestic users will have their DNS of internal machines that will never connect to the internet suddenly known about and catalogued by their ISP.
"Link-local address allocation is such a popular feature it got back-ported to IPv4 despite major performance loss on allocation, issues with IPv4 not coping with multiple IPs per machine interface, and the 169/8 range being globally routable adding security issues the IPv6 dedicated private range does not have."
-- 169/8 is globally routable but 169.254/16 is not!!! there is no security issue for a machine on 169.254/16 as it can't be routed too across the internet, just like the other rfc1918 addresses can't be routed to.
"You also aware of a little old protocol called ARP? The one which puts the word 'address' in the term "MAC address"."
-- ARP is what a machine does to find the MAC of a machine that can receive traffic for an IP. ARP is links L2 with L3. ARP doesn't put 'the word 'address' in the term "MAC address".' MAC address already has the word Address in it. On LAN 's machines connect via MAC addresses, IP's are needed to route from 1 LAN to another. IP's are an overlay to MAC's & provide a hierarchical method of organising & addressing machines and systems. IPv6 looks like its trying to be both an I & MAC in 1.
IPv4 has shown itself to be extensible and has had a long life far beyond what it should have had because of its extensibility.
IPv6 has not gained wide spread adoption because it has sought to be like IPv4 but better and has proven to be flawed and its proponents picking the wrong fights when it comes to privacy & usability.
IPv6 should be being advertised as greatly different to IPv4 and different toolsets being used to manage it.
if IPv6 is seen as a different thing that can run on the L2 underlay then it will likely gain more adoption.
Many provider still see it as a problem, not a solution. But Sky here in Italy announced its FTTH will be IPv6 first - MAP-T will be used for IPv4.
I had to use Hurricane Electric tunnel service to try it in my network. It handily give you a /48 prefix to play with if you ask, but it still means you're connection to the internet is in one of its tunnel gateways - although you can choose where from several locations, and you still tunnel all your IPv6 traffic through HE systems.
The worst thing till now is ISC DHCP support for IPv6, having to run two separate processes both needing updating Bind is not as smooth as it should be.
Back in 2011, before IANA ran out, we were going through more than one /8 per month. Three /8s sounds like a lot, but it's only a few months of global demand... and that's at 2011 levels. The internet has gotten bigger since then, and a lot of people are running low on addresses.
Of course, these addresses are restricted to just the APNIC region rather than being a global pool, but on the other hand it sounds like the majority of them still belong to someone and they'll just be asking politely if those owners would like to give them up, which many might not.
This isn't going to obviate the need to do IPv6.
0) Thank you all for the generous guidance, suggestion, comment, etc. through the past discussions on this forum. Below are URLs to two updates:
1) This is a discussion about the state of the IPv6 based on publicly available statistics.
2) This is a report on the possible use of the long-reserved but hardly-used 240/4 netblock and its implications.
3) We are keenly aware that our approach is rather unorthodox. However, please consider the proposed architecture as a newly created full spherical layer of cyberspace consisting of RANs (Regional Area Networks), between the current Internet proper and the subscriber premises. Each RAN is defined around one 240/4 netblock. Regarded as a private / independent environment, much of the existing Internet protocols, conventions, restrictions, etc. may be repurposed from a revised perspective in the RAN.
4) Hope you will enjoy exploring this new facility.
Feedback will be much appreciated.
Abe (2020-08-22 09:59 EDT)
Biting the hand that feeds IT © 1998–2021