|
|
From what has been discussed I can see it as a nice to have at least until the IPv4 addresses run out although this has been a topic for 20 years.
I was curious as I have never had a requirement to use it, I don't feel that not having it has restricted any internet usage and from a corporate perspective it is completely non-existent in our organisation, probably disabled for complexity or security reasons.
Well they have ran out, in some parts of the internet getting IPv4 is either extremely difficult or very expensive, the only way that can be solved is a internet wide switchover, but the isp's that have hoarded ipv4 with a lack of local urgency dont do the switch and here we are.
Web masters on many web sites of course have also not deployed v6, because they dont need to do so to keep their website online and keep it on google, so many do the absolute minimum, SSL take up didnt increase until google made it a requirement for max SEO.
I think the situation wont ever be solved, these altnets will eventually be gobbled up by someone with loads of v4 at which point those customers will get moved of CGNAT.
Edited by Chrysalis (Wed 16-Aug-23 23:08:59)
|
|
|
The ISP will keep records of which customer has which internal IP and CGNAT mapping for RIPA compliance.
Yes, but if you have a number of customers all using the same public IPv4 address, the only thing which distinguishes them is the TCP/UDP port number. The far end (the receiver of the connection) is unlikely to keep a record of this in their logs - only the source IP address. Plod then has a list of possible users, and may knock on the wrong door.
A related issue is the geolocation assigned to an IP address, which Plod tends to follow blindly:
https://arstechnica.com/tech-policy/2016/08/kansas-c...
|
|
|
Yes, but if you have a number of customers all using the same public IPv4 address, the only thing which distinguishes them is the TCP/UDP port number. I assumed under RIPA and its successors that the ISP had to keep such mapping logs for at least 12 months.
The far end (the receiver of the connection) is unlikely to keep a record of this in their logs - only the source IP address. Plod then has a list of possible users, and may knock on the wrong door.
Yes, a website such as Thinkbroadband only gets to see the source IP, but if Plod is involved they go to the ISP and ask the ISP to map that back to the user. The ISP then has the hard work of trawling the CGNAT logs, in the same way they used to look at their DHCP logs.
A related issue is the geolocation assigned to an IP address, which Plod tends to follow blindly:
https://arstechnica.com/tech-policy/2016/08/kansas-c...
Geolocation is a complete mess; and worse in some countries than others.
23 years of broadband connectivity since 1999 trial - Live BQM
|
|
Register (or login) on our website and you will not see this ad.
|
|
|
Yes, a website such as Thinkbroadband only gets to see the source IP
Actually, it gets to see the source port as well, but rarely would it bother to log it.
but if Plod is involved they go to the ISP and ask the ISP to map that back to the user. The ISP then has the hard work of trawling the CGNAT logs, in the same way they used to look at their DHCP logs.
Correct: but the point is, with CGNAT, at any given point in time *multiple users* will have been using that IP address. Not just one.
|
|
|
Correct: but the point is, with CGNAT, at any given point in time *multiple users* will have been using that IP address. Not just one.
Agreed, so the CGNAT system needs to log real user (internal IP) & public IP allocated & port and the time. Then at least some correlation can be done.
No real difference to how the mobile networks have been working for 10/15 years, almost all domestic users on a mobile (2G/3G/4G/5G) connection is behind CGNAT. Some corporate and special services provide public IP. (and some Three mobile broadband offerings).
23 years of broadband connectivity since 1999 trial - Live BQM
|
|
|
|
Adding my vote to this thread. I was very surprised that Swish doesn't support ipv6, incredible really.
|
|
|
A related issue is the geolocation assigned to an IP address, which Plod tends to follow blindly:
https://arstechnica.com/tech-policy/2016/08/kansas-c...
Geolocation is a complete mess; and worse in some countries than others.
There is a technical solution to this but it doesn't work because like all other technical solutions it requires people to follow it and nobody does:
https://geolocatemuch.com/
Instead of this simple Geofeed solution we have multiple different companies maintaining their own databases and selling this as a service to other companies which creates a whole mess and takes responsibility away from network operators.
|
|
|
|
Given the role of IP addressing in network routing an given the nature of the internet as a network, [not to mention the extension of the internet to mobile devices] geolocation is an extremely ill-conceived idea.
It deserves to fail.
|
|
|
Given the role of IP addressing in network routing an given the nature of the internet as a network, [not to mention the extension of the internet to mobile devices] geolocation is an extremely ill-conceived idea.
It deserves to fail.
It only has to be accurate enough to the city level. Providing the exact longitude/latitude or post code, etc, is absurd, but just the country and city is enough for Netflix to show you its UK catalogue and the BBC to deny you its catalogue you pay TV license for when you're abroad.
|
|
|
It only has to be accurate enough to the city level. Providing the exact longitude/latitude or post code, etc, is absurd, but just the country and city is enough for Netflix to show you its UK catalogue and the BBC to deny you its catalogue you pay TV license for when you're abroad.
Even at that level, it is absurd.
|
Pages in this thread: 
Print Thread
Chrysalis
