About 10 years ago, IBM used to use the 9.0.0.0/8 space in basically exactly the same way as one would use 10.0.0.0/8, for internal-only networking. Each workstation got its own 9.x.x.x IP, but it wasn't routable from outside. I hope they stopped doing that, but I doubt it.
IBM owned 9/8 so this is a legitimate use of address space. All hosts should have globally unique addresses, even if you want to use NAT to hide various things. IBM does multiple acquisitions per year. Imagine merging two corporate networks that both use 10/8; it's a nightmare.
Who is squatting IPv4 addresses?
161–170 of 208 posts
Re: Who is squatting IPv4 addresses?
#162Earlier quoted context omitted.
As for point one: I'm not talking about client/server access to services. I'm talking about the capacity for endpoints to talk to each other. IPv4 would be fine if we want a fully centralized computing infrastructure where everything is only a thin client, but that's a future with zero privacy or personal freedom. I don't think there's anything special about IPv4 in terms of DDOS mitigation. What you're probably seei…
I'm not clear that IPv4 doesn't offer at least one measure of reduction against a DDoS, and that's just one time hits every second from 1 quadrillion unique IPv6 addresses. You simply can't have that level of problem in IPv4. However, I have never been on the inside of a DDoS attack, so I don't speak from experience on this. In regards to mitigation, what we are talking about is an exclusive network with central cont…
When DDOS black holing is done the recipient will actually look up the BGP advertised prefix from which the attack is coming and black hole the whole thing. Many IPv6 prefixes are /32 and /48.
I am pretty deeply familiar with this stuff. There's nothing about IPv6 that makes current mitigation techniques much harder. The most logical explanation for IPv4-only in the DDOS protection world is just to limit the attack surface by picking the lowest common denominator address. That way you only have to defend in the IPv4 realm instead of in two addressing realms.
IPv6-only would give you the same effect but there are still too many edge devices without IPv6 addresses to use IPv6 alone for anything public facing. IPv6-only systems are sometimes used in private networks, as bastion boxes, etc.
Re: Who is squatting IPv4 addresses?
#163The author of trilema.com at some point boasted of having bought a /16 and then renting it out. Something I don't quite understand is why IPV6 is better, if anything sticking to IPV4 will lead to more "selective" use. Actually useful things get an IP, the rest, well, better get more useful
Re: Who is squatting IPv4 addresses?
#164Earlier quoted context omitted.
Should change to a montly fee. Then people would get rid of the ones not used.
> Should change to a montly fee. Then people would get rid of the ones not used. There is a fee. Your EUR 1,400 annual fee. For that money you get one IPv4 and one IPv6 (IPv4 subject to availability, obvs!). Above that they charge per resource assignment, 50EUR per annum per resource assignment ( defined as: "IPv4 and IPv6 PI assignments; Anycast assignments; IPv4 and IPv6 IXP assignments; and Legacy IPv4 resource re…
They would just start to put IPV4 blocks behind shell companies.
Re: Who is squatting IPv4 addresses?
#165About 10 years ago, IBM used to use the 9.0.0.0/8 space in basically exactly the same way as one would use 10.0.0.0/8, for internal-only networking. Each workstation got its own 9.x.x.x IP, but it wasn't routable from outside. I hope they stopped doing that, but I doubt it.
At MIT we had 18.0.0.0/8 until they sold a bunch of it to Amazon.
MIT Student Radio WTBS 1964-65.
https://www.youtube.com/watch?v=PI2Xx3XSTFw
WTBS "The Ghetto": Soul-Music Radio Show. Created by Black MIT students in 1970, this radio program gained popularity in the Cambridge/Boston area.
https://www.blackhistory.mit.edu/story/wtbs-ghetto
Promo for MIT BSU's "The Ghetto" (WTBS 88.1 FM)
Re: Who is squatting IPv4 addresses?
#166Earlier quoted context omitted.
As for point one: I'm not talking about client/server access to services. I'm talking about the capacity for endpoints to talk to each other. IPv4 would be fine if we want a fully centralized computing infrastructure where everything is only a thin client, but that's a future with zero privacy or personal freedom. I don't think there's anything special about IPv4 in terms of DDOS mitigation. What you're probably seei…
I'm not clear that IPv4 doesn't offer at least one measure of reduction against a DDoS, and that's just one time hits every second from 1 quadrillion unique IPv6 addresses. You simply can't have that level of problem in IPv4. However, I have never been on the inside of a DDoS attack, so I don't speak from experience on this. In regards to mitigation, what we are talking about is an exclusive network with central cont…
Re: Who is squatting IPv4 addresses?
#167Earlier quoted context omitted.
I applied in October last year, and at that time the waiting list was zero days, so I received our /24 instantaneously. Some nice data on the prices of IPv4 addresses: https://auctions.ipv4.global/prior-sales
Should change to a montly fee. Then people would get rid of the ones not used.
Re: Who is squatting IPv4 addresses?
#168It was "hot" so he couldn't just squat on it or sell it in the open, but he laundered it by trading it to some shady company in exchange for free network services for life.
If I were him, I would have printed out all the addresses on little slips of paper and taken an "IPV4 Address Bath" like Huell's scene in Breaking Bad:
https://www.youtube.com/watch?v=7HrmD_vIMIk
>(sexy lip bite) ... "I gotta do it, man!" ... "Mexico, all's I'm sayin'!"
Re: Who is squatting IPv4 addresses?
#169Earlier quoted context omitted.
> Should change to a montly fee. Then people would get rid of the ones not used. There is a fee. Your EUR 1,400 annual fee. For that money you get one IPv4 and one IPv6 (IPv4 subject to availability, obvs!). Above that they charge per resource assignment, 50EUR per annum per resource assignment ( defined as: "IPv4 and IPv6 PI assignments; Anycast assignments; IPv4 and IPv6 IXP assignments; and Legacy IPv4 resource re…
> And yes, I think the 50EUR should be put on a ladder scale so hoarders get charged exponentially more. ;-) They would just start to put IPV4 blocks behind shell companies.
Several people on the list, including someone who has made quite the sum from collecting allocations and then reselling them, were outraged that this is going on and deem it in violation of the spirit of the waiting list (that spirit being that it is supposed to give small entrants one last method of getting onto the Internet with their own /24). One of the proposed changes is to make transfers of IPv4 space received via the waiting list non-transferable retroactive to January 2021. Of course, transfers that have already happened wouldn't be reversed, since that wouldn't "be fair." Thus, all of the people who got the last "free" /22 blocks around November 2019 and who have gone on to resell those addresses (like the participant on the list who is now advocating that addresses shouldn't be allowed to be transferred) when the 24-month transfer delay expired get to keep their gains. Everyone who comes after who signed up under those rules gets the rules changed out from underneath them.
The whole system of dealing with IPv4 assignments and allocations has gotten completely out of hand. We have these large entities all acting like they are just volunteers doing a charitable thing keeping this polite ad-hoc system going, when that hasn't been true for at least a decade.
Re: Who is squatting IPv4 addresses?
#170Earlier quoted context omitted.
My old uni holds two /16s which are used for clients. Firewalled, but if you use the Wi-Fi there then you are getting your own "personal" IPv4 address per device.
You mean they are using IP's as they were originally designed? Those people probably have NO conflicts or problems with things like video conferencing, VPN's, etc. NAT is a cludge, not a security feature.