MIT no longer owns 18.0.0.0/8 • Hacker News Board

Re: MIT no longer owns 18.0.0.0/8

#121

Post by mjcl » Thu, Apr 20, 2017, 12:34 AM UTC

flyinghamster wrote: post #107

kiallmacinnes wrote: post #78

Earlier quoted context omitted.

Just to be sure, with a routed network being aimed at your house, you'll actually receive two subnets (well, a single address in one subnet, and a whole second subnet). You need to use DHCP-PD or similar to make use of it. The /128 sounds perfectly normal, and is used as a "point to point" link between the ISP and your modem. Over this, DHCP prefix delegation typically gives you a whole second subnet of /60 or so. If…

Hmmm, maybe I should revisit this. I first tried it a year or so ago with pfSense on a virtual machine. Between the IPv6 issues and generally poor performance (long story), I went back to my IPv4-only configuration with DD-WRT on a D-Link DIR-825. I'm now using a Ubiquiti EdgeRouter POE, and maybe I should give it another shot. I'm not convinced I'll actually get a prefix from Comcast, though.

It may differ by region, but in the SF Bay Area I've been getting a /60 from Comcast residential for a few years now using a MikroTik router. See here: http://i.imgur.com/2mrRk01.png

Re: MIT no longer owns 18.0.0.0/8

#122

Post by imjustapie » Thu, Apr 20, 2017, 12:44 AM UTC

feld wrote: post #10

imjustapie wrote: post #2

An updated list of MIT's IPv4 address ownership: https://whois.arin.net/rest/org/MIT-2/nets

Probably easier to just look at BGP http://bgp.he.net/AS3#_prefixes

Ah, it looks like BGP caught up. Yesterday AS3 was still announcing 18.0.0.0/8.

Re: MIT no longer owns 18.0.0.0/8

#123

Post by imjustapie » Thu, Apr 20, 2017, 12:45 AM UTC

knd775 wrote: post #82

imjustapie wrote: post #2

An updated list of MIT's IPv4 address ownership: https://whois.arin.net/rest/org/MIT-2/nets

Why'd they get of them in such odd intervals?

Well, I think some of the intervals are still actively used by building networks!

Re: MIT no longer owns 18.0.0.0/8

#124

Post by virtuallynathan » Thu, Apr 20, 2017, 12:47 AM UTC

ajross wrote: post #79

flyinghamster wrote: post #32

Earlier quoted context omitted.

I've tried to run IPv6 on my Comcast connection. Unsurprisingly, Comcast doesn't get the point of IPv6. They allocate one single IPv6 address to my router. That's right, ONE. LOUSY. ADDRESS. This seems to be a "feature" of owning your own modem rather than renting one of theirs. I hesitate to even try to get this fixed via their tech support. I guess if I want IPv6 connectivity, I'm either going to have to get a tunn…

Actually, Comcast supports DHCP prefix delegation (I'm led to believe everywhere, but it works at my home anyway), which is pretty much the Right Thing in this space. Your default DHCPv6 request is from a single client and it returns a single IP, as expected. And you can then do NAT across that if you want, which is what their default routers do. But if you want a subnet, you can make a second DHCPv6 request with pre…

Comcast will give out up to a /60 for residential if you hint for it in DHCPv6-PD.

Re: MIT no longer owns 18.0.0.0/8

#125

Post by virtuallynathan » Thu, Apr 20, 2017, 12:48 AM UTC

apostacy wrote: post #109

kiallmacinnes wrote: post #85

Earlier quoted context omitted.

This sounds like CloudFlare etc have just blacklisted hosting provider IPv6 blocks, as there a real easy way to get millions of "clean" IPs priced by the hour. CloudFlare etc certainly dont do this with SiXXS IPv6 addresses, or any traditional residential provider I'm aware of.

Well, I'd love to use ipv6. In fact, I can't wait. But, what am I supposed to do? * T-mobile: ipv4 * Office: ipv4 * Verizon: ipv4 * University: ipv4 * Linode: ipv6 (but flagged as hostile) Google will not just show me captchas, but will deactivate accounts if I use some of their apis. I had to reverify several accounts when I've tried using ipv6. I'm not being backward about ipv4, I'm really not. But we're in denial…

T-Mobile stopped handing out IPv4 addresses to handsets on iOS 10 -- IPv6 Only.

Re: MIT no longer owns 18.0.0.0/8

#126

Post by flyinghamster » Thu, Apr 20, 2017, 1:10 AM UTC

ajross wrote: post #79

flyinghamster wrote: post #32

Earlier quoted context omitted.

I've tried to run IPv6 on my Comcast connection. Unsurprisingly, Comcast doesn't get the point of IPv6. They allocate one single IPv6 address to my router. That's right, ONE. LOUSY. ADDRESS. This seems to be a "feature" of owning your own modem rather than renting one of theirs. I hesitate to even try to get this fixed via their tech support. I guess if I want IPv6 connectivity, I'm either going to have to get a tunn…

Actually, Comcast supports DHCP prefix delegation (I'm led to believe everywhere, but it works at my home anyway), which is pretty much the Right Thing in this space. Your default DHCPv6 request is from a single client and it returns a single IP, as expected. And you can then do NAT across that if you want, which is what their default routers do. But if you want a subnet, you can make a second DHCPv6 request with pre…

UPDATE: I now have IPv6 connectivity via my shiny new EdgeRouter POE. Now I just have to poke a few holes in the firewall where appropriate (now that I've set drop-by-default rules).

Getting it to work with the ER-POE was non-intuitive but not as bad as I was expecting. I'm even getting a /60...

Re: MIT no longer owns 18.0.0.0/8

#127

Post by lokedhs » Thu, Apr 20, 2017, 1:32 AM UTC

jagger27 wrote: post #38

flyinghamster wrote: post #32

Earlier quoted context omitted.

I've tried to run IPv6 on my Comcast connection. Unsurprisingly, Comcast doesn't get the point of IPv6. They allocate one single IPv6 address to my router. That's right, ONE. LOUSY. ADDRESS. This seems to be a "feature" of owning your own modem rather than renting one of theirs. I hesitate to even try to get this fixed via their tech support. I guess if I want IPv6 connectivity, I'm either going to have to get a tunn…

Rogers in Canada gives /64, which is lovely.

My old ISP gave me a /64 and it was anything but lovely. It's too small to break up into subnets, and they were unable to give anything larger.

This caused me to switch ISP's and now I have a static /48. That's much nicer.

Re: MIT no longer owns 18.0.0.0/8

#128

Post by niij » Thu, Apr 20, 2017, 2:14 AM UTC

tarboreus wrote: post #83

runjake wrote: post #65

Earlier quoted context omitted.

Guy who manages a two-dozen site v6 network with tens of thousands of clients here. What issues are you referring to?

Yeah if you want to make a cname for yourself in this field you'll have to bring your AAA game.

*AAAA

Re: MIT no longer owns 18.0.0.0/8

#129

Post by crisopolis » Thu, Apr 20, 2017, 2:27 AM UTC

apostacy wrote: post #109

kiallmacinnes wrote: post #85

Earlier quoted context omitted.

This sounds like CloudFlare etc have just blacklisted hosting provider IPv6 blocks, as there a real easy way to get millions of "clean" IPs priced by the hour. CloudFlare etc certainly dont do this with SiXXS IPv6 addresses, or any traditional residential provider I'm aware of.

Well, I'd love to use ipv6. In fact, I can't wait. But, what am I supposed to do? * T-mobile: ipv4 * Office: ipv4 * Verizon: ipv4 * University: ipv4 * Linode: ipv6 (but flagged as hostile) Google will not just show me captchas, but will deactivate accounts if I use some of their apis. I had to reverify several accounts when I've tried using ipv6. I'm not being backward about ipv4, I'm really not. But we're in denial…

Yeah, T-Mo is IPv6 on smartphones such as Android/iPhone.

Re: MIT no longer owns 18.0.0.0/8

#130

Post by mortehu » Thu, Apr 20, 2017, 3:21 AM UTC

mirashii wrote: post #112

mortehu wrote: post #100

Earlier quoted context omitted.

I'm not sure what you're referring to when you say "Thats not how it works". I was referring to people with broken IPv6 setups, mostly Teredo and 6to4 (old Mac OS X versions), for whom IPv4 worked but IPv6 was not actually routable. At the time, publishing AAAA records would lock out all of those users.

> If you can't access IPv6 enabled sites, you can't access Google He was referring to your lack of qualification in the above. An IPv4 user can indeed access Google, which is an IPv6 enabled site.

Oh, I see now. You're reading "Having IPv6 access is a necessary for accessing Google". What I meant was "If the presence of AAAA records means you can't access a site, then you can't access Google".

A user who only has IPv4 connectivity relies on their software recognizing this when accessing an IPv6 enabled site. Software's failure to do so is one of the main reasons for the slow rollout of IPv6.