Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

40% of Google users now connect via IPv6

google.com

Post reply on HN
Print view View original

321–330 of 353 posts

Re: 40% of Google users now connect via IPv6

#321

Post by withinboredom »

neilalexander wrote: post #63
withinboredom wrote: post #32

Earlier quoted context omitted.

Ooooh, this is actually fun. Your IP does not change, even when switching from LTE, 3G, 4G and 5G. There's an excellent post in the old Sprint forums where a company network engineer explains how it works. I can't find it though, but it's pretty wild.

In brief, it is because internet connectivity on most cellular networks goes through single gateways and devices are effectively "tunnelling" to those gateways. As long as the tunnel stays up and the mobile device is still identifiable on the RAN then the IP address assigned from the internet gateway doesn't need to change.

What I found especially interesting is how it keeps track of the device, especially when transferring from say a LTE capable station to a station that only supports 3G.

Re: 40% of Google users now connect via IPv6

#322

Post by Dagger2 »

lmm wrote: post #300
oarsinsync wrote: post #259

Earlier quoted context omitted.

Amen. I dual stacked my home network 10 years ago. 5 years ago I joined an ISP that gave me CGNATv4 and IPv6 and I opted to disable IPv6 at the router.

Why? If you have an IPv4 address, even a dynamic one, then IPv6 may not offer you a lot of practical benefit, but CGNAT-only sucks if you're at all technical.

It sucks if you're non-technical too, it's just harder for non-technical users to figure out the underlying source of any problems they have.

v6 also has better measured performance on webpage load times. Perhaps "pages load slightly slower than they could do" isn't a show-stopping problem, but faster would still be better, right?

Re: 40% of Google users now connect via IPv6

#323

Post by cassianoleal »

zeristor wrote: post #42
cassianoleal wrote: post #5

Earlier quoted context omitted.

If you have a large estate of IPv4 addresses, the more ubiquitous IPv6 gets, the less they're worth.

Not there yet it seems: https://ipv4marketgroup.com/ipv4-pricing/

We're still at a point where at least 60$ of users (according to Google) are still IPv4-only. I imagine we're still a little far from the tipping point where IPv4 becomes less valuable.

Re: 40% of Google users now connect via IPv6

#324

Post by Dagger2 »

roamerz wrote: post #137
spystath wrote: post #117

Earlier quoted context omitted.

I know it's been repeated a million times by now, but NAT is not a replacement for a firewall. Most residential routers are deny in by default so you get zero incoming connections from the internet unless you open the relevant ports, exactly as with NAT.

NAT is id10t proof though. It takes a concerted effort to set a static internal IP then NAT traffic to it and then allow that traffic through the firewall. The other advantage is that it obfuscates the internal addresses. IPV6 is is unnecessarily complex for what it solves. How hard would it have been to just add an additional octet? Pretty sure a large number of those that embrace it just love the opportunity to cha…

Quite hard, actually, since that's mostly what v6 already does and you can see how many things need to change to accommodate it. Most other parts of v6's design are the same as v4, so it's not really very complicated compared to what we've already got.

(Of course v6 adds more than just one octet, since one additional octet wouldn't be enough even for the current size of the Internet, let alone for future growth. It would be really stupid to go through all this effort, only to have to turn around and do it all again immediately afterwards because you forgot to add enough the first time around.)

Re: 40% of Google users now connect via IPv6

#325

Post by ajsnigrutin »

somat wrote: post #304
ajsnigrutin wrote: post #74

Earlier quoted context omitted.

> Firewall rules work exactly the same in IPv6 land as they do in IPv4 land. Yes, rules do work exactly the same, but with IPv4, you just let all the connections out through, and let just the established and connected ones back. > Indeed you shouldn't block ICMPv6, but that is not really making anything "more complex". But it is... you need a bunch of new rules to pass through, limit or block a bunch of ICMPv6 messag…

People keep saying "the sky is falling, with ip6 all the hosts are open to the internet" but not really it is usually one rule. on openbsd pf block outside connections from initiating connections to your hosts block in on $external_if from any to $ip6_network on ip4, if the world was just you would have the same rule(in ip4). however the world is not just and you usually only get one address so you have to pull some…

> it is usually one rule

But it is not, because you have to let ICMP pass through, for IPv6 to work (eg. for path MTU discovery to work (no more "classic" fragmentation in ipv6)).

So it's one rule to block incoming traffic, and a bunch of rules to properly allow ICMPv6 to pass through to the internal network (look at the RFC linked above)

Re: 40% of Google users now connect via IPv6

#326

Post by Dagger2 »

oefrha wrote: post #30
2Gkashmiri wrote: post #9

Why can't we paste aan ipv6 address into the browser address bar and have it go to that address? I'm forced to open [long brackets ] and the it works. Why!???? Ipv4 address just works without the hitch

Is "2001::0:8080" 2001::0 port 8080, or 2001::0:8080 port 80? There you go. Of course you can argue for automatic conversion when there's no ambiguity.

It's 2001::0:8080 port 80. Require the brackets for specifying a port number with a v6 address, and don't accept a port without the brackets. That way there's no ambiguity.

Re: 40% of Google users now connect via IPv6

#327

Post by Dagger2 »

bumblebritches5 wrote: post #176
NelsonMinar wrote: post #35

I sure wish Starlink supported IPv6. They're a brand new ISP planning for millions of customers and decided from the start they couldn't get enough IPv4 addresses for everyone. Fair enough! But we're stuck with Carrier Grade NAT and it is a drag. On Starlink it's impossible to host a server socket directly, which makes any peer to peer networking a PITA. Geocoding IP addresses doesn't work so I have to bend over back…

What's the point of Starlink? with 4G/5G home internet, there's no real point anymore.

Decent satellite internet for people who can't get anything else?

I don't see how 4G/5G home internet existing removes the point of it.

Re: 40% of Google users now connect via IPv6

#328

Post by withinboredom »

collsni wrote: post #132
withinboredom wrote: post #29

Earlier quoted context omitted.

NAT itself doesn't provide any protection at all. You can set up NAT in dozens of different configurations (1:1 NAT comes to mind), but in the way consumer routers generally set up NAT, I can see why you'd say that (despite there being standard ways to forward ports without any user intervention such as uPnP). There's nothing "secure" about NAT.

I think the argument is about the address origin being overwritten vs forwarded to the destination. Is about device tracking and privacy

I don't think that works. A router should decrement the TTL of the frame, and thus showing that there is a router between the host device. The linux default is 64 and windows is 128 IIRC, so you can easily deduce the OS just from looking at the TTL. This can tell you whether an ipv4 device is directly connected. From there, you just need to look at IP ID in the packet and figure out which ones are increasing independently to determine individual devices behind the NAT.

So, no. NAT gives you 0 privacy.

Re: 40% of Google users now connect via IPv6

#329

Post by Dagger2 »

BirAdam wrote: post #285
CharlesW wrote: post #110

Earlier quoted context omitted.

Hardly seems like “fear mongering”. https://en.m.wikipedia.org/wiki/IPv4_address_exhaustion

Any given resource, as it becomes more scarce, goes up in price thus preventing there ever being zero of the given resource. The “world ending” result that fear mongering press outlets pushed is therefore not possible.

So why did you ever pay any mind to them?

The world isn't going to end just because v4 addresses are scarce (...and I don't think anybody was arguing that it would...), but that doesn't make the problems any less real, expensive or damaging.

Re: 40% of Google users now connect via IPv6

#330

Post by Dylan16807 »

Dagger2 wrote: post #317
Dylan16807 wrote: post #306

Earlier quoted context omitted.

> NAT doesn't. > 1:many NAT does. This is technically correct but how often do you really see 1:1 NAT.

It's not even technically correct; it's just wrong. NAT doesn't make any choices on where a packet gets delivered. For packets that aren't part of an existing steam, NAT will simply not edit the packet. Unless there's a separate firewall that chooses to drop it, the packet will get delivered to whatever IP was already in the destination field, which could be the IP of one of your LAN machines.

> For packets that aren't part of an existing steam, NAT will simply not edit the packet.

A 1:1 NAT should generally just swap IP for IP and not know about streams or ports at all.

> Unless there's a separate firewall that chooses to drop it, the packet will get delivered to whatever IP was already in the destination field, which could be the IP of one of your LAN machines.

I would call that a routing rules error, even in the absence of a firewall.

Post reply on HN