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

The IPv6 Transition

potaroo.net

Post reply on HN
Print view View original

11–20 of 433 posts

Re: The IPv6 Transition

#11

Post by Dylan16807 »

hairyplanter wrote: post #10

I have fully implemented IPv6 in my home network. I have even implemented an IPv6-Only network. It fully works, including accessing IPv4 only websites like github.com via DNS64 and NAT64 at my router. The only practically useful thing about my IPv6 enabled network is that I can run globally routable services on my lan, without NAT port mapping. Of course, only if the client is also IPv6. Other than this one use case,…

NAT is mostly okay, but carrier grade NAT where you can't forward a port causes real problems.

IPv4 exhaustion is a real problem, it's just not enough to motivate people much.

Re: The IPv6 Transition

#12

Post by Uptrenda »

These charts that show IPv6 adoption really don't mean shit. The thing is: every single device out there isn't being used directly by a human bean (and a real hero.) They include things like sensors, smart lights, fridges, washing machines, a huge huge number of mobile devices, company networks, ... apparently even tooth brushes? Look at another sector and the story is ((quite horrible.)) I'm talking a regular fixed home network.

Start by looking at routers for IPv6 support. And what do you see? Total crap across the board. Here's some of the issues I've seen. Routers that have no IPv6 support (common for ISP provided routers.) Routers that have NO FIREWALL for IPv6. Routers that crash every 3 minutes after assigning an address. Routers that don't support the exact combination of network details to setup IPv6 on your network (there are multiple ways to deploy IPv6.)

What about if you want to use features like UPnP with IPv6 (something that would probably be useful for some software given that IPv6 is supposed to give you public addresses but firewall it on the router.) What I've found is there's really just one UPnP library that every router uses even though it sucks. miniupnpd. This is a library that can barely manage to handle different types of addresses. It's really a mixed bag whether an IPv6 firmware will have miniupnpd enabled and if its built for IPv6 (and if anyone bothered to test it.) The odds go down dramatically.

If you manage to get a router with IPv6 at home working alongside other useful Internet standards made for it (since 2010) color me impressed. You probably buy a lottery ticket at that point. Because if testing IPv6 deployments for the past 2 years has taught me anything: its that no one really cares about this shit. Present day, present time. You still hear people telling others to turn IPv6 off for some vague reason ('security', 'bad', 'problems.') These people don't really have a clue. It's all just a massive cope because they tried to get it to work and failed. And after the shit I've said I can't say I blame them. But I also want to note that their conclusions are BS.

Re: The IPv6 Transition

#13

Post by somat »

kijin wrote: post #7

I think the article's diagnosis is spot on. The urgency of IPv6 adoption was predicated on the assumption that every connected device, both server and client, needs a unique and stable IP address. Back when IPv6 was first discussed, you couldn't even host two HTTPS sites on the same IP/port combination! That was such a colossal waste of IP addresses. Another thing that changed on the server side was that, thanks to A…

it is unfortunate that tcp and ip are as interlocked as they are, by which I mean, there is no way to keep your tcp connection while swapping out the underlying ip addresses.

This is not actually a real problem, we do just fine without it, it can be solved at higher or lower layers. But it would have been nice to have.

Re: The IPv6 Transition

#14

Post by Dylan16807 »

AdamH12113 wrote: post #8

I’ve often wondered if going with 64-bit addresses with a dotted quad hex notation would have eased the roll-out. I remember a lot of resistance when IPv6 was first announced along the lines of “I can’t memorize/type in giant addresses and I don’t want to have to use DHCP and DNS everywhere.” It felt like IPv6 never recovered from a bad first impression.

Couldn't anyone in that position use 2xxx:yyyy:zzzz:ww::1, 2xxx:yyyy:zzzz:ww::2, etc. and get the same effect?

Re: The IPv6 Transition

#15

Post by saurik »

Dylan16807 wrote: post #11
hairyplanter wrote: post #10

I have fully implemented IPv6 in my home network. I have even implemented an IPv6-Only network. It fully works, including accessing IPv4 only websites like github.com via DNS64 and NAT64 at my router. The only practically useful thing about my IPv6 enabled network is that I can run globally routable services on my lan, without NAT port mapping. Of course, only if the client is also IPv6. Other than this one use case,…

NAT is mostly okay, but carrier grade NAT where you can't forward a port causes real problems. IPv4 exhaustion is a real problem, it's just not enough to motivate people much.

Have you tried using PCP to forward the port? I was under the (maybe-incorrect, and if so I would really like to learn) impression that most major CG-NAT setups supported it.

Re: The IPv6 Transition

#16

Post by alexchamberlain »

gorgoiler wrote: post #3

> In 2024 it’s estimated that 20 billion devices use the Internet, yet the Internet’s IPv4 routing table only encompasses 3.03 billion addresses … sharing each individual IPv4 address across an average of 7 devices. …but the graph below that text shows 40% of traffic is IPv6, so the v4 space is only shared across 12e9 devices? In my experience the big holdouts these days are corporate networks. All my domestic ISPs (…

FWIW, domestic ISPs in the UK are lagging on IPv6; I'm with Vrigin Media and, afaict, there is no immediate plan to deploy it either.

Re: The IPv6 Transition

#17

Post by kijin »

Dylan16807 wrote: post #11
hairyplanter wrote: post #10

I have fully implemented IPv6 in my home network. I have even implemented an IPv6-Only network. It fully works, including accessing IPv4 only websites like github.com via DNS64 and NAT64 at my router. The only practically useful thing about my IPv6 enabled network is that I can run globally routable services on my lan, without NAT port mapping. Of course, only if the client is also IPv6. Other than this one use case,…

NAT is mostly okay, but carrier grade NAT where you can't forward a port causes real problems. IPv4 exhaustion is a real problem, it's just not enough to motivate people much.

If it was a real problem, market pricing would reflect the increasing severity of that problem.

The truth is that people who care about port forwarding are such a small minority -- especially now that P2P file sharing has lost its hype -- that they don't make a visible dent in the rate of IPv4 exhaustion.

Re: The IPv6 Transition

#18

Post by kalleboo »

The internet stopped being a network of peers where everyone needed an address and is now a split into producers (a handful of large companies) and consumers (everyone else).

The consumers are not expected to need a public address where they can be reached - in fact, having a public address is actually a security and privacy risk.

Re: The IPv6 Transition

#19

Post by edf13 »

kijin wrote: post #7

I think the article's diagnosis is spot on. The urgency of IPv6 adoption was predicated on the assumption that every connected device, both server and client, needs a unique and stable IP address. Back when IPv6 was first discussed, you couldn't even host two HTTPS sites on the same IP/port combination! That was such a colossal waste of IP addresses. Another thing that changed on the server side was that, thanks to A…

Exactly this… which raises the question- do we need ipv6 at all?

Re: The IPv6 Transition

#20

Post by Animats »

China's IPv6 transition is 74% complete.[1] Conversion to IPv6 was specifically called out in China's 14th Five Year Plan, which gives the goal high visibility within the government and the Party. Conversion is quite far along. The current goal is everything IPv6 enabled by 2025, IPv4 turns off in 2030.

99% of the top 100 mobile applications in China are on IPv6. China Mobile's backbone is now IPv6 only.

[1] https://www.china-ipv6.cn/#/

Post reply on HN