Tell HN: IPv6-only still pretty much unusable • Hacker News Board

Re: Tell HN: IPv6-only still pretty much unusable

#321

Post by Latty » Wed, Dec 07, 2022, 8:50 PM UTC

mort96 wrote: post #110

arcade79 wrote: post #31

I was thinking about ipv6 the other day. I concluded in my head that adoption was just around 5-10%. Luckily I went to verify that with statistics. https://www.google.com/intl/en/ipv6/statistics.html While price of ipv4 addresses are increasing, the world has slowly been adopting ipv6. From the graph above, I'd say we cross over 50% in about 2-3 years time. At some point the "dash" to adopt ipv6 starts, and brave fol…

> At some point the "dash" to adopt ipv6 starts, and brave folks will drop support for ipv4. I wouldn't be sure about that. I don't see any "dash" to support v6 in our future, when the option to just keep working around issues with v4 is so much easier and cheaper in the moment. Really, what does anyone have to gain by switching to v6?

The thing is, it's not v6 or v4, it's v6, v4 with a price premium, or v4 with CGNAT.

CGNAT sucks. Stuff blocks you because you get lumped in with other users. You can't take inbound connections. Average users don't know that, but they get annoyed with side effects. Not being able to play multiplayer stuff or it being slow/high latency because of no inbound connection. Having to do extra CAPTCHAs, being straight out blocked, etc...

Price and annoyance are absolutely things people want to avoid.

I guarantee you at some point, some ISP is going to realise they can market themselves as the gamer ISP and sell IPv6 as the option for pro gamers to ensure the lowest latency in their games. CGNAT will be the congested roads, IPv6 the open motorway. (To be clear, it obviously isn't that simple, but I'd put money on that's how they'll sell it.)

Re: Tell HN: IPv6-only still pretty much unusable

#322

Post by uuddlrlrbaba » Wed, Dec 07, 2022, 8:51 PM UTC

kd913 wrote: post #265

cramjabsyn wrote: post #226

Earlier quoted context omitted.

> It’s just that transition is expensive It’s impossible to finish a transition when the old version has no end of life in sight.

>It’s impossible to finish a transition when the old version has no end of life in sight. The end of life is gonna happen when ipv4 addresses end up being cost prohibitive. They are already some $50 an ip address. That is gonna be cost prohibitive in developing Countries, who already are making a transition to ipv6.

Increased demand will drive the price up, but that's not end of life.

Developed countries will pay obscene amounts for ipv4 space. Just as they do for shorthand .com domains.

Re: Tell HN: IPv6-only still pretty much unusable

#323

Post by nightfly » Wed, Dec 07, 2022, 8:59 PM UTC

throw0101a wrote: post #164

redox99 wrote: post #32

IPv6 has been one of the biggest failures in the last couple of decades. And I don't mean adoption, I mean the standard itself. If IPv6 were IPv4 with more octets, then we would all have been using it for like a decade. Yes, I understand it would still require some breaking changes, but it would have been a million times easier to upgrade, as it would be a kind of superset of IPv4 (1.2.3.4 can be referred as 0.0.0.0.…

> And regarding compatibility, with ipv4+ if you have a 0.0.0.0.x.x.x.x ip address you would be able to talk to both ipv4+ aware and legacy ipv4 devices natively without any tunneling (because you also own the legacy, non quad 0 ip address). This exists: > Addresses in this group consist of an 80-bit prefix of zeros, the next 16 bits are ones, and the remaining, least-significant 32 bits contain the IPv4 address. For…

We've tried to roll out IPv6 at my work, and after several years of it causing more issues than it help we turned it all off again... We're gonna get back to it again this year, but I can see many organizations just not doing that until absolutely forced.

Re: Tell HN: IPv6-only still pretty much unusable

#324

Post by qwertywert_ » Wed, Dec 07, 2022, 9:05 PM UTC

throw0101a wrote: post #177

patrakov wrote: post #149

Earlier quoted context omitted.

> NAT isn't needed anymore. False. The most obvious case is multi-homing (for redundancy, fail-over, and policy-routing reasons) without an AS available and thus without BGP. In other words, a typical case when a user has a fiber connection and LTE as a backup. Then it is the router who should pick the correct source address, according to the link which is up. Another reason is to deal with dynamic addressing from th…

>> * NAT isn't needed anymore.* > False. "IPv6 Multihoming without Network Address Translation" Network Address and Port Translation (NAPT) works well for conserving global addresses and addressing multihoming requirements because an IPv4 NAPT router implements three functions: source address selection, next-hop resolution, and (optionally) DNS resolution. For IPv6 hosts, one approach could be the use of IPv6-to-IPv6…

I just did a quick read but I don't understand how this would help the case of your Gateway ethernet link going down temporarily and switching to Cellular WAN?

The client would still need some smart steering to select the correct route no? Does the gateway invalidate the ethernet address somehow? But with NAT you don't need to worry about it.

Re: Tell HN: IPv6-only still pretty much unusable

#325

Post by pm2222 » Wed, Dec 07, 2022, 9:07 PM UTC

No post body was provided.

Re: Tell HN: IPv6-only still pretty much unusable

#326

Post by Sesse__ » Wed, Dec 07, 2022, 9:08 PM UTC

est31 wrote: post #157

redox99 wrote: post #32

IPv6 has been one of the biggest failures in the last couple of decades. And I don't mean adoption, I mean the standard itself. If IPv6 were IPv4 with more octets, then we would all have been using it for like a decade. Yes, I understand it would still require some breaking changes, but it would have been a million times easier to upgrade, as it would be a kind of superset of IPv4 (1.2.3.4 can be referred as 0.0.0.0.…

One of the ideas of ipv6 was to reduce routing tables, those tables that backbone providers have to keep in memory and look up for incoming traffic. With ipv4's fragmented allocation scheme, these routing tables are huge. With ipv6, even huge companies like amazon only have a couple of global allocations. A "ipv4 with more octets" scheme would have kept that fragmentation around. That being said, Amazon currently has…

> One of the ideas of ipv6 was to reduce routing tables

That idea was abandoned about 20 years ago. One fairly quickly discovered that hierarchical routing does not work well in the real Internet, where redundancy is done on the IP level with everybody and their dog having provider-indepent IP space.

> like that the only standard for local ipv6 addresses

Which one is the only standard? Link-local, site-local, ULA, or using a non-routed netblock with or without DHCPv6 (or RA)?

Re: Tell HN: IPv6-only still pretty much unusable

#327

Post by bryanlarsen » Wed, Dec 07, 2022, 9:09 PM UTC

bombcar wrote: post #55

dboreham wrote: post #8

> The GitHub API and its code load endpoints are not reachable via IPv6 This is quite strange given the period around 2010 where there were government edicts that IPv6 must be supported. Perhaps GitHub wasn't mission critical back then.

There are a number of companies (not naming names) that do NOT support IPv6 for commoners but if you are on a government account/contract you have different government endpoints that ... work over IPv6. Infuriating, but perhaps understandable; they're willing to support the government over IPv6 because it's required and they pay more. But I wish there was a "I know what I'm doing and understand you'll make fun of me…

There's probably a ddos service in front of the public facing commercial service that isn't there on the government service.

Re: Tell HN: IPv6-only still pretty much unusable

#328

Post by Yuyudo_Comiketo » Wed, Dec 07, 2022, 9:11 PM UTC

Klonoar wrote: post #298

Yuyudo_Comiketo wrote: post #289

All ipv6 shortcomings discussion aside; What I think is the more vital problem to focus on is that the governments clearly don't want us mere mortals to expose our own servers running on our own hardware to the outside world (most often justifying that with "it's for your own security" mantra, for we're all deemed too dumb to figure that out for ourselves). ISP-imposed ipv4 double NAT (imposed on ISPs by the governme…

Can you provide a source for that government claim?

- Comment sections getting closed for anonymous replies almost everywhere over the course of the last 15 years

- Undisguised surveillance becoming the new normal

- Confinement of all communication to a handful of platforms

- Snowden's disclosures

- Huge datacenters built by NSA to tap into telecom

- Crackdown on p2p sharing

- Push for The Cloud

- Closure of Lavabit and other independent email providers

- Rabid push for phone-based 2fa

- Ongoing merger and conglomeration of everything into a venture-fund-owned megacorporation invisible only for those who call these obvious practices "conspiracy theories" with religious zeal

- Failure of everything initially claimed to be decentrallized to live up the name, including blockchains, IPFS etc

These and many other similar issues combined don't quite make for an illusion that the govenments are willing to allow us to communicate freely via a greater number of tapping and datamining points than they could possibly manage.

Now burn this heretic!

Re: Tell HN: IPv6-only still pretty much unusable

#329

Post by cranekam » Wed, Dec 07, 2022, 9:12 PM UTC

tlb wrote: post #277

est31 wrote: post #157

Earlier quoted context omitted.

One of the ideas of ipv6 was to reduce routing tables, those tables that backbone providers have to keep in memory and look up for incoming traffic. With ipv4's fragmented allocation scheme, these routing tables are huge. With ipv6, even huge companies like amazon only have a couple of global allocations. A "ipv4 with more octets" scheme would have kept that fragmentation around. That being said, Amazon currently has…

I'm not up to date, but when I knew about this stuff the cost of memory for routing tables was only a tiny, tiny fraction of the cost of a network. Most of the cost is burying and maintaining fiber. So unless something big has changed, it seems like a terrible choice to twist the whole system into uselessness to try to save a small amount of RAM cost.

According to this Stack Exchange post from last year a full IPv4 routing table requires on the order of a few hundred MBs of RAM. This is indeed a tiny fraction of the cost of maintaining the global internet infrastructure.

https://networkengineering.stackexchange.com/questions/76562...

Re: Tell HN: IPv6-only still pretty much unusable

#330

Post by ju-st » Wed, Dec 07, 2022, 9:12 PM UTC

nikanj wrote: post #271

I can trivially remember a bunch of ipv4 addressed, and recognize them at a glance (”oh that’s just our office vpn endpoint, that traffic is benign”). No hope of that with ipv6.

If your office has a /48 you could surely memorize the subnet itself and/or give your important endpoint the shortest possible IP address like: 2001:db8:a0b0::1 or even something funny like 2001:db8:a0b0::0ff1:ce