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

Tell HN: Hacker News now supports IPv6

news.ycombinator.com

Post reply on HN
Print view View original

381–390 of 396 posts

Re: Tell HN: Hacker News now supports IPv6

#381

Post by Dagger2 »

commandersaki wrote: post #380
Dagger2 wrote: post #379

Earlier quoted context omitted.

Can you tell us what the better transition story was? As far as I can tell, v6 is already approximately as interoperable with v4 as it's possible to be. I don't get your claim that it's a separate island at all. It's not. This machine is v6-only and isn't on an island, it has full access to the entire Internet.

TUBA had a sane transition story [1] although I do not know of the mailing lists these were discussed (probably done in in-person IETF meetings). For IPv6 transition mechanism you can look into the ngtrans mailing list regarding AutoIPv6 which is basically an extension of 6to4 but instead of connecting IPv6 networks over IPv4 it would extend to IPv4 hosts. There's also an archive link in my post history which you can…

Hm... according to that draft, TUBA's transition story is dual stack, which is the same as the main transition approach for v6. How come it's sane for TUBA but not sane when v6 does it?

> As for current v6 being interoperable with v4, that is patently false.

This machine is v6-only and can reach the entire Internet, including v4-only hosts. It clearly has more interoperability than you're claiming or this wouldn't be possible.

Re: Tell HN: Hacker News now supports IPv6

#382

Post by commandersaki »

Dagger2 wrote: post #381
commandersaki wrote: post #380

Earlier quoted context omitted.

TUBA had a sane transition story [1] although I do not know of the mailing lists these were discussed (probably done in in-person IETF meetings). For IPv6 transition mechanism you can look into the ngtrans mailing list regarding AutoIPv6 which is basically an extension of 6to4 but instead of connecting IPv6 networks over IPv4 it would extend to IPv4 hosts. There's also an archive link in my post history which you can…

Hm... according to that draft, TUBA's transition story is dual stack, which is the same as the main transition approach for v6. How come it's sane for TUBA but not sane when v6 does it? > As for current v6 being interoperable with v4, that is patently false. This machine is v6-only and can reach the entire Internet, including v4-only hosts. It clearly has more interoperability than you're claiming or this wouldn't be…

Interoperability would be the ability to use either IPv4 or IPv6 and talk to each other seamlessly. This is clearly not the case for IPv4 clients talking to IPv6 servers. This is clearly not the case for IPv6 servers accepting IPv4 connections.

As for TUBA, the transition plan incorporated tunneling, with a bit more thought put into this we could've had something similar to AutoIPv6. That document was just a starting point and outlined some important aspects of the criteria such as low administrative overhead for transitioning IPv4 networks to IPng, something which IPv6 falls completely flat on.

Re: Tell HN: Hacker News now supports IPv6

#383

Post by orangeboats »

commandersaki wrote: post #380
Dagger2 wrote: post #379

Earlier quoted context omitted.

Can you tell us what the better transition story was? As far as I can tell, v6 is already approximately as interoperable with v4 as it's possible to be. I don't get your claim that it's a separate island at all. It's not. This machine is v6-only and isn't on an island, it has full access to the entire Internet.

TUBA had a sane transition story [1] although I do not know of the mailing lists these were discussed (probably done in in-person IETF meetings). For IPv6 transition mechanism you can look into the ngtrans mailing list regarding AutoIPv6 which is basically an extension of 6to4 but instead of connecting IPv6 networks over IPv4 it would extend to IPv4 hosts. There's also an archive link in my post history which you can…

I think you are having a very romanticized view of 6to4 and its derivatives.

By the late 2000s it was clear as day that 6to4 wasn't working out (the design rationales of contemporary IPv4IPv6 transition technologies will tell you why [0]). By extension, AutoIPv6 which was building on 6to4 was also unlikely to work out. Even worse, AutoIPv6 relies at least partially on anycast 6to4 which was later deprecated due to operational problems [1].

The only surviving 6to4 derivative is 6rd and even that is mostly phased out.

>As for current v6 being interoperable with v4, that is patently false.

You need to define your scope. Are you talking about programming? Or ability for IPv6 clients to talk to IPv4 servers? Or...?

IPv6 is interoperable with IPv4 from a programmer's perspective, once you upgrade your sockets from accepting sockaddr_in to accepting sockaddr_in6, your program automatically receives both IPv4 and IPv6 packets with IPv4 addresses represented as ::ffff:x.y.z.w.

And from a client's perspective, IPv6 clients absolutely can connect to IPv4 servers through a border relay of some sort, typically NAT64. But you can do SIIT as well if you are feeling fancy. Note that this is not much different from 6to4 and its derivatives.

As for the perspective of a server, indeed this is unsolved but again AutoIPv6 doesn't solve the issue as well since the server still needs a public IPv4 address.

[0]: https://en.wikipedia.org/wiki/IPv6_rapid_deployment#Comparis...

[1]: https://datatracker.ietf.org/doc/html/rfc6343#section-3

Re: Tell HN: Hacker News now supports IPv6

#384

Post by ale42 »

bqmjjx0kac wrote: post #269
ale42 wrote: post #160

Earlier quoted context omitted.

Sounds like a technical problem. I don't see any reason they should de-prioritize IPv6 traffic...

Agreed, I don't think it's intentional. It feels like at peak times, some component of their IPv6 infra gets maxed out.

If they have separate infra for IPv6 (but why?) and it's undersized compared to IPv4, it would totally make sense to observe that.

Re: Tell HN: Hacker News now supports IPv6

#385

Post by rini17 »

iknowstuff wrote: post #372
rini17 wrote: post #361

Earlier quoted context omitted.

..until another Alice joins the network with her iphone. There are plenty of valid reasons to manage names centrally and configure addresses explicitly, but no I'm supposedly a dinosaur that is supposed to get extinct finally :(

It will rename itself to be alices-iphone-2.local

which one?

Re: Tell HN: Hacker News now supports IPv6

#386

Post by Dagger2 »

commandersaki wrote: post #382
Dagger2 wrote: post #381

Earlier quoted context omitted.

Hm... according to that draft, TUBA's transition story is dual stack, which is the same as the main transition approach for v6. How come it's sane for TUBA but not sane when v6 does it? > As for current v6 being interoperable with v4, that is patently false. This machine is v6-only and can reach the entire Internet, including v4-only hosts. It clearly has more interoperability than you're claiming or this wouldn't be…

Interoperability would be the ability to use either IPv4 or IPv6 and talk to each other seamlessly. This is clearly not the case for IPv4 clients talking to IPv6 servers. This is clearly not the case for IPv6 servers accepting IPv4 connections. As for TUBA, the transition plan incorporated tunneling, with a bit more thought put into this we could've had something similar to AutoIPv6. That document was just a starting…

I've talked from v4 clients to v6 servers before, and I've accepted v4 connections to v6 servers. How can it be so clear that it's not possible when I've done it before?

> As for TUBA, the transition plan incorporated tunneling

The transition plan for v6 also incorporates tunneling (6in4 is the same thing as the EON tunneling described in the draft you linked), so again I ask why TUBA counts as having a sane transition plan when doing the same things in v6 doesn't.

There's not much administrative overhead involving in transitioning a network to v6. For most people it involves doing exactly no extra work beyond what they already do to deploy v4. Of course, you do need to reconfigure anything that's been configured such that it can only work with v4 (and this might be quite a lot of stuff), but that was always going to be necessary no matter what you're transitioning to.

Re: Tell HN: Hacker News now supports IPv6

#387

Post by Dagger2 »

greyface- wrote: post #378
Dagger2 wrote: post #377

Earlier quoted context omitted.

And for $256,000/year you can... exhaust the entirety of 2000::/3? Yeah, that's not okay. Either the price needs to go up a lot or they need to include factors other than money in the allocation criteria.

> they need to include factors other than money in the allocation criteria They already do, and simply being able to pay the fee does nothing to qualify you for an allocation: https://www.arin.net/participate/policy/nrpm/#6-ipv6 At least, in theory. I'm not going to attempt to defend the Capital One allocation.

They might have included them in the published policy, but it seems they aren't in whatever policy they're actually following.

Re: Tell HN: Hacker News now supports IPv6

#388

Post by greyface- »

Dagger2 wrote: post #387
greyface- wrote: post #378

Earlier quoted context omitted.

> they need to include factors other than money in the allocation criteria They already do, and simply being able to pay the fee does nothing to qualify you for an allocation: https://www.arin.net/participate/policy/nrpm/#6-ipv6 At least, in theory. I'm not going to attempt to defend the Capital One allocation.

They might have included them in the published policy, but it seems they aren't in whatever policy they're actually following.

It seems more likely to me that ARIN is following the text of published policy, but Capital One found a way to interpret it that violates its spirit. My guess is that they're treating each customer debit/credit card as a serving site that qualifies for a /48. To address this, ARIN should fix the policy, not raise fees.

Re: Tell HN: Hacker News now supports IPv6

#389

Post by mike_d »

cesarb wrote: post #221
mike_d wrote: post #134

Earlier quoted context omitted.

Not at all. We could have taken one of the unassigned /8's at the time and allocated it to v6 transitional addressing (the failure to address 4->6 reachability is IMO why v6 failed). For the sake of this example lets use 53.0.0.0/8. All new addresses start with 00110101 followed by the first three bytes of the new v6 prefix. The prefix acts as a flag that indicates it is a new address and routers read an additional 5…

> The prefix acts as a flag that indicates it is a new address and routers read an additional 5 bytes from the beginning of what would be the data section of a traditional v4 packet to get the complete address. Your proposal, as usual for these kinds of proposals, fails to consider how would an "old world" endpoint talk to a "new world" endpoint. An "old world" endpoint wouldn't know about these "additional 5 bytes",…

I address that. Old hosts and networks only know about the first part of the address and the rest of the address resides in the data section so a new application can run on an old host even. When you cross from an old to a new network, the router can hand packets to a side box that runs CGNAT. This allows for new hosts and networks to talk to old without having to include their full address in the data payload.

Think of it like the mail service. For the majority of the journey your letter is only routed based on the first 3 digits of the zip code. When it reaches the last mile sorting facilities is when the full address becomes relevant.

Re: Tell HN: Hacker News now supports IPv6

#390

Post by mike_d »

Dagger2 wrote: post #319
mike_d wrote: post #134

Earlier quoted context omitted.

Not at all. We could have taken one of the unassigned /8's at the time and allocated it to v6 transitional addressing (the failure to address 4->6 reachability is IMO why v6 failed). For the sake of this example lets use 53.0.0.0/8. All new addresses start with 00110101 followed by the first three bytes of the new v6 prefix. The prefix acts as a flag that indicates it is a new address and routers read an additional 5…

Your scheme already exists, although instead of using the first octet of the IP address it uses a 41 in the protocol field, so it doesn't need to eat an entire /8. Routers can see the 41 as a flag and then read the v6 address from the beginning of the packet payload. You haven't managed to suggest anything that v6 doesn't already do, or anything that would solve any of the problems v6 has when doing it. I'm not sure…

Throw that packet at a non-v6 router and let me know what happens. v6 could have been an overlay extension to solve the address space issue without the migration complexities.

We are now 26 years in and v6 has only gotten serious adoption because mobile carriers came along too late in the game to get enough number resources.

Post reply on HN