Re: [IPv6] Second Working Group Last Call for <draft-ietf-6man-rfc6724-update>

[Ted Lemon <[e-mail]>]

But this is counterfactual. Rght now if you publish a ULA in the global
DNS, it will not cause a delay because no host with IPv6 connectivity will
try to connect to it. We only run into a problem if we decide to prefer all
ULAs over all IPv4 addresses. That /will/ in fact cause delays. Making the
known-local mitigation mandatory corrects that problem. So it gets you what
you actually want, whereas what you are currently advocating for makes
things worse for you, and is presumably behind your concerns about not
publishing ULAs in the DNS.

On Thu, Apr 11, 2024 at 8:22 AM Kyle Rose <[e-mail]> wrote:

> Globally publishing addresses that are unreachable means clients will try
> to connect to addresses that are known a priori to the address publisher to
> be unreachable from much of its audience. Anyone who cares about
> performance will not want to take the hit of clients waiting for a timeout,
> if the application even implements retries.
>
> Happy Eyeballs is a mitigation for configuration errors and transient
> connectivity problems: it's not part of the TCP handshake for a reason, and
> should not be required for a properly-configured client on a
> properly-configured and fully-working network to reach a
> properly-configured service.
> Kyle
>
> On Thu, Apr 11, 2024 at 8:02 AM Ted Lemon <[e-mail]> wrote:
>
>> Perhaps we could start by having you explain why it's a bad practice,
>> Kyle, since this does not appear to be universally accepted as true? :)
>>
>> On Thu, Apr 11, 2024 at 7:53 AM Kyle Rose <[e-mail]> wrote:
>>
>>> Also, MUST allows us to make ULA more useful than it is today. It is
>>>> *desirable* to be able to publish non-local ULAs and have hosts know what
>>>> is local and what is not. As a simple example: once all hosts implement the
>>>> MUST, it will be safe to publish local ULAs in the global DNS, because
>>>> hosts won't try to use them unless they are local.
>>>>
>>>
>>> Perhaps ironically, this is the best argument I've read in opposition to
>>> mandating the policy table be updated for known-local ULAs. Unreachable
>>> addresses should never be published in global DNS with the intent that
>>> clients figure it out, and the IETF should publish a standards track
>>> document explaining why this is a bad practice.
>>> Kyle
>>>
>>>