Re: Link-local IPv6 addresses in the DNS

[Brian E Carpenter <[e-mail]>]

Kerry,

> In the case of link-
> local adresses stored in AAAA records, the zone index would seem to
> indicate the corresponding interface with respect to the *server* (assuming
> a multi-homed server here).

Which is completely and utterly irrelevant to any host except
that one particular physical DNS server. So I don't see what
possible use it has for anyone.

Regards
   Brian

On 2011-11-23 10:56, Kerry Lynn wrote:
> On Tue, Nov 22, 2011 at 3:52 PM, Brian E Carpenter <
> [e-mail]> wrote:
> 
>> On 2011-11-23 05:34, Philip Homburg wrote:
>>> In your letter dated Tue, 22 Nov 2011 14:30:03 +1100 you wrote:
>>>> On a related issue to link locals in URI's, we don't currently have
>>>> a good method of supporting link locals in the DNS.  Sure we can
>>>> add them as AAAA records but they are essentially useless as the
>>>> scope information is lost.  People keep saying use LL for disconnected
>>>> but it just doesn't work without more support.
>> Other people keep saying "use ULA for disconnected". The fact
>> that you can put ULA into (er, local) DNS without any fancy
>> stuff is a distinct advantage.
>>
>> IMHO link-local should be used only for bootstrapping a host and
>> for diagnostic purposes. I guess I could statically configure a
>> printer on fe00::a%1 if I really had no choice.
>>
>> This is a distinctly different problem than the one that kicked off the
> link-
> local discussion.  In the web browser case, you know the link-local
> destination address of the server a priori but it only has validity with
> respect to a particular link, and there's no way to indicate the zone
> index to the browser (assuming multi-homed client here).  The once-
> existing capability was removed because there is no RFC support for it.
> 
> The DNS case seems like a server-side issue.  In the case of link-
> local adresses stored in AAAA records, the zone index would seem to
> indicate the corresponding interface with respect to the *server* (assuming
> a multi-homed server here).  What's more, the DNS server would need to
> keep track of the interface on which the query arrived and only respond
> with a link-local address if the client and server are on the same
> interface.
> 
> Some additional observations:
> - If you want to stay with link-local addresses then perhaps multicast
>   DNS is the best solution for you.
> - If you want to stay with unicast DNS, then ULAs have the advantage
>   of being routable and the problem goes away.
> - If you want to stay with link-local addresses *and* DNS, then perhaps
>   you need to engage dnsext WG to discuss the server-side issues.
> 
> -K-
> 
> 
>   Brian
>>> For disconnected operation, why not have getaddrinfo fill in the scope?
>>> Just set it to the interface over which the DNS reply arrived.
>>>
>>> I have to admit that this may become a bit tricky if the DNS resolver is
>> local
>>> or if interface information is lost in some other way.
>>>
>>>
>