Re: [DNSOP] Where in a CNAME chain is the QNAME?

[Ólafur Guðmundsson <[e-mail]>]

On Mon, Sep 26, 2016 at 8:33 AM, Peter van Dijk <[e-mail]
> wrote:

> Hello,
>
> On 23 Sep 2016, at 10:22, Stephane Bortzmeyer wrote:
>
> On Tue, Sep 20, 2016 at 06:13:50PM +0200,
>>  Stephane Bortzmeyer <[e-mail]> wrote
>>  a message of 68 lines which said:
>>
>> This issue was spotted by Peter van Dijk. It is about
>>> draft-ietf-dnsop-nxdomain-cut-05, recently approved by IESG. The
>>> problem is the definition of "QNAME" when there is a CNAME chain.
>>>
>>
>> OK, after reading the discussion, my opinion, as an author (but I'll
>> of course defer the decision to the working group, the WG chairs, the
>> RFC editor and the flying spaghetti monster):
>>
>> The re-definition of QNAME by RFC 2308 is awkward and does not match
>> the general usage, or the previous definitions. Therefore, I prefer to
>> keep the "common sense" usage "QNAME is the owner name of the record
>> in the Question Section". Which means that, in my example, the QNAME
>> is "www.afnic.fr" and the current text of
>> draft-ietf-dnsop-nxdomain-cut-05 is correct.
>>
>
> 2308 does not “redefine” QNAME. It clarifies that the usage in 1034 4.3.2
> is the definition we use in RFCs. 1035 4.1(.2) does not conflict with this;
> the QNAME there is just the initial QNAME.
>
> Many other RFCs need 2308: 2874, 4035, 4343, 4592, 4470, 4471, 5074, 5155,
> 6147, 6672 and most likely several others rely on the 2308 definition of
> QNAME. Without 2308, each of these RFCs would need extra text such as the
> text your draft has now. It is simply not necessary.
>
>
> Peter, is right
but the root cause is the role of "CNAME" and to lesser extent "DNAME".
Both is what I have part of DNS Navigation records that act like
"rewriters".
1034 was light on detail, 2308 might have uses a better notation but the
effect would have been the same,
The RCODE applies to the RRSET pointed to by the last CNAME in answer
section (or the missing one).

In hindsight the use of single RCODE with limited enumeration is the root
cause.
If one was designing this again one would allow multiple Questions in the
header and each RRset would have a header that says what its ReturnCode is,
but that is the advantage of hindsight.

Olafur