Re: [DNSOP] I-D Action: draft-ietf-dnsop-glue-is-not-optional-02.txt

[Geoff Huston <[e-mail]>]

> On 29 Jul 2021, at 10:33 am, Mark Delany <[e-mail]> wrote:
> 
> On 29Jul21, Geoff Huston allegedly wrote:
> 
>> For me it appears to depend on the actions of the resolver as to whether this would be faster
>> or not. If all resolvers blindly re-query using TCP for all UDP responses where TC=1 is seen in
> 
> I'm not sure I follow this bit. Are you merely implying that the resolver should first
> consider a larger edns0 bufsize before resorting to TCP?

Seems that the DNS Flag Day 2020 precluded that option, so I don’t think its available.

> 
> Or are you suggesting that responses with TC=1 should include as much of the answers as
> will fit and then the resolver can decide whether enough is enough?

Well no. Because RFC2181 says "When a DNS client receives a reply with TC set, it should
ignore that response, and query again, using a mechanism, such as a TCP connection, that
will permit larger replies.”, which, unless there is some later RFC that quietly countermanded
this directive then I don't think the resolver has a choice here.

(digression into the interpretation of the term “should” in a document that pre-dates normative 
language. Is that “should” used then in the way we use MUST today? Or is that a SHOULD? I interpreted
it as a MUST but as I don't write resolver code for a living or as a hobby (and you should all be thankful
for that!) I'm definitely not an authoritative sources as to how resolvers SHOULD or MUST interpret
TC=1 responses.


Geoff