Hey,

I also support killing ANY...

- I've heard that some people like to use ANY for cache snooping to see what's 
there
- Killing ANY from authoritative to recursive is much larger than killing ANY 
between client and resolver

BIND 9 now has MR in progress that does a mild step in the direction of killing 
ANY:

https://gitlab.isc.org/isc-projects/bind9/-/merge_requests/13103

- It just sends REFUSED to ANY queries with RD flag set
- The resolver (DNS server as a client) never sends outgoing ANY query now 
(that includes delv)
- The authoritative ANY answers are still provided[*]
- The recursive ANY answers with no-RD flags is still served[*]

There's one extra MR that implements the suggestion made by someone yesterday 
(Wes?):

https://gitlab.isc.org/isc-projects/bind9/-/merge_requests/13107

- ANY over UDP always just sets +TC and send empty response
- ANY over TCP is sent[*] (as we got reports that miscreants used ANY over TCP 
to cause harm)

* - minimal-any default is yes, so any ANY query just sends single RRset to the 
client

********

That said, if we can agree on killing ANY completely, I can whip up the MR that 
kills it in BIND 9 in next five minutes :-D.

Cheers,
Ondrej
--
Ondřej Surý (He/Him)
[email protected]

A gentle nudge is always appreciated if I take a little longer to reply.

> On 12. 11. 2026, at 12:13, Vladimír Čunát 
> <[email protected]> wrote:
> 
> I support killing QTYPE=ANY (at resolvers already).
> I don't really have a preference/survey comparing particular ways of doing 
> it, but NOTIMP seemed to have worked well enough as Knot Resolver default in 
> the part years.
> --Vladimir
> _______________________________________________
> DNSOP mailing list -- [email protected]
> To unsubscribe send an email to [email protected]

_______________________________________________
DNSOP mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to