MoiN!

On 11 Nov 2026, at 10:25, Andrew Sullivan wrote:
> I suppose I don't really care one way or the other, so if people really want 
> this I wouldn't object.  At the same time, while I recognize many, many 
> places where ANY is doubtless a terroble thing to support, one of the places 
> that it can be significantly useful (at least in my experience) is in 
> debugging the behaviour of iterative-mode resolvers, recursive servers, &c.

As shown in the slides I also did that ANY based debugging and therefore know 
that it can be useful. However if you run any DNS infrastructure and look at 
the query patterns you really don’t find these uses anymore. You either get 0 
queries with qtype ANY or it is so much that you need to do something about it. 
And the amount of work, in code and operation, spend defending against these 
attacks while still allowing in practice 0 users to do some debugging is 
something I don’t think can be justified anymore, especially as doing so 
introduced problems. Also with some authorities already implementing RFC8482 
the usefulness for debugging with ANY has decreased.

So I think it is better to stop this now.

So long
-Ralf
——-
Ralf Weber
_______________________________________________
DNSOP mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to