Re: [dnsext] DNSSEC, robustness, and several DS records

[Doug Barton <[e-mail]>]

On 5/12/2011 6:48 PM, Matt McCutchen wrote:
> On Thu, 2011-05-12 at 12:05 -0700, Doug Barton wrote:
>> [...] So if both work,
>> sure, disregard SHA-1. Otherwise, go with what works.
>
> While you have defended your position amply, in your description you
> continue to gloss over the issue.  The resolver does not know whether
> the SHA-256 DS is "working" (it has no way to distinguish a
> "non-working" SHA-256 DS from malicious replacement of the DNSKEY), so
> what you literally stated above is unimplementable.  What I believe you
> are really proposing, to honor the SHA-1 DS if the SHA-256 DS does not
> match /for any reason/, is only as secure as the second preimage
> resistance of SHA-1.
>
> I just wanted to make sure this is clear in everyone's mind.

Thanks for clarifying that. I was (obviously foolishly) thinking that 
there was a way to distinguish the 2 cases. In that case ignore 
everything I said, it should fail in Stephane's case.


Sorry for wasting everyone's time,

Doug

-- 

	Nothin' ever doesn't change, but nothin' changes much.
			-- OK Go

	Breadth of IT experience, and depth of knowledge in the DNS.
	Yours for the right price.  :)  http://SupersetSolutions.com/