Re: [dns-privacy] Ben Campbell's Yes on draft-ietf-dprive-dns-over-tls-07: (with COMMENT)

[Sara Dickinson <[e-mail]>]

> On 15 Mar 2016, at 14:23, Stephane Bortzmeyer <[e-mail]> wrote:
> 
>> Do (or should) the profiles have anything to say about clear-text
>> fallback if a client cannot connect to the server's DNS-over-TLS
>> port, or the TLS handshake fails? I infer that such fallback should
>> not occur with the pinned profile, but what about the opportunistic
>> profile?
> 
> I don't want to reopen discussions on
> draft-ietf-dprive-dns-over-tls-07 but may be we can improve it for
> draft-ietf-dprive-dtls-and-tls-profiles-00: its definition in section
> 5 does not address this issue. May be have two opportunistic
> profiles, one with fallback to clear text if necessary and another
> without?

Hi Stephane/Ben, 

Note that the discussion in draft-ietf-dprive-dtls-and-tls-profiles-00 section 4.2 quotes RFC7435 directly which states that Opportunistic Security is described as

“... the use of cleartext as the baseline communication
         security policy, with encryption and authentication negotiated
         and applied to the communication when available.”

So at the moment that draft describes 2 profiles, a Strict one which requires authentication (or failure), or an Opportunistic one as described above. The use case for an ‘in-between’ profile that requires encryption but not authentication is debatable since it only provides limited protection from attacks.


Sara.