Re: [secdir] Security review of draft-ietf-dnsop-onion-tld-00.txt

[Mark Nottingham <[e-mail]>]

Applied in <https://github.com/mnot/I-D/commit/61535215c7b2fcf864faf08ada7b6117bf9a88d9>.

Thanks,


> On 1 Sep 2015, at 2:39 pm, Christian Huitema <[e-mail]> wrote:
> 
> On Monday, August 31, 2015 5:48 PM, Mark Nottingham wrote
>> 
>> On 1 Sep 2015, at 2:59 am, Christian Huitema <[e-mail]> wrote:
>>> 
>>> A legacy client may inadvertently attempt to resolve a ".onion" name
>> through
>>> the DNS. This causes a disclosure that the client is using TOR to reach
> a
>>> specific service. Malicious resolvers could be engineered to capture and
>>> record such leaks, which might have very adverse consequences for the
>>> well-being of the TOR user. This issue is mitigated if the client's TOR
>>> software is updated to not leak such queries, or if the client's DNS
>>> software is updated to drop any request to the ".onion" TLD.
>> 
>> Is that just a drop-in replacement for this?
>> 
>> """
>> If client software attempts to resolve a .onion name, it can leak the
> identity
>> of the service that the user is attempting to access to DNS resolvers,
>> authoritative DNS servers, and observers on the intervening network. This
> risk
>> is mitigated when the recommendations in {{onion}} are followed, but are
> still
>> present when using systems that are not updated.
>> """
> 
> You challenged me to produce some text, so I did... But I should have taken
> a closer look at your change. We are definitely expressing the same idea. My
> text is perhaps a bit more explicit, but it lacks the pointer to the
> specific section in which the recommendations are made. Your text maybe
> lacks the motivating part that fixing DNS software is part of "defense in
> depth."
> 
> -- Christian Huitema
> 
> 
> 

--
Mark Nottingham   https://www.mnot.net/