Re: [secdir] sectdir review of draft-ietf-isis-node-admin-tag-08

[Catherine Meadows <[e-mail]>]

Hannes:

My apologies, I could have been a little more clear.

The security considerations section begins with a security consideration:

 
Node administrative tags may be used by operators to indicate
   geographical location or other sensitive information.  The
   information carried in node administrative tags could be leaked to an
   IGP snooper.

This doesn’t come with a citation of any other document, so I assumed it was “new”, and was confused when the
next sentence said the document does not introduce any new security concerns.   If this security consideration also applies to 
RFC5130 and RFC5305, you could cite those (although they do not address this particular issue in their security
concerns section it would still show that you are not introducing any new security concerns).


Cathy 




Catherine Meadows
Naval Research Laboratory
Code 5543
4555 Overlook Ave., S.W.
Washington DC, 20375
phone: [phone]
fax: [phone]
email: [e-mail] <mailto:[e-mail]>
> On Apr 22, 2016, at 3:34 AM, Hannes Gredler <[e-mail]> wrote:
> 
> catherine,
> 
> i fail to see how this does introduce *new* security concerns.
> 
> background:
> 
> we have in IS-IS two existing tagging mechanism's:
> 
> 1) admin-groups for links
>   https://tools.ietf.org/html/rfc5305#section-3.1
> 
> 2) prefix-tags for prefixes
>   https://tools.ietf.org/html/rfc5130#section-3.1
> 
> both "tagging-spaces" are local administered and have
> no global significance.
> 
> the purpose of draft-ietf-isis-node-admin-tag-08 is to
> add support for a third tagging method which allows tagging
> of an entire node using local administered tagging space.
> (same operational spirit as in 1) and 2) )
> 
> now my question(s):
> 
> - why is adding of an local-administered tag with no implied
>  semantics considered to be security relevant ?
> 
> - what specific (apparently mis-)wording in
>  draft-ietf-isis-node-admin-tag makes you think
>  it is security relevant ?
> 
> /hannes
> 
> On 4/21/16 19:36, Catherine Meadows wrote:
>> I have reviewed this document as part of the security directorate's
>> ongoing effort to review all
>> IETF documents being processed by the IESG.  These comments were written
>> primarily for the
>> benefit of the security area directors.  Document editors and WG chairs
>> should treat these comments
>> just like any other last call comments.
>> 
>> This draft describes an extension to the IS-IS routing protocol, that
>> allows tagging and grouping of nodes
>> in an IS-IS domain.  This makes it possible to increase the efficiency
>> of route and path selection, since the tags
>> give information about a router’s capabilities.
>> 
>> The Security Considerations section correctly identifies one of the main
>> security risks of using such tags:  they may leak
>> sensitive information about, e.g., geographical location.   However, I’m
>> confused by the statement following that:
>> 
>> “This document does not introduce any new security concerns.  Security
>> concerns for IS-IS are already addressed in
>> [ISO10589], [RFC5304], and [RFC5310] are are applicable to the
>> mechanisms described in this document.”
>> 
>> As far as I can tell, this document *does* introduce new security
>> concerns, because the tags may reveal sensitive
>> information that may not have been made available otherwise.  Moreover,
>> RFCs 5304 and 5310 concern authentication, not
>> secrecy, and so do not address information leakage at all.  My own
>> suggestion
>> for a recommendation would be that implementors should weigh the
>> benefits of putting certain kinds of information on tags
>> versus the risk of its being used by an attacker, and make their
>> decisions accordingly.  This would not be a SHOULD a MUST
>> recommendation by the way, but simply advisory.
>> 
>> I’m not sure what is meant by the last sentence in this paragraph:
>> 
>>  Extended authentication mechanisms described in [RFC5304] or [RFC5310]
>> SHOULD be used in
>>    deployments where attackers have access to the physical networks and
>>    nodes included in the IS-IS domain are vulnerable.
>> 
>> Is this addressing the problem of sensitive information on tags?  If so,
>> you need to say how.   If it is addressing
>> spoofing of tags, it should be given its own paragraph, and the threat
>> you are talking about should be made clear.
>> 
>> In the last paragraph, on the misattribution of tags from different
>> domains, what would you recommend for mitigating against
>> this problem?  Also, since this is in the security considerations
>> section, you should say something about how an attacker
>> could take advantage of it.
>> 
>> In my opinion, the Security Considerations section needs a major
>> revision.  However,
>> I consider this document Almost Ready, because the purpose of the
>> revision would be mainly to make the section more clear, not to address
>> any overlooked security problems.
>> 
>> 
>> 
>> 
>> Catherine Meadows
>> Naval Research Laboratory
>> Code 5543
>> 4555 Overlook Ave., S.W.
>> Washington DC, 20375
>> phone: [phone]
>> fax: [phone]
>> email: [e-mail]
>> <mailto:[e-mail]>
>>