Re: Stephen Farrell's Discuss on draft-ietf-6man-stable-privacy-addresses-16: (with DISCUSS and COMMENT)

[Fernando Gont <[e-mail]>]

On 01/23/2014 12:09 PM, Eliot Lear wrote:
> 
> On 1/21/14, 8:08 PM, Fernando Gont wrote:
>>
>>> If keeping it, I'd say give the example and then add a
>>> security consideration that that interface might be
>>> vulnerable (e.g. 'cat /proc/net/eth0/rfcxxx-secret'
>> How about rather noting that the secret key should only be accessible by
>> the system administrator? (i.e., non-RFC2119 recommend that implementers
>> do the right thing :-) )
> 
> I agree with the requirement but I think Stephen raises an important
> point, which is that it should be highlighted that the information is
> sensitive.  As such, implementations should constrain access to the
> information, to the extent practicable.  Furthermore, I understood
> Stephen's point also to be that the private key information should not
> be used for any other purpose.

It's not actually a private key in the sense of private/public key..
Just a secret key.

But yeah, I have no objections to adding some words on the topic. How
about this:

"Since the privacy of this system relies on the secrecy of the
secret_key parameter, implementations should constraint access to it to
the extent practicable (e.g., require superuser privileges to access
it). Furthermore, in order to prevent leakages of the secret_key, this
parameter should not be used for any other purposes than being a
parameter to the scheme specified in this document"

?

Thoughts?

Thanks!

Best regards,
-- 
Fernando Gont
SI6 Networks
e-mail: [e-mail]
PGP Fingerprint: 6666 31C6 D484 63B2 8FB1 E3C4 AE25 0D55 1D4E 7492