Re: [dtn] AD review of draft-ietf-dtn-bpsec-default-sc-02

[Mehmet Adalier <[e-mail]>]

Dear Martin,

I am not the author but please see a couple of answers below:

 

From: dtn <[e-mail]> on behalf of Martin Duke <[e-mail]>
Date: Wednesday, May 12, 2021 at 2:42 PM
To: <[e-mail]>
Cc: Zaheduzzaman Sarker <[e-mail]>
Subject: Re: [dtn] AD review of draft-ietf-dtn-bpsec-default-sc-02

 

Hello DTN,

 

As Zahed is still on medical leave, I have conducted a partial review of draft-ietf-dtn-bpsec-default-sc-06 to check that it fully complies with the outcome of his AD review. I apologize for any errors due to my insufficient context -- I have not read the entire draft yet.

 

(3.3.2) I appreciate the new Key Management section, but I don't want to lose part of Zahed's original review -- does the key wrapping [AES-KW] include enough metadata that the receiver can select the correct cipher and key to use without having to do trial decryption?

mA>>If one implements the AES-KW as indicated in the referenced NIST SP800, there should be no interop issues (i.e., no need to do a trial decryption). The AES-KW cipher and any “metadata” are pre-determined, but one can use either 128 or 256-bit keys.

 

(3.3.3) IMO the discussion about a registry for Integrity Scope Flags petered out with a clear resolution -- are the authors of the opinion that use of more of these flags is simply inconceivable? A registry would provide guidelines for how to assign more of these in future.

 

(4.1) In item (3), it states "The use of the Galois/Counter Mode produces cipher-text with the
       same size as the plain text". My understanding of GCM is that there is a 128 or 256 bit authentication tag appended to it that makes the whole of the cipher text larger than the plaintext.  Is this a semantic thing, am I wrong, or are you wrong? 

mA>>The Authentication Tag is not considered as part of the ciphertext. With BPsec the tag is stored separate from the ciphertext. On receiving a ciphertext +tag, the AES-GMC process includes calculating the Authentication Tag on the ciphertext and comparing it to the received Tag, which is separate from the decryption algorithm.

 

(6.1) I fear that you successfully scrubbed sc-02 for lower-case normative words, but then added new stuff here. Specifically:

 

"must be able to perform the following activities"

"This may include pre-sharing of key encryption keys"

 

Please verify that these should not be capitalized.

 

Regards,

Martin

 

 

 

_______________________________________________ dtn mailing list [e-mail] https://www.ietf.org/mailman/listinfo/dtn