Last Call Review of draft-ietf-lamps-hash-of-root-key-cert-extn-03
review-ietf-lamps-hash-of-root-key-cert-extn-03-secdir-lc-montville-2019-01-08-00

Request Review of draft-ietf-lamps-hash-of-root-key-cert-extn
Requested rev. no specific revision (document currently at 07)
Type Last Call Review
Team Security Area Directorate (secdir)
Deadline 2019-01-10
Requested 2018-12-27
Draft last updated 2019-01-08
Completed reviews Secdir Last Call review of -03 by Adam Montville (diff)
Genart Last Call review of -03 by Joel Halpern (diff)
Genart Telechat review of -05 by Joel Halpern (diff)
Secdir Telechat review of -05 by Adam Montville (diff)
Assignment Reviewer Adam Montville
State Completed
Review review-ietf-lamps-hash-of-root-key-cert-extn-03-secdir-lc-montville-2019-01-08
Reviewed rev. 03 (document currently at 07)
Review result Ready
Review completed: 2019-01-08

Review
review-ietf-lamps-hash-of-root-key-cert-extn-03-secdir-lc-montville-2019-01-08

This draft is ready. It's a clever (though not foolproof) way to prime the pump for root certificate updates. I'm not an ASN.1 expert, so I can't really opine on the structure in Section 3, but from what I can tell it looks sane. Operational considerations seems sane. Security considerations rely on those from RFC5280, and additionally addresses: 1) analysis before the next-generation root certificate is released, 2) key strength considerations (equal or greater than current), 3) unforeseen cryptoanalytic advances, 4) typical hash pre-image attacks, and 5) early release of the next-generation public key.

One area in the security considerations that could be enhanced is the recommended action to take in the case of an early next-generation public key release. The language in the draft states: "The second transition occurs when the Root CA is confident that the population of relying parties have completed the first transition, and it takes the Root CA to the freshly generated key pair." The question that came to mind was: What might bring about such confidence? I'm not sure that it's possible to generalize an answer to that question, however.

Kind regards,

Adam