Skip to main content

Minutes IETF122: rats: Mon 06:00
minutes-122-rats-202503170600-00

Meeting Minutes Remote ATtestation ProcedureS (rats) WG
Date and time 2025-03-17 06:00
Title Minutes IETF122: rats: Mon 06:00
State Active
Other versions markdown
Last updated 2025-03-17

minutes-122-rats-202503170600-00

RATS Agenda, Monday Mar 17th - Session II (1 of 1 meetings)

Room: Boromphimarn 4 -
https://datatracker.ietf.org/meeting/122/floor-plan?room=boromphimarn-4

Notes Takers

  • Muhammad Usama Sardar (MUS)
  • Michael Richardson (MCR)
  • Mike Ounsworth (MO)
  • Thomas Fossati (TF)

Session Notes

13:00 - 15:00 ICT (6:00 - 8:00 UTC)

13:00 - 13:05 Agenda Bash & Logistics

(5 min) Kathleen Moriarty (KM), Ned Smith (NS)

A virtual interim meeting will be scheduled for the remaining topics
that didn't make it on the agenda this time. Look for announcements on
the RATS list.

13:05 - 13:15 Conceptual Message Wrappers (CMW) (in 2nd WGLC)

(10 min) Thomas Fossati (TF) - https://datatracker.ietf.org/doc/draft-ietf-rats-msg-wrap/ - Laurence Lundblade(LL): On [Issue#169](https://github.com/ietf-rats-wg/draft-ietf-rats-msg-wrap/issues/169): Drop "cmw" and define more semantically precise claims ("submods", etc.) in another document. - Hannes T.(HT): CMWs in X.509.  The problem with dumping the whole AR or Evidence (using `id-pe-cmw`) is privacy.  How can only a selection of claims be disclosed? Selective disclosure? - Laurence (LL): [Issue#154](https://github.com/ietf-rats-wg/draft-ietf-rats-msg-wrap/issues/154): Proposes to drop `id-pe-cmw`.  Do it in a separate I-D - MIKE O.(MO): isn't this what the LAMPS CSR attestation draft is about? NO. There are significant privacy concerns with putting any of the Attestation Results into a PKIX certificate. - MO: idP and other stuff... is in PKIX attest which is next. - Henk B. (HB): CWM is a conceptual message, and can contain all forms of concepual messages, so this OID should be defined somewhere, defining OIDs for evidence and Attestation Results (AR) is something different.
  • The RATS chairs will monitor the RATS list until the issues and
    discussion has subsided.

13:15 - 13:25 EAT Measured Component

(10 min) Thomas Fossati - https://datatracker.ietf.org/doc/draft-ietf-rats-eat-measured-component/
  • Carsten Bormann(CB): don't get into the habit of dumping some bytes
    where you can't figure out what they are. Use a CDDL socket instead.
  • Thomas F (TF): In this case, the EAT profile is the global type
    which defines what the bytes mean
  • Monty Wiseman (MW): has any looked at this and made sure it's
    compatible with the TCG's canonical event log? .... MW will review
    this... until the review is done, will not comment whether this is
    done. (ACTION MW)
  • Discussion on SVN to be done on the list (ACTION TF)

13:25 - 13:45 Concise Reference Integrity Manifest (WGLC)

(15 + 5 min) Yogesh Deshpande (YD) - https://datatracker.ietf.org/doc/draft-ietf-rats-corim/ - Requests for review and WGLC and directorate reviews - **ACTION**: reviews are to be done before WGLC: Monty (MW), MCR, ... (others are welcome). The time frame for completion is April 15, 2025.
  • ACTION: The RATS chairs will start a WGLC on the RATS mailing
    depending on what reviewer feedback reveals.

13:45 - 14:05 PKIX Key Attestation

(20 min) Mike Ounsworth - https://datatracker.ietf.org/doc/draft-ietf-rats-pkix-key-attestation/ - Scope is HSM-only - PKIX definition based on editors' opinion (not based on literature) - Usama (MUS) asks us to look into the literature for terms for "key attestation", 
- How is the key attestation different from workload attestation?
- we ask for what literature we should look into?
- Kathleen offers to help with the terminology alignment. - Yaron Sheffer(YS): the definition is clear, but unclear if it's useful.  The Presenter is often ... the TLS stack... then your security depends more on the stack than on the key.  Signing a PDF, then the operation of the signer is more important than where the key is.
- Is there value in defining key attestation without the rest of the stack? - Is attestation key the key being attested?
- the HSM exists to hold all the application secret keys.
- HSM comes with a pre-configured attestation key, which is used to sign the claims.
- Unlike other devices, the purpose of HSM is only to store the keys. - Jon (J?): Key Hierarchy  - MO: Yes, this would be in-scope for our data model. We welcome feedback.

14:05 - 14:20 Reference Interaction Models (WGLC)

(10 + 5 min) Henk Birkholz - https://datatracker.ietf.org/doc/draft-ietf-rats-reference-interaction-models/  - How many people have read this? in the room 4, + 2ish who made issues on the list but not in the room
  • Usama (MUS), Monty (MW), Yogesh (YD) offer to review. 2-3 weeks and
    WGLC afterwards.

  • ACTION: The RATS chairs will monitor reviewer feedback and start
    a WGLC on the RATS mailing list.

14:20 - 14:30 Event Stream Subscription (WGLC)

(5 + 5 min) Henk Birkholz - https://datatracker.ietf.org/doc/draft-ietf-rats-network-device-subscription/ - Ned (NS): Asked question if there is anyone who wants to review the draft or needs more time to review it. (Room consensus appeared to be NO)
  • ACTION: The RATS chairs will start a WGLC on the RATS mailing
    list.

14:30 - 14:40 Evidence Transformations (Adoption)

(5 + 5 min) Ned Smith (NS) - https://datatracker.ietf.org/doc/draft-smith-rats-evidence-trans/ - Provide evidence *semantics* to CoRIM verifiers. - Thomas (TF): This is a requirement that CoRIM creates for all evidence formats that want to be appraised by a CoRIM verifier.  There is going to be a long list of similar documents (per evidence format / CoRIM profile).
- Scope: what belongs (what evidence formats) belongs to the document?
- Adoption... in the more general, what do you want to do with future documents like this?
- Ned (NS): this is the set that industry cares about (outside of the IETF)... support that other work.
- If the DMTF is using CORIM, then it needs to define which profile it wants to use?
    - They define the measurement block, but not CORIM.
    - OCP has evidence formats that they are using.
- do we continue to do this Endorsement mapping in the endorsement drafts?
    - There may need profile specific mappings that extend the standard... then put the mapping in the profile for that extension.  This draft is dealing with what is already standard. - Usama (MUS): Is attested TLS not in scope?  - Monty (MW): Is the TPM Canonical Event Log (CEL) in scope? - Ned (NS): It could be included or a peer draft to this. Monty will review options - Henk (HB): We are preparing a CORIM profile for Linux IMA. Maybe that helps a bit? We have one for a - - TCG Evidence which is not DICE.
- Maybe they can share formats... we should coordinate with the Linux kernel maintainers that work on this.  Include IMA maintainer: Mimi Mimi Zohar also include David Safford who is also working it.
  • Adoption poll: 15 yes, 0 no. 9 no-opinion

  • ACTION: The chairs will start an adoption call on the RATS
    mailing list.

14:40 - 14:50 MUD-Based RATS Resources Discovery (Adoption)

(5 + 5 min) Henk Birkholz - https://datatracker.ietf.org/doc/draft-birkholz-rats-mud/
  • Thomas (TF) seems in favour.

  • Adoption poll: 11 yes, 0 no, 10 no opinion

  • Wait to adopt poll: 9 for wait = yes, 6 for now = no, 7 no opinion

  • ACTION: The chairs will contiue discussion on the RATS mailing
    list.

14:50 - 15:00 EAT Attestation Results (Adoption)

(5 + 5 min) Thomas Fossati - https://datatracker.ietf.org/doc/draft-fv-rats-ear/ - MCR: A bunch of your slides say "AR4SI" and some say "EAR". - TF: Yes, I have mixed two topics. - MCR: EAR is very important to NASR, we need to figure it out. - Henk (HB): AR4SI is an adopted draft, but it can't work without EAR - Kathleen (KM): the "posture" I-D would use it.
  • Adoption Poll: 12 yeses, 0 noes, 3 no-op

  • ACTION: The chairs will start an adoption call on the RATS
    mailing list.