Dime WG notes from IETF#83 ** WG status update: Glen zorn: PS for end2end security as WG no problem statement Glen Zorn: draft-ietf-dime-erp - date for WGLC? Jouni K: To be announced 1st April WGLC for draft-ietf-dime-erp Glen Zorn: draft-ietf-dmer-rfc4005bis - issue tracker updated Jouni K: Next step is proto writeup. ** draft-ietf-dime-pmip6-lr-10 : Netext to review. Glen Zorn: Timeline for review? Dan Romascanu: Ralph discuss netext review. Cross reference drafts between netext and dime required. Lionel M: Review related netext drafts, and can be referenced in this. Marco L: Normative reference? Dan R: No informative reference ** MIB and cc-appl-mip draft: Dan R: Waiting for editor, move them out of WG charter. Jouni: Announcement on the mailer to move it out of WG charter/std track Glen Zorn: Co-author promised to revise before taipai. Doesn't care about removal from charter. Dan: AD review 700 days unanswered cc-appl-mip draft. Benoit Claise: In X days delay for the documents to be removed from WG charter. Dan: Can be made dead and can be reactivated when required. Can be taken off from WG charter when the charter is revised. ** RFC3588bis update: Robert Sparks: Ok to close Discuss with security consideration updated. Glen: Security consideration to be updated. Robert Sparks: Encourages WG to make peer discovery mechanism to be made mandatory, decide on making it mandatory or otherwise. (Section 5.2 rfc3588bis - manual config MUST be supported, while peer discover is a MAY) Ben Campbell: Supports making Peer discovery to be made mandatory. In large scale deployment helps and make deployment easy. It should latest be a SHOULD. Lionel: Peer discovery - should be "SHOULD" not a "MUST", as MUST will break existing implementations? Dan: Implementation MUST support this Jouni: Inclines towards SHOULD Dan: Propose to add a paragraph clarifying this. Ben: Deployment should be given a choice to use peer discovery. Implementation MUST support peer discovery. Glen: understands the change - MAY to MUST for peer discovery. Mark Jones: its a bis, having it as a must breaks 3588 implementation Glen: Existing text not clear, deploying peer discovery is optional. Implementation MUST support peer discovery. Ben: 3588bis compliance should make this change. Dan: Agrees, change in this text is clarification to remove ambiguity. non-complaince to be documented due to disambiguating this. Frank Brockeners: This change to be taken back/announced to the list regarding this uncomplaince of existing implementation. Jouni/Glen: change the text to : Must implement ** Diameter Design guidelines: Presenter Lionel Morand. Lionel: Will submit a new revision on 1st april, request for help from WG to progress the document. Or conclude that this draft may not be needed and can be removed from WG charter. Jouni: Opinions? Frank: Its a useful document. Should we add additional information: Is diameter a transport protocol or a AAA protocol? Lionel: Formal guidelines to IETF and other SDOs to use and design diameter application. Frank: WGLC, no major changes/open questions Dan: scope of the protocol is important to document. Robert Sparks: Regarding - M bit AVP in a command to be replaced/removed guideline. Glen: M bit means it is to be understood not that its mandatory option in the command. Robert Sparks: Will reread, might have confused M bit with "Required" in command ABNF. ** Diamter NAT Control Application (DNCA): Presenter Frank Brockeners. #1 Atomicity of the controller for an endpoint Dan: Must consider multiple types of controllers. Proposed text seems reasonable. Frank: propose to send it to the WG list Robert Jones: ok with the text, does this need new error codes? #2 timer based nat binding cleanup upon unexpected termination Lionel: Ok with the text, but it is not specific to DNCA, Diameter base handles this commonly to all applications. Frank: Agree, but for NAT control this was requested to be made explicit. Lionel: Ok with this. jouni: list to review. Frank: till when? Jouni: few days. ** individual presentation: Diameter group signaling Mark jones presenter. Frank: Will this be redefining diameter base? Mark Jones: not an update to 3588, new application to allow group commands. Lionel: minimum disruption of application, a backward compatible solution. could be done using end to end capability feature documented in guidelines document. Application to handle grouping. Will look into this direction. Mark Jones: Could not find a solution that eliminated a need to define a new application to support group signalling for an existing application. Open to alternate that prevents defining new app, but not open to redefine Diameter base. Marco: supports the work Frank: can individual and group command be mixed in the current proposal? Mark jones: yes. Mark jones: looking for collaboration and contribution. Will start the thread on WG. Jouni: Will wait for discussion before adoption but may not wait till IETF84. Glen Zorn: Proposes to adopt the draft in WG in the current form as this work is of interest. Ben Campbel: clarification regarding the version of the document. Mark jones: this is the -01 version Mark Jones: propose to take input regarding group management solution and document the solution rather than have all the solutions considered and documented. Jouni: Lets go ahead with WG adoption call on the mailer list. Glen: opposes going to mailer list just announce it as a WG list. Jouni: prefers to consult WG list before adoption. Dan: Supports chair, WG should be consulted. Mark Jones; will wait for adoption and feed back. Jouni: to announce on mailer on 1st April. Marco: treat the draft as a template the problem space and do a WG adoption, solution/specification can be discussed/changed after adoption. Lionel: Agrees, but a complete change in the specification after adopting is not possible. Wait for a discussion. Glen Zorn: I am not sure what the quality requirements for WG draft are. Blaims Bernand Aboba for starting the editor / individual submission to change according to WG desire as a pre condition for adoption. This violates several different laws. Dan: Quotes note well, no copy right, only change control right. Address quality level for WG draft - consensus is all that is needed. Mailer list to be informed before adoption. ** Diameter e2s security: Presenter Jouni Glen Zorn: Are you carrying 2 copies of each AVP? Jouni: AVP and a hash of the AVP. Mark jones - hashing - shows payload not header. Jouni: header contains the algorithm that is used. The solution makes use of IETF Jose WG's JSON Web Signature and Encryption work. Glen Zorn: original Diameter security adapted email security to diameter and expired due to its complicity. Key management etc way too complicated and WG lost interest. Now this one doesn't include key management etc, it is follows web style security. Hopes this one progressed faster. Prefers method used in RADIUS to send encrypted attribute. RADIUS has the problem of fragmentation, Diameter doesn't. Jouni : would you write it up? Glen Zorn: sure Mark Jones: 2 byte structure not too bad, can u talk about benefits? (Mark commented he doesn not appreciate the advantages of JSON over the old fashioned Diameter grouped AVPs). Jouni: has no more to add that was stated earlier about reusing other ongoing IETF work. Mark jones: Reason to adopt this approach is to decouple key distribution with this? Jouni: Yes Mark Jones: was key distribution the reason earlier approach died? How does RADIUS do that? No answer to Mark's question was provided. ** Open discussion Mark Jones: 3GPP trusted wan access specification coming up should we define a draft to define the AVPs. Lionel: new draft to define AVPs and restrict to only listing definition. Mark Jones: Yes, can pull something together quickly. Suggestion to add those into existing 802.11 draft in RADEXT.