Minutes captured by "Bertz, Lyle T [CTO]" ** DMM Working Group Agenda ** Date: Wed, July 19th, 2017 Time: 9:30 AM to 12:00 PM, CEST Location: IETF99 - Prague, Czech Republic Meeting Chairs: Dapeng Liu (Alibaba) & Sri Gundavelli (Cisco) 9:30 AM: Title: Administrivia & Intro, WG organization & milestones Time: 15 minutes Description: Agenda, Note-taker negotiation and WG Progress Update Presenters: Chairs 9:45 AM: Title: Protocol for Forwarding Policy Configuration (FPC) in DMM Time: 25 minutes Presenter: Lyle Bertz Description: Document Status and Changes Draft: https://tools.ietf.org/html/draft-ietf-dmm-fpc-cpdp-07 - have been working weekly since Chicago - focus on topology and policy - implementation from Verizon (demo'd at MWC) - ONOS and OpenDayLight. - this is an information model (abstract) - enhanced type model - DPN type - topology is only for DPN selection Suresh (AD): how soon before deadline? Lyle: Aiming for end of Sept. Sri: when are reviewers assigned? Suresh: when you tell me its ready from YANG perspective. These specs are obscure for those outside the area. Let's avoid last-minute surprise. Lyle: understaood. Dapeng: volunteers within this WG? (no one volunteered) Suresh: without review, this document doesn't leave WG. Wants to see non-author read. Lyle: invite anyone to pop into the weekly calls. SSuresh: we need people within the group to review. Wants to see more discussion on list. Lyle: we have write-up internally, will see if impact on mobility. Determines how much before 10:10 AM: Title: MN Identifier Types for RFC 4283 MN Id Option Time: 15 minutes Presenter: Charlie Perkins (and Suresh Krishnan on IESG feedback) Description: IESG DISCUSS Status Draft: https://www.ietf.org/id/draft-ietf-dmm-4283mnids-04.txt - Lots of discussion around types and issues on mailing lists wrt the draft - Security issue - Believe we have possible resolutions for issues that were raised - Presented many parts of the mailing list discussion in the presentation - RFID Types explained further - Noted adding more types for LPWAN - Reviewed commentary from mailing list - Privacy concerns - MNIDs by their nature are privacy issues - recommend encrypting all MNIDs - Why so many MNIDs? - people ask for them - this proposal is a registry - MUST encrypt (proper security measures) can help with the - Low energy on mailing list on what is needed - push registry to drive the expert review - ensure IANA considerations in the registry are strong - confirmed - only really for newer MNID types - plan is to change IANA policy, get approval, designate the experts and move on - Next steps - straw proposal - keep id types and make further considerations and re-submit for last call - No hard delivery date for the next steps 10:25 AM: Title: Distributed Mobility Anchoring Time: 15 minutes Presenter: H Anthony Chan Description: Update on the changes related to last call comments Draft: https://tools.ietf.org/html/draft-ietf-dmm-distributed-mobility-anchoring- 06 - 4 reviews resulting in 3 version updates; 1 review pending - Reviewed each change - In 06 - deleted slicing from the document - deleted security management - Removed description of the forwarding table - Noted the draft does not propose a solution but describes the mobility anchor and parameters used in communication/signaling - Described in detail the many different examples by which DMA can be achieved - Clarified scope of document and organized it so that you don¹t have to read everything - Chair wants to ensure that all 4 prior reviewers approve the changes in the latest spec version 10:40 AM: Title: On Demand Mobility Management Socket Extensions Time: 15 minutes Presenter: Danny Moses Description: Update on changes to the draft since IETF 98 Draft: https://www.ietf.org/id/draft-ietf-dmm-ondemand-mobility-11.txt - reviewed l(blocking issue) - noted request from chairs to address socket blocking (setsockopt) raised in IETF 98 by Erik Kline - 3 alternatives proposed and put on list - Noted that there was no mailing list response - Dave Dolson made a comment (off list) but did not select one of the 3 options nor discussed it on the mailing list - Selected 2nd alternative - Chair asked if issue originator was satisfied by the solution? - What is the guarantee of the SetSc return? session has been successfully allocated and assigned to the client - v6 Only noted several times - Requests for accepting 2nd alternative - As long as the function is documented that it blocks - For POSIX, a non-blocking version or call back version is async probably should be provided - AD - Do we need another abstract function? - EK - depends on OS - In general, bind is hard to use cause the AF must be known - More than 1 V6 address in the return? no - SetSc uses return address - Another approach raised - set preferences then do something else - Michael - the text is not clear that the APIs are abstract; needs to make clearer - does not mind but concerned by issues raised in other IETF 99 meetings - Suresh - although we don¹t do language bindings; goal is to stay abstract - Lyle - Suggested to change any reference of Œcode¹ to pseudo code to make it more apparent it is not a language binding. This seems to address the concern - Suresh - this work and privacy access are orthogonal; not sure how the flags are considered or merging (RFC 5014 and this work) - Need to check between specs and make some new considerations - Support of future on demand types - add new continuity type similar to 3GPP SSC mode 3 - time limited session continuity - valid and preferred lifetime - Suresh - already signaled in the framework - What is the graceful replacement use case (will ask on mailing list) 10:55 AM: Title: DMM Deployment Models and Architectural Considerations Time: 5 minutes Presenter: Seil Jeon Description: Update on the draft status Draft: https://www.ietf.org/id/draft-ietf-dmm-deployment-models-01.txt - quick update - Chair suggest to authors to fined reviewers (offline) for the document 11:00 AM: Title: SRv6 for Mobile User-PlaneTime: 15 minutes Presenter: Satoru Matsushima Description: Applicability of SRv6 (Segment Routing IPv6) to user-plane of mobile networks Draft: https://tools.ietf.org/html/draft-matsushima-spring-dmm-srv6-mobile-uplane -00 - Context; a new proposal - Showed current example of mobile network and tunneling solution - Showed SRv6 impact - Gave Srv6 in a nutshell review - Showed examples of how this solution supports - related to FPC yang as a possible way to implement the solution Questions - Xingpeng- Does this work use the prefix types for on demand mobility? - Don¹t think this is binding to the MN address assignment. - Marco - May add more items to the paths that are not representative of roles - How is QoS handled? It is mentioned in the draft - Dapeng - Why this vs. any other solution? Gets rid of tunnels - Very good work item but not in the charter. Keep progressing and will discuss with AD. - Author hopes to update 11:15AM: Title: Network-based and Client-based DMM solutions using Mobile IP mechanisms Time: 15 minutes Presenter: Carlos Jesús Bernardos Cano Description: Use of MIP protocol in DMM architecture Draft (s): https://www.ietf.org/id/draft-bernardos-dmm-cmip-07.txt https://www.ietf.org/id/draft-bernardos-dmm-pmip-08.txt https://www.ietf.org/id/draft-bernardos-dmm-distributed-anchoring-09.txt - reviewed the various documents - noted prior demos @ IETF - noted open source code availablity - ask WG if this work should continiue and will follow this up on the maling list 11:30 AM: Title: Anchor-less Mobility Management Time: 10 minutes Presenter: Xinpeng(Jackie) Wei Description: Draft: https://tools.ietf.org/id/draft-wei-dmm-anchorless-mm-01.txt - reviewed proposal - described MEC use case (smart relocation) and changes for DMM - has the author looked at LISP? Yes 11:40AM: Title: Router Advertisement Prefix Option Extension for On-Demand Mobility Time: 10 minutes Presenter: Wu-chiX Feng Description: Extensions to ND for indicating Mobility Service Type Draft: https://www.ietf.org/id/draft-feng-dmm-ra-prefixtype-00.txt - reviewed - suggested it was brought to 6man (was done) - no conflict with X bit but there is another proposal to burn anohter bit which eats 2 of the 3 R-bits - Suresh noted an extension mechanims for the exhaust or control through a registry - Alex - Who makes the demand? - Authors: We don't want to change the semantics of RS with this proposal. Network will provide alternatives for the 'who'? - If there is an argument to modify RS we could try but authors were discouraged. - PBD option may be good for this. 11:50 AM: Title: FORCES for FPC Time: 5 minutes Presenter: Jamal Salim Description: Proposal for using FORCES on FPC as a south bound protocol Draft: TBD - Lyle - As an author we are interested in finishing the information model - as an implementor interested in the protocol 11:55 AM: Adjourn