IETF 96 Working Group High Level Summary

Page created for the IETF 96, July 17 - 22 2016, in Berlin.


  • Met on Wednesday with 49 local participants and 6 remote, including one presenter remote.
  • Enhanced attendance also resulted in 8 new list subscriptions (both are recent records).
  • Almost all WG drafts advanced at this meeting
    • 2nd WGLC for Datacenter and NFVI Benchmark Considerations
    • 1st WGLC for SDN Controllers and OPNFV VSPERF Summary.
  • IPv6 ND was re-directed at last meeting, but no progress.
  • Several new work proposals, EVPN will benefit from consolidating with PBB draft
  • Two invited talks - very successful and relevant to the WG, much discussion!


  • DIME met on Friday morning with around 10 participants in the room.
  • The Diameter overload related work is reaching the end. Documents are either in past WGLC or entering one.
  • New big topic in Dime is RFC4006bis:
    • Update of the RFC4006bis work with currently identified topics was given.
    • draft-bertz-dime-rfc4006bis-01 - will be adopted as a WG item. To be ratified in the mailings list.
  • Response LS to 3GPP SA5 (CC CT3/CT4) regarding their question on RFC4006 will be sent back before 3GPP SA5 meeting in August. The response indicates DIME will update RFC4006.
  • A series of four new individual I-Ds were presented mostly around extending policy control and QoS.
  • A proposal to produce RFC7423bis (application design guidelines) was given. The first item would be clarifying the answer messages and related procedures when the 'E' bit is set.
  • Chairs solicit more activity to complete the end to end AVL level security milestone.
  • Chairs opened the discussion of potentially not having meetings in coming IETFs. The work could be done through emails and possible virtual interim meetings.


  • The Chairs were most pleased with the draft-woodworth-bulk-rr draft, as it was the authors first draft, and their first IETF, and was received positively.
  • There was also a short updates on the 6761 problem statements, and the current plan is to ask the working group to adopt one of the two drafts to move forward.



  • Did not meet in Berlin
  • Sole WG draft is in WG last call
    • Please review it!


  • Layer Independent OAM Management in the Multi-Layer Environment
  • LIME met on Thursday afternoon during the IETF96 week.
  • Three main topics:
    • Connection-Oriented model: WGLC comments
      • Expect new revision based on normalizing the address ME/ fields
    • Connectionless model: Post interim review
      • Similarly, expect new revision with specific updates
      • Also, suggestion to split data model from RPCs, take to the list.
    • Applicability table at the wiki.


Met on Friday 7/22. Main focus of the agenda was discussing open issues related to the three WG documents in WGLC. Good progress in solving the open issues due also to the f2f work done during the week by the editor and WG participants. Comments expected until 7/31, new version in August, followed by 2nd LC. Virtual interim in September if necessary. Proposals for new work are expected on the mail list. WG will meet in Seoul only if there is a consistent number of new proposals, or critical items that need to be solved f2f. The WG also received a presentation from the EC project and a demo of the Jacobs University implementation.


  • Approx. 120+ participants in the 2.5 hour NETCONF session.

  • Andy Bierman will provide an update for RESTCONF and YANG Patch within 2 weeks.
  • “YANG Push” drafts were discussed (Eric Voit, The WG agreed to adopt the three drafts 5277bis, NETCONF Transport and RESTCONF & HTTP Transport. The next version of the drafts will be accepted as draft-ietf.
  • System Keychain Model, SSH/TLS Client Server Models, TLS/NETCONF Client Server Models, Zero Touch (Kent Watsen): Issues will be addressed and drafts will be updated as discussed. Client models will be addressed in the next version.
  • I2RS strawman proposal (S. Hares): Sue presented the I2RS requirements documented in ephemeral state draft. There was no further comments. The chairs asked the attendees to comment and discuss if they see an issue. Otherwise they will are seen as accepted. Sue agreed to provide an overview on which requirements are addressed in which draft and where an action is necessary.
  • Refined YANG datastores with Meta-data (R. Wilton): The datastore conceptual discussion is continuing in NETMOD WG. It has been proposed that the architectural discussion should be continued in NETMOD WG and the protocol part as well as the necessary changes to the datastore definition to NETCONF RFC should be done in NETCONF WG.


  • Two 2-hour NETMOD sessions (Monday and Tuesday)
  • Roughly 100 participants in both NETMOD sessions
  • The first session was dedicated to opstate and schema mount
  • The second session reviewed 9 chartered drafts and 3 unchartered drafts
  • Drafts nearing last call include 6087bis, routing-cfg, and acl-model
  • Ops AD stressed the urgency of unblocking other drafts ASAP
  • WG needs to provide clear guidance on how current drafts should handle 1) applied configuration and 2) derived state for system-generated objects in the interim while waiting for the rollout of a complete opstate solution


  • Three items on the on the agenda, all with an IPv6 orientation
  • WG is trying to progress -Operational Security Considerations for IPv6 Networks- after 4 years of existence to closure
  • WG meeting progressed positively and no notifications needed to highlight further to OPS ADs


  • Very high attendance by the standards of a radext meeting (19 people on blue sheets)
  • Productive discussions on all presented drafts
  • four individual drafts waiting to become WG documents; mostly looking on-topic and useful
  • The vivid discussions made us go overtime, will request a 90 minute slot for IETF97 to accomodate


  • SUPA's Information Model (GPIM) and Data Model (EPIM) are now WG drafts, work on them is progressing slowly.
  • New Framework draft, replaces the former Architecture and Value Proposition drafts. The meeting supported its adoption as a WG draft (now announced on the SUPA list).
  • Problem Statement draft: vigorous discussion of whether it is needed, maybe it should be replaced by an Applicability draft. Further discussion was left for he SUPA list.
  • Rakesh Kumar gave a brief presentation on work in the I2NSF WG, that work may also be relevant for SUPA.


Monday afternoon session, 120 minutes. 41 attendees in the room + 5 remote participants in the room.

  • Signaling design team presented document status, including several open issues that discussed. The document is almost ready, waiting the final checking against all other documents.
  • Bootstrap design team presented document status, including several open issues. The document is almost ready, but requires more reviews.
  • Reference model document presented work status. Progress since last IETF includes, etc. It is remaining for discussion how to handle the out-of-charter contents.
  • Two use cases documents (Prefix management in large-scale network & stable connectivities) are steady, but wait for depended documents to be sent to IESG together.

Thursday afternoon, session 2, 120 minutes. About 29 attendees in the room + 3 remote participants.

  • Autonomic control plane draft presented work status. The document is almost ready, but requires more reviews.
  • Information Distribution over GRASP was presented.
  • Autonomic Functions Coordination was presented with a concreted example.
  • Autonomic network intent concept and format is presented. Just after last meeting, there are a number of active discussions regarding to this topic in the mail list. The WG showed great interests on this topic.
  • Autonomic functions life-cycle and management was presented.
  • GRASP API was presented.


  • OPSAWG met in Tuesday morning joint with OPS Area meeting. 64 participants.
  • MUD was presented with updates since last meeting. OPSAWG is hosting the MUD related work. Hum in favor none really opposed. Will call for adoption when new draft arrives.
  • draft-wu-opsawg-service-model-explained-01 explained the service model concept and service architecture.
  • A requirement of operator facing VPN model is proposed. While L3SM is customer facing, the composed VPN model helps operators to monitor, diagnos, and operat the PE based VPN deployment.
  • draft on IPFIX IE extension for BGP community. It's the first one based on basicList. It deserves special attention, review, and potential documentation as an RFC.
  • Automatic attachment goes a little further than the bootstrap process work.
  • In Band OAM was presented for comments from OPS.It's an interesting generalized "in band" method. We hummed to see if this work fit in OPS, some in favor none opposed.
Last modified 6 years ago Last modified on 28/07/16 17:30:01