IETF Working Group 91 High Level Summary

Page created for the IETF 91, November 09 - 14 2014, in Honolulu.


This meeting is the first ANIMA meeting. It first clarified the WG charter. It currently focuses on the reusable autonomic network infrastructure that helps to reduce the OpEx? for "on professionally-managed networks". Six individual drafts have been presented in the meeting. Each of them matches a WG work item. They have invoked active discussion. Potential relevant/overlap works, including Homenet and zero-touch from netconf, have also been introduced and discussed.

The WG has not made any decision to adopt any draft as WG document yet. We are open for competing proposals/drafts. We also encourage future discussion regarding to more validation use cases or autonomic service agents.


  • We had approximately 25 participants in the 2.5 hour BMWG session on Thursday afternoon/evening, using the entire 2.5 hours, and had a lively discussion over new WG items.
  • Existing documents are healthy and new drafts look promising; we are on track with our milestones.
  • Note takers were Lucien Avramov and Bill Cerveny. Bill Cerveny would have been the jabber scribe, but there were no jabber participants this meeting.

-- SIP Drafts are IESG Approved.

-- BGP Dataplane drafts are in IETF last call.

-- Traffic Management draft is adopted in the WG and sent to WGLC.

-- ISSU Draft is adopted and expected to go into WGLC shortly after IETF-91.

-- Virtual Network Function and SDN work well discussed, including 2 new drafts within the working group.

-- The “Benchmarking Methodology for Virtualization Network Performance” given by Rong Gu (China Mobile) was particularly engaging, and quite a Q/A session at the microphone ensued. There’s a lack of clarity on how to trust the results obtained from the testing, but the authors are committed to reviewing the feedback and revisiting these questions for IETF-92.

-- The new IPv6 Transition Benchmarking draft presented by newcomer Marius also elicited a healthy amount of feedback and refinement, with a commitment from Marius to come back to the list with a revised draft that addresses the TCP/UDP questions, ad the 6-to-4 versus 4-to-6 comments/concerns.


Dime session on Tuesday 11/11, 0900-1130 Approx 20 people in the room, with the key contributors in the room. So the location of this meeting was not an issue for Dime.

  • WG documents:

-- Design guidelines (draft-ietf-dime-app-design-guide-28): in AUTH48.

-- Overload Control (draft-ietf-dime-ovli-04): new version of the draft -04 has received a lot of constructive comments and the next version should reach a WG consensus. WGLC should be launched quite soon before an IETF LC and IESG review in December. This should meet 3GPP deadline for a normative reference in their Diameter specification at the end of the year.

-- Congestion flows attributes (draft-ietf-dime-congestion-flow-attributes-01): No remaining issue after publication of -01. WGLC launch right after the IETF meeting

-- Group Signalling (draft-ietf-dime-group-signaling-04: )Revision needed after comments received just before and during the meeting. Mainly clarifications required. If no more comments received after the next version, WGLC will be launched.

-- Diameter E2E security requirements (draft-ietf-dime-e2e-sec-req-01): Draft has expired but Jouni will reactivate the draft. Should be ready for WGLC and IETF LC after submission of the next version.

  • Charter items:

Two new drafts related to the overload topic have accepted during the session: Agent overload (draft-donovan-dime-agent-overload-03) and rate control alogorithm (draft-donovan-dime-doc-rate-control-02). Approval will be confirmed on the mainling list.

  • Individual submissions:

the draft defining AVP for Provisioning Customer Equipment Supporting IPv4/IPv6 Transitional Solutions (draft-zhou-dime-4over6-provisioning-05) is stable enough to become WG document and launch a WGLC. A new milestone is required but no charter update. A document on load architectural considerations (draft-campbell-dime-load-considerations-00) has been discussed but no WG adoption for now. Further discussion is needed.


Two drafts are heading to WGLC - draft-ietf-dnsop-edns-chain-query and draft-ietf-dnsop-dnssec-roadblock-avoidance. Four documents are moving to a call for adoption: draft-eastlake-dnsext-cookies; draft-dickinson-dnsop-5966-bis, draft-livingood-dnsop-negative-trust-anchors, draft-wkumari-dnsop-root-loopback. The last document has some more discussion needed. A lot of discussion about draft-howard-dnsop-ip6rdns, and more discussion is needed before this can move forward. The time was spent pushing more documents through the mill.


During the inagural meeting, there was a lot of discussion around the problem statement draft-ietf-dprive-problem-statement, and more work will be done on editing and reviewing before WGLC, which will happen shortly. There is also a desire for an evaluation of methods. Three different proposals were presented: draft-hoffman-dprive-dns-tls-https (in variations), draft-hzhwm-dprive-start-tls-for-dns, and draft-hallambaker-privatedns. The plan is to adopt one or more documents, with the idea that only one would head to WGLC.


This meeting grow had 2 presentations. One was a proposed new piece of work -- draft-van-beijnum-grow-controlled-deagg -- was discussed at length during the meeting, centred around geoloc routing, and hierarchical advertisement. Based on the discussed from the group, further work will need to be done before the draft is brought back to GROW. The second presentation on draft-sriram-route-leak-problem-definition-00 was a follow up from the toronto ietf. The work has progressed and interest was expressed during the meeting to adopt this draft as a working group item. The request for adoption has been put out on the mailing list, and the expectation is this document will be a working group item.

Outside of IETF or coinciding with it, we have received an update on draft-ietf-grow-ops-reqs-for-bgp-error-handling allowing the group to move forward with progressing the document. draft-ietf-grow-bmp is promised to be updated soon. draft-ietf-grow-ix-bgp-route-server-operations has moved to the RFC Editor queue, draft-ietf-grow-irr-routing-policy-considerations is in IESG. Lastly draft-ietf-grow-bgp-gshut is ready for submission to the IESG.


LMAP met on Thursday morning. There were around 50 attendees. Some of the key contributors could not attend the meeting, but two of them attended remotely via Meetecho. It was not perfect but it worked.

Framework document - needs work to resolve comments from AD, especially refering how passive monitoring is reflected in the framework

Information Model - in progress

Protocol contributions - three out of four contributions were presented, including demo. There is a need for criteria for protocols selection, Tim and Barbra volunteered to draft proposal

Three presentations for non-chartered items - contributors are encouraged to continue to work and update the WG

The WG will hold three virtual interim meetings - first one in the week of December 15.


  • We had approx. 85+ participants in the 2 hour NETCONF session,
  • We reviewed the status of the WG,
  • We had a discussion on all 6 chartered documents.
  • Note takers were: Lada Lhotka and Balazs Lengyel. The Jabber scribe was Rolf Winter.

Many thanks to the note takers and jabber scribe for volunteering. The session agenda is available at: NETCONF co-chairs will ask for a room for 100 for the next meeting.

Following are the decisions taken per humming in the NETCONF session. If there is no strong objection we will implement as proposed.

RESTCONF Call home support: humming supported this approach, and the WG wants to do it now.

Separate ports for callhome for SSH, TLS, RESTCONF: The WG supports with a large humm multiple/separate ports.

Focus on the use of X.509 for TLS (rfc5539bis): supported by a large humm

Describe algorithm for extracting a user name out of the X509 attributes: loud humm in favor

Use if-feature for server-model instead of waiting for an YANG 1.1 solution: loud humm in favor, certainly we do not want to wait

Should we generate ssh host keys/509-certificates via a yang model or generate them out of band: big humm for yes, i.e. we postpone and do it later.

Report hostkeys/certificates via yang data model: Humm to take them out and not report them via a data model

RESTCONF state monitoring would require RESTCONF sessions: ACTION item: Andy will take the issue to list

RESTCONF Default handling: ACTION item: Andy will take it to the list

SHOULD or MUST use HTTP 1.1: humm in favor of SHOULD

The issue on the I2RS ephemeral state has been summarized by the I2RS co-chair Jeff Haas. The discussion will be continued in NETCONF virtual meetings.

The two drafts on fragmentation support will be merged and provided to NETCONF maillist. The aim should be to raise a discussion and to get the support of NETCONF WG memebers.

Virtual Meetings:

  • The WG will continue with virtual meetings.
  • In the first meeting we will discuss I2RS ephemeral state.
  • Action item: Mehmet to send out a new doodle poll to chose the best time for all.

Potential requirements on zerotouch from ANIMA WG:

  • Action item: WG chairs to communicate with ANIMA WG co-chairs that they must agree on requirements and provide with a draft to NETCONF WG.


The NETMOD meeting ...


The meeting focused on cycling through all the drafts which are currently in the works, and on a rechartering discussion at the end. Almost all drafts which were under active discussion had a document author or editor present, with the exception of CoA-Proxy and the Datatypes draft; there we simply flashed the slides and had a Jabber conversation with the main author. I could not see detrimental effects of the chosen meeting location (there were a few people less in the audience, but key people were there).

The proposed text for rechartering was accepted by all participants with surprisingly few discussion. The new charter will bring three new drafts into the WG (coa-proxy, datatypes, populating-eapidentity) and is generic enough to allow possible new work in without needing rechartering again. For document track changes, RFC6613, RFC6614 to STD was welcomed, and surprisingly, moving CoA from EXP to STD was also on the table. People unanimously felt that MAJOR changes to that one would be needed though, so this will not be done in the immediate future. As a consequence, coa-proxy will also be EXP.


No mayor items to report. WG items under control.

There were three drafts discussed at the WG. In general the OPSEC WG is relatively silent and the ongoing work items progressed without mayor blocks. A document regarding secure operation of an IPv6 network which was found ready for WGLC, while the two other documents will be worked upon inside the working group.

create you own -

Last modified 4 years ago Last modified on Apr 9, 2015, 12:49:56 PM