IETF-92 Routing Area Open Meeting ================================= https://trac.tools.ietf.org/area/rtg/trac/wiki/WikiStart ADs: Alia Atlas (akatlas@juniper.net), Alvaro Retana (aretana@cisco.com), Deborah Brungard (db3546@att.com) Scribe: Jon Hardwick (jonathan.hardwick@metaswitch.com) Location: Gold Meeting Room, Fairmont Hotel, Dallas, Texas, USA Time: March 26, 2015, 1520-1720 (3:20pm-5:20pm) ------------------------------------- 1. Administrivia ---------------- Agenda bashing -------------- Alvaro opened with the note-well notice. Reminder about the directorate and wiki. Reviewed the agenda. There will be an update on the status of the WGs. There were no BoFs this time. There are no additions or changes to the agenda. 2. Update on Initiatives ------------------- a. Area Tuning for 3 ADs ------------------------ Alia reviewed the overall area status. ForCES has concluded. There are still some remaining drafts to be published as RFCs and these will be AD-sponsored. Alia thanks Jamal and the rest of the ForCES WG participants for doing a good job. Since the last IETF the ADs have rechartered mpls, ccamp, pce, i2rs. The ADs have created charters for two new WGs: teas, bier. Three new WGs have been added to rtgarea from intarea: l2tpext, lisp, trill. They belong more properly in rtgarea and this will offload the intarea ADs. Many thanks to Adrian Farrel for 6 years of service to the routing area and IETF as area director. There are now three ADs for the routing area (as above). Alia showed a slide which shows how the WGs are partitioned between the new ADs. This information should be reviewed online. b. WG Chair training -------------------- Alia has been running a WG chair training program since June 2014. The aim is education and team building amongst WG chairs. WG training sessions are recorded and available to view online. All rtgarea participants are encouraged to view them. The session on consensus is particularly recommended to all participants. It is really critical for undserstanding how the IETF works. There is an upcoming session on YANG/netconf. There is an upcoming session on charters / milestones. There is an upcoming session for document shepherds. RTGAREA participants are encouraged to volunteer to be document shepherds. This is good experience and takes workload off the WG chairs. RTGAREA participants are encouraged to keep up the work / momentum between physical IETF meetings. 3. YANG Updates --------------- a. update from rtg-yang coordinators ------------------------------------ Dave Sinicrope and Qin Wu are the RTGAREA YANG coordinators. Dave gave an overview of what they are doing. They have set up an area of the RTGAREA wiki to record what YANG models are being worked on, who is working on them and what state they are in. Please can the WG chairs get the folks who are writing the YANG models to keep the tables on the wiki up to date. Some of the information in the tables is out of date - some drafts have expired. Please can the WG chairs figure out who needs to review each model and assign reviewers in the table. There is another area on the wiki for people who are thinking of creating a model to post their thoughts and solicit feedback and collaborators. There is a set of links to the OPS area's YANG compiler information. As part of the review process for the YANG models please could people compile their YANG. Alia: This is a bg job for Dave and Qin. There are >80 drafts with YANG models in and this is growing. YANG authors must take responsibility for updating the wiki. Alia discussed how to deal with Cross-WG YANG models. Some models will cover multiple working groups. This is fine as operational requirements often straddle WGs. Primary ownership will be assigned to one WG. The WG chairs will coordinate. There are two motivations. 1) RTGWG cannot be over-filled with YANG drafts. It is busy doing other things. 2) There is no RTGAREA yang model WG because then we would create a centre for YANG expertise, not routing expertise. b. L3SM Working Group Intro --------------------------- (The presenter was not present in the time slot so this section was deferred until later in the meeting and handled out of order.) Adrian Farrel presented for Benoit Claise. This is a WG that was formed "by stealth" without a BoF. It was chartered by the IESG, but not in time for a meeting at this IETF. It aims to be a short-lived working group for less than a year. Milestone is to complete by Jan 2016. The chairs are Adrian Farrel and Qin Wu. The work comes out of the operator-led design team that created draft-l3vpn-service-yang. The scope of work is the interface between the operator and the operator's customer. Can a group of operators agree on the description of a service model. L3VPN was a good starting point. Adrian was sceptical as operators describe this very differently. But the operators came up with a draft very quickly. The aim is to produce a service model description for L3VPN. It is not an L3VPN configuration model. It is for talking to a service orchestrator, or like an automated purchasing system between customer and operator. It will involve a magic unicorn box thing to map the service request to the L3VPN provisioning model. It will start by considering adopting draft-l3vpn-service-yang. Please don't write docs to compete with this. Instead, bring your comments on the doc to the WG. Comments from the floor. Eric Osborne: I have a stone to throw at you. When I spoke about this in the OPS area meeting I did not understand this and was vehemently against it as it sounds like an attempt to get into the ITU-T style service delivery business. I am less against it now as I see this is not the intent. But why is this standards track and not experimental track? It is a science experiment. Benoit: It sounds like Eric is afraid that this will appear in RFPs. I understand the concern. This is more about making sure we have a complete picture of requisite YANG models in the IETF and have done our homework. Adrian: Propose to work on the draft, and if it turns out that people are building it then it will be strandards track, otherwise we may re-charter to change its status to experimental or informational. We can ask Benoit to recharter nearer the time. Adrian: Service providers please review this. If you know enterprise customer that might like to contribute to this then please let them know about it, Adrian would like to have them involved. 4. ANRP Presentation -------------------- "OpenNF: Enabling Innovation in Network Function Control" - Aaron Gember-Jacobson Aaron described his work in Network Function Control which demonstrated a need to add an extra ingredient to SDN and NFV in order to control dynamic transfer of workload from one VNF to another. See the slides for details. A longer version of this talk was presented at the IRTF open meeting. There were a few questions and comments from the floor. Linda Dunbar: This is a very broad problem. Will there be a new IETF work item? Aaron: No, this is for information, it is our research project. Alia: ANRP is an opportunity for researchers to liaise with the IETF / IRTF and share ideas. Linda: We have I2NSF (interface to network security function) which is similar but smaller in scale, you may wish to view the demo at bits-n-bytes. Acee: Can you use the state transfer part for HA? Aaron: No, but see pico-repliocation from IBM for HA, which is similar. 5. Working Group and BoF Reports -------------------------------- Please refer to the slides for details. BESS (Martin Vigoureux): Attendance of the group is stable at ~100 people. Activity on the list is bursty but constructive. Two new RFCs since IETF 91. Working on EVPN, MVPN, cleaning up IANA registries. BFD (Jeff Haas): Jeff is serving year 11 of hos 1 year WG. published RFCs 7492 and 7419. Seamless BFD is progressing, but slowly. rfc5884bis is ready for WGLC - please review. BFD YANG team has published first I-D - please review. BIER (Tony P): initial meeting, some drafts already at version 3, animated discussions on the list. Seen drafts have been called for adoption. BGP, OSPF, ISIS extensions. Adopted OAM requirements draft. We have no YANG model. CCAMP (Daniele Ceccarelli): we were affected by the reorg but the room was still full; Two new RFCs 7487 and 7447. There are three YANG model drafts. ForCES (Adrian Farrel for ForCES chairs): They are closed, published 2 RFCs, they have 2 WG drafts hanging over that will be passed to Alia to do an AD-sponsored RFC, and some 8 drafts they will write in future to document what has been implemented. I2RS (Sue Hares): Making good progress; thanks to our shepherds Mach Chen and Qin Wu. Looking for volunteers to join protocol requirements design team and FB-RIB design team. 2+ calls per week over 2 weeks. 6 interims are planned. 1.5 hours, please join if interested. IDR (Sue Hares): Meeting room was way too small. Since IETF 91 many drafts have gone through WG last call and some to the IESG, some in the RFC editor queue. Eight new WG drafts adopted and multiple adoption requests pending. ISIS (Hannes Gredler): No new RFCs, one new document: MRT for IS-IS. Four active WG docs. The segment routing draft is particularly important, please review it. L2TPEXT (Alvaro Retana): New WG, did not meet this week, we did not get an update from them. LISP (Joel Halpern): We are going to try to finish off our existing documents before taking any new ones on. MANET (Justin): 12 drafts; 6 drafts submitted to IESG for publication; 3 being last called; 2 will be last called soon; 1 (AODVv2) is pending a document update. We will have interim meetings to get it ready for last call. If not ready for LC at the next IETF then we will change it to experimental. MPLS (Ross Callon): Updated charter, some work moved to TEAS. 7 new RFCs, 5 in RFC editor queue. 5 more with AD (now Deborah) or with the IESG. We talked a lot about making rings haveless configuration. Lots of work in progress. YANG models will be discussed. NVO3 (Benson): Had 3 interim meetings each focussed on a topic; working well. 2 smaller in-person meetings this week and that results in fewer meeting conflicts. OSPF (Acee Lindem): 1 new RFC (TE metric), 1 doc with RFC editor (v2 security). Want to last call the prefix/link attributes draft. Would like a volunteer to write up the implementation report - several implementations. YANG model moving well. OSPFv3 extended LSAs going well but needs implementations. PALS (Stewart Bryant): Thanks to AF for the excellence of his reviews over the last 6 years. Two new RFCs. 7436 published as historic after 12 years as an IETF document. 4447bis is nearly cooked. PCE (Julien Meuric): Lots of people in the meeting. 3 RFCs published. Started discussion of the new PCE WG charter scope. Focus on 6tisch needs, which may overlap also with CCAMP and TEAS. Lou: the discussion was not specific to 6tisch, it was focussed on detnet. PIM (Alvaro Retana for PIM chairs): In process of being rechartered. Forming a YANG design team - please talk to the chairs if you want to be involved. ROLL (Alvaro Retana for ROLL chairs): No meeting, no Ines or Michael. RTGWG (Jeff Tantsura): MoFRR is being published, rLFA published, OAM for LFA is WGLC complete. SFC (Alia Atlas for SFC chairs): Problem statement in RFC ed Q, arch draft in Alia Q, NSH encap is finally adopted. Control plane discussions in progress. SIDR (Sandy Murphy): AS migration draft has been sent to IESG for publication; 1 draft in RFCEd Q, has been there since last IETF. Tangled in IETF Trust process of how to deal with templates. This has now cleared so should be published. New work adopted to define new transport protocol for rPKI - people have concerns about rsync performance so a new protocol is being defined. Discussed WGLC comments on BGPSEC protocol. Discussion about deployment (or lack thereof) of RPKI and DNSSEC. SPRING (John Scudder): Problem statement draft is through WGLC, architecture document should be WGLC soon. There is related work all over the routing area, so there is a lot of coordination required. TEAS (Vishnu Beeram): 3 docs in RFC ed queue (inherited); lots of yang discussion; discussion on ACTN; intent to prodce a requirements document for TEAS. TRILL (Sue Hares): Props to Donald Eastlake for being a phenomenal secretary. Multi-level / multi-topology trill are next on the agenda for this WG. 6. Routing Area Directorate --------------------------- Jon Hardwick gave an overview of the performance of the routing area directorate over the past year. Over 40 documents have been reviewed. The directorate are good at finding issues and add value to documents. They are also lightly loaded so WG chairs are encouraged to make more use of them. Alia thanked the directorate for their efforts. Alia reminded the participants that the directorate do not just do IETF last call reviews, they also do QA reviews. These can be requested by WG chairs whenever documents have been adopted by a WG. The aim is to provide directorate feedback earlier in the document's lifecycle. Chairs should make requests for QA reviews to the directorate secretaries. Stewart Bryant and Ross Callon suggested that review requests be made visible to the entire directorate. Jon and Alia agreed that review requests should be copied to the directorate email list. Acee Lindem: has found that recently the drafts he is reviewing are more relevant to his experience which is good. Alia: please could the directorate make sure that they fill in the relevant wiki page with their experience and interests. 7. Open Discussion / Any other business --------------------------------------- Alvaro: please give the ADs feedback on how well the area is working. Stewart: can we slim down the administrative plenary? It is overlong. Alia closed the meeting at 17:10.