Routing Area Open Meeting (rtgarea) IETF 96 (Berlin) =============================================================================== Area Directors: Alia Atlas (akatlas@juniper.net) Deborah Brungard (db3546@att.com) Alvaro Retana (aretana@cisco.com) Area Secretary: Jonathan Hardwick (jonathan.hardwick@metaswitch.com) Wiki: https://trac.tools.ietf.org/area/rtg/trac/wiki/WikiStart Scribe: Jonathan Hardwick (jonathan.hardwick@metaswitch.com) Location: Potsdam I, Intercontinental Hotel, Berlin, DE Time: July 19, 2016, 1620-1820 (4:20pm-6:20pm) ------------------------------------------------------------------------------- 1. Administrivia ---------------- Alvaro: Please review the documents in your WG, and also review a document from another WG. IETF document qualilty depends on this. Nobody bashed the agenda. Ross Callon is retiring. The ADs thanked him for all his work in his many roles during his long involvement with the IETF (beginning at the first ever IETF meeting!) ------------------------------------------------------------------------------- 2. RTG Technology Presentation - Mobile AdHoc Networks ------------------------------------------------------ Justin Dean/Stan Ratliff Alvaro: MANET was just rechartered - please go along to the meeting tomorrow. Justin: New items: MANET management document; multicast work (building block apppproach); DLEP extensions for more metric types. George Swallow: Building large packets by putting small packets together; large packets more likely to get corrputed; how about sending smaller packets? Justin: Smaller packets will get through more, which is not always good as you don't want to find neighbours that you have a poor link to. Chris Bowers: Why is the 2-hop neighbour important? Justin: Many of the algorithms require you to know your 2-hop neighbours. Chris: So unlike IS-IS, you need to be careful to constraing flooding? (Yes) Stan: RFC 5820 uses a 2-hop neighbourhood for flooding in OSPFv3. Jehan Tremback: Does DLEP replace Hello in OLSR protocols? Justin: Not completely as the Hello messages require TLVs. ------------------------------------------------------------------------------- 3. Keep it Simple: The cost of (too many) Standards --------------------------------------------------- Ross Callon Jeff Haas: See RFC 1925 point 12. Charlie Perkins: It will get worse fast with IoT. 50 billion devices in 5-6 years? How can the IETF make a concerted effort? Reducing protocols = reducing validation of interesting ideas. Alvaro: We all have to think about simplifying what we do. ------------------------------------------------------------------------------- 4. CodeMatch Overview --------------------- Christian O'Flaherty Rick Taylor: Licensing intentions: are you looking for code with BSD, GPL license? Christian: It is not a requirement to be open source. Rick: But of the code itself? Alia: This applies to closed-source implementations as well. Sandy Murphy: Interop is supposed to be between independent implementations. Two implememtor's independenty read ther standard and did the same thing. Won't the mentor overly influence the implementation with their knowledge / experience of the draft? Alvaro: These implementations would not usually count in the IETF as interoperable implementations and that's not what this is intended to achieve. We're talking about schools, not commercial grade software. We're hoping these people may help us find bugs in the documents. ------------------------------------------------------------------------------- 5. Working Group and BoF Reports -------------------------------- In general, see the notes in the wiki for the reports. Additional comments noted below. Babel (Donald Eastlake) Bess (Thomas Morin) Bfd (Jeff Haas) All seamless BFD drafts are now published. Bier (Greg Shepherd) We have a stack of drafts close to last call but we need to do the due diligence in other WGs before we move them forward. We need some implementations and interop. The WG is looking at selecting another data plane encap, besides MPLS. There are a number of contenders but the charter allows only one to be adopted. We've agreed to progress this discussion through a design team. Ccamp (Daniele Ceccarelli) Detnet (Deborah for Chairs) Busy with terminology and data plane alternatives. I2rs (Sue Hares) Idr (John Scudder) Excellent operator engagement in this session. Flow spec has been the subject of many interims and this has been helpful in driving consensus. Isis (Chris Hopps) L2tpext (Deborah for chairs) Lisp (Luigi Iannone) Manet (Stan Ratiff) Mpls (Ross Callon) Nvo3 (Matthew Bocci) Ospf (Abhay Roy) Pals (Stewart Bryant) Those wishing to read the new congestion considerations RFC should look at the PDF version, not the text file! Authors - please reply promptly to IPR polls: it really holds up the process. Pce (Jon Hardwick) Pim (Mike McBride) PIM WG is not the PIM protocol but now contains all multicast protocol work. Roll (Ines) Rtgwg (Jeff Tantsura) Sfc (Alvaro for chairs) Alvaro publicly shamed the chairs for not filling their details into the wiki! Sidr (Sandy Murphy) Spring (Bruno Decraene) Teas (Vishnu Pavan Beeram) Trill (Sue Hares) ------------------------------------------------------------------------------- 6. Open Discussion / Any other business None