IETF 67 MANET WG Agenda MANET Session 1 (2.5hours) Monday, Morning Session I Room Name: Harbor Island III San Diego, CA You can find the agenda and presentation material at https://datatracker.ietf.org/public/meeting_materials.cgi?meeting_num=67 + Agenda Bashing Joe Macker introduced the draft agenda and opened it up for discussion. Co-chair Ian Chakeres was absent from IETF 67. Thomas Clausen asked that his presentation times be contiguous since all material was in one presentation. Charlie Perkins asked to add a small presentation on DYMO path accumulation. Requests were accommodated to adjust the agenda accordingly. +WG progress IPR reminder - RFC 3979 http://datatracker.ietf.org/public/ipr_disclosure.cgi Joe Macker provided a brief overview of WG progress and issues. At present chartered working group documents are moving forward. DSR as a remaining legacy charter EXP RFC track document is now in the final stage of the RFC editor queue. Focus of the next WG milestones is to begin preparing documents for Standards track submission. Authors were reminded to reexamine requirements for standard track documents. Fenner mentioned that previous router requirement ID referenced was NOT in effect. There was encouragement to carefully review documents and send comments to the list and the authors. Joe Macker mentioned that packetbb and NHDP need to be quickly matured so other documents using them can progress. +OLSR Interop/Workshop 2006 -Thomas Clausen See slides. Clausen mentioned that 6 different OLSRv1 implementations and 4 different OLSRv2 implementations were tested over 2.5 days. +Packetbb -Thomas Clausen Fragmentation was removed from packetbb since the last IETF. The address-block- semantics field needs some rearranging. The approach is to move the text to IANA section to discuss private/std usage allocation. Joe Macker questioned if the security section in its current state was needed since this only a container specification. Clausen agreed with Joe that this section might not be needed in its current state. Perkins mentioned that packet efficiency matters and would like people to consider related packetbb issues. Open group discussion on the tradeoffs of efficiency and flexibility in packetbb then took place. Sar and Bormann both suggested that networks in need of more efficiency could use a modified version of packetbb. The possibility of 6lowpan using packetbb was mentioned. Clausen stated that despite the discussion regarding efficiency that packetbb is a solid document and the group really needs to take it forward. There appeared to be rough consensus with that direction. Dean also mentioned that it has been demonstrated at the Interop that in some cases packetbb messaging is more efficient that the original packet formats used by OLSR. +NHDP -Clausen NHDP is awaiting packetbb finalization and needs review with SMF functionality. Clausen would like to explore the idea of a document of TLVs which we have found to be useful. Perkins asks if NHDP needs to be timer/proactively driven. Adamson responded that the messages allow for reactive messaging but that there may be left over text which hasn't been removed. Dean mentioned that we need to make sure protocols using NHDP do not cause bad behaviors when working in the same network. +OLSRv2 -Clausen Awaiting packetbb and NHDP finalization and lingering text needs to be removed. Discussion occurred regarding the maturity of the multiple interface protocol behavior and it was mentioned that three different implementations already had multi-interface support but no thorough testing had taken place. +SMF -Macker Version 3, draft-ietf-manet-smf-03.txt, has been posted since the last IETF. Changes since the last version include: NHDP modifications, gateway issues expanded, applicability section modified, model of attached non-SMF nodes discussed, and removed TLVs that belong in other documents. There are currently at least three different implementations of SMF. +DYMO -Perkins Freshness of routing information, and naming categories was discussed and anyone with better naming scheme was encouraged to comment to the list and authors. Timeouts currently have clear state changes but having one timer to remove old information is being looked into. There are at least 2 full implementations for DYMO. +Path Accumulation -Perkins When the network has a high PDR then path accumulation will not help, but under congestion path accumulation can help. Results were discussed and value was scenario dependent. +Reliable Flooding -Perkins Study results were presenting on reliable flooding using FEC encoding and results so far are ambiguous. +MANET IANA needs -Macker The ID authored by Chakeres on MANET IANA requirements was discussed. Does MANET need common port assignment? Is one or more link local MANET multicast address needed for router neighbor signaling? Should site local multicast address(es) be requested? Discussion on what happens to LL packets with TTL values of greater than 1, with general agreement that it is not the intention of the MANET working group to forward LL packets. LL packets only reach directly connected neighbors at L3. At the IP layer, site local multicast could be appropriate multi-hop multicast contained to the MANET. Draft-fenner-iana-exp-2780-05.txt has relevant IANA number assignments for ip addresses and their uses during experimentation ONLY. People may wish to know about and use these approaches prior to formal assignments. +DYMO Implementations -Oh Implementation uses packetbb and nhdp and TLVs defined in the various documents. Packetbb was a little difficult to implement and there was some processing overhead for various TLVs in one address block. +Open Discussion -Macker Wrap-up of discussion. Urged WG to please review packetbb and NHDP and send comments to the list and authors within the next few weeks.