DTN WG Minutes IETF 82 March 26, 2015 ======================= Administrativia ------------------------ Ed Birrane volunteered as WG Secretary Ed Birrane volunteered to take minutes for the meeting. Agenda Bashing ------------------------ No changes to the agenda proposed. WG Way Forward ------------------------ * Marc Blanchet reviewing initial slides. A total list of potential working group items was displayed. * General agreement that the use cases have been agreed upon and that there was no need to go through them in detail in the context of the working group meeting. * Subset of potential working group items put forward as initial working group items. Request to the working group to provide comments, suggestion proposals, or changes for presented milestones. Questions (Q), Answers (A), and Comments (C): ------------------------ Q: Kevin Fall asked if the proposed work items culminate in some related consensus documentation. For example, hop count was listed under a revised 5050, but what if it is determined that hop count is not a good idea. A: Marc Blanchet noted that the list of sub-bulleted items represent things that should be considered. C: Rick Taylor (Airbus) commented that there should be a general intention to keep documents ?short and sweet?, rather than putting too many items into a revised 5050 and creating monolithic tomes. This was generally agreed as appropriate. C: Scott Burleigh (JPL) commented on his support for the proposed work items. Q: Nabil Benamar asked about static routing. Should this working group tackle routing protocols used in DTN and different environments, such as vehicular ad-hoc networks? A: People interested in this, but how will we do it. Many people not sure yet on routing specifics. We could pick up some of this with the DTNRG, so we have a path to take work out of the DTNRG and bring it forward. Q: Bob Cole (Us Army) asked about the item for convergence layer adapter requirements. He sees two scenarios being deploys: space and terrestrial networks. Is there a bigger requirements discussion on the tradeoff between large/small bundles, error correction, converge on noise environment versus long-distance high delay, and how do we make those trades? A: Brian Haberman (APL): These requirements are not meant to cover those cases. This work item describes the minimum set of things that the Bundle Protocol would expect to have from a convergence layer. A: Scott Burleigh: Specific requirements to motivate the selection of one CL protocol over another are interesting topics, but different from this work item. C: Rick Taylor offered support for the proposed set of working group items. Q: Rick Taylor asked what is the next step to starting work? A: Identify milestones and volunteer for work items. C: Rick Taylor: Would like to do enhanced neighbor discovery. C: Fred Templin: Would like to volunteer for address architecture work. C: Nabil Benamar: Would like to work on static routing. C: John (Airbus): Support this working group list and would like to work on static routing. C: Scott Burleigh: Happily volunteer on revision to RFC5050. C: Ronald In 'T Velt: Volunteer to work on revision to RFC5050. C: Kevin Fall: Volunteer to work on revision to RFC5050. Also, there is a legacy thing in enhanced neighbor discovery that is worth discussing. Will send out information. C: Fred Templin: Volunteer to work on investigating security key management, with analysis and design. Q: Will Ivancic (remote) : What is the difference between node neighbor discovery and enhanced neighbor discovery? A: Enhanced provides other options, such as offerings from Manet, see whether they provide value in this context. C: Rick Taylor. There is no dependency graph yet. A revision to RFC5050 is important piece of work. Need to find a way to understand the dependencies. Some working items must make progress faster than others. C: Scott Burleigh: Along those lines, if it is envisioned to have a document for addressing architecture, possible that might be developed before a revision to RFC5050. Volunteering for that one as well, it is very important. C: Marc B. will take on registry of service identifiers. C: Marc B. noted that there were almost two editors per document and no objections to the proposed set of work. Q: Rick Taylor: Question for the room. The intention was to discuss network management with the OPS area. Are people happy with that approach and saying we will not work on it in this cycle? A: Marc B: A network without management doesn?t make sense, so it must eventually be addressed, but we don?t know how to put it all together just yet. A: Brian H: To elaborate on an earlier discussion, on first blush, the existing IETF model of NM as SNMP plus Netconf using YANG doesn?t seem to be a good fit for this. To make that point clear, one of the things is to have a discussion within the ops area. Q: Rick Taylor: Do we need a position paper on network management for DTN? Something more formal than chats with Ads. Happy to have a network management discussion on the DTNWG wiki. Agreed this doesn?t look like SNMP or netconf. A: Ed BIrrane: Happy to pull some of this initial work together. Q: Bob Cole: Do we have a timeframe for accomplishing this work? Is this a 2 year effort? A: Brian H: Believes this work should be accomplished within 18 months. A: Marc B: At the end, a graph of dependencies between documents. We expect to work them in parallel. To finish in this decade. Q: Rick Taylor: Did anyone put hand up for security? A: Ed Birrane did. SBSP Presentation ------------------------ Ed Birrane presented current status of SBSP and the underlying security model for DTN. Question asked on how to we proceed with SBSP? Consensus answer seemed to be to say that SBSP is the security protocol for a revised RFC5050. Closing Documents ------------------------ Brian Haberman: Way ahead, for people who volunteered, say how long it will take to write such a document, consider what your dependencies would be, and then Marc and Brian can take a cut at a more global list of deadlines that we can put into milestones and bring it back to a list discussion on the working group.