[14:51:23] jishac joins the room [14:52:25] jishac has set the subject to: Delay-Tolerant Networking Research Group [14:57:25] MarkFeldman joins the room [14:59:01] jishac is now known as jishac (MOD) [14:59:01] jishac (MOD) is now known as jishac [15:01:53] jishac is now known as jishac (MOD) [15:01:53] jishac (MOD) is now known as jishac [15:03:14] sftcd joins the room [15:06:54] jishac is now known as jishac (MOD) [15:06:54] jishac (MOD) is now known as jishac [15:07:14] samo.grasic joins the room [15:11:54] jishac is now known as jishac (MOD) [15:11:54] jishac (MOD) is now known as jishac [15:13:08] Quick discussion about the prolific nature of the DTNRG. [15:14:03] Aaron Falk, IRTF chair, asked if the plethora of documents are destined to be RFCs. Many are on that track. [15:15:05] Currently going down list of drafts and their status, specifically older/expired documents. [15:15:34] Authors are stepping up to the mike and describing their status/applicability. [15:16:55] jishac is now known as jishac (MOD) [15:16:55] jishac (MOD) is now known as jishac [15:17:18] Kevin Fall and Stephen Farrell, DTNRG Co-Chairs, are providing context. [15:19:49] Discussing process for moving drafts from "done" to RFCdom. Steve Crocker recommends taking notes along the way so that the new process can be better understood and improved. [15:20:21] [It would help is speakers identified themselves at the start of their comments] [15:21:56] jishac is now known as jishac (MOD) [15:21:56] jishac (MOD) is now known as jishac [15:22:16] Aaron Falk is now discussing the shepherding of documents. Delays often caused by dropped tokens and convergence after reviews. Author and shepherd must work together. Responsibility is shepherd's. Poke your shepherd if things seem to slow. [15:22:45] Now discussing documents for completion/"last call." [15:23:48] And what "last call" means in the ITRF/DTNRG context. [15:24:27] Wesley Eddy & Aaon Falk are asking clarifying questions. [15:26:07] Andrew Sullivan joins the room [15:26:09] Susan Symmington wants to see the previous hotblock. [15:26:17] Steve Crocker joins the room [15:26:45] Scott Burleigh identified a missing document for completion. [15:26:56] jishac is now known as jishac (MOD) [15:28:50] Avri Doria says decision on prophet doc will be made in 15 minutes. [15:29:28] Susan Symmington sats Bundle Security will be done end of CY08. [15:30:12] Lloyd Wood sats Bundle-checksum needs some reworking. Will discuss reasons later. [15:31:28] Remaining documents ready now (Sdnv, Previous-hop block, Compressed bundle header encoding). [15:31:39] Prophet will be ready end of CY08. [15:31:57] jishac (MOD) is now known as jishac [15:32:52] Elwyn Davis asks about a wiki page for tracking documents. Aaron Falk says we can use IRTF document tracker. We may need to add states. [15:33:34] "Any News?" slide up. No takers. [15:33:40] Andrew Sullivan leaves the room [15:33:46] Andrew Sullivan joins the room [15:35:31] Lloyd Wood giving talk named "Reliability, further points for discussion." Lloyd says presentation shows that bundle checksum needs to be reworked to cover all bases. [15:36:20] ... Comparing bundling to legacy (IPv4/IPv6) networking. [15:37:06] jishac is now known as jishac (MOD) [15:37:44] Slides available at http://www3.ietf.org/proceedings/08nov/slides/DTNRG-1.ppt [15:38:14] Overall agenda, including links to submitted presentations, at http://www.ietf.org/proceedings/08nov/agenda/DTNRG.html [15:38:38] avri joins the room [15:38:47] Steve Crocker leaves the room [15:38:49] Steve Crocker joins the room [15:41:48] PCB -- payload confidentiality block. Quick side conversation discussing appropriate terminology for cryptographic-based security services. [15:42:02] jishac (MOD) is now known as jishac [15:45:25] Q&A for the presentation. [15:45:31] Andrew Sullivan leaves the room [15:45:37] Andrew Sullivan joins the room [15:46:56] Steve Crocker leaves the room [15:46:57] Steve Crocker joins the room [15:47:03] jishac is now known as jishac (MOD) [15:47:03] jishac (MOD) is now known as jishac [15:47:54] Question is whether specification allows for multiple security suites. PIB requires nodes along the way to have shared key to verify integrity. Checksums or signatures on the encrypted bundle should be allowed to permit checking by parties along the way, shortening the control loop in case of failure. [15:50:06] Steve Crocker asks about protection against various modes of failure (natural vs. malicious). Any of the described techniques work against natural failure. Only public key-base digital signature can protect against malicious nodes, including nodes having a shared secret. [15:51:17] Lloyd's slides are primarily concerned with performance with respect to natural faults. An external checksum is sufficient for this. [15:51:36] mark - thanks for taking such good notes! [15:52:03] jishac is now known as jishac (MOD) [15:52:04] jishac (MOD) is now known as jishac [15:52:26] Kevin Fall pints out that DTN can use FEC and therefor retransmission is not the only form of reliability. [15:52:35] if anyone's on the audio let us know if Lloyd's audible [15:53:20] Andrew Sullivan leaves the room [15:53:26] Andrew Sullivan joins the room [15:53:35] samo.grasic leaves the room [15:54:07] Steve Crocker leaves the room [15:54:10] Steve Crocker joins the room [15:55:26] Discussing convergence layer support for reliability. [15:56:39] Wes gives a pointer to Jim Wylie's badisc document (I think) [15:56:50] Convergence layers have different capabilities/properties with respect to bundle size, reliability, etc. [15:57:05] jishac is now known as jishac (MOD) [15:57:05] jishac (MOD) is now known as jishac [15:57:29] Steve Crocker leaves the room [15:57:34] Steve Crocker joins the room [15:58:02] Steve Crocker leaves the room [15:58:05] Steve Crocker joins the room [15:59:00] Hans Krusse & Kevin Fall discussing options. [16:00:02] Multiple parties discussing requirement for UDP convergence layer. [16:01:02] Steve Crocker leaves the room [16:01:07] Steve Crocker joins the room [16:01:15] draft-irtf-dtnrg-udp-clayer-00 is new and should be read. [16:01:17] http://tools.ietf.org/html/draft-irtf-dtnrg-udp-clayer-00.html [16:02:06] jishac is now known as jishac (MOD) [16:02:31] Thumbs up from Christopher Small and others for the orbital internet work and Time article. [16:03:31] Steve Crocker leaves the room [16:03:32] Steve Crocker joins the room [16:03:54] Discussing the non-mandatory (optional?) nature of reliability, security, and other features since there are many alternatives in DTN. [16:04:27] marc.blanchet.qc joins the room [16:04:53] Now on to Hans Kruse's first presentation, http://www.ietf.org/proceedings/08nov/slides/DTNRG-2.pdf [16:06:09] Steve Crocker leaves the room [16:06:10] Steve Crocker joins the room [16:07:07] Hans noted that the ION TCP CL is not the same as the I-D [16:07:08] jishac (MOD) is now known as jishac [16:07:08] jishac is now known as jishac (MOD) [16:07:50] Single slide presentation. For the powerpoint-challenged: dtnbone Test Systems at OU • Public facing servers, network management • ION • ION end point • Bundle receipt and Debug Logs via Web Page • FireDTN • ION Firefox plugin, ION web proxy • ION Bundle Relay Server • Wiki to manage end-point-identifiers • DTN2? • Caveat: Waiting to complete NASA Grant Agreement • Mailing List: http://korgano.eecs.ohiou.edu/mailman/listinfo/dtnbone [16:11:25] (slide is PDF, not powerpoint) [16:12:06] jishac (MOD) is now known as jishac [16:12:40] Need to address routing issues. Should we have a dtn-deploy mailing list for those wishing to run DTN? [16:13:19] Hans is bringing some networking to the table. Would like too connect to others. [16:14:24] Steve Crocker leaves the room [16:14:24] Steve Crocker joins the room [16:16:30] Approximately 10 folks running DTN nodes. Same group is interested in connecting to each other over the Internet. [16:16:43] I think that was more than 10:-) [16:17:05] The more the better. [16:17:06] jishac is now known as jishac (MOD) [16:18:05] Need to have some applications other than just ping. [16:18:15] Steve Crocker leaves the room [16:18:16] Steve Crocker joins the room [16:19:06] Aaron discussing additional benefits of such testing: Use of multiple, interesting networks and not just multiple applications. [16:19:20] Steve Crocker leaves the room [16:19:25] Steve Crocker joins the room [16:20:05] Steve Crocker leaves the room [16:20:05] Steve Crocker joins the room [16:20:14] Kevin Fall discussing DTN on test/emulation networks (e.g., planetlab) and examples of interesting higher-level applications/protocols (voip). [16:20:38] Avri discussing connecting European networks to others. [16:22:05] jishac (MOD) is now known as jishac [16:22:05] jishac is now known as jishac (MOD) [16:22:30] Chairs asking about goal for DTNBONE for March meeting and even for end of CY08. [16:22:41] Six sites is being discussed. [seems low to me] [16:23:08] Multiple implementations, convergence layers, etc. [16:24:12] Now on to Hans' second presentation: http://www.ietf.org/proceedings/08nov/slides/DTNRG-3.pdf [16:24:53] If anyone is remote and would like to say something in the mic just prepend MIC: to your message and I'll relay it [16:25:17] Also a single-slide PDF: UDP Convergence Layer LTP and Bundle Protocol • -00 draft to be formally posted right after IEFT73 • Questions: • Bundle Protocol over UDP, why? • Does native UDP as a convergence layer meet "pseudo-reliable" assumption? • Ack & Retransmission? • Fragmentation? • UDP • Checksum could fail, security extensions? • Congestion Control [16:26:42] Andrew Sullivan leaves the room [16:26:48] Andrew Sullivan joins the room [16:27:08] jishac (MOD) is now known as jishac [16:27:32] fyi, great presentation on transport protocols this week by Lars and Magnus: [16:27:33] http://people.nokia.net/~lars/talks/ietf73-wgchairs-training.pdf [http://people.nokia.net/~lars/talks/ietf73-wgchairs-training.pdf] [16:27:35] Elena joins the room [16:28:07] Bob Cole commented on use of UDP. Missed the comment:-( [16:28:56] Cole said that you might want a UDP CL for cases where the receiver wants to not acknowledge the packet [16:29:15] Kevin then added that some multipath routing things might also etc. [16:29:47] Anonymity. [16:30:18] Long line at the mike for UDP CL layer. [16:31:00] Why use UDP and not just IP? [UDP & TCP are understood by existing firewall/nat boxes] [16:32:07] Scott Burleigh would like a hint mechanism in the bundle layer that can help make CL decisions. [16:32:09] jishac is now known as jishac (MOD) [16:32:09] jishac (MOD) is now known as jishac [16:33:09] Andrew Sullivan leaves the room [16:33:15] Andrew Sullivan joins the room [16:33:52] Lloyd Wood: Turning off UDP checksum would be unwise. [16:34:58] Underlying reliability should not be imposed. DTN nodes may have the best understanding of link issues. [16:35:45] udp checksum is mandatory in ipv6... [16:36:01] Joerg Ott: What errors are being discussed? Bit errors? UDP Lite is an alternative. [16:36:11] samo.grasic joins the room [16:36:49] Elena leaves the room: Replaced by new connection [16:36:49] Elena joins the room [16:36:49] UDP-Lite is RFC3828. http://www.ietf.org/rfc/rfc3828.txt Not supported by all nodes. [16:37:10] jishac is now known as jishac (MOD) [16:39:03] Joerg Ott: Congestion control. Just shut up if there's an end-to-end UDP issue? [16:39:25] DCCP is being discussed. [16:39:40] Elena leaves the room: Replaced by new connection [16:39:40] Elena joins the room [16:40:48] Unidirectional links require UDP. DCCP was designed for the terrestrial Internet. [16:41:16] UDP/UDP-Lite/DCCP et al were considered in saratoga. [16:42:03] Now on to Scott Burleigh's presentation. Not on web site. "Space-based DTN." [16:42:11] jishac (MOD) is now known as jishac [16:43:19] Actually that's Will Ivancic - Kevin put up the wrong slides (my fault though) [16:44:00] Thanks. Still learning names and can't read badges at a distance. [16:44:33] Slides are now up at: http://www.ietf.org/proceedings/08nov/slides/DTNRG-5.ppt [16:44:58] Steve Crocker leaves the room [16:45:00] Steve Crocker joins the room [16:45:00] Elena leaves the room: Replaced by new connection [16:45:01] Elena joins the room [16:46:05] Steve Crocker leaves the room [16:46:06] Steve Crocker joins the room [16:46:34] jishac is now known as jishac (MOD) [16:46:43] Steve Crocker leaves the room [16:46:43] Steve Crocker joins the room [16:47:02] Steve Crocker leaves the room [16:47:14] jishac (MOD) is now known as jishac [16:47:14] jishac is now known as jishac (MOD) [16:48:36] Steve Crocker joins the room [16:48:48] Steve Crocker leaves the room [16:48:49] Steve Crocker joins the room [16:49:27] Steve Crocker leaves the room [16:49:43] Steve Crocker joins the room [16:49:43] Discussion with the chairs on addressing. Space-based addressing may be optimized differently than terrestrial addressing. DTN supports multiple, simultaneous forms of addressing, so a single format is not required. [16:50:41] Steve Crocker leaves the room [16:50:41] Steve Crocker joins the room [16:52:15] jishac (MOD) is now known as jishac [16:53:13] Now on to Avri Doria's presentation: http://www.ietf.org/proceedings/08nov/slides/DTNRG-6.ppt [16:53:31] rvdp joins the room [16:53:54] "Update on Prophet" [16:54:08] Almost [16:54:10] read [16:54:11] y [16:57:14] jishac is now known as jishac (MOD) [16:57:14] jishac (MOD) is now known as jishac [16:59:06] Now on to Bundle encapsulation discussion with Susan Symmington. No slides. [16:59:48] There was a telecon about encapsulation and this talk is to make what was discussed available to all. [17:00:55] Example is taking a cached multicast bundle and embedding it as a unicast bundle so it can be sent to a new member without sending to the entire group. [17:01:06] Another example is tunneling. [17:02:02] Encapsulation is an application process and doesn't involve the bundle protocol directly. [17:02:14] jishac is now known as jishac (MOD) [17:03:35] The embedded bundle is merely the payload. The bundle protocol needs a way to divert bundles to be embedded and we need to be able to inject de-embedded bundles back into the node as if it was just received. [17:05:21] Kevin Fall is asking about perceived problem with using multicast with new members (one of Susan's examples of the need for encapsulation). [17:06:14] Need an ID for encapsulation that is an application protocol and need changes to the bundle protocol that allows for diversion and injection of bundles. [17:07:16] jishac (MOD) is now known as jishac [17:07:45] Steve Crocker is asking about multicast assumptions: Suppose all attempts at multicast as partially successful. Assume that at any point a certain set of members are unavailable. [17:08:57] Doug Otis joins the room [17:09:39] Avri Doria suggests that routing above the bundle layer may be a general requirement in nodes along the path and not just at end nodes for encapsulation/decapsulation. [17:10:04] Scott Burleigh asks if encapsulation could be a CL. [17:10:38] Scott Burleigh is now making his presentation: http://www3.ietf.org/proceedings/08nov/slides/DTNRG-4.ppt [17:12:16] jishac is now known as jishac (MOD) [17:12:16] jishac (MOD) is now known as jishac [17:13:27] rvdp leaves the room [17:14:49] This is the really cool deep impact stuff [17:17:15] jishac is now known as jishac (MOD) [17:17:15] jishac (MOD) is now known as jishac [17:17:50] Elena leaves the room: Replaced by new connection [17:17:50] Elena joins the room [17:22:17] jishac is now known as jishac (MOD) [17:23:07] Elena leaves the room: Replaced by new connection [17:23:07] Elena joins the room [17:24:39] Andrew Sullivan leaves the room [17:24:46] Andrew Sullivan joins the room [17:27:19] jishac (MOD) is now known as jishac [17:27:19] jishac is now known as jishac (MOD) [17:28:31] Elena leaves the room: Replaced by new connection [17:28:31] Elena joins the room [17:28:51] Elena leaves the room [17:28:55] Malcolm Airst wants to know about bandwidith studies based on stacks & options. Scott says bundle & LTP overhead is in the noise compared to the FEC in DINET. [17:30:23] Malcom is working on the next generation timing spec and wants to understand/include DTN requirements for timing. Scott says sync was 5-10 seconds and worked. [17:30:46] DTN requirements will vary depending on application. [17:31:27] Andrew Sullivan leaves the room [17:31:33] Andrew Sullivan joins the room [17:31:36] DINET gets multiple awards: longest distance, worst routing:-) [17:32:19] jishac (MOD) is now known as jishac [17:33:12] marc.blanchet.qc leaves the room [17:34:13] Scott responds to question from Will Ivancic about limiting changes to contact plan to help ensure success. Dynamic routing vs. management based on origin of routing information. [17:34:33] Meeting time has run out. [17:34:46] Additional questions will be offline. [17:34:46] avri leaves the room [17:34:57] Andrew Sullivan leaves the room [17:35:05] Meeting adjourned at 11:35 CST. [17:35:07] MarkFeldman leaves the room [17:35:46] jishac leaves the room [17:35:47] jishac (MOD) joins the room [17:36:11] The sound level is a bit low. [17:36:41] Steve Crocker leaves the room [17:37:53] jishac (MOD) leaves the room: Computer went to sleep [17:41:43] samo.grasic leaves the room [17:42:47] samo.grasic joins the room [17:52:45] sftcd leaves the room [17:53:52] Doug Otis leaves the room [18:01:42] samo.grasic leaves the room [18:02:12] samo.grasic joins the room [18:02:50] samo.grasic leaves the room [18:23:44] sftcd joins the room [18:26:56] sftcd leaves the room [19:04:26] samo.grasic joins the room [19:39:17] samo.grasic leaves the room