IETF 88 Vancouver, Canada Minutes of the Geonet BoF Nov. 5 2013 1:00 to 2:00 PM pacific time Note taker: Mike Kallas BoF Chairs : Melinda Shore, Alexandru Petrescu. Meetecho log: http://ietf88.conf.meetecho.com/index.php/Recorded_Sessions#GEONET Jabber log: http://www.ietf.org/jabber/logs/geonet/2013-11-05.html ------------------------------------------------------------------ Agenda shown Note Well slide displayed to room. Margaret could not make it Agenda shown Webex is up Alex: Introductions. Internet wide geo networking Explained other related work by other groups Disseminating IP packets to geographical areas. Or receiving IP packets from a geographical area. Problem statement draft presented by Georgios Karagiannis Georgios: Challenges presented in the problem statement is not addressed by IETF or any other bodies Lost protocol: Location to service mapping Two solutions presented have actual implementations Suggestion: List all the related working groups Answer: no working group working on the exact same problem Question: both use cases similar to ETSI scope. I don't see a wider coverage than the ad hoc implementation. Why not use ETSI? Answer: short range only. Source has to be close to the area. This is internet wide solution: provide that ability to any source on the internet to send information to geographic areas. Question: Bob (Verizon, ex-Chrysler) this is how things work right. What about when things go wrong? What about malicious events. Like to see attack review, failure nodes scenarios. Answer: Security and privacy are very important. We have to focus on it. Next on the Agenda: implementation Interest: presentation. Last topic: IPv6 over 802.11p. Presentation End of presentation part. Interest, clarifications. Question: problem not well defined. Is it: enumerate IP addresses that are currently in a geographical area to send to. Answer: its is one of the potential solutions. Answer: need to better define the problem statement. Need to learn geographic proximity of certain devices and leave it at a high level (not just IP) other options are possible. Answer: looking at a relationship between IP addresses and locations of Access Routers. Question: requirements around trust model on the geographic database? Are the set of networks all part of the same organization or people who don't trust each other? Multiple network providers? Could affect trust model. Answer: both are relevant. Security is very important. Trust with respect to positioning is different from security as defined in IETF. Precision is also a factor. Trust of position: precision should be part of the discussion. Main concern not who owns IP but where it is. Comment: trust model and the uncertainty of locations. Question: have you looked at ILNP? Answer: no, but will. Comment: see the document referred to on jabber (draft-hain-ipv6-pi-addr-10). Question: uniqueness vs non uniqueness and indirection in addressing. For example who is moving and who is not. ILNP tries to do indirection. Other issues are privacy: present information without identifying person. You may not want symmetric behavior V2x different mechanism from x2V. Answer: thank you. Agree on non symmetry. Need to take ILNP into account in addition to others. DNS, LOST Question: Guarantees. Are you only sending to the area. Is it ok it reaches outside the area? Are you expecting error messages if the message does not reach? Answer: Reliability of delivery important, reliability. Question: one of the use cases was querying? How do you identify what nodes you want to query? Answer: example fire detectors, temperature, wind Question: translation between geo and IP very important. What about translating from IP address to geo? May be nice to have the solution for both. Answer: yes Question: IEEE 1609 does not make the documents available. We need to make them available for this work here. Because it is related. Answer: this is an issue and we need help with the liason with IEEE. Liaison: we will address it. Comment: HIP MAP server. Question: skeptical about problem. Geographical routing vs optimal routing. Answer: an optimal IP routing path may be different from optimal geographic 'routes'. Optimal geographical route may not be possible, whereas the other can, and vice-versa. Stop discussion: need to figure out what will happen next How many people read the problem statement: only a few (7) How many saw the charter: a few Is there work appropriate to the ietf: a few, some reluctance Not appropriate for IETF: one person only (the reason has to do with explicit reference to DNS). Scope seems to still be too broad. Is it ready? Who is interested in working on this? Several people (20) There is a mailing list including discussing scope http://ietf.org/mailman/listinfo/its Question: there is interest but are we ready to say go? We should not drop it. Is the next step another BOF? next BoF - another BoF? Interesting in a broad sense but the scope is too wide. Need to narrow scope. Is there application level stuff needed. Who is capable of handling the received data. Ted Lemon (Area Director in charge of this BoF): have heard there is interest, haven't heard there is a go, but there honestly this is not a drop issue, next step is maybe another bof. Last word by Eliot Lear (IAB member following this BoF): thanks to BOF chairs. Couple of very interesting problems: actions to move it forward: Crisp problem statement, a little of solutioneering, iteration on problems and solutions need to be gone through. More detailed problem statement more specifity. Look for any existing work. And help limit the scope of things needed to be done.