IP over IEEE 802.16 Networks Working Group (16ng) Minutes Meeting : IETF72, Wednesday July 30 2008. Location: Citywest hotel, Convention 3, 15:10-16:10 Afternoon Session II. Chairs : Gabriel Montenegro , Soohong Daniel Park . Minutes : Ji Hoon Lee ========================================================= AGENDA A. Agenda bashing B. IP transmission over Ethernet CS C. IP transmission over IPv4CS D. Closing ========================================================== Meeting Report A. Agenda bashing - Chairs we currently have 3 RFCs and 2 working group drafts: - RFC4968, - RFC5121, - RFC5154, - draft-ietf-16ng-ip-over-ethernet-over-802.16-06 AD Evaluation, - draft-ietf-16ng-ip-v4over-802-dot-16-ipcs-03 Active. B. IP transmission over Ethernet CS (25 mins) - Sangjin Jeong draft-ietf-16ng-ip-over-ethernet-over-802.16-06 status: AD review material: http://www3.ietf.org/proceedings/08jul/slides/16ng-1.ppt Jari Arkko said that the authors should clearly declare that this spec has no effects on the hosts and the other parts of a network, if Ethernet over 802.16 behaves like standard Ethernet. More discussion is needed for the bridging behavior (comment#3). David Johnston indicated that there is no document addressing GPCS in IEEE 802.16, yet. Jari Arkko suggested that the authors could add a referece of Ethernet CS from IEEE 802.16, and then we could easily make an amendment to the RFC in the future, since it's just a reference problem. More work is needed to specify flooding behavior (the second, comment#11). Jari Arkko said that he thought that a reasonable way was without requiring one box approach. Jari Arkko proposed getting some technical comments from DSL guys. Gabriel Montenegro also suggested that the authors clarify the difference of wimax from DSL, and some reasons that make such a difference. C. IP transmission over IPv4CS (25 mins) - Samita Chakrabarti draft-ietf-16ng-ipv4-over-802-dot-16-ipcs-03 status: comment resolution after WGLC material: http://www3.ietf.org/proceedings/08jul/slides/16ng-2.ppt Peter Lothberg argued about why default MTU is not 1500, but 1400 for IPv4CS (actually this arguement has been discussed many times). There were some comments on this: - Gabriel Montenegro said that wimax forum has already decided to use 1400. Whatever we do here, such problems that are caused by MTU 1400 can still exist. - Jari Arkko said that this is only temporary situation and the wimax forum will reference this RFC. - David Johnston recalled that the maximum size of 802.16 PDU is 2048, which is not related to the SDU size. And he also said that default MTU 1400 is reasonable since wimax forum analyzed newtorks considering a number of deployment scenarios and decided default MTU to be 1400. - Jari Arkko said that the authors can use 1500 instead of 1400 if they think 1400 is bad, even though wimax forum has decided to use 1400 in the past. Jari Arkko indicated that DHCP MTU option can only be helpful if the host is an actual end host. If it is a router, an actual host behind that router may suffer fragmentation over wireless. D. Closing - Chairs