PPP Extensions Working Group Minutes 45th IETF-Oslo, Norway Wednesday, July 14, 1999, 1530-1730 Chair: Karl Fox karl@extant.net Reported by Matt Holdrege matt@ascend.com Discussion of PPPEXT Standards Track Document Status Karl Fox Karl explained the rules for moving documents along the standards track. Then we reviewed each protocol. To PS L2TP L2TP MIB AODI To DS IPCP BCP CCP ECP PPP over ISDN BACP LCP Internationalization LCP Extensions X.25 Frame Relay AAL5 Appletalk CP To STD CHAP MP LQM To HISTORIC DNCP SNA Reliable Mode XNS First phase is taking L2TP MIB and AODI to proposed standard. Second phase is taking IPCP, IPXCP, BCP, VJ, CCP, ISDN, X.25 and BACP to Draft Standard. CHAP and MP and LQM to full standard. SDP will be punted to the IPSEC working group. Future Directions of L2TP draft-ietf-pppext-l2tp-16.txt Mark Townsley Draft 16 was approved as a proposed standard but is still in the editors queue. Clerical errors are still being accepted. The RFC number should be published in about a month. For promotion to draft standard, we must prove multiple interoperable implementations. We can also identify items that might need to be changed or added. It was noted that it is not in anyone's best interest that we rush to draft standard status. Realistically, if there are changes, it will take a year to complete testing before we move to draft standard. Should we have a separate UDP port for control and data channels? It would require a new port registered with IANA. All sessions within a tunnel would use the same ports. Alternatively, we could use ADC which is more complex, but individual sessions within a tunnel may use different ports. Flow control issues: Limiting number of packet in transit to help with statefull protocols. Congestion Management BOF draft-balakrishnan-cm-00.txt may be useful for the L2TP control channel, but is not advisable for the data channel. What to do if an AVP that is defined for one message is received in another? The same AVP may have a different meaning in two different messages. Thus the M-bit rules apply. If you want to do this, an extensions draft must be written and approved as if it were a brand new AVP. PPP over Simple Data Link (SDL) using raw lightwave channels with ATM-like framing draft-ietf-pppext-sdl-pol-00.txt James Carlson Enrique Hernandez-Valencia Nevin Jones Concerns were expressed about the CRC since it wasn't specified in the draft. It was explained that the existing SDL draft made it clear. It was expressed that existing PPP over SONET was more efficient than SDL at the existing Internet packet size samples. Also that that SDL was addressing an error scenario that doesn't impact users. Another concern was expressed about the lightwave transport and that this draft doesn't address it. We could not determine the status of this document here and that discussion will move to the list. Mobile PPP draft-ietf-pppext-mppp-00.txt M. Chuah This draft was presented last meeting, but its purpose wasn't explained well. The draft was changed in respect to L2TP flow control and is dependent on future changes to L2TP flow control. L2TP works fine in a mobile network EXCEPT when moving from one Radio node to another. This draft adds a new AVP to L2TP to enable this function. This Mobile AVP is an option for LNS's that wish to support this feature. If user auth is required by an operator upon handoff, this new draft supports it. This draft is also consistent with Mobile IP as used in wireless networks. It was pointed out that ACCM should be in the SLI rather than the SRRQ. The author will consider that change. L2TP MIB draft-ietf-pppext-l2tp-mib-04.txt Evan Caves The big change in this draft is about payload flow control being removed. There will be another draft to make a few minor corrections. The IP tunnel MIB will be moving to proposed standard soon which is good since the L2TP MIB is an extension of this draft. L2TP ATM Access Network Extensions draft-ietf-pppext-l2tp-atmext-01.txt Yves T'Joens Draft 01 adds a few new extensions and will be updated now that L2TP has been approved and is stable. Always On Dynamic ISDN (AODI) draft-ietf-pppext-aodi-01.txt Matt Holdrege [5 minutes] Not many updates; mostly small issues. New draft went to the list. One comment: PPP over X.25 never included the D-channel consideration. Trying to move this forward quickly. Providing QoS in Multilink PPP M. Chuah Enrique Hernandez-Valencia MCML only specifies how to interleave packets on multiple links. But how to schedule interleaving events. Did not mention how to map PPP QoS to different links carrying traffic with varying characteristics. PPP over Virtually Concatenated SONET/SDH Low and High Order Channels Nevin Jones The draft will be submitted in the near future. The authors wish to communicate to the transport standards bodies that the IETF sees this as desirable so that they can see the need for adding support for combining lower and higher order channels in their specifications. Irfan Ali fia225@email.mot.com presented draft-pazhyannur-avt-pppmux-00.txt It was noted that AAL2 muxing went through the same issues and it found a lot of complexity. But this draft tries to avoid those problems with a very simple approach. Karl Fox Key fingerprint = 5B15 7260 D55E D680 0B93 4953 8A3B AB0E C05D 77A3