A few clarifications, most of which are probably courtesy of the difficulty in taking notes on a free flowing discussion:
OLD GUE Encap (Tom Herbert): ------------------------ Debate begins on checksums David Black: Update on checksum for IPv6: The checksum should be MUST USE FULL STOP for IPv6 based on data we have seen in testing. NEW GUE Encap (Tom Herbert): ------------------------ Debate begins on checksums David Black: Update on UDP checksum for IPv6: The checksum should be MUST USE FULL STOP for IPv6. This is for generality based on design team experience in specifying conditions under which zero UDP checksum is ok to use with IPv6 for MPLS/UDP and GRE/UDP in compliance with RFC 6935 and RFC 6936. OLD VxLAN QOS (Frank Xia): ----------------------------------- David Black: RFC 2474 - Penalty Flag Flown: respecified that the 6bits must be dealt with as a whole, not messed around with using 3+3 type stuff. NEW VxLAN QOS (Behcet): ----------------------------------- David Black: RFC 2474 - Penalty Flag Flown: This 3+3 bit messing around is not DiffServ and is not compatible with RFC 2474. The entire 6 bit field must be dealt with as a whole. Thanks, --David From: nvo3 [mailto:[email protected]] On Behalf Of Sam Aldrin Sent: Monday, November 24, 2014 12:53 PM To: [email protected] Cc: Benson Schliesser; Bocci, Matthew (Matthew) Subject: [nvo3] IETF91 NVo3 Meeting minutes Hi, IETF91 meeting minutes have now been uploaded. Thanks to Jon Hudson for taking minutes. Really appreciate that. Do find them at <http://www.ietf.org/proceedings/91/minutes/minutes-91-nvo3> Please do send in any missing minutes/notes or any corrections needed to be made. Would be glad to entertain them. cheers -sam
_______________________________________________ nvo3 mailing list [email protected] https://www.ietf.org/mailman/listinfo/nvo3
