NTP WG Meeting Minutes 13:00 - 15:00, Wednesday March 12, 2008 The meeting was chaired by Karen O'Donoghue. Malcolm Airst took the minutes and Yaakov Stein acted as Jabber scribe. There were no changes to the agenda as presented. NTPv4 Document (Jim Martin) ============== Document: draft-ietf-ntp-ntpv4-proto-09.txt. Slides: ntp-2.pdf Jim has produced another revision of this document in response to Vancouver discussions. The following significant changes were made: Stewart/Yakov/Greg’s text for timestamp location text was added. It was proposed to add a paragraph to the text. After some discussion, Karen suggested removing the timestamp material from this draft and producing a small informational RFC. Mark Townsley pointed out that the TICTOC WG could produce this RFC. Yaakov Stein asked why this would be an informational versus standards track RFC. Also, if this material is for NTPv4, why separate it? Removing this note would allow us to move forward since this paragraph is not needed for NTPv4. The Chair asked for clarification: Can we do NTPv4 maintenance items in TICTOC? There’s a gap between the NTP WG Charter and the TICTOC Charter. There was broad agreement to pull the text referenced above from the draft. Clarifying text for the KOD Packet Timestamps was added. Mutiple editorial tweaks were performed. There is one open item remaining. There are references to the appendix that are arguably normative. To ensure clarity and cohesiveness, text will be pulled into the main document. There are approximately 10-20 references. It is believed that there is an IETF requirement to avoid normative references in an appendix. In other words, proper operation of the protocol is not dependent on the Appendix. Chair asked and obtained a few volunteers to help. The draft would go to WGLC again when the Appendix text is adjusted. NTP Autokey Specification (Brian Haberman via Karen O’Donoghue) ========================= Document: draft-ietf-ntp-autokey-01 Slides: ntp-0.pdf The -01 update was published. This update is pending security advisor review. Updated document as autokey exists today and tightened up text. Greg Dowd recommended the following wording: “Every NTP association has two variables which maintain the liveness state of the protocol, the 8-bit reachability register defined in [7] and the watchdog timer.” The reason for this text was that although the Autokey doc is only informative, there is no definition of a watchdog in NTP. Therefore, stating that this is a new feature of NTPv4 means that we either have to redefine NTPv4 to include this or, more simply, just remove the offending text. Also, we need to add acknowledgements and IANA considerations. Hold the next revision pending security review. NTPv4 MIB (Chris Elliott) ========= Document: draft-ietf-ntp-ntpv4-mib-04.txt. Slides: ntp-0.pdf Chris Elliott was unable to attend the meeting. The only technical change from the last discussion is the addition of statistics table on a per-packet mode. This document is ready for working group last call. NTP server option for DHCPv6 (Benoit Lourdelet, Richard Gayraud) ============================ Document: draft-gayraud-dhcpv6-ntp-opt-01.txt. Slides: ntp-1.pdf Network operators like to centrally configure IP services through DHCP. DHCPv6 options are added to offer NTPv4 client configuration. The requirements include the ability to advertise NTP server locations and address both public and private network requirements. Draft -01 adds FQDN with an applicability statement, clarifies the SNTP issues, and removes all parameters except NTP server location. FQDN can specify NTP server location in addition to IP address. This is seen as an efficient tool to prevent or mitigate inadvertent flooding. After some discussion with the AD, it was agreed that the authors would update the draft and resubmit it as a working group item. TICTOC Requirements Discussion: ============================== What items should be in the next generation of NTP? What about NTP maintenance items? Greg Dowd: TICTOC scope focused on improving performance of NTP and improving IEEE 1588. Others at NTP WG should voice their opinions on NTP improvements from a user point of view. Identify issues and enhancements via the mailing list. Those interested in timing issues should attend the TICTOC meeting since TICTOC is now a WG. The meeting was adjourned.