Markdown Version | Session Recording

Session Date/Time: 28 Mar 2023 08:00

ntp

Summary

The Network Time Protocol (NTP) working group meeting covered document status updates, a presentation on DNS Tesla synchronization vulnerability, updates on NTP five development, deployment updates from K-net node, and discussions on NTP for PTP, NTP over PTP, and the Rough Time document. The meeting also included discussions on future directions and potential hackathons.

Key Discussion Points

• Document Status: Updates were provided on documents in the publication channel and ongoing drafts. Kronos publication has been requested, interleave modes deferred to NTP five, and registry documents with the AD, enterprise profile documents with me, and Kronos has just recently been publication has been requested.
• DNS Tesla Synchronization Vulnerability (Jason Anderson):
  • Presentation on a vulnerability related to clock offsets in Tesla-enabled GNSS authentication.
  • The vulnerability allows attackers to forge messages if a receiver's clock offset is greater than the key release delay.
  • A simple test was presented to identify and secure outliers.
  • Suggestion to avoid revealing the t1 timestamp in NTP queries and to make queries non-periodic.
  • Concerns about the data minimization draft not addressing the acceptance of forged messages and focusing more on privacy.
• NTP five:
  • Review of the use cases and requirements document.
  • Discussion of technical changes including a new version field and the use of secondary receive times.
  • Recommendation to conduct consensus calls regarding mode one to five before a working group last call.
• K-net node Deployment Update (Krista):
  • Discussion of traffic increases (thousands to millions of hits per hour) on their NTP service.
  • Observations of traffic patterns correlated with daylight cycles.
  • Identification of a high percentage of unique IP addresses with relatively few requests per address.
  • Speculation about the source of traffic and the potential for a custom client implementation.
• NTP for PTP (Martin Lange):
  • Overview of the nt for okey protocol and goals for securing Pt 2.1
  • Update was that there has been no much action.
  • Consolidating all documents, mails and feedbacks, to push the content to to go further
  • Need to define the communication between the key server And so time server side, and Nt for anti pieces of limitation.
  • Plan to prepare a short document and a second, more detailed document for the working group.
• NTP over PTP:
  • The sequence ID is now required to be increment to not break hardware drivers it might be into this field.
  • Discussion about adding clarification about using Pt p frames to transmit Nt p
• Rough Time:
  • There has not been much work done as of recently.
  • Asking for common on message format that multiple people have asked about.
  • The Rough Time does not use network by order, and is it too late to change the order?

Decisions and Action Items

• Jason Anderson:
  • Action Item: Contact authors of the data minimization draft to revive it.
  • Action Item: Develop text about acceptance of forged messages for inclusion in data minimization.
• Mai (NTP five specification):
  • Action Item: Switch the draft over to a working group document.
• Mai (NTP over PTP):
  • Action Item: Clarify that the service will not work with PTP in the document.
  • Action Item: Then resubmit as a working group document.
• Martin Lange (NTP for PTP):
  • Action Item: Prepare two documents for the booking group: the short version and the more detailed version.
• Marcus (Rough Time):
  • Action Item: Update the draft with changes.
  • Action Item: Consider whether or not to change the code to using network by order.

Next Steps

• Schedule a virtual interim meeting in January or February.
• Plan to have hack projects for NTP five, Rough Time, and other relevant areas.