Minutes of the IPPM meeting.
============================

Date:    18/Nov/2008
Chair:   Henk Uijterwaal
Minutes: Notes were taken by Christopher J. White and Al Morton, then
         turned into minutes by Henk Uijterwaal.


1. Administrativia.
   Henk opened the meeting.  Matt Zekauskas sent his apologies as he had
   to be at another meeting.

2. Current work.

2.a. Status of Draft and Milestones.
   The TWAMP draft has been published as RFC 5357.  The traceroute draft
   has been cleared for publication and is in the RFC editor queue.  Three
   drafts passed WGLC (Duplicate, MultiMetrics and Packet DV AS).  These
   are waiting for a write-up by the chairs.  Unfortunately, the chairs
   didn't have as much time for IPPM work as they had hoped, so this hasn't
   happened even though it should.  They promised to do this in the coming
   weeks.

2.b.1. TWAMP extensions. (Al Morton).
   Draft-ietf-ippm-more-twamp-00.   OWAMP - errata - include "MUST" for
   matching modes.  This is a simple fix supported by the WG.  It can
   be taken to the list for WGLC.

2.b.2. TWAMP Reflect OCTETS Feature (Al Morton)
   Al presented the draft.
   Open question: consider change the structure of the TWAMP structure
    to TLV type structure to make it more extensible.

   Questions from the floor: Roman Kraznowski said he had some things he'd
   like to add.  He'd post to the list.

   Kaynam Hedayat asked if the extensions backward are compatible?  Al
   answered, Yes, no problem on either the server or client

2.b.3. TWAMP Features - Individual Session Control
  Al Morton presented the draft and will take the proposal to the list
  to see if there is interest.

2.b.4. TWAMP TLS (Al)
  Al presented the draft.  The open question is if TLS will provide better
  security than the present mechanism.  Walt Stevenson commented that he
  does not see this.  The security community has a preference for TLS but
  so-far, they haven't pointed out why.

  The chair will take this question to the list, if there isn't sufficient
  feedback, it will be dropped.

2.c. Reporting Draft
  None of the authors were in the room to present the draft.  The chair
  will ask the list if there is interest to continue to this effort.

  Comment by Al Morton - scope renamed to implications for long term reporting,
  scope was initially very short term, and should remain short.

2.d.  Composition
  Al Morton gave a brief overview of the draft.  Roman Kraznowski and Yakov
  Stein voluntered to read the draft.  After that, Al thinks it is finished.

  Questions from the flow.
  * Roman K: Forum 2 years back of how to provide metrics for
    intercontinental, he would provide a reference to work by Kevin Mason.
  * Roman K: Asked about other temporal / spatial aggregation
    drafts as they are important for SLAs.  Al answered that this is a
    framework draft, with the details in other drafts.
  * Yakov Stein said that spatial aggregation is difficult problem
    mathematically

3. Future work.
   Henk gave a brief overview of the new charter.  He had received some
   comments by mail and will send out a new version shortly after the
   meeting.   Comments from the floor:

   * Lars Eggert: Recharter incrementally, add small number of things
     at a time, rather than large recharting adding multiple things.
   * Lars: Passive measurements are the most complex, with only a few
     people interested.  Perhaps this should be dropped.
   * Henk said that there was interested in versions of the IPPM
     metrics for passive measurements from IPFIX in the past.  Brian
     Tremmel commented that there was currently no active interest.

   Possible new drafts:
   * New Network Layer Metrics for Packet Loss, Delay, Delay Variation.
     Nick Duffield briefly gave a reminder of work shown earlier.
   * Comments:
     Al Morton: Delay quantiles are a good thing to estimate, but RFC3550
      is a good relative measure but not worth pursuing a detailed
      statistical pursuit of relative jitter
     Henk if the goal is standardization then we should get authors from
      other groups active in this field.  Reference to their work was in
      the slides. Nick agreed.
     Henk said that this can be picked up as part of the new charter.  This
      requires the new charter to be approved first.

4. AOB
  Al Morton: consider liasons to other groups to help with problem
  areas.  Take more active role to make sure IPPM is understood
  properly elsewhere.  The chair agreed and asked everybody to report
  if they see areas where our metrics are being used and support with
  the interpretation can be useful.

5. The chair then closed the meeting and hoped to see everybody at the
   next meeting or on the mailing list.