Minutes of the IPFIX meeting at IETF 85 in Atlanta
About 22 people present
Scribes: Chris Inacio, Nevil Brownlee
Meeting chair: Juergen Quittek, Nevil Brownlee

Juergen opened the meeting, Nevil gave the IPFIX status update:
 - Flow Selection draft is in AD Review, progressing
 - IE-Doctors in AD Followup: in response to DISCUSS, IANA Actions
   will be moved to 5102bis, IE-Doctors will need normative
   reference to it
 - Aggregation draft: on IESG telechat agenda, will need a new
   revision before going to RFC Editor
 - No work done on Mediation-Protocol since IETF 84, expect
   some progress by IETF 86 (Orlando)

Brian presented our 5101bis and 5102bis drafts
- Benoit Claise commented that IPR needs to be re-asserted in 
  bis documents, Cisco legal will do that for these two
- 5102 (IPFIX Information Model)
  - WGLC ran from 14 to 22 October, lots of comments
  - Section 5, Information element Categories, is no longer
    needed, proposedto replace it with stub pointing to IANA
    registry
  - Benoit suggests a summary/history as well as that pointer,
    including text about categories of IEs, so that 5102 remains
    a complete description of our model
  - Meeting agreed to this, to be done in next revision
- 5101 (IPFIX Protocol)
  - No updates since IETF 84, authors working on next revision
  - AD comment: we don't have to interop all features of
    protocol to advance standards level
  - List comment about byte ordering, Brian will add a reference
    for that
  - WGLC to start when document is stable

Paul Aitken presented (via Meetecho) the Link Layer IEs draft.
- WGLC for this started 15 Oct 12; it will end after next week's
  (bi-monthly) IEEE 802.3 meeting
- Juergen commented that the draft has a VXLAN I-D as an
  Informative Reference
- Brian pointed out that several IPFIX IEs cover VLANs, we
  need to check that the proposed VXLAN IEs will not cause
  interoperation with them
- Paul will find a VLAN-tag expert to check that
- Will probably need a new revision, when that's stable Nevil
  will submit to IESG

Paul presented the MIB Variable Export draft
- Most issues have now been resolved
- 'Extended Field Specifiers' introduced as a good way to
  represent OIDs in an IPFIX template.  Clear agreement
  in the room on this
- Field Lengths in the EFS were discussed.  Brian asked
  how variable-length fields would be handled?
- New version needed, WGLC when draft is stable

Paul presented two new individual drafts
- IE Equivalence
  - Addresses question "how does a collector know that a
    Private IE is equivalent to an IANA IE?"
  - Four possible methods explored, Paul proposes exporting
    an 'equivalence table'
  - How long does an exporter need to keep sending the table?
  - Benoit suggests using URIs instead, as Chris proposed
    at IETF 84.  Brian wants to see Chris' draft about that
    completed first
- Recent IANA IE Registry Changes
  - Summarises changes to the registry over about the last 
    18 months; this is a useful document
  - Brian pointed out that IE 353 should have an upper-case 
    O in 'octets'
  - Benoit asked whether these changes meet the guidelines
    set out in IE-Doctors?
  - A general review of IE semantics would be 'a good thing'
  - Nevil will ask for volunteers on the list

Any Other Business discussions
- IPFIX Milestones
  - Exporting MIB Objects and Mediation Protocol
    changed to July 2013
  - Aim for Mediation Protocol WGLC in April 2013
- Interop
  - Interop testing is useful whether or not we need it
    to advance standards level
  - Brian's Open Source implementation (ripfix) is now 
    available, he will post instructions on how to use it
    to the list
 - IPFIX work in other WGs
   - Benoit reported on RADIUS Accounting draft
   - Brian commented that ALTO have a new draft that describes
     exporting values as text strings; he would like others
     to read it and comment on the IPFIX list

The meeting finished at 1435