When Monday 24 July 2023, 13:00-15:00 PDT

Where: Golden Gate 6,
https://meetings.conf.meetecho.com/ietf117/?group=ippm

Chairs: Tommy Pauly & Marcus Ihlar

Working Group Documents

Chair Slides (slides-117-ippm-chair-slides)

• Note well
• Memoriam - Al Morton
• status of draft-ietf-ippm-capacity-protocol to be continued by
  co-author

draft-ietf-ippm-pam

Precision Availability Metrics for SLO-Governed End-to-End Services

• observations
• updates
• discussion items
• next steps
• Questions
• Poll (show of hands)
  How many poeple read this document ?
  7

Start WGLC after meeting

draft-ietf-ippm-ioam-data-integrity

Integrity of In-situ OAM Data Fields

• status-06
• draft-ietf-ippm-ioam-data-integrity-06
• Header protecrion
• Integrity protection

• Next
  Q(Justin, as presenter): Secdir review ? WGLC? RFC 9326bis should be
  created?
  A(chair): New RFC should indicate that it updates RFC9326, no need
  for 9326bis.
  Sec directorate review will be done first then depending on outcome,
  next steps (such as WGLC) will be taken.

• Poll (show of hands)
  How many read this document?
  1

draft-ietf-ippm-encrypted-pdmv2

IPv6 Performance and Diagnostic Metrics v2 (PDMv2) Destination Option

• fields in PDMv2: S-S
• proposal to pass PSNTP in clear
• current vs poposed packet layout
• Question for WG
  Q(Nalini, as presenter): To discuss the SECDIR to get a review ?
  A(chair): Yes

Q(): Did we do any initial security review ?
A: some initial reviews were done for using HPKE.

Q(Martin Duke - Google): What's the difference between the epoch field
and allocating reserved bits?
A: unanswered

Q(Tommy): Is it a monotonically increasing number, so HPKE can use it as
the counter?
A: yes

draft-ietf-ippm-ioam-yang

YANG Data Model for In-Situ OAM
presenter: Tianran Zhou

• comments from IETF116

Q(Gregory Mirsky - Ericson): Assign role for IOAM functionality on the
node
A: discuss offline to put a default action.

• Next
  Q(presenter) : Continue ahead?
  A(chair) Conclude discussion with greg, then decide to move forward.

draft-ietf-ippm-responsiveness

Responsiveness under Working Conditions
presenter: Christoph Paasch

• Implementation status
• RPM-community updates
• invitation to collab and join open source work or slack for weekly
  meeting

Q(presenter): Next move ?
A(chair): Discuss moving forward after next revision.

• Poll (show of hands)
  How many read this document?
  5

Proposed Work

draft-olden-ippm-qoo

Quality of outcome
presentor : Bjørn Ivar Teigen

• challenges in network Quality

• design goals that is useful for

  • end users
  • applications
  • operators

• proposed solution

• building on TF-452 (QED)
• middleround for QoE, Quality of outcome, QoS
• Formula
• multiple prespective
• example graph on CDF vs latency

Q(Martin Duke - Google): Representation from percentile ?
A : Choice of depiction is to show arbitary CDF for perfection and
uselessness

Q(chair): Attributes that define the curve to be tolerant or not at
various percentiles?
A: The graph is generated with self created data. Starting points were
perfect(ideal) CDF and then more relaistic

Q(Martin Duke - Google): Example may be useful along with graph.
A: Agree

Q(Dave Taht): sampling rate ?
A: Instaneous metrics are used for this work

Q(Gregory Mirsky - Ericsson): QED documents are available.
A:

• Implementation status for the C library and Go
• how it works links
• Poll (show of hands)
  How many read this document?
  4
• Poll (Meetcho)
  should we adpot Quality of outcome ?
  Y: 11, N:5

draft-mirsky-ippm-asymmetrical-pkts

presenter : Gregory Mirsky

• Reflected Test Packet control TLV pac ket structure
• performance monitoring in a multicast network
• Address Group Sub-TLV in Layer 2 and Layer 3 packets structure
• Next steps and invitation for collaboration

Q(Rakesh Gandhi - Cisco): clarification question- Reflected packet with
diff size wasnt clear.
A : Can clarify examples such as when requestor packets will be large.

Q(Rakesh Gandhi - Cisco): When we want only 1 leaf node to reply ?
A: example prefix length based match to network address, thus defining
only subset of system that would respond.

Q(Rakesh Gandhi - Cisco): How would sender know each leaf node replied ?

A: Source Address by Route reflector could be used. More in details to
come.

Q(Martin Duke- Google): Security in congestion. Set limits ?
Misconfiguration risks?
A: Must be authticated to set boundaries for this field.
To look deeper into amplication issues.
Number of reflected packets in multiucast mode == 1

Q(Xiao Min -ZTE Corporation): Can it be established for peering
enviornment ?
A: Yes

• Poll (Meetcho)
  Do you think this document should be adopted by IPPM ?
  Y:13, N:2

Lightning Talks

draft-fz-ippm-alt-mark-deployment

Alternate Marking Deployment Framework
presenter: G. Fioccola

• How to deploy Alt Marking

draft-wang-ippm-stamp-hbh-extensions

STAMP Extensions for Hop-by-Hop OAM Data Collection
presenter: G. Fioccola

• STAMP (RFC8762) Extensions for Hop-by-Hop OAM Data Collection

Suggestion(Gregory Mirsky - Ericson): Good idea to meaure Hbh but this
information can be collected by underlay rather than on STAMP. Such as
given by ipv6 plane.
A: can discuss based on consensus

Q(Rakesh Gandhi - Cisco): How would this work? Describe STAMP
processing.
A: It is active measurememnt. Intermediate nodes needs to handle STAMP
TLV option. To be discussed.

draft-wang-ippm-ipv6-distributed-flow-measurement

Distributed Flow Measurement in IPv6
presenter: Haojie Wang (China Mobile)

• motivation
• overview
• Extension of the flow monitor option

draft-wang-ippm-ipv6-flow-measurement

presenter: Haojie Wang (China Mobile)

• overview
• Running code: lab interop test status and Deployment status
• Next step

Q(Rakesh Gandhi - Cisco): Extension at transport layer such as ipv6
defines destination option too. This draft needs clarification.
A: _ (TBD: ask later)

draft-he-ippm-integrating-am-into-ioam

Integrating the Alternate-Marking Method into In Situ Operations,
Administration, and Maintenance (IOAM)
presenter: Xiaoming He (China Telecom)

• Motivation and objective
• Benefits of the integration
• Extended DEX Option - Type Format

Q(Gregory Mirsky - Ericson): Delay meaure and transit delay measurement
is already possible. The proposed idea uses seq number. Directed
conflicts.
A : discuss offline

Q(Justin Lurman - University of Liege) : Do we need another option type
? Annother 2 bits needed
A: discuss offline