Session Date/Time: 20 Nov 2025 13:00

IPPM

Summary

This was a joint interim meeting of the BMWG and IPPM working groups, following the pattern established at IETF 124 to facilitate cross-group exchange. The session focused on updates to existing documents, presentation of nine new work proposals, and a dedicated discussion on the potential for joint chartering of the two working groups. Significant technical discussions occurred on various IOM (In-situ OAM) extensions, benchmarking methodologies for new network paradigms like CATS and network slicing, and CNI performance. The joint chartering discussion addressed participant feedback from IETF 124, highlighting commonalities, resource management, and interaction with other SDOs. A poll indicated a strong interest in continuing the joint chartering discussion.

Key Discussion Points

Administrative & Joint Session Context
- The session began with the IETF Note Well reminder.
- The chairs reiterated the goal of joint BMWG/IPPM meetings: to facilitate exchange, encourage cross-reviewing, and continue the joint chartering discussion initiated at IETF 124.
- The distinct charters of BMWG (lab-based, controlled stimuli) and IPPM (live, operational networks) were re-emphasized.
- Matt voluntarily offered to help with note-taking.
Working Group Document Updates
- IPPM: Capacity Protocol (approved by ISG), Asymmetrical Packets & Hybrid Two-Step (passed to ISG, received AD review), Quality of Outcome (publication requested), IOM Integrity Yang (WG adoption call concluded, adopted).
- BMWG: MLR search (submitted to ISG, now with RFC editor), Network Tester Configuration YANG model (new revision to address Yang doctor review), Power Benchmarking (new document adopted), RFC 2544 update (discussion resumed).
- Liaison Updates: SR Benchmark (liaison to IQT Study Group 11 pending), Quality of Outcome (liaison sent to BroadBand Forum, planned for ITOT Study Group 12 with positive prior interactions).
IPPM Asymmetrical Packets (Greg)
- Presented updates addressing comments from AD review and IANA.
- Key changes include new terminology, clarification of reflected test packet control (TOV), extension to multicast environments, IANA considerations for registry hierarchy, and editorial improvements.
- The document defines a new TOV in the reflected test packet control to manage the number, interval, and size of reflected packets.
- Matt (AD) confirmed that comments were clear and appreciated, looking forward to the next version for Last Call.
IOM Direct Exporting of Alternate Marking (Xiaoming)
- Proposed augmenting IOM DX options by incorporating the alternate marking method for performance measurements, addressing issues like loss inclusion and overhead.
- Proposed extending reserved fields (D, L bit, measurement tiered number field), and improving setting 2-bit Flow ID for better collision probability and identifying measurement blocks.
- Discussion: Greg raised concerns about backward compatibility if the encapsulation format changes for an already published IPv6 extension header. Tal and Justin clarified that the draft defines a new IOAM type using an existing IPv6 option type, so six-man WG involvement might not be necessary.
BMWG CATS Benchmarking Methodology (Kehan)
- Introduced a benchmarking methodology for Compute-Aware Traffic Steering (CATS) solutions, focusing on instance selection, benefits/trade-offs of deployment models (distributed, centralized, hybrid), and effectiveness from system utilization, latency, and session continuity perspectives.
- Updates include new co-authors, alignment with CATS framework, detailed reporting indicators (control, forwarding, application planes), and a new test for load balancing variance.
- Discussion: Joseph emphasized the importance of continued collaboration between CATS WG and BMWG given the specialized nature. Carsten noted the document is a 0.0 version and asked for plans to deepen the methodology details before requesting adoption. Joseph requested more information on NRP implementations.
IOM Trace Option Extensions for Alternate Marking (Xiaoming)
- Similar to the previous draft, this proposes augmenting IOM trace options by incorporating the alternate marking method.
- Proposes extending reserved fields (D, L, F, SM bits, Flow ID, SN, MPS fields).
- Discussion: Tal questioned the need for a D-bit when timestamp is already in the trace option, and suggested adding a counter data field. Rakesh asked for clarification on when to use this approach vs. existing alternate marking (RFC 9343) or IOM (RFC 9486) IPv6 options.
IOM Trace Option Extensions for Carrying Fixed Size Data (Xiaoming)
- Proposed extending the IOM trace option to carry fixed-size data by defining a "fixed size flag".
- This flag enables an adapted format with an "IAM fixed data trace type" (bitmap) to indicate fixed-size aggregate data (e.g., min available bandwidth, max link/buffer utilization).
- Discussion: Tal suggested using an existing IOM option type or collaborating with his template draft to achieve similar goals, arguing against a new flag for entirely different functionality. Xiaoming expressed concern that Tal's template design might be too complex for data plane implementation.
IOM Templates (Tal)
- Proposed a new IOM option type with a simple shim header that allows for different fixed-length template IDs.
- Templates can be changed by intermediate IOAM nodes, and the length is fixed throughout the path, not a function of hop count, simplifying packet processing.
- Use cases include aggregating data (min, max, average) and accumulating delay.
- Discussion: Thomas commented on the term "template" potentially being misleading from an IPFIX perspective, where it implies schema advertising. He suggested considering an alternative term. Giuseppe supported the proposal as an interesting way to generalize different options.
Network Slice Service Benchmarking (Chidong)
- Presented a 0.0 draft for benchmarking methodology for network slicing, focusing on key functionalities and effectiveness, particularly for Network Resource Partitions (NRP) and enhanced VPN services.
- Proposed test cases for NRP creation, topology alignment, policy association, and performance under background traffic.
- Discussion: Joseph asked for more information on available NRP implementations. Carsten requested more details on the benchmarking methodology itself in future revisions.
Telco Cloud Benchmarking Consideration (Tina)
- Discussed benchmarking considerations for Kubernetes CNI (Container Network Interface) in heterogeneous and multi-cluster environments.
- Highlighted the lack of standardized methodology, leading to non-reproducible and vendor-specific outcomes.
- Identified key aspects (core, extended performance, QoE metrics), best practices, and considerations for multi-cluster environments (topology, security, overhead).
- Mentioned "Coded" as an open-source benchmarking environment.
- Discussion: Tina plans to share information about an upcoming workshop in February 2026 on the mailing list. Joseph noted existing discussions with another author (Ming Onk) to define boundaries between related drafts.
Joint Chartering Discussion (Chin)
- Recapped the AD's initiative for BMWG/IPPM consolidation and feedback from Montreal.
- Motivation: Improve synergy, better structure ID work, enhance cross-peer review.
- WG Health: Noted insufficient cross-peer review and a significant drop in BMWG participation/new work growth.
- Issue 1: Common Framework:
  - Feedback: Some see commonality (metrics, terminology, RFCs 6390/8911); others see different expertise. Concerns about leadership imbalance.
  - Discussion: How to encourage cross-peer review if WGs remain separate? Carsten noted peer review is a generic IETF problem. Ali Safar argued merging doesn't magically increase participation or force reviews; it could even be less productive due to fragmented interest.
- Issue 2: Resources and Time Management:
  - Feedback: Fear of congestion if WGs combine; concern that one session is insufficient.
  - Discussion: Chairs can prioritize, schedule multiple sessions, or use interim meetings more effectively. Tal stressed the importance of sufficient IETF meeting time for discussion-heavy drafts. Ali Safar emphasized that session allocation is a management issue, independent of merging. Tal suggested experimenting with joint sessions for a year before merging. Giuseppe noted other WGs successfully use interim meetings for progress.
- Issue 3: Working Group Dispatcher:
  - Leverage the revived Performance Metric Directorate for publication reviews.
  - Optionally dispatch new work to technology-specific WGs.
  - Principle: Technology-specific work still goes to specific WGs; joint WG for "homeless" tech-specific work.
  - Discussion: Ali Safar viewed this as a management aspect, not a driver for merger.
- Issue 4: Interact with other SDOs:
  - Simplify interaction, appoint liaisons for common interests (e.g., ITU-T SG12). Limit scope of cross-SDO collaboration to IETF expertise.
  - Discussion: Ali Safar again saw this as a management aspect. Matt (AD) clarified the discussion aims to identify problems and solutions, not just whether to merge.
- Open Discussion:
  - Questions on high-priority performance/benchmarking topics.
  - Carsten (independent tester) preferred keeping WGs separate due to differing goals (benchmarking vs. development) and skillsets.
  - Tim raised protocol-specific aspects of measurement design (e.g., IPv6 extension headers), suggesting a need for standardized design approaches across areas.
  - Thomas pondered if the core difference lies in how to measure vs. what metrics are defined.
  - Rakesh reiterated that organizations often have separate teams for benchmarking and performance, indicating different skill sets and supporting the view against merging.
- Poll: A quick poll was taken on continuing the discussion on joint chartering. A clear trend indicated interest in continuing the discussion.

Decisions and Action Items

Note-Taking: Matt volunteered to assist with note-taking.
IPPM Asymmetrical Buckets: Greg will submit a new version of the document addressing AD and IANA comments shortly.
IOM Direct Exporting of Alternate Marking: Xiaoming will continue the discussion on the mailing list regarding IPv6 extension header compatibility and use cases.
BMWG CATS Benchmarking Methodology: Kehan will provide more details on existing NRP implementations and further elaborate on the benchmarking methodology in subsequent revisions.
IOM Trace Option Extensions for Alternate Marking: Xiaoming will continue discussions on the mailing list regarding the D-bit, L-field, and clarification of use cases relative to existing RFCs.
IOM Trace Option Extensions for Carrying Fixed Size Data: Xiaoming and Tal Mizrahi will continue discussions on the mailing list, exploring collaboration or alternative approaches to address concerns about complexity and functionality.
IOM Templates: The authors will consider using an alternative term for "template" to avoid confusion with IPFIX.
Network Slice Service Benchmarking: Chidong will check for sharable information on NRP implementations and improve the methodology section in future revisions. They will also consider involving the TEAS WG in future discussions.
Telco Cloud Benchmarking Consideration: Tina will share the link for the February 2026 standardization workshop on the IPPM mailing list.
Joint Chartering Discussion: Chin will continue to manage the discussion on the mailing list, seeking volunteers to help coordinate the effort.

Next Steps

Document Progression: Authors of new proposals are encouraged to address feedback, revise their drafts, and potentially seek working group adoption in future iterations.
Joint Chartering Discussion: Continue the discussion on the IPPM and BMWG mailing lists to further explore the issues, benefits, and potential paths forward for the working groups, building on the feedback received during the session.
Interim Meetings: The chairs will continue to evaluate the effectiveness of joint interim meetings and session allocation strategies to ensure sufficient time for technical discussion and progress.