Session Date/Time: 06 Nov 2025 22:00

CCAMP

Summary

The CCAMP session covered a working group status update, including IPR polling, document dependencies, and prioritization for adoption and last calls. Several working group documents and individual drafts were discussed, focusing on WDM interface parameters, WDM tunnel models, pluggable WDM transceivers, CMIS/control YANG, performance management streaming, unified optical networks and AI computing (UNACO), and microwave interface YANG. Key technical discussions revolved around refining YANG models, addressing dependencies, and identifying suitable architectural approaches for new problem spaces. The session concluded with an update on the proposed CCAMP charter revision and a call for feedback.

Key Discussion Points

Administrative Matters:
- The "Note Well" was highlighted, emphasizing professional conduct and IPR disclosure.
- Meetecho was used for recording participation, replacing traditional blue sheets.
- The link for collaborative note-taking was provided.
- Reminder for authors and contributors regarding IPR processes.
Working Group Status Update:
- The Chairs provided an update on the wiki page, which tracks IPR polls, draft statuses, dependencies, and WG priorities.
- Noted significant dependencies on other working groups, particularly the TE YANG (draft-ietf-teas-yang-te, draft-ietf-teas-yang-te-types) documents which have now been submitted for publication.
- The last draft pushed to IESG was 1993bis, with optical impairment topology next.
- No new RFCs since the last meeting, but a long list of drafts are in the editor queue or IESG processing (1923bis L1 Types, L1CSM, OTN topology YANG, optical impairment topology).
- WG last call priorities: FlexGrid YANG, DWDM interface, OTN tunnel model.
- WG adoption priorities: WDM pluggables (IPR polling ongoing), data model for client signal monitoring, data model for fine-grain optical transport network.
- Updates on other WG drafts (not on agenda):
  - ACTN optical transport management: New version in May, considered a living document.
  - Client signal YANG model: No update since July, evaluating interest to split.
  - Ethernet client TE Topology YANG: New version in October, needs alignment with packet TE topology and other models.
  - Flexi-Ethernet YANG model: Stable since v03, authors believe it's ready for WG last call.
  - OTM path computation YANG model: Refreshed in October, requires alignment with latest models.
  - OTN tunnel model: No update since Madrid, needs checking for changes before progressing.
  - YANG slicing: Needs review comments addressed before WG last call.
  - TSB mode signaling: Awaiting final versions of optical impairment and Layer 0/1 interfaces YANG modules.
- Liaison Statements: Two liaison statements from ITU-T (IMT 2020 related to CG and SG15) were acknowledged.
  - Action: Update a second document for ITU-T consideration to provide fresh information.
- The high activity of the CCAMP WG was highlighted, with numerous weekly/bi-weekly calls on various topics (Flexi-grid, microwave, OTN slicing, optical plug-able, service path computation, optical environment, topology, and transport NBI).
- A side meeting on Layer 0/1 boundary scenarios for WDM tunnel modeling was held during the week. Organizers were encouraged to provide updates to the mailing list.
DWDM Interface Parameter YANG (draft-ietf-ccamp-dwdm-interface-parameters-yang) presented by Roberto Manzotti:
- This draft augments ietf-interfaces to provide DWDM-specific management parameters and operational data.
- It is a joint work with optical impairment topology and RFC 1993.
- Most GitHub issues from upstream and YANG doctor reviews are closed, with two remaining.
- Enhanced the model with operational parameters like chromatic dispersion, polarization mode dispersion (PMD), and polarization-dependent loss (PDL).
- Changed the primary key for the Threshold Crossing Alert (TCA) list from TCA-type to TCE-ID to allow multiple thresholds per TCA type.
- Split the DWDM interface into three containers: configuration, performance monitoring, and operational state.
- Duplicated central frequency and transmit power into target-central-frequency and target-transmit-power (configuration) and transmit-channel-power and central-frequency (operational status).
- The authors believe the draft is stable and ready for WG last call after addressing the remaining issues.
WDM Tunnel YANG (draft-ietf-ccamp-wdm-tunnel-yang) presented by Iva Bojic:
- Updates were made to align with RFC 1993bis (ongoing WG last call) and to improve modeling for regenerators.
- Added a section describing regenerator types, covering single/multi-carrier scenarios and different types of transponders acting as regenerators.
- Defined base identities for regeneration layer, digital layer termination, and inverse multiplexing type; full identity list to be completed after Layer 0/1 boundary discussions.
- Open issues: addressing comments from routing area director and YANG doctor, improving text for WDM tunnel provisioning, updating relationships with dependent models (optical impairment, 993bis, optical path computation), and adding constructs for optical operations (e.g., automatic power equalization).
- Dependency on the gap analysis for optical pluggables was noted.
- Discussion: Clarified that WDM Path Computation depends on this draft, not vice versa. The pluggable work could impact this draft, but authors feel there's time to address gaps if identified.
Pluggable WDM Transceiver YANG (draft-roza-ccamp-pluggable-wdm-transceiver-yang) presented by Reza Rokui:
- This draft is a top priority for WG adoption, with IPR polling currently running.
- Two parallel efforts:
  1. Refining the attribute list for capability configuration and PM stats in a Google Sheet.
  2. Performing gap analysis to identify missing attributes from existing IETF models.
- The approach is to reuse existing IETF YANG models as much as possible, complementing existing drafts (DWDM interface, tunnel, impairment).
- The proposed solution is to augment the DWDM interface parameter model with attributes missing from the Google Sheet.
- The draft will provide guidelines and solutions for identified gaps, not just state them.
- The work can potentially extend to other areas like the WDM Tunnel model.
- The Google Sheet is open for community suggestions and contributions.
- Emphasis on completing the work in a timely manner.
YANG Data Model for CMIS and Control (draft-homma-ccamp-yang-cmis-control) presented by Shunsuke Homma:
- Concept: Proposes a YANG model to handle CMIS registers directly, augmenting ietf-interfaces. This aims to address issues with network OS dependency on specific transceivers and the lack of a standardized way to handle vendor-specific CMIS custom pages.
- Use Cases: Centralized control of pluggable DCO modules and leveraging CMIS custom pages (e.g., for high-accuracy metrics).
- Updates: New co-author, enhanced use case description (e.g., remote monitoring), extended YANG model coverage (more RPCs/monitoring), and editorial fixes.
- Key Discussion Points:
  - Should standard CMIS pages be handled? There are challenges in exposing low-level CMIS data to external entities, but also benefits. The authors argue it does not violate CMIS definition if done via the network OS.
  - How to handle CMIS custom pages? This is a challenging aspect, and input on better approaches is sought.
- Open Discussion in Queue:
  - Reza Rokui suggested merging this work with the pluggable-wdm-transceiver-yang draft to create a cohesive IETF solution, particularly for "opaque attributes" (custom pages). He expressed concerns about exposing low-level CMIS registers directly due to operational complexity.
  - Roberto Manzotti raised concerns about the network OS losing visibility and control if binary CMIS data is directly passed through. He also worried about allowing access to standard pages with "clear-on-read" registers, potentially putting the network OS out of sync. He noted the challenge for open-source systems to support new transceivers.
  - Fatai Zhang agreed with converging efforts with the pluggable draft. He also mentioned that ITU-T SG15 is starting a new project on similar CMIS-related topics and suggested holding off for a while.
  - Swami (Nokia) acknowledged the valid problem statement but suggested that solutions for custom capabilities should be handled in a more abstract, CMIS-agnostic way to ensure future compatibility.
  - Nigel Davis emphasized distinguishing between non-invasive monitoring/states and configurations. While some monitoring might be acceptable, direct CMIS exposure for control (especially for things like reset-on-read counters) is problematic. He agreed the problem needs solving but not via direct CMIS exposure.
  - Daniel King acknowledged the validity of the problem (addressing cutting-edge WDM technology features) but reiterated the concern about direct CMIS exposure.
YANG Data Model of Performance Management Streaming (draft-yun-ccamp-pm-streaming-yang) presented by Minyong Yun:
- Aims to stream server performance data to clients for applications like network digital twin and AI/ML.
- Combines ITU-T G.7710 concepts (e.g., measurement methods like count, snapshot, tight marks) with IETF YANG push mechanisms.
- Applicable across various transport equipment types (OTN, Ethernet, MPLS-TP).
- The new version adds a pm-capability YANG model, following RFC 9196, to advertise capabilities like supported sampling and measurement intervals.
- ITU-T G.7710 is undergoing revision in SG15 to include PM streaming requirements, supporting variable intervals and streaming with subscriptions.
- The procedure involves clients discovering server capabilities and then subscribing to metrics.
- Discussion: The work was deemed valuable. Fatai Zhang suggested inviting operators to review and contribute, especially those active in ITU-T SG15 Q14. The chairs noted good support and interest, and would consider it for WG adoption after further socialization.
Unified Optical Network and AI Computing (UNACO) Use Cases (draft-liu-ccamp-unaco-usecases) and Framework (draft-liu-ccamp-unaco-framework) presented by Chaoyun Hu:
- Motivation: Address the urgent need for deep collaboration between optical networks and AI due to latency, bandwidth, and isolation requirements of distributed AI.
- UNACO Concept: A framework enabling bi-directional awareness, resource abstraction, and joint orchestration across the compute-optical boundary to optimize efficiency and resource utilization.
- Use Cases: Distributed AI training (syncing parameters across data centers), distributed inference (distributing AI model layers), and accessing remote AI services.
- Problem Statement: Current compute and network resources are managed in silos, lacking unified control and performance measurement standards, leading to suboptimal resource allocation.
- Requirements: Integrated control/measurement architecture, unified abstraction/evaluation framework, and scheduling algorithms for end-to-end resource allocation.
- Control & Management Architecture: Proposed a logical architecture with Service Orchestrator (SO), UNACO Orchestrator (UCO) as the core, Optical Network Controller (ONC), and Compute Provider System (CPS). Defined SUI, UOI, and UCI interfaces.
- Discussion:
  - Daniel King questioned why the orchestrator was "unified compute and optical" rather than a more generic "unified compute and transport" orchestrator (covering IP, OTN, optical).
  - Stephen expressed a similar concern, finding the draft too narrow. He suggested the general problem involves managing compute, networking, and storage, and hinted at a recursive management solution, not just a point solution for optical.
  - Adrian Farrel (Katz co-chair) invited the authors to review the Katz architecture (currently in WG last call) and provide feedback. He also raised questions about whether UNACO assumes dynamic provisioning or planning, and how it handles provisioning paths (which Katz does not explicitly cover).
Microwave Interface YANG (draft-ietf-ccamp-microwave-yang-bis) presented by Italo Busi:
- This draft is an update to obsolete RFC 8561, addressing gaps identified by an ETSI micro millimeter IESG review.
- Manages microwave interfaces, including carrier and radio link termination, and aggregation of multiple carriers.
- Most editorial issues are closed; technical open issues have agreed solutions but await YANG model updates (e.g., band/kind aggregation for remote nodes, cross-polarization interference cancellation, mismatch detection).
- Ongoing discussions on: new coding for modulation (e.g., 1024 BPSK), configuring maintenance windows (RPC vs. state), and managing scenarios with one transmitter and multiple parallel receivers.
- IANA early review comments need to be addressed.
- The draft is discussed in weekly calls and is considered ready for WG adoption.
CCAMP Charter Update:
- Chairs presented a proposed update to the working group charter. A link to the GitHub repository containing the new proposed charter was shared on the mailing list.
- No significant changes to the core scope (non-packet Layer 0, 1, and 2).
- Explicit mention of optical pluggables and OTN slicing was added.
- The chairs called for feedback on the mailing list by the end of November.

Decisions and Action Items

Decisions:
- The DWDM Interface Parameter YANG (draft-ietf-ccamp-dwdm-interface-parameters-yang) is a top priority for Working Group Last Call.
- The Pluggable WDM Transceiver YANG (draft-roza-ccamp-pluggable-wdm-transceiver-yang) is a top priority for Working Group adoption (IPR polling currently ongoing).
- There is strong interest in the YANG Data Model of Performance Management Streaming (draft-yun-ccamp-pm-streaming-yang), and it will be considered for Working Group adoption after further socialization.
Action Items:
- Chairs: Push RFC 1993bis and optical impairment topology for IESG publication.
- Chairs: Set a deadline (end of November) for community feedback on the proposed CCAMP charter update, then submit it to the IESG.
- Shunsuke Homma (and co-authors): Clarify the problem statement for the CMIS/control YANG draft, investigate gap analysis, and consider merging efforts with the pluggable-wdm-transceiver-yang work for a cohesive solution.
- Minyong Yun (and co-authors): Socialize the PM Streaming YANG draft with operators and other active Working Group participants to gather further feedback.
- Chaoyun Hu (and co-authors): Review the Katz architecture and engage with the Katz Working Group/mailing list. Consider generalizing the UNACO concept to cover broader transport technologies beyond just optical.
- Italo Busi (and co-authors): Address IANA early review comments for the microwave-yang-bis draft.
- Working Group Participants: Review the proposed CCAMP charter update on GitHub and provide comments on the mailing list by the end of November.
- All Authors: Follow IESG guidelines for authors and contributors to ensure efficient IPR polling and document processing.
- Organizers of CCAMP side meetings/calls: Provide updates and summaries to the CCAMP mailing list to keep the wider Working Group informed.

Next Steps

The Chairs will continue to shepherd drafts prioritized for WG Last Call and IESG publication.
Authors of YANG Data Model for CMIS and Control are encouraged to engage in further discussion on the mailing list and consider alignment with the Pluggable WDM Transceiver YANG work.
YANG Data Model of Performance Management Streaming will undergo further socialization with operators and other stakeholders in preparation for potential WG adoption.
Authors of UNACO drafts will continue to refine their work, taking into account feedback regarding generality and alignment with existing architectures like Katz.
The Microwave Interface YANG draft will proceed with addressing open issues and comments, aiming for WG adoption.
Feedback on the proposed CCAMP charter update will be gathered via the mailing list, with the aim of submitting the updated charter to the IESG by the end of November.