Hi Linda, Thank you for your review. And sorry for the very delayed response.
Please see some comments inline … From: Linda Dunbar via Datatracker <[email protected]> Date: Wednesday, 22 October 2025 at 20:16 To: [email protected] <[email protected]> Cc: [email protected] <[email protected]>, [email protected] <[email protected]> Subject: draft-ietf-netmod-sub-intf-vlan-model-17 early Opsdir review Document: draft-ietf-netmod-sub-intf-vlan-model Title: Sub-interface VLAN YANG Data Models Reviewer: Linda Dunbar Review result: Has Nits Summary: The draft defines YANG models for VLAN and flexible encapsulation on sub-interfaces. It is technically solid and aligns well with existing interface and VLAN models. Questions: - Given that most commercial routers already implement mature sub-interface VLAN configurations, and are unlikely to change their operational behavior, could the authors clarify whether the intent of this model is purely to provide a standardized YANG abstraction for management systems rather than to influence device-level configuration semantics? This model is primarily intended to be a standardized device level configuration model for configuring VLAN services that terminate at L3, or for connecting into IETF's L2 bridging constructs, e.g., Pseudowires, VPLS or EVPN. You are right that devices already implement their own configuration models. In some cases, those configuration models will align quite closely to this model, in other cases, there may be different abstractions. This is true and a common concern of any YANG model standardised in the IETF or OpenConfig and isn't specific to this model in any way. - How is this model intended to coexist with the bridge-based VLAN model (IEEE 802.1Q / RFC 8919)? Can both be used on the same device or interface? This model is different from the IEEE 802.1Q YANG model because its goal is to provide the necessary interface extraction to allow IETF L2VPN/EVPN services to be configured and to terminate VLAN tagged traffic to L3 services, but in many cases the two different configuration models can be used to configure similar L2 forwarding constructs. I think that which model is supported may well depend on whether a device is predominantly a L2 switching based product or a L3 routing based product. I think that it is entirely valid for both models to be used on the same device., if they are being used for different interfaces. Possibly they could also be used on the same interface, but I believe that is (and should be) out of scope for this draft. Guidance on how the two models work together could be a future informational draft. But I don't think that we should delay publishing the base YANG models for this. Minor comments: - Need to clarify operational behavior when devices support both bridge-based VLAN configuration (per IEEE 802.1Q / RFC 8919) and sub-interface–based VLAN configuration (as defined in this draft), particularly if the same physical port could be referenced by both models. Thanks, we will add an informative comment to the draft, perhaps in an appendix, indicating that this model can be used alongside the IEEE bridging YANG models, particularly when used on different physical or LAG interfaces, but any further interoperability between the two models on the same device, or same interface, is out of scope for this document. Kind regards, Rob Best Regards, Linda Dunbar
_______________________________________________ netmod mailing list -- [email protected] To unsubscribe send an email to [email protected]
