IETF Last Call Review of draft-ietf-netconf-distributed-notif-19
review-ietf-netconf-distributed-notif-19-secdir-lc-migault-2026-06-08-00

Hi,

I have reviewed this document as part of the security directorate's ongoing
effort to review all IETF documents being processed by the IESG.

I have no particular expertise in YANG or NETCONF, so please bear with me if I
am missing some context.

As I understand it, the document introduces a shift from the traditional
model—where a single route processor aggregates data from internal Components
and publishes it to the Receiver—to a distributed model where multiple
Publisher Agents communicate directly with the external Receiver.

If my understanding is correct, this seems to carry a few security implications
that might be worth expanding upon in the Security Considerations section:

1. Expanded trust surface at the Receiver. In the traditional model, the
Receiver establishes trust with a single entity (the route processor). In the
proposed model, the Receiver would need to trust multiple Publisher Agents
independently, which increases the number of identities to manage and
authenticate.

2. Internal communication becomes external. In the traditional model, Agents
provide their data to the route processor over an internal path. This draft
appears to move that data flow from an internal Agent-to-Processor exchange to
an external Agent-to-Receiver exchange, which would increase the overall
exposure of the system.

3. Reduced centralized control. The route processor traditionally could serve
as a control point for outbound telemetry. With Agents publishing directly to
Receivers, some of that centralized oversight may no longer apply.

It might be helpful to briefly describe these architectural trade-offs.