Document: draft-ietf-rtgwg-dst-src-routing-revive Title: Destination/Source Routing Reviewer: Linda Dunbar Review result: Not Ready
I have reviewed this document as part of the Ops area directorate's ongoing effort to review all IETF documents being processed by the IESG. These comments were written primarily for the benefit of the Ops area directors. Document editors and WG chairs should treat these comments just like any other last-call comments. Summary: The draft proposes Destination/Source Routing (Dst-Src Routing), where routers use both the destination and the source address of a packet during forwarding lookups. Major: While the draft makes clear that “source ranges” are attached to routes, it does not explicitly warn against using per-host (/128) source prefixes. From an operational standpoint, this could be misread as requiring host-specific routing entries across the network, which would be infeasible for FIB scaling and hardware TCAM limits. I suggest adding an explicit note that destination-source routing is intended to operate on aggregated source prefixes (e.g., ISP-assigned /48 or /56), and that installing per-host source-qualified routes is not recommended. This clarification would help prevent misinterpretation and reassure operators about scalability. A natural place would be in Section 3.2 or Section 4. Insert a note like: Destination-source routing is designed to operate on aggregated source prefixes (e.g., ISP-assigned /48 or /56). It is not intended for installing per-host (/128) source prefixes across the network, which would be operationally prohibitive and could overwhelm router FIB resources. Implementations and operators SHOULD avoid generating or propagating per-host source-qualified routes. Minor: - The draft uses dst-src, source-destination, destination-source, suggest selecting one primary term (e.g., destination-source routing) and noting others only once in introduction. - Figures: ASCII art is helpful but lacks captions explaining behavior. Suggest adding labels such as “D/S-aware area” vs “Legacy area.” Nits / Editorial Issues Section 2.1: “redundandy” → redundancy Section 2.1: “neccessary” → necessary Section 5.6: “choosen” → chosen Section 5.1: “recusion” → recursion Section 6.2: “neccessary” → necessary Section A.2: “destiation” → destination Section 11: “dessiminating” → disseminating Ward regards, Linda Dunbar _______________________________________________ rtgwg mailing list -- [email protected] To unsubscribe send an email to [email protected]
