Skip to main content

IETF Last Call Review of draft-ietf-snac-simple-13
review-ietf-snac-simple-13-dnsdir-lc-ma-2026-11-02-00

Request Review of draft-ietf-snac-simple
Requested revision No specific revision (document currently at 14)
Type IETF Last Call Review
Team DNS Directorate (dnsdir)
Deadline 2026-11-04
Requested 2026-10-21
Authors Ted Lemon , Jonathan Hui , Esko Dijk
I-D last updated 2026-11-11 (Latest revision 2026-11-11)
Completed reviews Genart IETF Last Call review of -13 by Pete Resnick (diff)
Dnsdir IETF Last Call review of -13 by Di Ma (diff)
Tsvart IETF Last Call review of -13 by Brian Trammell (diff)
Assignment Reviewer Di Ma
State Completed
Request IETF Last Call review on draft-ietf-snac-simple by DNS Directorate Assigned
Posted at https://mailarchive.ietf.org/arch/msg/dnsdir/2tY-IPZQ_22QxsbDJyR3kAlDD4s
Reviewed revision 13 (document currently at 14)
Result Almost ready
Completed 2026-11-02
review-ietf-snac-simple-13-dnsdir-lc-ma-2026-11-02-00
Comments on the registrar discovery paragraph in §5.5.2.

1. The "either ... or ..." construction seems to frame the two discovery
approaches as mutually exclusive options for the SNAC router. Per RFC 9665
§3.1.1 and §3.1.2, the applicable discovery mechanism is determined by host and
network type: general-purpose hosts use the dnssd-srp SRV discovery procedure,
while constrained nodes follow a discovery mechanism defined by their
constrained node network’s specification. Where both classes of hosts coexist
on the stub network, the SNAC router MUST support both discovery pathways,
rather than selecting only one.

2. The Thread example risks creating an unbounded escape hatch for proprietary
discovery schemes. In its current form, it may be read as endorsing direct IPv6
address distribution outside the DNS-SD/SRP framework. RFC 9665 §3.1.2 permits
alternative registrar discovery only when a separate constrained-node-network
specification defines that mechanism. The text should clarify that
stub-network-specific discovery mechanisms must be formally defined by the
corresponding network specification and cannot be arbitrarily invented by
implementations. Without this boundary, stable conformance testing and
consistent interoperability cannot be guaranteed for this Standards-Track
document.