Skip to main content

Last Call Review of draft-ietf-rtgwg-qos-model-11

Request Review of draft-ietf-rtgwg-qos-model-11
Requested revision 11 (document currently at 12)
Type Last Call Review
Team Internet Area Directorate (intdir)
Deadline 2023-12-01
Requested 2023-10-26
Requested by Jeff Tantsura
Authors Aseem Choudhary , Mahesh Jethanandani , Ebben Aries , Ing-Wher (Helen) Chen
I-D last updated 2023-12-26
Completed reviews Tsvart Last Call review of -11 by Colin Perkins (diff)
Intdir Last Call review of -11 by Dr. Joseph D. Touch (diff)
Rtgdir Last Call review of -11 by Adrian Farrel (diff)
Yangdoctors Last Call review of -03 by Jürgen Schönwälder (diff)
Yangdoctors Last Call review of -06 by Jürgen Schönwälder (diff)
Dear Directorate,

In preparation for the last call, we would like to request your review.

RTGWG chairs
Assignment Reviewer Dr. Joseph D. Touch
State Completed
Request Last Call review on draft-ietf-rtgwg-qos-model by Internet Area Directorate Assigned
Posted at
Reviewed revision 11 (document currently at 12)
Result Ready w/nits
Completed 2023-12-26
This document defines a YANG model for configuring router QoS parameters.

Understandably, it does not provide a single, definitive approach, but instead
defines modules that can be supplemented as needed to represent the parameters
of specific devices. Although that’s consistent with the YANG model approach,
it also seems to somewhat undercut the basic idea of a common model. This is a
deficiency in the IETF’s QOS approach, however, rather than being specific to
this document.

I did not note any areas of concern to INTAREA.

From a YANG viewpoint, I found the sole use of “leafref” (queue policy) a bit
concerning; it seems like there ought to be more than just a single case of a
named object whose names are referred to elsewhere in ways the YANG model can
verify. Almost anything with a name whose name is referenced elsewhere within
this model would be a candidate as leafref, e.g.

Minor nits:

The Traffic-Group definition appears misworded (remove “is a”) and is somewhat
incomplete (what kinds of meta data?):

   *  Traffic-Group: is a metadata stored in a packet buffer.

There are some parts where the terminology and explanation yield redundant or
tautological statements; it would be useful to rephrase these – the first of
which is “vendors might want to extend the model to add their own extensions,
such as to extend Traffic Policy module”.