ietf
[Top] [All Lists]

Rtgdir early partial review of draft-ietf-rtgwg-routing-types-02

2017-05-03 09:54:37
Review is partially done. Another review request has been registered
for completing it.

Reviewer: Stewart Bryant
Review result: Has Issues

Firstly I should say that I am not an expert in YANG, hence my
suggestion that you may wish to assign an additional reviewer.

However in reviewing this a number of questions and issues arose:

2.  Overview

   This document defines the following data types:

SB> Accessibility note - the order of types seems random. It might
SB> be more helpful to the reader if they were, in a systematic order
SB> for example alphabetical and/or dependency order.

=======

 router-id
      Router Identifiers are commonly used to identify a nodes in
SB> s/a nodes/nodes/

=======

   route-target-type
      This type defines the import and export rules of Route Targets,
as
      descibed in Section 4.3.1 of [RFC4364].  An example usage can
be
SB> s/descibed/described/
      found in [I-D.ietf-idr-bgp-model].

========

SB> I am surprised that IP multicast addresses are here, but IP
addresses are not.
SB> I would have thought that both should be in the same place.

========

SB> In some protocols we use the NTP and the 1588 timer types, I
assume
SB> they are defined elsewhere.

========

   mpls-label
      The 20 bits label values in an MPLS label stack entry,
specified
      in [RFC3032].  This label value does not include the encodings
of
      Traffic Class and TTL (time to live).  The label range
specified
      by this type covers the general use values and the
special-purpose
      label values.  An example usage can be found in
      [I-D.ietf-mpls-base-yang].

SB> I am surprised that you don't start with label and then define the

SB> other label definitions in terms of existing definitions.
SB> The obvious order being label, sp-label, gp-label, generalized

========

S/ "This identity represents IPv4 address family.";/"This identity
represents the IPv4 address family.";

========

     //The rest of the values deinfed in the IANA registry
SB> s/deinfed/defined/

SB> However a question arises, the list stops at mt-v6
SB> Why do you stop at this point in the IANA list?
SB>
https://www.iana.org/assignments/address-family-numbers/address-family-numbers.xhtml
SB> It cannot be because some of the later ones are less relevant, as
some of the 
SB> included ones are rather rare. One the other hand there are some
later ones that 
SB> seem modern and useful.
SB>
SB> Also why do you have types in this list that you do not later
define in detail?

===========

SB> You have generalized label in the list top, but not in
SB> the YANG model itself.

===========


<Prev in Thread] Current Thread [Next in Thread>