Re: [netmod] review of draft-acee-netmod-rfc8022bis-05

Vladimir Vassilev <vladimir@transpacket.com> writes:

> Hello,
>
> I have reviewed draft-acee-netmod-rfc8022bis-05. My conclusion is that 
> the YANG modules part of the draft have been successfully modified in 
> accordance with sec. '4.23.3 NMDA Transition Guidelines' of 
> draft-ietf-netmod-rfc6087bis-14. The modifications are coherent with the 
> ietf-interfaces@2017-08-17.yang module in 
> draft-ietf-netmod-rfc7277bis-00 and ietf-ip@2017-08-21.yang module in 
> draft-ietf-netmod-rfc7277bis-00.
>
> Vladimir
>
>
> Review of draft-acee-netmod-rfc8022bis-05.
> Vladimir Vassilev
> 2017-10-30
>
> 'Abstract':
> 'Introduction 1':
>   - Both Abstract and Sec 1. contain duplicated text which can be removed
> from Abstract. The text in Sec 1. can be simplified:
>
> OLD:
>     This version of these YANG modules uses new names for these YANG
>     models.  The main difference from the first version is that this
>     version fully conforms to the Network Management Datastore
>     Architecture (NMDA).  Consequently, this document obsoletes RFC 8022.
> NEW:
>     This version of the Routing Management data model supports the Network
>     Management Datastore Architecture (NMDA) 
> [I-D.ietf-netmod-revised-datastores].
>
>
> '7.  Routing Management YANG Module':
>
>   - Why should address-family identity be different e.g. mandatory 
> "false"; for system created RIBs? I think this needs some explanation 
> (Page 21):
>             ...
>             uses address-family {
>               description
>                 "Address family of the RIB.
>
>                  It is mandatory for user-controlled RIBs.  For
>                  system-controlled RIBs it can be omitted; otherwise, it
>                  must match the address family of the corresponding state
>                  entry.";
>               refine "address-family" {
>                 mandatory "false";
>               }
>             }
>             ...
>

Following the logic of RFC 8022, address-family has to be a mandatory
parameter for all RIB entries in <operational>, so there seems to be no
other choice than make it "mandatory true" in the schema, as NMDA only
has one schema for all datastores.

Lada

>   - Suggested change of 'base address-family;' -> 'base 
> rt:address-family;' for identity ipv4 and ipv6 (ref. 
> draft-ietf-netmod-rfc6087bis-14#section-4.2):
>
>      o The local module prefix MUST be used instead of no prefix in
>      all "default" statements for an "identityref" or "instance-identifier"
>          data type
>
> '8.  IPv4 Unicast Routing Management YANG Module' 
> (ietf-ipv4-unicast-routing@2017-10-14.yang):
> '9.  IPv6 Unicast Routing Management YANG Module' 
> (ietf-ipv6-unicast-routing@2017-10-14.yang):
>
>
>   - The ietf-ipv4-unicast-routing and ietf-ipv6-unicast-routing modules 
> import the ietf-routing without revision (ref. rfc6087#section-4.6):
>
>
>      o The revision-date substatement within the imports statement SHOULD be
>      present if any groupings are used from the external module."
>
>
> 'Appendix D. Data Tree Example':
>
>   - The example in the Appendix D. has not been updated and it must be 
> extended in order to demonstrate a usecase of operational datastore of 
> configuration data with different origin (intended, system, etc.) 
> similar to the 'Appendix C. Example Data' of 
> draft-ietf-netmod-revised-datastores-05.
>
>
> Nits:
>   - s/Figures 1/Figure 1/
>   - s/systemindependently/system independently/
>
> _______________________________________________
> netmod mailing list
> netmod@ietf.org
> https://www.ietf.org/mailman/listinfo/netmod

-- 
Ladislav Lhotka
Head, CZ.NIC Labs
PGP Key ID: 0xB8F92B08A9F76C67