Reviewing the l2nm module, I made the comments below which are perhaps of a 
more general application.

Looking at the tree diagram snippet, I see a seamless list of nodes in e.g. 
section 7.3.  In fact, there are four 'uses' tucked away in there, three of 
them imports from the module vpn-common, so it is some effort to track down the 
source YANG statements that the tree diagram references   I see no way of 
telling from the tree diagram what is 'native' and what is the result of 'uses'.

Second, again using section 7.3 as an exemplar, there is text giving 
description and usage of key nodes for both native and imported nodes.  The 
imported nodes also have their own description  and usage in RFC9181 where 
vpn-common is defined.  It seems to me that the descriptions for the  common 
module in RFC9181 are rather thin so while having a more comprehensive 
description in the l2vpn I-D may help, it may also confuse if the two are not 
in line; and of course, the author of another I-D which imports vpn-common may 
decide to create their own and different 
descriptions.

As ever, I am better at problems than solutions but it seems that when we 
create a separate RFC for  a common module, then the descriptions in the YANG 
need to be really comprehensive, more so than when the descriptions can be 
amplified by the text in the RFC in which a module appears, be it common or no.

Tom Petch


From: OPSAWG <[email protected]> on behalf of tom petch 
<[email protected]>
Sent: 12 May 2022 17:08

This I-D imports and uses a number of YANG grouping from RFC9181,
ietf-vpn-common. This fact gets a mention in section 5 but you would not
know this from the descriptive section, section 7, where imports and
native YANG are seamlessly merged.

This I-D contains text on the meaning of the nodes, their usage and such
like regardless of where they are defined.  Looking at RFC9181 it is
rather short, or very short, on such text (e.g. in vpn-profile-cfg,
vpn-description).  It would seem then that each document that imports
RFC9181 will have to incorporate its own, potentially divergent, text on
the meaning of the nodes and their usage while if ietf-vpn-common gets
updated, the ripples may spread.

This seems unfortunate.

Tom Petch

On 29/04/2022 15:40, The IESG wrote:
>
> The IESG has received a request from the Operations and Management Area
> Working Group WG (opsawg) to consider the following document: - 'A YANG
> Network Data Model for Layer 2 VPNs'
>    <draft-ietf-opsawg-l2nm-15.txt> as Proposed Standard
>
o

_______________________________________________
netmod mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/netmod

Reply via email to