Re: [DNSOP] Call for Adoptions: draft-lhotka-dnsop-iana-class-type-yang

[Normen Kowalewski <[e-mail]>]

Dear Paul, Benno,

thanks for your replies.

> On 7. Oct 2019, at 19:31, Paul Wouters <[e-mail]> wrote:
> 
> On Mon, 7 Oct 2019, Benno Overeinder wrote:
> 
>> Questions to WG:
>> 
>> 1) iana-class-type-yang document to OPSAWG?
> 
> I would assume most people here will the same about the document,
> wherever it is discussed ? So this option seems odd.
> 

We should IMHO be as close to the DNS experts as possible, to me that feels rather like DNSOP than like OPSAWG.  
A "YANG doctors” expert review is part of the publication process, so this will happen anyway. It’s also worth nothing that one author is one of these experts himself.


>> 2) follow-up work on YANG data models for DNS servers in DNSOP?
> 
> Speaking for myself, as long as we are not populating RFCs with
> obsoleted DNS data or just create RFC with copies of IANA registries,
> I'm fine with helping on a document. But not if it is a blind copy
> and paste from IANA (whether at DNSOP or OPSAWG)
> 
> Paul

In order for the IANA registry to be re-used by other YANG modules, a YANG modelled version of the registry is needed. There is precedence for doing this for other registries. e.g. The IANA ifType registry has a corresponding YANG module in RFC8343. A number of other registries also have corresponding YANG modules published or in draft (iana-identity-mib, bfd-parameters, smi-numbers).

Updating and maintaining the contents of the IANA registry as a whole is an orthogonal question to creating a YANG representation of an existing registry and should be handled as a separate task.

Finally, one reason why we would like to see this draft adopted is because we’d like to use it within a real world DNS server implementation.

BR, 

Normen