Re: [netmod] Fw: New Version Notification for draft-bjorklund-netmod-snmp-cfg-00

[David Spakes <[e-mail]>]

On Thu, 21 Oct 2010, Andy Bierman wrote:

> I can't commit to working on the details on this module,
> but I don't mind if it is published.  I agree with Juergen
> that having specific modules published in RFCs is forward progress.
> This also shows how much more powerful YANG is than SMIv2, which is
> important.

In time, the capabilities of YANG that are beyond SMIv2 to model
application data will be commonly exercised in new documents that
are created.  However, IMHO, the wise course of action for the short
term would be to take small steps.

Instead of redesigning "MIB-II" objects in the first go around, I
would suggest writing YANG documents that mirror the existing RFCs
as closely as possible.  Think about vendors who would want to add a
NETCONF server in their product, and then consider how many might
delay that decision because they would have to overhaul their
code that interacts with the stack and kernel to exchange data.


> It turns out that what modules to standardize and in what order
> is rather important.  IMO, interface naming has to be first.
> Almost every YANG module, including 1 in progress in the IPFIX WG,
> needs to refer to interface instances.

I agree with Andy here.  At the very least, the working group should
produce a document that is the YANG equivalent to RFC 2863.  IMHO,
it should exactly mirror the existing objects, at least the ones
that are not deprecated.

-dss


-------------------------------------------------------------
 David Spakes                       email:   [e-mail]
 SNMP Research                      voice:   [phone]
 3001 Kimberlin Heights Road          fax:   [phone]
 Knoxville, TN  37920-9716  USA      http://www.snmp.com
-------------------------------------------------------------