William Lupton <[email protected]> wrote:
> So could “being writable” be associated with a feature (which might
> potentially apply to multiple such nodes)? W.

We can put a if-feature statement on the node 'mfg-name' in the config
tree, but not in the state tree.

But the issue is that there are two conflicting use cases:

  1)  use the writable nodes for validation with the real hw
  2)  use the writable nodes as a "post-it" function

Currently the only such node is 'serial-num', and it is used as (2).

But maybe we can actually combine these two cases.

  If the node has a value in the config, and the system can detect the
  real value from the hardware, then they MUST be equal, otherwise the
  config is not applied. (case (1) above)

  If the node has a value in the config, and the system cannot detect
  the real value from the hardware, the config is applied and the
  state value will be the config value. (case (2) above)

Maybe this violates the principle of least astonishment though...


/martin


> 
> (and if so, what’s the cleanest way to achieve this? it seems that
> “if-feature” can’t be used with “config” but that it could be used by
> refining a grouping?)
> 
> > On 22 Dec 2016, at 05:14, Randy Presuhn
> > <[email protected]> wrote:
> > 
> > Hi -
> > 
> > On 12/21/2016 6:56 AM, Martin Bjorklund wrote:
> > ...
> >> This procedure would change how serial-num currently is handled.  I
> >> don't have a strong opinion on this, but I note that serial-num is
> >> writable in the ENTITY-MIB.  Maybe someone can comment on why it is
> >> writable in the MIB, and if this is a use case that we want to
> >> support?
> > 
> > I *think* this is an example of what were known as "post-it note"
> > objects.  Their values could be set by management action
> > to accommodate implementations in which, for example, the hardware
> > provided no way for software to access the information printed on an
> > inventory sticker affixed to a card.
> > 
> > Randy
> > 
> > _______________________________________________
> > netmod mailing list
> > [email protected]
> > https://www.ietf.org/mailman/listinfo/netmod
> > 
> 
> _______________________________________________
> netmod mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/netmod
_______________________________________________
netmod mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/netmod

Reply via email to