Hi Jürgen,

> I am not sure I follow. If I replace the line card, I may have to
> update the type of the interface config. Why would this be disallowed?

Nothing is being disallowed, by this proposal.  There is no new server 
behavior.  The proposal only enables a server to programmatically describe to a 
client what data it considers immutable.


> In the NC world, config is applied to resources. Validation is against
> the data model, not against the resources currently present on a
> system. See for example RFC 6241 section 8.6.1. with starts with:
> 
>   Validation consists of checking a complete configuration for
>   syntactical and semantic errors before applying the configuration to
>   the device.
> 
> This is why we have the notion of <running> -> <intended> ->
> <applied>.  Initially, we did not expose <applied> but operationally
> the possible difference between <running> and <applied> is important
> enough that we started to expose it. But the point here is that the
> resources available do not influence whether <running> is valid. A
> valid configuration satisfies the data model constraints, it does not
> necessarily satisfy the resource constraints.

Agreed.

> 
> /js

K.


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

Reply via email to