On Wed, May 7, 2014 at 3:56 AM, Susan Hares <[email protected]> wrote:

> Andy:
>
>
>
> Thank you for detailed discussion with Tom Petch/Jeff/Juergen/Tom Nadeau:
>
>
>
> <snip>
>
> I imagine I2RS will be completely separate from NETCONF and it should have
> its own datastore -- so "i2rs-config" is appropriate because I2RS is the only
> protocol using that datastore. The combined "operational state" is not
> editable.
>
>
>
> Yes, that makes sense. Then if at some point in time you want to save  the
> running config (i2rs), the system has to explicitly copy it over. But until
> then, its separate.  The only question is: what is the complete running
> config represented as? And what if Netconf wants to modify the running
> config (i.e.: and the RIB in particular)?
>
>
>
> That copy-to-local-config feature would be extra, outside the scope of the
> i2rs-config.
>
> </snip>
>
> -----
>
>
>
> Why only is the copy-to-local-config feature out of scope for the
> I2rs-config?   How can the work be interoperable if this feature (required
> or optional) does not work the same in different vendor's boxes?  If the
> implementations are different, this is great (differentiated products =
> good); but IMHO this feature actions has to be interoperable... that is give
> the same effective response in the different vendors.
>
>
>
> Can you explain further?
>
>
>

Writing to the local config is the charter of the NETCONF WG, not I2RS.
In order to copy from the I2RS data model to the configuration data model,
a detailed mapping is required.  Since the local-config is out of scope,
the config data corresponding to the I2RS data is not required to exist.
If it does exist, it can be vendor-specific, right? So the mapping is
vendor specific.

I would agree with you if I2RS was planning on standardizing data models
for configuration and RIB editing.  It that were the case, then I would
re-ask
the question Ron Bonica raised in London -- why would we want to write
config data models with 1 language and I2RS data models with a different
language?


> <snip>
>
>
>
> Right. If we take that approach we need to have a way of the i2rs
> "protocol" to signal that the config
>
> needs to be saved "soon".  That is, its not done in real time, but rather
> as a background process.
>
> Its probably. This can either be through a single keyword that is added
> under the i2rs-config section, or
>
> per element or even per-element, if we decide to be more granular.
>
> </snip>
>
>
>
> IMHO it is important to know when the write to local-config is done even
> if done in background.   Are you indicating there is no
> write-config/ack-complete-write config interaction?
>
>
>


> <snip>
>
> You seem to be suggesting that the I2RS behavior of forgetting all
> injected state
>
> if the router happens to reboot is not that useful, and therefore it needs
> to be
>
> saved in the background (best-effort).
>
>
>
> I think it is up to all the I2RS clients to always be around to receive
> some
>
> sort of agent-reboot event, and re-install any required I2RS config.
>
> I can't imagine any use cases where the operator thinks "I only need to
>
> install this RIB data until that router accidentally reboots".
>
> </snip>
>
>
>
> If you re-install any required I2RS config, are you sure this is what the
> I2RS client wants?
>

The protocol request would indicate what the client wants (e,g,
copy-to-local config flag set).
Since notifications are unreliable, a client will poll the 'agent-boots'
counter frequently to
see if needs to re-install lost state.  This design does not scale well and
introduces significant
latency.  It is much more convenient to have the agent deal with
reinstall-after-reboot.



>
> Sue
>
>
>

Andy
_______________________________________________
i2rs mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/i2rs

Reply via email to