Differentiation of inventory data vs monitoring data

[oscargdd]

In my view, the yang model augmenting the inventory should contain only "static" information, That is, the capabilities of the device or the components to save energy, For example, the part of +--rw energy-saving-modes {energy-saving}? is clear resides in the inventory.

However, the metrics (raw or cooked in some form of KPI), should go to a different model. The base structure of nodes could be similar, but I expect quite a different usage.

[williamchern]

Hi, Oscar, as discussed with Qinwu, it is the goal to keep only static information in the inventory, while other metrics goes to other models. Could you please give a more concrete example here, thanks.

[boucadair]

However, the metrics (raw or cooked in some form of KPI), should go to a different model. The base structure of nodes could be similar, but I expect quite a different usage.

Currently those are sperate in the module: e.g., augment "/ni:network-elements/ni:network-element" vs. augment "/nw:networks/nw:network/nw:node"

I guess your comment is to keep only augments to ni in one module and nw augments in a separate module.

that's would be cleaner and easy to do.

[boucadair]

Please see #27 for a proposal to split the module into three main areas:

• common
• ntw augmentation
• ni augmentation

I'm not sure yet how/where to put the component container

[billwuqin]

The proposed split make sense to me, the only comment I have is you remove Energy efficiency ratio from energy-consumption-data grouping, which is also a metric or dyanmic data we need to incorporate in.