search

ietf-ivy-wg / network-inventory-yang

Other
2 stars 4 forks source link

Terminology and use cases clarification #16

Open YuChaode opened 1 year ago

YuChaode commented 1 year ago

This is the slide we discussed in last week Meeting Material_02.08.pptx

italobusi commented 1 year ago

2023-02-08 Network Inventory weekly call

Reviewed slides from Chaode

Agreed to clean-up the terminology in the draft and provide some description of the use cases in the scope of our draft

The figure describing the different OSS systems can be simplified to focus only on the OSS systems which are relevant for the inventory use cases.

oscargdd commented 1 year ago

The ITU-T definition seems to cover wider scope. Please be clear in the draft what is the scope here. Is there a term such as hardware inventory?

italobusi commented 10 months ago

2024-01-17 Network Inventory weekly call

Need to check for existing definitions in TMF and ITU-T

italobusi commented 9 months ago

2024-01-31 Base Network Inventory weekly call

A 'physical thing' can be seen as something that can be measured with a ruler and potentially weighed with scales

A NE is not a physical thing: inventory can consider also useful 'grouping/structure of physical things' like network elements (sometimes called device)

Note: in case of geographically distributed components, with the current model it is possible to group these distributed components as multiple NEs or as a single NE.

Note: it is important to also consider how the information about the physical things is accessed (this is often an aspect of NE)

The 'functional behavior' is outside the scope of the inventory work: the topology, interface and other YANG models are covering the functional behavior. Some navigation capability is needed and the definition of the navigation can be in the scope.

Need to clarify that the term inventory in our context is not an unconstrained 'list of anything' but be more precise about the scope of what we are listing. In our context for this initial work, inventory is a representation of the structure of physical things (i.e., not just a list and not all things).

There are informational things. Software is an 'informational thing' and it is in the scope of the work (although perhaps not in the first stage). Note that software is not {{Italo: I added not here as I think that that is what you meant to type and it is certainly what I was trying to say}} intended to be the running software which instead is representing the emergent 'functional behavior'.

A traditional NE is an assembly of physical things, informational things, the functionality emergent when powered along with the method of access. A virtual NE does not directly have physical components. There are also hybrid NEs with some virtual components and some physical components.

It has been suggested also to look at the definitions in draft-palmero-ivy-ps-almo (in particular the asset definition) to see if there are also some concepts that can be used to defined what is in the scope for IVY

italobusi commented 2 months ago

Some proposal sent to the mailing list:

https://mailarchive.ietf.org/arch/msg/inventory-yang/IvgK1ca7Sq8TTOfBv9-JbcqzcI0/

italobusi commented 2 months ago

2024-09-11 Base Network Inventory weekly call

Italo has sent some thoughts to the IVY WG mailing list before the IVY WG session at IETF 120:

https://mailarchive.ietf.org/arch/msg/inventory-yang/IvgK1ca7Sq8TTOfBv9-JbcqzcI0/

italobusi commented 2 days ago

2024-11-20 Base Network Inventory weekly call

Discussed in the IVY WG meeting during IETF 121:

Keep silent in this draft about the'energy-object' component type: IVY Mailing List Discussion

In order to be able to provide some text proposal, it has been agreed to keep the RFC 8348 definitions for port, container, chassis and stack and update them based on other comments resolution. A note can be added to clarify to the readers that the definitions of these components is under review and may be changed in future versions of the draft.