Type of LTP

[italobusi]

In a multi-technology network topology instance, there is a need to identify the type of LTP (e.g., OTN, flexi-grid, ...)

There are two options to define the type of LTP: 1) implicit based on the type of Links terminated on the LTP 2) explicit based on some presence containers and the interface-switching-capability list 3) implicit when the LTP terminates one or more Links or explicit when the LTP does not terminate any Link (e.g., at the edge of a network domain)

Notes:

• if solution 1) is adopted:
  1. inter-domain links MUST always be reported as open-ended Links (to report the type of Link)
  2. the interface-switching-capability list, defined in RFC8795, is always empty
• if solution 2) is adopted there will be duplication of information between the type of Link and the type of LTP when the LTP terminates one or more Links

See: https://github.com/ietf-ccamp-wg/draft-ietf-ccamp-flexigrid-yang/files/10510430/new-ml-topology-models-00.txt

Need to decide which option to adopt and change the YANG model(s) accordingly

[italobusi]

Some slides describing the two options as discussed during the call on 2023-04-27: inter-domain-ltp-00.pptx

[italobusi]

2023-05-25 Weekly call

With option 2 there is duplication of information in the data schema as well as in the data instance (at least for intra-domain links). With option 3 there is duplication only in the data schema but not in the data instance (the information is reported either on the LTP or on the Link terminated by the LTP but not on both).

With option 3 we need to describe that the information in the LTP cannot be present in the data instance when the LTP is terminating a Link: need to check if this condition can be enforced with YANG syntax (e.g., with a when statement), see issue ietf-ccamp-wg/transport-nbi#107

Advantages of option 1 and 3 is to avoid duplicated information in the data instance which can be inconsistent.

Advantages of option 2 and 3 is that they allow not reporting open-ended links, which could be hard to define in case the LTP in unconnected.

Agreement to propose option 3 to CCAMP WG

• [ ] @italobusi : update the YANG model in PR ietf-ccamp-wg/draft-ietf-ccamp-flexigrid-yang#65

[sergiobelotti]

Some slides describing the two options as discussed during the call on 2023-04-27: inter-domain-ltp-00.pptx

I think we should update the slides considering also Option 3. We can exploit today's meeting to adjourn that together.

[sergiobelotti]

2023-05-25 Weekly call

With option 2 there is duplication of information in the data schema as well as in the data instance (at least for intra-domain links). With option 3 there is duplication only in the data schema but not in the data instance (the information is reported either on the LTP or on the Link terminated by the LTP but not on both).

With option 3 we need to describe that the information in the LTP cannot be present in the data instance when the LTP is terminating a Link: need to check if this condition can be enforced with YANG syntax (e.g., with a when statement), see issue ietf-ccamp-wg/transport-nbi#107

Advantages of option 1 and 3 is to avoid duplicated information in the data instance which can be inconsistent.

Advantages of option 2 and 3 is that they allow not reporting open-ended links, which could be hard to define in case the LTP in unconnected.

Agreement to propose option 3 to CCAMP WG

• [ ] @italobusi : update the YANG model in PR Updates to support multi-layer topology instances #65

Option 3 means that in case of inter-domain link and then the need to use LTP information you need presence container or just using interface-switching-capabilitity-list grouping or both?

[sergiobelotti]

Some slides describing the two options as discussed during the call on 2023-04-27: inter-domain-ltp-00.pptx

@all : I've tried to add a slide of summary and explanation , one slide for legenda and one more since the options are 3. @italobusi : please check if the message is correct inter-domain-ltp-01 (2).pptx

[danielkinguk]

Need to develop text that describes how the multi-technology single instance will be used and we can then decide where to place the text, in the flexi-grid topo I-D, another I-D, all technology (Ethernet, OTN, MPLS, et al.) I-Ds?

[sergiobelotti]

I've updated the slide for what I remembered form the last week discussion. inter-domain-ltp-02.pptx

[italobusi]

We have simplified a bit the slides to focus the description only to intra-domain and inter-domain links without describing how the MDSC is processing the information:

inter-domain-ltp-03.pptx

Co-authored-by: sergio belotti sergio.belotti@nokia.com

[danielkinguk]

This is now referenced by ietf-ccamp-wg/transport-nbi#107