- Aihua Guo
- Fabio Peruzzini
- Italo Busi
- Jeff Bouquier
- Julien Meuric
- Nigel Davis
- Sergio Belotti
- Yu Chaode
None
- May 3rd at 3pm CEST / 9am NA EDT / 9pm CST
- May 10th at 3pm CEST / 9am NA EDT / 9pm CST
Note: on May 3rd Nigel will present the T-API Equipment model
Alternative option identified could be Monday 3pm CEST, starting from May 8, 2023
- Italo: check with other participants if Monday 3pm CEST is ok
Feedbacks from Fabio: he is not available on monday and friday
The only option left would be to start 1h later, i.e., at 4pm CEST, starting from May 10th
- Italo: check with other participants if Wednesday 4pm CEST is ok
- Jeff/Fabio: send a mail to the inventory mailing list expressing the urgent need from operators perspective
We can leverage on the operators' feedbacks to push the WG to address the hardware and software inventory models in a step-by-step approach
- Italo/Chaode: Check whether these examples are already covered with the examples in #55
Slides from Italo: https://github.com/ietf-ccamp-wg/ietf-network-inventory/files/10843102/multi-rack-NE-questions-00.pptx
Nigel: there are examples of multi-rack NEs in the old switching technologies. There are also cases where a NE is delivered as a single rack. The key issue is whether the NE has details information about the rack: in particular if the rack has some circuitry embedded in the rack with some serial-number
There are cases where also the chassis is not part of the NEs where multiple NEs can be hosted in the same chassis: in this case, the NE will not know in which slot it is plugged in
We need to be careful in order not to overlay constraint the inventory model
Detectable information is information that can be checked by the NE (e.g., the slot a board is located within a shelf that belongs to the NE) while non-detectable information is information that cannot be checked by the NE (e.g., the room location) and the boundary may vary depending upon the specific construction of the NE (note that the boundary can also be different if the same type of NE is deployed differently)
We can discuss next week how this flexible structure has been partly supported in TAPI and investigate even further improvements
Italo: maybe it is better to decouple the splitting based on the type of NE to avoid proliferation of ways to manage the same NE type depending the way it is installed. Some additional information can be added to the information that is not always detectable to indicate if it has been detected or not
Sergio: can we say that what is detectable can be subject of software updated?
Maybe the difference could be between what is detectable by the software internal information and what can be detectable only through external information if available