Dear all,
Below is a summary of the meeting on Nov 24, 2020.
Richard proposed some approaches to improve the productivity and to engage
people outside the group. Gang and Luis also proposed some principles to be
followed.
Sabine gave a summary of the entity property map document and was asking the
co-authors to review the document.
Richard raised the problem of whether entity property map is sufficient to
express CPU/storage. In a more general setting, information in the network
which can be modeled as a graph can be associated with nodes (entities) but
also with edges (relations). Is entity property map capable of expressing the
properties of the edges?
Kai mentioned that edges can also be treated as entities, as how graphs are
being stored in relational databases.
Richard argued that this approach was too hacky and proposed to use syntax to
enforce semantics.
In the discussion on distinguishing the recharter items related to cellular
information, Kai proposed to divide the current first-hop recharter item into
the northbound part (identifying and specifying cellular metrics) and the
southbound part (information collection using 3GPP services, etc.) which can be
put into the operation automation item. Richard suggested that defining the
formats of metrics is too simple and we should tackle the difficult problem of
identifying which metrics are useful and practical.
Gang suggested that the metrics should be motivated by real use cases. As Gang,
Chunshan and Sabine had proposed use cases in previous drafts, it is agreed
that they would present these use cases and lead the discussion on the cellular
information in the next meeting.
Open issues:
- Richard/Kai/Jensen will provide reviews to the unified property document
- Gang/Chunshan/Sabine will present/lead the discussion on cellular information
in the next meeting. One item is to present the information flow in the current
4G/5G networks.
- Jensen will also present the ideas on operation automation
Please let me know if anything is missing or misinterpreted.
Best,
Kai
_______________________________________________
alto mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/alto
