Skip to end of metadata
Go to start of metadata

 | 9am PDT |  noon EDT | 16:00 UTC | 18:00 CEST | 21:30 IST 


ZOOMhttps://us02web.zoom.us/j/89069708424?pwd=aGJOZm54eTUxd0FXR0VCU1N0ejBrUT09
Meeting ID: 890 6970 8424

Attendees

Informed

Discussion items

TimeItemWhoNotes
00:00Admin

Next meetings: 

2021-07-14: John Keeney

2021-07-21: Martin Skorupski

2021-07-28: John Keeney 

2021-08-04: Martin Skorupski

00:05O-RAN-SCO-RAN-SC PTLs

O-RAN-SC ToC - D-release was approved today


Status reports:

00:20NexusAlex Stancu

How to use O-RAN-SC Nexus and what is the difference in ONAP

Current consequence: manually changes of the environment variables

Subject for future improvements???? 

trick:

  • use in helm always released images and for testing an overwrite helm

Proxy on LFN level

00:30SMO deployment options

SMO deployment options

  • no tested helm for OAM in O-RAN-SC D-Release - idea update the one from C-release
  • for Non-RT-RIC a new helm is available - still some updates to be done
  • Please see presentation within  REQ-887 - Getting issue details... STATUS

Meeting? on Friday 6/25

https://lists.onap.org/g/onap-meetings/viewevent?repeatid=37969&eventid=1192141&calstart=2021-06-25


Activity

  • kubernetes updates
  • issues with certs addressed
  • reuse of O-RAN-SC helm 
  • Winriver lab


<Martin Skorupski needed to leave. John Keeney continues minutes. See also: 2021-06-23 Meeting notes - Joint OAM / NONRTRIC / SIM SCRUM meeting >


INFO from Christophe - 2021-06-30:

Just a few notes:

  • We have agreed with ONAP and other members of O-RAN SC to keep an ad-hoc call on Friday to discuss technical details for now (will remove this meeting once all details are sorted out and use the joint meeting to report status)
  • Main comment from John is that we cannot consider an SMO deployment without having a way to deploy Non-RT-RIC as well as ONAP components (hence a debate on how this should materialize : extend OOM? )
  • We have progressed on the POC in the meanwhile, Sebastien had an idea and has pushed a small repository on github to show what we are doing for now to create a unique installation for O-RAN, trying to implement the following idea:
    • Started from IT/DEP repo, containing some Helm charts for O-RAN (including Non-RT-RIC)
    • Seems an embryo of using ONAP charts to deploy SMO components but sounded like a fork of ONAP charts
    • Instead of copy pasting OOM charts, he create a git submodule referring to oom ONAP charts repo (to keep ONAP untouched)
    • Rework the O-RAN charts to map them to the OOM structure (like OOM'ized O-RAN charts)
    • Try and build (created some makefiles, inspired from ONAP) then push them to a single chart repository
    • Merged ONAP override and O-RAN override to have a single file
    • Created an Install script to deploy the charts in a single Namespace
    • … it worked more or less

We plan to show that maybe on Friday call (https://wiki.onap.org/pages/viewpage.action?pageId=103422272)

--

Are the following pages related?

  • CNF/VNF deployments

2021-07-07:

way forward


Q/A

00:16

<End>



END






00:00Use Case

@all

OSC Proposed e2e integration use case: O-RU FH connection recovery

<bridge dropped recording stopped at this point. Most participants reconnected once bridge started again>

  • Topology ??: (contd)
    • John Keeney Lots of options for how this should be done properly in ONAP.
    • Swaminathan Seetharaman ONAP usecases already have some plans for this e.g. slicing, oof,  etc
    • Andy Mayer A model should be agreed in ONAP. Lots of previous discussions in ONAP - e.g. what goes in A&AI and CPS ... but that was a while ago.
    • Martin Skorupski SDNR has some topology model too - not standards based.
    • @Sven John Keeney Martin Skorupski Lots of issues with different topology models ... and most existing hierarchical models break down in a 5G/CloudRan environment
    • Lots of work ... Andy Mayer Martin Skorupski Can create a JIRA in ONAP MODCOM ...
    • ... but will keep it simple here for this use case.
  • @? Will other O1 usecases be effected? Martin Skorupski i No. O1 functions continue as before.
  • Martin Skorupski Please add feedback if any



Action items