Skip to end of metadata
Go to start of metadata

CNTT Working Model

Problem Statement: the current working model for CNTT to affect change within OPNFV is not effective.

  • OPNFV has projects. OPNFV projects have contributors, committers, and a PTL. Effective change happens in an OPNFV project.
  • Creating markdown in Github does not help see those desired changes end up in OPNFV.
  • How many CNTT folks do we see as contributors, committers, or PTL within OPNFV projects?
  • How about OPNFV folks engaging CNTT? (This is most common situation at the moment)
    • OPNFV Functest project is a great example as you have the PTL engaging in both communities
    • Problem is that the PTL has to work twice: once within CNTT via meetings or github and again within the OPNFV functest project
    • This is why we often hear comments about more developer resources needed from CNTT in OPNFV projects
  • Looking at the current model, how does a CNTT "requirement" become incorporated into an OPNFV project?
  • Lab Requirements Use Case
    • CNTT RI1: Lab Requirements (Ch.04) defines changes to the Pharos specification within OPNFV
      • Increase memory to 512Gb RAM
      • Additional disks (at least 3 more)
      • Additional node (6 instead of 5)
      • Leaf/Spine switch topology for HA
    • OPNFV Pharos project deals with developing an OPNFV lab infrastructure via the Pharos Specification and Community Labs
    • How do we go about dealing with this change? What is the workflow? Process?
      • There was no engagement with the OPNFV Pharos project or PTL within OPNFV (just OPNFV folks within CNTT)
      • We don't know why the requirement was needed or any justification for it
    • Has this requirement been implemented? What is the expectation?
      • Community labs: ask member organizations to re-architect labs to incorporate changes?
      • Upstream (Openstack) / Adjacent (ONAP) communities: ask them to also change their lab infrastructure
      • OVP Badging: do we need to make a change here as well?
      • End Users: do end users also need to instantiate these lab requirements in production?
      • How do we confirm compliance? What do we do with non–compliance?
  • How does OPNFV gather requirements?

Focus Areas

Does CNTT have a formal process to gather requirements?

  • For example, collect requirements → community feedback (flush out concepts) → POC validation → Vote/Agreement/Acceptance → Release

Does CNTT have enough in the roadmap?

  • CNTT RI1(RC1) - Openstack based + CNTT RI2 - Kubernetes based → What's next?

Do we have a single source of truth for CNTT? OPNFV?

  • Central documentation instead of multiple sources - LFN wiki, OPNFV wiki, Github
  • Can we make this an OPNFV space? (create a new wiki space for housing all CNTT documentation - call it CNTT)
  • For OPNFV, see project to CNTT chapter mapping idea below

Next Steps in Collaboration

  1. OPNFV Requirements Work Group
    1. proposal to create a work group gather and discuss requirements
  2. OPNFV Project to CNTT chapter mapping
    1. proposal to identify OPNFV projects that map or are impacted by CNTT chapters
  3. TBD

  • No labels