Minutes of Technical Community Discussion on June 09, 2016
Date and Time: 6:00am PDT / UTC 13:00, Thursday June 09, 2016
Convener: Bin Hu (AT&T)
- Alok Gupta (AT&T)
Brian Skerry (Intel)
Bryan Sullivan (AT&T)
Carlos Goncalves (NEC)
- Christopher Price (Ericsson)
Dan Radez (Redhat)
Dave Neary (Red Hat)
Gerald Kunzmann (DoCoMo)
Juha Oravainen (Nokia)
Larry Lamers (VMWare)
- Mark D Gray (Intel)
- Mark Szczesniak (Casa)
- Michael Kearns
- Narinder Gupta (Canonical)
- Olga Havel (Huawei)
Pasi Vaananen (Artesyn)(Stratus)
- Pierre Lynch (Ixia)
Prakash Ramchandran (Huawei)
- Randy Levensalor (Cable Labs)
Stuart Mackie (Juniper)
Ulrich Kleber (Huawei)
- Yingjun Li (Huawei)
Active Project Proposal Discussion
Yingjun (Huawei) introduced OPNFV-MANO proposal, and indicated that it focuses on NFVO requirements for integration. Some revisions are made based on comments from last week, e.g. collaboration with other OPNFV projects. Requirement and integration is limited to P1 interface (green in slides), OPEN-O is out of scope. Integration based on P2 will be considered in next step.
One question was what API test means here.The proposal says we are a requirement project, Prakash (Huawei) said that it is a paper exercise, e.g. documenting test cases, and there is no code.
Bryan (AT&T) indicated that testing should be done on other dependent projects.
Team also discussed the dependencies on Promise, Doctor, Models, MOVIE, Domino, VES and ETSI standard.
Gerald (DoCoMo) indicated that the proposal is confusing. Doctor and Promise are not NFVO. Different things seem to be mixed together in the “project dependencies” slide. Moreover, the alignment between ETSI NFV and OPNFV is already looked into in the Doctor and Promise projects on the related operations.
Chris (Ericsson) raised a concern that this proposal seems different from other proposals. Other proposals are supported by SMEs. This proposal covers very broad area - the entire MANO area. Chris wonders if MANO indeed fits the scope of a single project. Chris suggested to look into WG model. And Chris also suggests to deccompose it into several smaller project with clear scope and concrete deliverables.
Carlos (NEC) has 3 concerns - (1) the scope is too broad (2) committers and contributors are only from a couple of companies, mainly are OPEN-O members. (3) OPEN-O is yet not an open community. The project is still being formed by company members in closed-doors. As such, it is my personal understanding that it would not be a good time to approve a MANO project that regards OPEN-O just yet until it’s open (as in open governance, open source code, etc).
Prakash said OPEN-O was open 2 days ago (with reference to Thursday June 9). Carlos said he needed to board a plane and would check later. (PS: Carlos shared his findings in this email thread on June 11).
Dave (RedHat)'s concerns include broader scope, difficulty for others to participate because it is heavily associated with OPEN-O. Dave further suggested to consider leveraging Doctor, VES etc. as a starting point, and consider a common service across projects. And also he suggested to break it into several more focused on projects, because of lack of resources while trying to address large scope.
Gerald indicated that DoCoMo is neither a committer nor a contributor. He requested that please remove DoCoMo's name from the list.
Bin agreed to Chris and Dave, and suggested a bottom up approach by starting from several more focused projects.
Dave asked if the proposal can be raised to the TSC without consensus on the community call, and if the project does not change and it is raised to the TSC for a vote what will happen next.
Chris said it is up to the project to decide. According to our processes a proposal may be brought to TSC with or without community consensus after 2 weeks of community discussion. When asked if proposals had been rejected by the TSC, Chris stated that proposals have been rejected in the past, and can be rejected in the future, elaborating that community consensus is an important component of the TSC’s decision process.
Chris said he will bring a topic of MANO WG in next TSC meeting.
Bryan also indicated that MANO WG is different from particular projects. Convergence is good. But we need particular topics of convergence, and focuses of each project.
Prakash said that he would change the project name to OPNFV-OPEN-O.
Gerald indicated that changing the name doesn't affect the project scope. We need to re-scope and make the project more focused.
Bin indicated we only have 2 minutes left. Bin suggested to remove DoCoMo's name according to Gerald's request. Bin also suggested to list only the names of current committers and contributors. Although the good intention is to invite additional proposed committers and contributors, it is good to communicate with them offline instead of listing them publicly, which may be misleading.