Re: MSO code flow to OpenStack for Orchestration

DeWayne Filppi

No, you're not misreading it.  A layer of mediation is a requirement of multi-VIM.  There is however, no "requirement" (using the term strictly), for SO to use it, particularly if it jeopardizes SO, and therefore the entire system.  Having said that, obviously if multi-VIM is available as planned in R1, and it provides all the needed functionality required by the use cases, then it should be used.  This only makes sense because the MSO already has Openstack capability.  It wouldn't make sense if a brand new interface had to be developed as an alternative.  IOW, SO needs multi-VIM, but has a fallback.


On Tue, Jun 27, 2017 at 5:08 PM, NOSHPITZ, CLAUDE <cn5542@...> wrote:
Apologies if I am misreading this -- it seems like interposing at least some layer of mediation in front of "raw" OpenStack calls is a pretty clear requirement for R1.

If that's already covered in MultiVIM,  then would the question be more about minimizing technical debt due to non-future-proof  workflows or other tooling?


> On Jun 27, 2017, at 3:55 PM, DeWayne Filppi <dewayne@...> wrote:
> I'd say the ideal would be for the MSO/SO to not directly call VIM APIs, but would not want to jeopardize the SO project based on an external dependency that frankly isn't necessary to reach the use case goal for R1.  Just my 2 cents.
> _______________________________________________
> onap-discuss mailing list
> onap-discuss@...

DeWayne Filppi, Director, Solutions Architect

M: +17145121706


Join to automatically receive all group messages.