Open Stack

Howdy everyone,

Those of us involved in the placement extraction process sat down 
together today to discuss the plan for openstack/placement governance. 
We agreed on a set of criteria which we will use to determine when we 
will switch the openstack/placement governance from the compute project 
to its own project. I'd like to update everyone with a summary of the 
plan we agreed upon.

Attendees: Balázs Gibizer, Chris Dent, Dan Smith, Ed Leafe, Eric Fried, 
Jay Pipes, Matt Riedemann, Melanie Witt, Mohammed Naser, Sylvain Bauza

The targets we have set are:

- Devstack/grenade job that executes an upgrade which deploys the 
extracted placement code
- Support in one of the deployment tools to deploy extracted placement 
code (Tripleo)
- An upgrade job using any deployment tool (this might have to be a 
manual test by a deployment tool team member if none of the deployment 
tools have an upgrade job)
- Implementation of nested vGPU resource support in the xenapi and 
libvirt drivers
- Functional test with vGPU resources that verifies reshaping of flat 
vGPU resources to nested vGPU resources and successful scheduling to the 
same compute host after reshaping
- Lab test with real hardware of the same ^ (xenapi and libvirt)

Once we have achieved these targets, we will switch openstack/placement 
governance from the compute project to its own project. The 
placement-core team will flatten nova-core into individual members of 
placement-core so it may evolve, the PTL of openstack/placement will be 
the same as the openstack/nova PTL for the remainder of the release 
cycle, and the electorate for the openstack/placement PTL election for 
the next release cycle will be determined by the commit history of the 
extracted placement code repo, probably by date, to include contributors 
from the previous two release cycles, as per usual.

Thank you to Mohammed for facilitating the discussion, we really 
appreciate it.

Cheers,
-melanie