Open Stack

On 8/7/15, 07:19, "Nicolas Trangez" <nicolas.trangez at scality.com> wrote:

>On Fri, 2015-08-07 at 01:21 -0400, Nikhil Komawar wrote:
>> 3. ease of addition of newer drivers for the developers -- drivers
>> are
>> only being removed since.
>
>The OpenStack team at Scality developed a Glance Store driver for RING,
>currently out-of-tree to get to a first fully-working version (
>https://github.com/scality/scality-glance-store), but with the
>intention from day 1 to propose this driver for inclusion in the
>upstream glance-store project, following the standard OpenStack
>processes.
>
>Whilst as you say new drivers haven't been proposed since the split,
>this could be explained by the fact the way Glance is designed now
>explicitly supports out-of-tree drivers (in glance-store or elsewhere)?
>
>We have a couple of questions related to this proposal:
>
>- Would folding glance-store back into glance have any impact on the
>process (or reluctance) to include new third-party drivers?
>
>- Will the glance-store core reviewer team be merged back into glance,
>focusing on the store drivers?
>
>Nicolas


Thanks for responding Nicolas. I would agree that it's very plausible that
new drivers have not been proposed because glance_store presently allows
for drivers to be loaded via entry-points
(https://git.openstack.org/cgit/openstack/glance_store/tree/setup.cfg#n26).
 All one would need to do is have their package register an entry point on
glance_store.drivers and they will then be loaded by glance_store. In
other words, there's absolutely no need to merge the scality driver
upstream. It's already should work perfectly.

That said, I think you struck on a point accidentally. Currently the
glance_store core reviewer team and the glance core reviewer team are one
and the same. I think there are reviewers for the glance_store project
that would make great *glance_store* core reviewers but still need to be a
little bit more active in glance before becoming Glance core reviewers.
They contribute primarily to specific drivers but also provide reviews for
other drivers as well. Given the overall hesitancy of the Glance project
members to use a trust model (like Neutron and others are experimenting
with) to add new cores for specific subsystems, this might be a good
micro-experiment for us - add driver specific cores who are trusted to
only +2/+2+W on their specific driver in the beginning. I think this would
provide a lot of value to us as an overall project.