Open Stack

On 24 August 2015 at 05:14, Shamail Tahir <itzshamail at gmail.com> wrote:
> Hi John,
>
> The team reviewed the existing nova priorities[1] that will probably be
> continuing forward into Mitaka along with those items that did not have a
> clearly defined priority[2].  The team agreed that the current list of nova
> priorities looks great and that it makes sense for them to continue to
> remain the same (or similar) moving forward.

Thank you for talking a look at those. Sounds good.

> those items that did not have a clearly defined priority[2].

These are basically things were we think they are high priority for
the project, but have been unable to find developers to work on them.
Help getting folks to work on those, would be very welcome :)

> The team noticed that most of the existing priorities align with stability,
> scalability, and/or resiliency of the service and we realize how crucial
> these themes are to the experience of the operators.

+1

> The only one that the
> team couldn't directly tie to these themes was the scheduler work, however
> it will be a great enhancement to the architecture when fully realized
> (based on information that was shared with some folks at the mid-cycle).

The current scheduler is not very reliable, and proving hard to maintain.

A long standing feature request is being able to schedule using data
from Cinder volumes, Nova compute resources, and Neutron IP
availability, all at the same time. This work is a requirement to do
that in a maintainable way.

Also, we have identified this as a part of the code that could become
its own project or sub-project.

> The team wanted to say a collective thanks to the nova team and leadership
> for doing phenomenal work!

A big thank you to all who are allowing their developers to work on
these priorities.

Also a big thank you to all who give there developers time to review code:
https://wiki.openstack.org/wiki/Nova/Mentoring#Why_do_code_reviews_if_I_am_not_in_nova-core.3F

With out that support, this list would be meaningless.

> Would you and the nova team be able to provide a sentence or two on each of
> the "priorities without a clear plan"?  The team seemed to be interested in
> the following items:
>
> A plan for the future of flavors, image properties and host aggregates
> Simplify Quotas
> Revisit how we talk to Glance, Neutron and Cinder APIs
>
> However, we could not say for certain that these require backlog-specs in
> the next release without having a better context for each one.

The initial plan here was to get some folks to help with these
efforts, and help with that documenting work.

Anyways, more detail might help, so here is my first quick stab at
fleshing those out a little:
https://review.openstack.org/#/c/216167/

> On that note, are you still planning to join our weekly meeting tomorrow?

Yes.

One quick question though, do you have a UK toll free number for that
voice bridge, or a digital way to join? I don't have a work phone, and
a call to the US gets expensive quite quickly.

Thanks,
John

> [1]
> http://specs.openstack.org/openstack/nova-specs/priorities/liberty-priorities.html
> [2]
> http://specs.openstack.org/openstack/nova-specs/priorities/liberty-priorities.html#priorities-without-a-clear-plan
>
> --
> Thanks,
> Shamail Tahir
> t: @ShamailXD
> tz: Eastern Time