Open Stack

Armando M. <armamig at gmail.com> wrote:

>
>
> On 4 March 2016 at 08:50, Ihar Hrachyshka <ihrachys at redhat.com> wrote:
> Hi all,
>
> currently we have both py27 and py27-constraints tox targets in neutron  
> repos. For some repos (neutron) they are even executed in both master and  
> stable/liberty gates. TC lately decided that instead of having separate  
> targets for constrained requirements, we want to have constraints applied  
> to default targets (py27, docs, …), unconditionally; we also want to use  
> those ‘default’ targets in gate; and we also want to eventually get rid  
> of those -constraints tox targets.
>
> To achieve that, I sent a set of patches spanning neutron, neutron-*aas,  
> and project-config repos:
>
> https://review.openstack.org/#/q/status:open+branch:master+topic:neutron-constraints
>
> For the very least, we want to get our mitaka gate switched to ‘default’  
> (but constrained) tox targets before final release, so that we have a  
> solid foundation in the stable/mitaka branch that would reflect TC  
> desires.
>
> Those important patches are (in order of merge):
>
> for mitaka:
> - https://review.openstack.org/286778: makes ‘default’ tox targets  
> constrained;
> - https://review.openstack.org/286777: switches mitaka gate to using  
> ‘default’ targets;
> - https://review.openstack.org/288516: cleans up -constraints targets;
>
> for liberty:
> - [not proposed yet; waiting for 286778]: makes ‘default’ tox targets  
> constrained;
> - https://review.openstack.org/288506: switches branch back to ‘default’  
> targets;
> * we probably don’t want to drop old targets since some external users  
> may already rely on them
>
> There are also patches to constrain remaining gate jobs (releasenotes,  
> cover) too:
> - https://review.openstack.org/288517: neutron
> - https://review.openstack.org/288472: lbaas
> - https://review.openstack.org/288470: fwaas
> - https://review.openstack.org/288443: vpnaas
>
> ...though those depend on some project-config work:
> - https://review.openstack.org/288451: releasenotes
> - https://review.openstack.org/288455: coverage
> * note those also depend on another patch for zuul-cloner
>
> Thanks for attention and reviews,
>
> This is mainly a question of timing: when shall we pull the trigger on  
> all of these? I am happy to do it today, but it's already Friday  
> afternoon in some parts of the world and changes span multiple projects…

I would not advice to pull it till Monday. I will revise the patches,  
including gate votes, early on Monday; then once everyone from US timezones  
is online, we may push first pieces in.

In the meantime, it would be great to see it validated by reviewers  
nevertheless.

Ihar