Open Stack

On Fri, Sep 1, 2017 at 8:27 AM, Marios Andreou <mandreou at redhat.com> wrote:
> I don't think this ^^ patch is what causes/ed that bug, at least I haven't
> found or seen that this is the case. As we discussed a bit on Wednesday
> evening, if you see the elastic-recheck query @
> http://status.openstack.org/elastic-recheck/#1713832 the issue was seen
> twice in the last 24 hours and 5 times since the revert landed early
> Wednesday EU morning. I think that's roughly the same rate it was being seen
> before and after my patch was reverted.
>
> Do you think we can now consider landing that again with
> https://review.openstack.org/#/c/499116/ please ?

At that point I would say no.
We have been merging so many stuffs lately because of release that we
don't even know why this thing broke.
Is it in Zaqar? In Swift? In TripleO?

So yeah maybe it's not your patch but it's maybe related or maybe not.
Since nobody is able to tell it, I would suggest to not revert the
revert until we find the root cause and we fix it.
Because if we don't do that, we'll merge your patch again and
eventually increase the number of hits in the gate which is something
we don't want at this stage.

When we discussed you told me this patch wasn't a requirement to
upgrade but an enhancement. At this stage of the cycle we are looking
for stability and not for enhancements, I hope you understand that.
It's a difficult choice to make for me but I'll prefer us to fix bugs
before we continue to merge new features from now.

Keep in mind feature freeze and why we're doing this.

> I also see you assigned me that bug - I'll try and have another look at it
> next week but we may want to reach out to someone from zaqar see if they
> have any ideas about why that happens.

Thank you, that would be very helpful.
-- 
Emilien Macchi