Awesome - thanks for that info! If you ever want someone to help be a
tester and experiment, I'm happy to be that person!

Cheers,
Leah

On Wed, Oct 2, 2019 at 10:21 PM Kamil Breguła <[email protected]>
wrote:

> Hello,
>
> We know that there is some problem with the stability of some tests
> and we try to limit this problem.
> * migration from nose to pytest. In the meantime, a large number of
> sideeffects will be deleted.
> * migration from Travis Ci to GitLab Ci on GKE. This will increase the
> computing power that we will have available.
> * migration Kubernetes tests to KinD. This can reduce the race conditions.
>
> Thanks,
> Kamil
>
> On Thu, Oct 3, 2019 at 6:22 AM Leah Cole <[email protected]>
> wrote:
> >
> > Hi all,
> >
> > I submitted a small PR today
> > <https://github.com/apache/airflow/pull/6246> that's
> > been approved, but am having trouble with Travis - the first time the job
> > ran, the "Tests postgres kubernetes python 3.6 (git)" task failed.
> Someone
> > generously reran it, and this time the "Tests mysql python 3.7" task
> > failed.
> >
> > To try to debug, I added Travis to my fork, and there, I had failures on
> > "Tests mysql python 3.7", "Tests sqlite python 3.5" and "Tests postgres
> > python 3.6" tasks.
> >
> > Anyone have any ideas as to why
> > 1) Travis running on the apache/airflow repo fails on different tasks in
> > different runs
> > 2) Travis running in my fork has different failures than upstream
> >
> > And if there are any other ways to debug?
> >
> > Thanks a million,
> > Leah
>


-- 

Leah Cole | Developer Programs Engineer | [email protected] | (925)
257-2112

Reply via email to