On Mon, Nov 9, 2015 at 4:58 PM, Julien Wajsberg <[email protected]> wrote:
> I usually have a look at the blues as well when I look at my pull request > runs. And most of the time it's an issue that's not related with the test. > I think this account for most of the issues we see, actually. Do we know > why this happens ? I started this thread specifically about Gij (integration tests). Gu (unit tests) historically have had less flakiness, but you're right that there has been some weirdness there recently too. I'm not looking at Gu yet since I think Gij is in worse shape, and I didn't want to conflate this thread with Gu runner issues. Now, concerning Gij, to address your statement here: "And most of the time it's an issue that's not related with the test. I think this account for most of the issues we see, actually." I thought this might be the case, and indeed I started this thread so we could have more awareness of any harness issues in Gij. But it turns out I think the (excellent) work Gareth and Aus have been doing over the last year might have addressed our biggest Gij harness issues (knock on wood as hard as possible). In any case, I have been able to fix most of the intermittents surfaced by bug 1222215 in the tests themselves. I'm hoping this continues to be the case :)
_______________________________________________ dev-fxos mailing list [email protected] https://lists.mozilla.org/listinfo/dev-fxos
