Open Bug 1374446 Opened 3 years ago Updated 2 years ago

Intermittent browser/components/translation/test/browser_translation_yandex.js | A promise chain failed to handle a rejection: can't access dead object

Categories

(Firefox :: Translation, defect, P5)

defect

Tracking

()

Tracking Status
firefox56 --- disabled

People

(Reporter: intermittent-bug-filer, Unassigned)

References

Details

(Keywords: intermittent-failure, leave-open, Whiteboard: [stockwell disabled])

Attachments

(1 file)

Duplicate of this bug: 1374472
in 2 weeks we will only be running e10s and this is a non-e10s only failure- that doesn't mean we should ignore it.

I am going to try to bisect down the autoland/m-c changes to see if we can narrow this down- thanks for reducing this down to a single merge :dbaron.

:felipe, in the meantime, is there anyone on the firefox::translation team that could pick up investigating this bug?
Flags: needinfo?(felipc)
Whiteboard: [stockwell needswork]
this is a mix of bc3 and bc4:
https://treeherder.mozilla.org/#/jobs?repo=autoland&filter-searchStr=win%20opt%20tc%20mochitest-browser-chrome-&fromchange=3791e8357839e9721bec9969a77fbf848ddee0ee&tochange=9516080c78a2aa7282ee279fd9dcb18449977d3d&selectedJob=113521377

I didn't retrigger all of them, but maybe every 3rd or 4th commit 15x.  I will see if there are patterns and make another pass in an hour.
Coming at this another way, the earliest report of this failure was Jun 19, and the test was updated that day, in bug 1242505.
(In reply to Geoff Brown [:gbrown] from comment #12)
> Coming at this another way, the earliest report of this failure was Jun 19,
> and the test was updated that day, in bug 1242505.

https://treeherder.mozilla.org/#/jobs?repo=mozilla-inbound&filter-searchStr=windows%20browser-chrome&tochange=3b544d62b76798ecdd553cba7a87be92012dd939&fromchange=4e2d838994772117314888a99bc71fc4ee822f30 -- it looks like the failures started here.
(In reply to Joel Maher ( :jmaher) (UTC-8) from comment #8)
> in 2 weeks we will only be running e10s and this is a non-e10s only failure-
> that doesn't mean we should ignore it.
> 
> I am going to try to bisect down the autoland/m-c changes to see if we can
> narrow this down- thanks for reducing this down to a single merge :dbaron.
> 
> :felipe, in the meantime, is there anyone on the firefox::translation team
> that could pick up investigating this bug?

Hmm, given Photon/Quantum Flow work, and that Translation was never really released, I don't know if anyone can take much time to investigate this.. Since it's non-e10s only I'm fine with just disabling this test for non-e10s.  If it was both e10s and non-e10s I could take a closer look to see what's happening..
Flags: needinfo?(felipc)
Attachment #8887206 - Flags: review?(jmaher)
Comment on attachment 8887206 [details] [diff] [review]
skip on windows, non-e10s

Review of attachment 8887206 [details] [diff] [review]:
-----------------------------------------------------------------

nice and simple.
Attachment #8887206 - Flags: review?(jmaher) → review+
Pushed by gbrown@mozilla.com:
https://hg.mozilla.org/integration/mozilla-inbound/rev/e8b9498c874a
Skip test browser_translation_yandex.js on windows, non-e10s; r=jmaher
Keywords: leave-open
Whiteboard: [stockwell needswork] → [stockwell disabled]
Pushed by kwierso@gmail.com:
https://hg.mozilla.org/integration/mozilla-inbound/rev/23f7cdcb0263
Followup to fix manifest syntax a=bustage
On Beta, this is failing like crazy on OSX non-e10s as well. Given the history of this bug, I'll just skip there as well.
Priority: -- → P5
You need to log in before you can comment on or make changes to this bug.