1836204 - Re-enable browser.translations.enable for automation

Status: RESOLVED FIXED

(Firefox :: Translations, defect, P2)

Description

[Julian Descottes [:jdescottes]]

In bug 1836093 we disabled translations for automation (webdriver, CDP, ...) because this was triggering timeouts on early beta (https://bugzilla.mozilla.org/show_bug.cgi?id=1835822).

We should investigate the reason for those timeouts. The tests did not seem to hang in a specific spot but just became slower overall and long tests which would take less than 20s now take more than 80s and reach the timeout.

Two important things to note:

• disabling the preference after starting the browser did not help, we had to disable it before the browser starts to avoid the timeouts
• nightly was not affected by the timeouts, despite having the preference as well

Comment 1

[Julian Descottes [:jdescottes]]

Greg: just a quick ping, in case you want to investigate this issue we filed this bug.

Comment 2

[BugBot [:suhaib / :marco/ :calixte]]

Set release status flags based on info from the regressing bug 1831963

Comment 3

[Greg Tatum [:gregtatum]]

Thanks, I'll look more into performance work and revisit.

Comment 4

[BugBot [:suhaib / :marco/ :calixte]]

Set release status flags based on info from the regressing bug 1831963

Comment 5

[Julian Descottes [:jdescottes]]

Blocking Bug 1820261 because this prevents from checking the performance impact of the translations feature in CI

Comment 6

[BugBot [:suhaib / :marco/ :calixte]]

The severity field is not set for this bug.
:nordzilla, could you have a look please?

For more information, please visit BugBot documentation.

Comment 7

[BugBot [:suhaib / :marco/ :calixte]]

The severity field is not set for this bug.
:nordzilla, could you have a look please?

For more information, please visit BugBot documentation.

Comment 8

[Greg Tatum [:gregtatum]]

I am just waiting for Bug 1836974 to land before re-attempting this.

Comment 9

[Pascal Chevrel:pascalc]

Greg, should we reenable it now? Thanks

Comment 10

[Greg Tatum [:gregtatum]]

Attachment: Bug 1836204 - Re-enable translations in automation; r?nordzilla!

Comment 11

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/790939d4314e
Re-enable translations in automation; r=nordzilla,webdriver-reviewers,whimboo

Comment 12

[Henrik Skupin [:whimboo][⌚️UTC+2]]

The upstream puppeteer PR is: https://github.com/puppeteer/puppeteer/pull/10522

Comment 13

[Noemi Erli[:noemi_erli]]

Backed out changeset 790939d4314e (Bug 1836204) for causing wd leak failures

Push with failures: https://treeherder.mozilla.org/jobs?repo=autoland&group_state=expanded&selectedTaskRun=IPJ94NowRFqkNRYjliklsQ.0&resultStatus=testfailed%2Cbusted%2Cexception%2Cretry%2Cusercancel&searchStr=os%2Cx%2C10.15%2Cwebrender%2Cdebug%2Cweb%2Cplatform%2Ctests%2Ctest-macosx1015-64-qr%2Fdebug-web-platform-tests-wdspec%2Cwd&revision=790939d4314e8c4842bc44eb1f554b1dcc242b11

Log: https://treeherder.mozilla.org/logviewer?job_id=421902018&repo=autoland

Backout: https://hg.mozilla.org/integration/autoland/rev/67087abbf476604d7a738021132b56b6446bb6ca

Comment 14

[Greg Tatum [:gregtatum]]

Attachment: Bug 1836204 - Fix memory leak with the ViewShown listener; r?nordzilla!

This was racing with the popup being shown and the browser window being
closed. If it lost this race, then it would cause a leak.

Depends on D182926

Comment 15

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/0ab331879e6d
Re-enable translations in automation; r=nordzilla,webdriver-reviewers,whimboo
https://hg.mozilla.org/integration/autoland/rev/1316e023af44
Fix memory leak with the ViewShown listener; r=nordzilla

Comment 16

[Serban Stanca [:SerbanS]]

https://hg.mozilla.org/mozilla-central/rev/0ab331879e6d
https://hg.mozilla.org/mozilla-central/rev/1316e023af44

Comment 17

[Sebastian Hengst [:aryx] (needinfo me if it's about an intermittent or backout)]

Backed out from autoland (Gecko 118) for causing frequent marionette / bidi failures:
https://hg.mozilla.org/integration/autoland/rev/5274e9eec10fd09a512295f81cf0a9aec373b119

The Try push in bug 1845518 indicates the issue would persist with that patch.

Comment 18

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/f21e171c96b3
Fix prettier lint failures. r=fix CLOSED TREE

Comment 19

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/122bcfb45013
Re-enable translations in automation; r=nordzilla,webdriver-reviewers,whimboo
https://hg.mozilla.org/integration/autoland/rev/14debb6cb956
Fix memory leak with the ViewShown listener; r=nordzilla CLOSED TREE

Comment 20

[Sebastian Hengst [:aryx] (needinfo me if it's about an intermittent or backout)]

This relanded because the backout caused test failures. There was a conflict backing out bug 1845428.

Comment 21

[Joel Maher ( :jmaher ) (UTC -8)]

this has caused a regression bug 1805613;

Comment 22

[Joel Maher ( :jmaher ) (UTC -8)]

another note here related to the WDSPEC failures- this only fails on fission (linux + windows); the no-fission (linux + android) seems ok. Also the osx tests seem ok.

Comment 23

[Erik Nordin [:nordzilla]]

Sebastian, can you clarify what you meant here?

Does this mean that you tried to back out Bug 1845428?

I'm extremely confused and quite concerned about the test failures that you linked, e.g.:

There should be 1 telemetry events of type OpenPanel - Got 38, expected 1
...
There should be 2 telemetry events of type OpenPanel - Got 77, expected 2

I have no idea what would be causing over-reporting to Telemetry in this way, and I don't know if it's coming from Glean somehow or from the code itself.

I haven't observed that locally, and those tests were all passing on both locally and on the try servers when that patch landed.

Comment 24

[Natalia Csoregi [:nataliaCs]]

https://hg.mozilla.org/mozilla-central/rev/f21e171c96b3
https://hg.mozilla.org/mozilla-central/rev/122bcfb45013
https://hg.mozilla.org/mozilla-central/rev/14debb6cb956

Comment 25

[Sebastian Hengst [:aryx] (needinfo me if it's about an intermittent or backout)]

I did not try to back out bug 1845428 but there was a conflict for the backout of https://hg.mozilla.org/integration/autoland/rev/14debb6cb956 which got resolved as https://hg.mozilla.org/integration/autoland/rev/5274e9eec10fd09a512295f81cf0a9aec373b119#l1.1 but caused the test failure (please let me know if the conflict got resolved incorrect). Because of the failure, this bug (bug 1836204) got relanded.