Closed Bug 1085247 Opened 10 years ago Closed 9 years ago

Intermittent mediasource-duration.html | Test seek starts on duration truncation below currentTime - assert_equals: Event types match. expected "seeking" but got "timeupdate"

Categories

(Core :: Audio/Video, defect, P5)

defect

Tracking

()

RESOLVED FIXED
mozilla38
Tracking Status
firefox36 --- wontfix
firefox37 --- fixed
firefox38 --- fixed
firefox-esr31 --- unaffected

People

(Reporter: cbook, Assigned: karlt)

References

(Depends on 1 open bug, )

Details

(Keywords: intermittent-failure)

Attachments

(1 file)

Ubuntu VM 12.04 x64 mozilla-inbound opt test web-platform-tests-4

https://treeherder.mozilla.org/ui/logviewer.html#?job_id=3126966&repo=mozilla-inbound

21:54:11 WARNING - TEST-UNEXPECTED-FAIL | /media-source/mediasource-duration.html | Test seek starts on duration truncation below currentTime - assert_equals: Event types match. expected "seeking" but got "timeupdate"
https://treeherder.mozilla.org/ui/#/jobs?repo=try&revision=d881c9e782ef

01:37:34     INFO - PROCESS | 1726 | MARIONETTE LOG: INFO: Timeout fired
01:37:34     INFO - TEST-TIMEOUT | /media-source/mediasource-duration.html | Test seek starts on duration truncation below currentTime - Test timed out
01:37:34     INFO - TEST-TIMEOUT | /media-source/mediasource-duration.html | Test endOfStream completes previous seek to truncated duration - Test timed out
01:37:34     INFO - TEST-FAIL | /media-source/mediasource-duration.html | Test setting same duration multiple times does not fire duplicate durationchange - assert_true: adjusted duration expected true got false
01:37:34     INFO - TEST-TIMEOUT | /media-source/mediasource-duration.html | took 10155ms

I saw these messages in a green W3C Web Platform Tests chunk-4.
I wonder why the test failure didn't mark the chunk orange.

I try to run mediasource-duration.html on my PC and always get timeouts.
Flags: needinfo?(ryanvm)
I'm not an expert in the webplatform tests, but IIUC from bug 1066467, these tests were supposed to be disabled for being unstable.
http://mxr.mozilla.org/mozilla-central/source/testing/web-platform/meta/media-source/mediasource-duration.html.ini

I'm wondering if the recent upstream update somehow re-enabled them?
Flags: needinfo?(ryanvm) → needinfo?(james)
So consistently getting timeouts isn't an unexpected result. Hence it doesn't mark the run orange. You only get orange if you got a timeout but expected something else (e.g. PASS).

I think the recent update actually re-disabled tests that were supposed to be enabled (bug 1087165).
Flags: needinfo?(james)
This test was enabled, about 4 days before the first report in comment 0, in
https://hg.mozilla.org/integration/mozilla-inbound/rev/2839cfaaa883

I don't know of any changes that may have made this unstable, but we can disable the particular subtest.
Assignee: nobody → karlt
Blocks: 1066467
OS: Linux → All
Hardware: x86 → All
https://hg.mozilla.org/integration/mozilla-inbound/rev/822005621954
Flags: in-testsuite+
Keywords: leave-open
Whiteboard: [disabled]
https://hg.mozilla.org/mozilla-central/rev/822005621954
Status: NEW → RESOLVED
Closed: 9 years ago
Resolution: --- → FIXED
Target Milestone: --- → mozilla38
Do we need to enable these on aurora/beta too?
Flags: needinfo?(karlt)
Yes, we would like this fix uplifted. Nominating patch from comment #31.

Approval Request Comment
[Feature/regressing bug #]: MSE
[User impact if declined]: Less consistent testing (including intermittant failures)
[Describe test coverage new/current, TreeHerder]: Landed on m-c.
[Risks and why]: Test only, so the risk is just more orange on treeherder.
[String/UUID change made/needed]: None.
Attachment #8552582 - Flags: approval-mozilla-beta?
Attachment #8552582 - Flags: approval-mozilla-aurora?
Comment on attachment 8552582 [details] [diff] [review]
Enable remaining mediasoruce-duratin subtests

Test-only changes don't need approval.
Attachment #8552582 - Flags: approval-mozilla-beta?
Attachment #8552582 - Flags: approval-mozilla-aurora?
(In reply to Ryan VanderMeulen [:RyanVM UTC-5] from comment #32)
> Do we need to enable these on aurora/beta too?

I would wait a couple of days to check that these really are no longer intermittent.  I don't know exactly what would have changed behavior, and it is hard to verify on try because the failures were not frequent.
Flags: needinfo?(karlt)
I see these were just re-enabled on aurora.  That's fine too.
Fun times, I'd already pushed them by the time I saw that comment. Oh well, we'll see I guess. Also, they're fully disabled on beta, so we'll need a different patch there.

https://hg.mozilla.org/releases/mozilla-aurora/rev/1711d9b3f9f4
Flags: needinfo?(karlt)
(In reply to Ryan VanderMeulen [:RyanVM UTC-5] from comment #37)
> Also, they're fully disabled on beta, so we'll need a different patch there.

See bug 1098126.  The patch here should apply fine on top of that.
Flags: needinfo?(karlt)
https://hg.mozilla.org/releases/mozilla-beta/rev/d918f7ea93fe

I merged it manually to make it look like central; hopefully that will work.
Flags: qe-verify-
Depends on: 1125776
(In reply to Wes Kocher (:KWierso) from comment #40)
> I had to back this out in
> https://hg.mozilla.org/releases/mozilla-beta/rev/a3202b0fe6a7 on beta for
> wpt-4 bustage:
> 
> https://treeherder.mozilla.org/logviewer.html#?job_id=235734&repo=mozilla-
> beta

That looks similar to changes in results from from bug 1096089 comment 25 and bug 1116007.  We can leave things disabled on beta 36.

jya has changes in bug 1125776 to address recent m-c intermittents, which seem different from those originally filed here.
Flags: needinfo?(karlt)
This will have been fixed by bug 1125776, bug 1130826 and bug 1130839
Status: REOPENED → RESOLVED
Closed: 9 years ago9 years ago
Resolution: --- → FIXED
See Also: → 1144509
You need to log in before you can comment on or make changes to this bug.