Closed Bug 1349051 Opened 3 years ago Closed 3 years ago

Intermittent browser_advanced_siteData.js | This test exceeded the timeout threshold. It should be rewritten or split up. If that's not possible, use requestLongerTimeout(N), but only as a last resort. -

Categories

(Firefox :: Preferences, defect)

defect
Not set

Tracking

()

RESOLVED FIXED
Firefox 55
Tracking Status
firefox55 --- fixed

People

(Reporter: intermittent-bug-filer, Assigned: jaws)

References

Details

(Keywords: intermittent-failure, Whiteboard: [stockwell fixed])

Attachments

(1 file)

This has been failing for a few days but failure frequency spiked markedly yesterday, around the time the test was updated in bug 1335907. Failures are all Linux32/64-Debug.

Test durations for browser/components/preferences/in-content/tests/browser_advanced_siteData.js on mozilla-central,mozilla-inbound,autoland between 2017-03-14 and 2017-03-28
linux32/debug-chunked:          61.99 s (37.03 s - 89.69 s over 849 runs)
linux32/debug-e10s:             71.47 s (50.91 s - 89.69 s over 496 runs)
linux32/opt-chunked:            18.87 s (11.58 s - 31.23 s over 1110 runs)
linux32/opt-e10s:               23.91 s (17.84 s - 31.23 s over 394 runs)
linux64/asan-chunked:           35.56 s (21.82 s - 54.33 s over 792 runs)
linux64/asan-e10s:              42.61 s (28.36 s - 54.33 s over 409 runs)
linux64/debug-chunked:          60.63 s (35.73 s - 89.05 s over 1012 runs)
linux64/debug-e10s:             67.78 s (48.32 s - 89.05 s over 695 runs)
linux64/opt-chunked:            19.25 s (11.12 s - 38.34 s over 620 runs)
linux64/opt-e10s:               23.01 s (16.28 s - 38.34 s over 316 runs)
linux64/pgo-chunked:            16.51 s (9.69 s - 29.34 s over 653 runs)
linux64/pgo-e10s:               19.87 s (12.76 s - 27.08 s over 344 runs)

It looks like the test was already running for about 70 seconds on average on linux32/64-debug before the recent change - it would not have taken much delay to start hitting the 90 second timeout frequently.

Zachary - Can you have a look at this? If the test cannot be optimized or split, a requestLongerTimeout(2) call should fix this.
Blocks: 1335907
Flags: needinfo?(herrickz)
Whiteboard: [stockwell needswork]
We should split the test.

Looking at the changes from the commit, https://hg.mozilla.org/mozilla-central/diff/069b6372cc15/browser/components/preferences/in-content/tests/browser_advanced_siteData.js, there is nothing that shouldn't be there. The test now opens the privacy pane, which is longer than the advanced pane was and thus will take longer to load, hence the change in timeout rate.
Flags: needinfo?(herrickz)
Assignee: nobody → jaws
Status: NEW → ASSIGNED
Comment on attachment 8853523 [details]
Bug 1349051 - Split browser_advanced_siteData.js to browser_siteData.js and browser_siteData2.js since it was running too long.

https://reviewboard.mozilla.org/r/125578/#review128324

Looks good. Thanks for helping on this.
Attachment #8853523 - Flags: review?(fliu) → review+
Pushed by jwein@mozilla.com:
https://hg.mozilla.org/integration/autoland/rev/c83ec472c2f4
Split browser_advanced_siteData.js to browser_siteData.js and browser_siteData2.js since it was running too long. r=Fischer
Whiteboard: [stockwell needswork] → [stockwell fixed]
https://hg.mozilla.org/mozilla-central/rev/c83ec472c2f4
Status: ASSIGNED → RESOLVED
Closed: 3 years ago
Resolution: --- → FIXED
Target Milestone: --- → Firefox 55
:jaws - Your change to split in-content/tests/browser_advanced_siteData.js was very effective - thank you! - but there are still some time outs from in-content-old/tests/browser_advanced_siteData.js. Should the same change be applied there, or could that test be skipped perhaps?
Flags: needinfo?(jaws)
We should do the same split for the in-content-old test too, but at this point it should be in a separate bug. I didn't do it in this bug because I thought the timeouts were only in the combined (re-org'd) preferences.
Flags: needinfo?(jaws)
I filed bug 1361170 for splitting up the in-content-old version.
You need to log in before you can comment on or make changes to this bug.