facebook story failing on mobile builders on chromium.perf |
||
Issue descriptionThe facebook story seems to be failing on multiple benchmarks on these builders: - Android Nexus5 Perf: https://build.chromium.org/p/chromium.perf/builders/Android%20Nexus5%20Perf - Android Nexus6 Perf: https://build.chromium.org/p/chromium.perf/builders/Android%20Nexus6%20Perf - Android Nexus7v2 Perf: https://build.chromium.org/p/chromium.perf/builders/Android%20Nexus7v2%20Perf https://console.developers.google.com/m/cloudstorage/b/chrome-telemetry-output/o/profiler-file-id_1-2017-06-22_16-30-5593069.png is an example screenshot. https://chromium-swarm.appspot.com/task?id=36e906120fcb3710&refresh=10&show_raw=1 is an example task. Not sure who exactly owns that story.
,
Jun 23 2017
Timing out while waiting 180s for HasReachedQuiescence. It started failing since May 15 on all these bots, so something must be changed around then.
,
Jun 23 2017
Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8975955470788472496
,
Jun 23 2017
This was also seen on the Android One Perf bot. I started a bisect. The story was failing for all recorded history for this bot; as soon as I migrated it to swarming, it was failing. So my bisect range is about 1000 CLs :/
,
Jun 24 2017
I think it might be easier trying to rerecord the story, tbh
,
Jun 24 2017
>c5 facebook is very frequently updated website. Updating the archive sgtm. Let's see if we hit the issue in the rerecord
,
Jun 24 2017
=== BISECT JOB RESULTS === Bisect failed for unknown reasons Please contact the team (see below) and report the error. Bisect Details Configuration: android_one_perf_bisect Benchmark : v8.mobile_infinite_scroll_tbmv2 Metric : memory:chrome:all_processes:reported_by_chrome:v8:allocated_by_malloc:peak_size_min/facebook To Run This Test src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=facebook v8.mobile_infinite_scroll_tbmv2 Debug Info https://chromeperf.appspot.com/buildbucket_job_status/8975955470788472496 Is this bisect wrong? https://chromeperf.appspot.com/bad_bisect?try_job_id=5887815943979008 | O O | Visit http://www.chromium.org/developers/speed-infra/perf-bug-faq | X | for more information addressing perf regression bugs. For feedback, | / \ | file a bug with component Speed>Bisection. Thank you!
,
Jun 25 2017
Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8975860557719134512
,
Jun 26 2017
=== BISECT JOB RESULTS === Bisect failed for unknown reasons Please contact the team (see below) and report the error. Bisect Details Configuration: android_one_perf_bisect Benchmark : v8.mobile_infinite_scroll_tbmv2 Metric : memory:chrome:all_processes:reported_by_chrome:v8:allocated_by_malloc:peak_size_min/facebook To Run This Test src/tools/perf/run_benchmark -v --browser=android-chromium --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=facebook v8.mobile_infinite_scroll_tbmv2 Debug Info https://chromeperf.appspot.com/buildbucket_job_status/8975860557719134512 Is this bisect wrong? https://chromeperf.appspot.com/bad_bisect?try_job_id=5887815943979008 | O O | Visit http://www.chromium.org/developers/speed-infra/perf-bug-faq | X | for more information addressing perf regression bugs. For feedback, | / \ | file a bug with component Speed>Bisection. Thank you!
,
Jun 26 2017
OK, let's rerecord the story. The docstring of LoginWithMobileSite() says that 'facebooktest' is the recommended account, but there's no entry for 'facebooktest' in credentials.json. Hmm? https://cs.chromium.org/chromium/src/tools/perf/page_sets/login_helpers/facebook_login.py?type=cs&q=LoginWithMobileSite&l=19 ..and 'facebook3' seems to have been locked again :(
,
Jun 26 2017
Hmhh, somehow https://codereview.chromium.org/2381113002 erased the facebooktest account. I will add it back.
,
Jun 27 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/e34c109769fb7d8cc37192a041d3893b7742d37e commit e34c109769fb7d8cc37192a041d3893b7742d37e Author: nednguyen <nednguyen@google.com> Date: Tue Jun 27 01:46:07 2017 Add facebooktest account back To test this, go to depot_tools, then run: ./gsutil.py cat gs://chromium-telemetry/<hash of the sha1 file> BUG= 736073 Review-Url: https://codereview.chromium.org/2960643002 Cr-Commit-Position: refs/heads/master@{#482508} [modify] https://crrev.com/e34c109769fb7d8cc37192a041d3893b7742d37e/tools/perf/page_sets/data/credentials.json.sha1
,
Jun 27 2017
The following revision refers to this bug: https://chromium.googlesource.com/chromium/src.git/+/9be57a21992e089f32cb7f1958ff19fd5b9ac353 commit 9be57a21992e089f32cb7f1958ff19fd5b9ac353 Author: Kunihiko Sakamoto <ksakamoto@chromium.org> Date: Tue Jun 27 09:11:10 2017 Rerecord Facebook story in loading.mobile and reenable it on Nexus7 Facebook story was timing out on perf bots while waiting 180s for HasReachedQuiescence. This patch updates Facebook recording, and re-enable the story on Nexus7 bot to see if if fixes the timeout. Bug: 736073 , 736877 Change-Id: I2e0e0c92a36dfb1f0059d2f8e5affd558f726edb Reviewed-on: https://chromium-review.googlesource.com/549538 Reviewed-by: Kouhei Ueno <kouhei@chromium.org> Commit-Queue: Kunihiko Sakamoto <ksakamoto@chromium.org> Cr-Commit-Position: refs/heads/master@{#482578} [modify] https://crrev.com/9be57a21992e089f32cb7f1958ff19fd5b9ac353/tools/perf/page_sets/data/loading_mobile.json [add] https://crrev.com/9be57a21992e089f32cb7f1958ff19fd5b9ac353/tools/perf/page_sets/data/loading_mobile_002.wpr.sha1 [modify] https://crrev.com/9be57a21992e089f32cb7f1958ff19fd5b9ac353/tools/perf/page_sets/loading_mobile.py
,
Jun 28 2017
Rerecorded Facebook story seems to be working fine. |
||
►
Sign in to add a comment |
||
Comment 1 by nedngu...@google.com
, Jun 22 2017Status: Assigned (was: Available)