New issue
Advanced search Search tips
Note: Color blocks (like █ or █) mean that a user may not be available. Tooltip shows the reason.

Issue 653213 link

Starred by 1 user
Status: WontFix
Owner:
tommycli@chromium.org
Closed: Dec 2016
Cc:
tommycli@chromium.org
█ mwlodar@chromium.org
majidvp@chromium.org
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression



Sign in to add a comment

27.1%-53.1% regression in page_cycler_v2.typical_25 at 422902:423020

Project Member Reported by majidvp@google.com, Oct 5 2016 
See the link to graphs below.
Project Member

Comment 3 by 42576172...@developer.gserviceaccount.com, Oct 6 2016 


===== BISECT JOB RESULTS =====
Status: failed


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@422947  518.036  27.2203  8  good
chromium@422957  591.737  143.149  8  good
chromium@422962  542.05   14.5659  5  good
chromium@422966  784.455  47.8715  8  bad
chromium@422984  804.746  52.9669  5  bad

Bisect job ran on: win_perf_bisect
Bug ID: 653213

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --also-run-disabled-tests page_cycler_v2.typical_25
Test Metric: timeToFirstContentfulPaint_avg/pcv1-warm/http___www.ign.com_
Relative Change: 54.89%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/win_perf_bisect/builds/6975
Job details: https://chromeperf.appspot.com/buildbucket_job_status/8999617656570918608


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5346528141508608

| 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 Tests>AutoBisect.  Thank you!
Project Member

Comment 4 by sheriffbot@chromium.org, Oct 7 2016

Labels: Hotlist-Google

Comment 6 by tdres...@chromium.org, Oct 11 2016

Cc: mwlodar@chromium.org
mwlodar@, any chance it was https://chromium.googlesource.com/chromium/src/+/1cfe9539a19cb6cdea82c91f78f9341418fa2d13?

It's the most suspicious in the range given by the last bisect (though I've launched another bisect to be sure).
Project Member

Comment 7 by 42576172...@developer.gserviceaccount.com, Oct 11 2016 


===== BISECT JOB RESULTS =====
Status: started


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@422962  309.761  16.186   5  good
chromium@422964  307.348  18.7148  5  good
chromium@422966  483.756  40.6408  5  bad

Bisect job ran on: linux_perf_bisect
Bug ID: 653213

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests page_cycler_v2.intl_ar_fa_he
Test Metric: timeToFirstContentfulPaint_avg/pcv1-warm/http___www.islamweb.net_
Relative Change: 56.17%
Score: 0

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/linux_perf_bisect/builds/6776
Job details: https://chromeperf.appspot.com/buildbucket_job_status/8999065256165157152


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5224223495159808

| 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 Tests>AutoBisect.  Thank you!
Project Member

Comment 9 by 42576172...@developer.gserviceaccount.com, Oct 26 2016 

Bisect failed: http://build.chromium.org/p/tryserver.chromium.perf/builders/linux_perf_bisect/builds/6798
Failure reason: the build has failed.
Additional errors:
The metric was not found in the test output.
Either of the initial "good" or "bad" revisions failed to be tested or built.
Project Member

Comment 12 by 42576172...@developer.gserviceaccount.com, Dec 17 2016

Cc: tommycli@chromium.org
Owner: tommycli@chromium.org

=== PERF REGRESSION ===


=== Auto-CCing suspected CL author tommycli@chromium.org ===

Hi tommycli@chromium.org, the bisect results pointed to your CL, please take a look at the
results.


===== BISECT JOB RESULTS =====
Status: completed


===== SUSPECTED CL(s) =====
Subject : [HBD] Hardcode SEI threshold to 30 (for testers)
Author  : tommycli
Commit description:
  
Normal users will have it configured to the real value via Finch. This
hardcoding is only for ease of testing in case the user does not have
a Finch config.

BUG=650903

Review-Url: https://codereview.chromium.org/2391823003
Cr-Commit-Position: refs/heads/master@{#422966}
Commit  : ef97eaf65cfaaa1728bfeebb0fa82701a4d5481c
Date    : Tue Oct 04 22:25:24 2016


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@422947  836.254  3369.33  9  good
chromium@422957  686.495  2209.82  9  good
chromium@422962  854.642  3661.4   9  good
chromium@422964  894.315  3988.15  9  good
chromium@422965  849.316  3656.64  9  good
chromium@422966  1469.22  5400.12  9  bad    <--
chromium@422984  1385.41  4770.24  6  bad
chromium@423020  961.22   2274.98  6  bad

Bisect job ran on: linux_perf_bisect
Bug ID: 653213

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=http...www.fifa.com. page_cycler_v2.typical_25
Test Metric: timeToFirstContentfulPaint_avg/pcv1-warm/http___www.fifa.com_
Relative Change: 115.28%

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/linux_perf_bisect/builds/6945
Job details: https://chromeperf.appspot.com/buildbucket_job_status/8993073729998839936


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5812585163653120

| 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 Tests>AutoBisect.  Thank you!
Project Member

Comment 13 by 42576172...@developer.gserviceaccount.com, Dec 17 2016 


===== BISECT JOB RESULTS =====
Status: completed


===== SUSPECTED CL(s) =====
Subject : [HBD] Hardcode SEI threshold to 30 (for testers)
Author  : tommycli
Commit description:
  
Normal users will have it configured to the real value via Finch. This
hardcoding is only for ease of testing in case the user does not have
a Finch config.

BUG=650903

Review-Url: https://codereview.chromium.org/2391823003
Cr-Commit-Position: refs/heads/master@{#422966}
Commit  : ef97eaf65cfaaa1728bfeebb0fa82701a4d5481c
Date    : Tue Oct 04 22:25:24 2016


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@422881  875.179  82.2392  6  good
chromium@422928  886.862  87.536   6  good
chromium@422951  906.123  32.2925  6  good
chromium@422963  912.219  91.838   6  good
chromium@422965  862.238  51.1005  6  good
chromium@422966  1311.14  109.626  6  bad    <--
chromium@422969  1302.42  118.786  6  bad
chromium@422974  1253.33  163.195  6  bad

Bisect job ran on: win_perf_bisect
Bug ID: 653213

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests --story-filter=http...terra.es. page_cycler_v2.intl_es_fr_pt-BR
Test Metric: timeToFirstContentfulPaint_avg/pcv1-warm/http___terra.es_
Relative Change: 43.21%

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/win_perf_bisect/builds/7075
Job details: https://chromeperf.appspot.com/buildbucket_job_status/8993073745925820624


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5778241531412480

| 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 Tests>AutoBisect.  Thank you!

Comment 14 by tommycli@chromium.org, Dec 22 2016

Status: WontFix (was: Untriaged)
This has gone back down without my intervention, so I'm assuming that this is a false alarm.
►

Sign in to add a comment