New issue
Advanced search Search tips

Issue 696078 link

Starred by 1 user

Issue metadata

Status: WontFix
Owner:
Closed: Mar 2017
Cc:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug-Regression



Sign in to add a comment

3.3% regression in memory.top_10_mobile_stress at 452170:452248

Project Member Reported by m...@chromium.org, Feb 24 2017

Issue description

See the link to graphs below.
 

Comment 1 by m...@chromium.org, Feb 24 2017

All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=696078

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDgxKmNtwkM


Bot(s) for this bug's original alert(s):

android-nexus6
Project Member

Comment 3 by 42576172...@developer.gserviceaccount.com, Feb 25 2017


=== BISECT JOB RESULTS ===
NO Perf regression found

Bisect Details
  Configuration: android_nexus6_perf_bisect
  Benchmark    : memory.top_10_mobile_stress
  Metric       : memory:chrome:all_processes:reported_by_chrome:v8:effective_size_avg/background/after_http_www_baidu_com_s_word_google

Revision             Result                 N
chromium@452169      4293754 +- 357229      21      good
chromium@452248      4336836 +- 373527      20      bad

Please refer to the following doc on diagnosing memory regressions:
  https://chromium.googlesource.com/chromium/src/+/master/docs/memory-infra/memory_benchmarks.md

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 memory.top_10_mobile_stress

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8986733867612437584

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=4931755397611520


| 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!

Comment 5 by m...@chromium.org, Feb 25 2017

Kicked-off another bisect run.
Project Member

Comment 6 by 42576172...@developer.gserviceaccount.com, Feb 26 2017


=== BISECT JOB RESULTS ===
NO Perf regression found

Bisect Details
  Configuration: android_nexus6_perf_bisect
  Benchmark    : memory.top_10_mobile_stress
  Metric       : memory:chrome:all_processes:reported_by_chrome:v8:effective_size_avg/background/after_http_www_baidu_com_s_word_google

Revision             Result                 N
chromium@452169      4294901 +- 381876      21      good
chromium@452248      4312337 +- 367311      21      bad

Please refer to the following doc on diagnosing memory regressions:
  https://chromium.googlesource.com/chromium/src/+/master/docs/memory-infra/memory_benchmarks.md

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 memory.top_10_mobile_stress

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8986649261705375984

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5901722490241024


| 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!

Comment 7 by m...@chromium.org, Mar 1 2017

Owner: perezju@chromium.org
Status: Assigned (was: Untriaged)
The bots can't seem to find a regression, but the graphs show a fairly clear regressing trend: https://chromeperf.appspot.com/group_report?bug_id=696078

Given trunk's memory performance is already much worse than the reference version, do we care about this small regression (assuming someone is working on the larger problem at-hand)? I'll leave it to you to decide whether to find someone to look into this issue, or WontFix.
Cc: simonhatch@chromium.org
+simonhatch can you take a look at why bisect doesn't repro?
Graph looks a little spiky around the alert, I'll try widening the range for now in case the alert just isn't quite in the right spot. There's quite a few alerts for other pages where the first spike is, they've been assigned to crbug.com/696073 which is marked as Fixed but those pages haven't recovered (ie. look at the wikipedia page).

=== BISECT JOB RESULTS ===
Perf regression found but unable to narrow commit range

Build failures prevented the bisect from narrowing the range further.


Bisect Details
  Configuration: android_nexus6_perf_bisect
  Benchmark    : memory.top_10_mobile_stress
  Metric       : memory:chrome:all_processes:reported_by_chrome:v8:effective_size_avg/background/after_http_www_baidu_com_s_word_google
  Change       : 2.69% | 4204409.90476 -> 4317713.71429

Suspected Commit Range
  41 commits in range
  https://chromium.googlesource.com/chromium/src/+log/ec69818af043485daccd4d4f43718390e9ddcb08..e4dd371c464f77b6fbef56d212f7de71e03bf002


Revision             Result                 N
chromium@451824      4204410 +- 131091      21       good
chromium@452027      4216307 +- 171270      14       good
chromium@452028      ---                    ---      build failure
---                  ---                    ---      too many build failures to list
chromium@452067      ---                    ---      build failure
chromium@452068      4366636 +- 201938      6        bad
chromium@452083      4364297 +- 129827      6        bad
chromium@452137      4316253 +- 310318      14       bad
chromium@452248      4317714 +- 441795      21       bad

Please refer to the following doc on diagnosing memory regressions:
  https://chromium.googlesource.com/chromium/src/+/master/docs/memory-infra/memory_benchmarks.md

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 memory.top_10_mobile_stress

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8986282966343359456

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5281101125255168


| 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!
Labels: Performance-Memory
Project Member

Comment 14 by 42576172...@developer.gserviceaccount.com, Mar 23 2017


=== BISECT JOB RESULTS ===
NO Perf regression found

Bisect Details
  Configuration: android_nexus6_perf_bisect
  Benchmark    : memory.top_10_mobile_stress
  Metric       : memory:chrome:all_processes:reported_by_chrome:v8:effective_size_avg/background/after_http_www_baidu_com_s_word_google

Revision             Result                 N
chromium@452169      4295704 +- 357336      21      good
chromium@452248      4336118 +- 354112      21      bad

Please refer to the following doc on diagnosing memory regressions:
  https://chromium.googlesource.com/chromium/src/+/master/docs/memory-infra/memory_benchmarks.md

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 memory.top_10_mobile_stress

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8984378584788327488

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=4931755397611520


| 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!
I think it's fine to WontFix this one. If anything, we should look at the later spike at 452914:453030. Filed 705481 and started bisect there.
Status: WontFix (was: Assigned)
Project Member

Comment 18 by 42576172...@developer.gserviceaccount.com, Mar 30 2017


=== BISECT JOB RESULTS ===
NO Perf regression found

Bisect Details
  Configuration: android_nexus6_perf_bisect
  Benchmark    : memory.top_10_mobile_stress
  Metric       : memory:chrome:all_processes:reported_by_chrome:v8:effective_size_avg/background/after_http_www_baidu_com_s_word_google

Revision             Result                 N
chromium@452169      4318618 +- 403826      21      good
chromium@452248      4267184 +- 323111      21      bad

Please refer to the following doc on diagnosing memory regressions:
  https://chromium.googlesource.com/chromium/src/+/master/docs/memory-infra/memory_benchmarks.md

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 memory.top_10_mobile_stress

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8983687250233820848

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=4931755397611520


| 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!

=== BISECT JOB RESULTS ===
NO Perf regression found

Bisect Details
  Configuration: android_nexus6_perf_bisect
  Benchmark    : memory.top_10_mobile_stress
  Metric       : memory:chrome:all_processes:reported_by_chrome:v8:effective_size_avg/background/after_http_www_baidu_com_s_word_google

Revision             Result                 N
chromium@452169      4284103 +- 404410      21      good
chromium@452248      4321822 +- 391103      21      bad

Please refer to the following doc on diagnosing memory regressions:
  https://chromium.googlesource.com/chromium/src/+/master/docs/memory-infra/memory_benchmarks.md

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 memory.top_10_mobile_stress

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8983033684308080896

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=4931755397611520


| 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!

Sign in to add a comment