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

Issue 612140 link

Starred by 2 users

Issue metadata

Status: Assigned
Owner:
Last visit > 30 days ago
Cc:
EstimatedDays: ----
NextAction: ----
OS: Mac
Pri: 2
Type: Bug

Blocked on:
issue 612779
issue 624355

Blocking:
issue 589726



Sign in to add a comment

[system-health] load:media:soundcloud ChromeClockSyncError in system_health.memory_desktop.reference on Mac HDD Perf (3)

Project Member Reported by petrcermak@chromium.org, May 16 2016

Issue description

The error occurred twice in the last ~10 runs:

  https://build.chromium.org/p/chromium.perf/builders/Mac%20HDD%20Perf%20%283%29/builds/1531/steps/system_health.memory_desktop.reference/logs/stdio
  https://build.chromium.org/p/chromium.perf/builders/Mac%20HDD%20Perf%20%283%29/builds/1535/steps/system_health.memory_desktop.reference/logs/stdio

Error message:

  [ RUN      ] load:media:soundcloud
  (INFO) 2016-05-14 20:14:37,197 chrome_tracing_agent._CreateTraceConfigFile:243  Trace config file string: {"trace_config":{"excluded_categories": ["*"], "included_categories": ["disabled-by-default-memory-infra"], "record_mode": "record-as-much-as-possible"}}
  (INFO) 2016-05-14 20:14:37,205 tracing_backend.StartTracing:124  Start Tracing Request: {'params': {'transferMode': 'ReturnAsStream', 'options': 'record-as-much-as-possible', 'categories': 'disabled-by-default-memory-infra,-*'}, 'method': 'Tracing.start'}
  (INFO) 2016-05-14 20:14:37,210 cache_temperature.EnsurePageCacheTemperature:29  PageCacheTemperature: any
  (ERROR) 2016-05-14 20:14:46,014 system_health_stories._TakeMemoryMeasurement:319  Unable to get a memory dump for load:media:soundcloud.
  Traceback (most recent call last):
    File "/b/build/slave/Mac_HDD_Perf__3_/build/src/third_party/catapult/telemetry/telemetry/internal/story_runner.py", line 86, in _RunStoryAndProcessErrorIfNeeded
      test.Measure(state.platform, results)
    File "/b/build/slave/Mac_HDD_Perf__3_/build/src/third_party/catapult/telemetry/telemetry/web_perf/timeline_based_measurement.py", line 279, in Measure
      trace_result = platform.tracing_controller.StopTracing()
    File "/b/build/slave/Mac_HDD_Perf__3_/build/src/third_party/catapult/telemetry/telemetry/core/tracing_controller.py", line 39, in StopTracing
      return self._tracing_controller_backend.StopTracing()
    File "/b/build/slave/Mac_HDD_Perf__3_/build/src/third_party/catapult/telemetry/telemetry/internal/platform/tracing_controller_backend.py", line 107, in StopTracing
      self._IssueClockSyncMarker()
    File "/b/build/slave/Mac_HDD_Perf__3_/build/src/third_party/catapult/telemetry/telemetry/internal/platform/tracing_controller_backend.py", line 197, in _IssueClockSyncMarker
      self._RecordIssuerClockSyncMarker)
    File "/b/build/slave/Mac_HDD_Perf__3_/build/src/third_party/catapult/telemetry/telemetry/internal/platform/tracing_agent/chrome_tracing_agent.py", line 181, in RecordClockSyncMarker
      sync_id, record_controller_clock_sync_marker_callback)
    File "/b/build/slave/Mac_HDD_Perf__3_/build/src/third_party/catapult/telemetry/telemetry/internal/platform/tracing_agent/chrome_tracing_agent.py", line 164, in _RecordClockSyncMarkerAsyncEvent
      raise ChromeClockSyncError()
  ChromeClockSyncError

  (INFO) 2016-05-14 20:14:46,109 exception_formatter.PrintFormattedException:19  Try printing formatted exception: None None None

  Exception raised when cleaning story run: 

  Traceback (most recent call last):
    _RunStoryAndProcessErrorIfNeeded at /b/build/slave/Mac_HDD_Perf__3_/build/src/third_party/catapult/telemetry/telemetry/internal/story_runner.py:108
      test.DidRunStory(state.platform)
    DidRunStory at /b/build/slave/Mac_HDD_Perf__3_/build/src/third_party/catapult/telemetry/telemetry/web_perf/timeline_based_measurement.py:293
      platform.tracing_controller.StopTracing()
    StopTracing at /b/build/slave/Mac_HDD_Perf__3_/build/src/third_party/catapult/telemetry/telemetry/core/tracing_controller.py:39
      return self._tracing_controller_backend.StopTracing()
    StopTracing at /b/build/slave/Mac_HDD_Perf__3_/build/src/third_party/catapult/telemetry/telemetry/internal/platform/tracing_controller_backend.py:107
      self._IssueClockSyncMarker()
    _IssueClockSyncMarker at /b/build/slave/Mac_HDD_Perf__3_/build/src/third_party/catapult/telemetry/telemetry/internal/platform/tracing_controller_backend.py:197
      self._RecordIssuerClockSyncMarker)
    RecordClockSyncMarker at /b/build/slave/Mac_HDD_Perf__3_/build/src/third_party/catapult/telemetry/telemetry/internal/platform/tracing_agent/chrome_tracing_agent.py:181
      sync_id, record_controller_clock_sync_marker_callback)
    _RecordClockSyncMarkerAsyncEvent at /b/build/slave/Mac_HDD_Perf__3_/build/src/third_party/catapult/telemetry/telemetry/internal/platform/tracing_agent/chrome_tracing_agent.py:164
      raise ChromeClockSyncError()
  ChromeClockSyncError

  Locals:
    has_clock_synced                             : False
    record_controller_clock_sync_marker_callback : <bound method TracingControllerBackend._RecordIssuerClockSyncMarker of <telemetry.internal.platform.tracing_controller_backend.TracingControllerBackend object at 0x1093223d0>>
    sync_id                                      : '1ffe3f32-e55c-44a0-aefd-9766a25c6e69'

  [  FAILED  ] load:media:soundcloud (8915 ms)

Next action: I would wait for a few more builds to see whether this failure is a flake or not.
 
Blocking: 589726
Blockedon: 612779
Status: Assigned (was: Untriaged)
Marking assigned to petrcermak@.
Project Member

Comment 5 by bugdroid1@chromium.org, May 19 2016

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/330cd00d2ce9cc10d34ba1264d4dcc66c40916b1

commit 330cd00d2ce9cc10d34ba1264d4dcc66c40916b1
Author: catapult-deps-roller <catapult-deps-roller@chromium.org>
Date: Thu May 19 06:37:01 2016

Roll src/third_party/catapult/ 73dc5f4ee..9085ddba7 (15 commits).

https://chromium.googlesource.com/external/github.com/catapult-project/catapult.git/+log/73dc5f4ee8e4..9085ddba7182

$ git log 73dc5f4ee..9085ddba7 --date=short --no-merges --format='%ad %ae %s'

BUG= 352807 , 352807 ,612140, 581716 , 602701 

TBR=catapult-sheriff@chromium.org

Review-Url: https://codereview.chromium.org/1994023002
Cr-Commit-Position: refs/heads/master@{#394699}

[modify] https://crrev.com/330cd00d2ce9cc10d34ba1264d4dcc66c40916b1/DEPS

Cc: petrcermak@chromium.org primiano@chromium.org
 Issue 620390  has been merged into this issue.
Neither of the recent failures (see comment #7) contains a "Failed to record clock sync marker ..." message (https://cs.chromium.org/chromium/src/third_party/catapult/telemetry/telemetry/internal/platform/tracing_agent/chrome_tracing_agent.py?l=163).

This means that by the time the story runner tries to stop tracing, there are no more inspector backends. That in turn implies that the browser/renderer has crashed during the story.

I won't get to try to reproduce the issue locally until next week. In the meantime, we have 2 options regarding the specific benchmark:

  1. Disable system health benchmark on mac-reference.
  2. Keep the benchmark enabled on mac-reference.

and 2 options how to simplify triaging issues like this:

  1. Modify story runner to print screenshots upon exceptions (the functionality is currently in SharedPageState).
  2. Modify story runner to print browser standard output and log upon exceptions (similarly to https://codereview.chromium.org/2072713003/).

My personal preference (definitely biased) would be to keep the benchmark enabled (there were 2 failures in the last 20 builds on each bot, which is below the 1/5 ratio suggested by the sheriffing guide for disabling tests: https://chromium.googlesource.com/chromium/src/+/master/tools/perf/docs/perf_bot_sheriffing.md#Handle-Test-Failures) and get screenshots+logs on all Telemetry failures.

Ned: WDYT?
Petr: keep the benchmark enabled & add more logging data sounds great to me as it also helps the tot one in the future. 
Project Member

Comment 10 by bugdroid1@chromium.org, Jun 20 2016

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/a604c7548e4b1ed89dfc8a8de3f0b5a501214b17

commit a604c7548e4b1ed89dfc8a8de3f0b5a501214b17
Author: petrcermak <petrcermak@chromium.org>
Date: Mon Jun 20 15:50:32 2016

[memory-infra] Add DumpStateUponFailure method to MultiBrowserSharedState

This is necessary for https://codereview.chromium.org/2074183002/
(Telemetry patch in the Catapult repository).

BUG=612140

Review-Url: https://codereview.chromium.org/2085483002
Cr-Commit-Position: refs/heads/master@{#400687}

[modify] https://crrev.com/a604c7548e4b1ed89dfc8a8de3f0b5a501214b17/tools/perf/page_sets/dual_browser_story.py

Project Member

Comment 11 by bugdroid1@chromium.org, Jun 21 2016

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/649d1016cfc6fcd4bcec9ade105f9c384c0fbdf4

commit 649d1016cfc6fcd4bcec9ade105f9c384c0fbdf4
Author: catapult-deps-roller <catapult-deps-roller@chromium.org>
Date: Tue Jun 21 09:32:29 2016

Roll src/third_party/catapult/ bed4eb00e..2c7c14263 (1 commit).

https://chromium.googlesource.com/external/github.com/catapult-project/catapult.git/+log/bed4eb00e376..2c7c14263b4e

$ git log bed4eb00e..2c7c14263 --date=short --no-merges --format='%ad %ae %s'

BUG=612140

TBR=catapult-sheriff@chromium.org

Review-Url: https://codereview.chromium.org/2083653004
Cr-Commit-Position: refs/heads/master@{#400930}

[modify] https://crrev.com/649d1016cfc6fcd4bcec9ade105f9c384c0fbdf4/DEPS

Comment 12 Deleted

Blockedon: 624355
Owner: hjd@chromium.org
Cc: -petrcermak@chromium.org
Cc: -petrcermak@chromium.org

Sign in to add a comment