New issue
Advanced search Search tips

Issue 715716 link

Starred by 1 user

Issue metadata

Status: Fixed
Owner:
Closed: Aug 2017
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Linux
Pri: ----
Type: ----



Sign in to add a comment

v8.infinite_scroll-classic_tbmv2 failing on chromium.perf/Linux

Project Member Reported by sullivan@chromium.org, Apr 26 2017

Issue description

v8.infinite_scroll-classic_tbmv2 failing on chromium.perf/Linux Perf

Builders failed on: 
- Android One Perf (2): 
  https://build.chromium.org/p/chromium.perf/builders/Android%20One%20Perf%20%282%29


This is really strange. If you look at the builder history, from build 545 on, the test fails with discourse page. Before that, it fails with facebook page. But if you look at the logs, it's only running one page:

https://luci-logdog.appspot.com/v/?s=chrome%2Fbb%2Fchromium.perf%2FLinux_Perf%2F604%2F%2B%2Frecipes%2Fsteps%2Fv8.infinite_scroll-classic_tbmv2_on__102b__GPU_on_Linux%2F0%2Fstdout
And there is no "Dashboard upload" step: https://uberchromegw.corp.google.com/i/chromium.perf/builders/Linux%20Perf/builds/604

But the perf dashboard has data from build 545 and beyond:
https://chromeperf.appspot.com/report?sid=8f08a0b4b5ec320dee9945f831bfbd50ee6e4c9103248ada09b72e6dc682ea21

Where is that data coming from?

It seems like the logs look the same way on linux before the discourse failure, only failing on facebook instead.

martiniss: any clue what's happening here? How could the perf dashboard have data for this test? Did I miss something here?
 
Summary: v8.infinite_scroll-classic_tbmv2 failing on chromium.perf/Linux (was: v8.infinite_scroll-classic_tbmv2 failing on chromium.perf/Android One Perf (2))
Update: the perf dashboard is showing data for the ref build. The non-ref has no data since build 529 on April 4.
Kicked off a bisect, disabling the benchmark in the short term.
Project Member

Comment 5 by 42576172...@developer.gserviceaccount.com, Apr 27 2017


=== BISECT JOB RESULTS ===
NO Test failure found

Bisect Details
  Configuration: linux_perf_bisect
  Benchmark    : v8.infinite_scroll-classic_tbmv2
  Metric       : memory:chrome:renderer_processes:reported_by_chrome:v8:effective_size_avg/memory:chrome:renderer_processes:reported_by_chrome:v8:effective_size_avg

Revision             Exit Code      N
chromium@458335      0 +- N/A       20      good
chromium@461634      0 +- N/A       20      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests v8.infinite_scroll-classic_tbmv2

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8981221691810648112

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5720522815438848


| 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 6 by u...@chromium.org, Apr 27 2017

Cc: hablich@chromium.org
hablich@, do we still need classic, turbo variants of the benchmarks?

Description: Show this description
Cc: -hablich@chromium.org
Owner: hablich@chromium.org
Assign to hablich@ for answering #6
Project Member

Comment 10 by bugdroid1@chromium.org, Apr 27 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/7a40043b942115c4285eec43109681e0f7c82515

commit 7a40043b942115c4285eec43109681e0f7c82515
Author: sullivan <sullivan@chromium.org>
Date: Thu Apr 27 14:07:29 2017

Disable failing v8.infinite_scroll benchmarks on linux.

BUG= 715716 

Review-Url: https://codereview.chromium.org/2844793002
Cr-Commit-Position: refs/heads/master@{#467666}

[modify] https://crrev.com/7a40043b942115c4285eec43109681e0f7c82515/tools/perf/benchmarks/v8.py

Project Member

Comment 11 by 42576172...@developer.gserviceaccount.com, Apr 27 2017


=== BISECT JOB RESULTS ===
NO Test failure found

Bisect Details
  Configuration: linux_perf_bisect
  Benchmark    : v8.infinite_scroll-classic_tbmv2
  Metric       : memory:chrome:renderer_processes:reported_by_chrome:v8:effective_size_avg/memory:chrome:renderer_processes:reported_by_chrome:v8:effective_size_avg

Revision             Exit Code      N
chromium@458335      0 +- N/A       20      good
chromium@461634      0 +- N/A       20      bad

To Run This Test
  src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests v8.infinite_scroll-classic_tbmv2

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8981155474196555152

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5825775955083264


| 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 12 by u...@chromium.org, May 3 2017

Owner: u...@chromium.org
Status: Assigned (was: Available)
Since hablich@ is OOO, I'll take over this bug.

Annie, why is v8.infinite_scroll_tbmv2 is disabled on Linux? I see that the classic version fails, but don't see failures of v8.infinite_scroll_tbmv2.

The classic version fails with:

TracingTimeoutException: Exception raised while sending a Tracing.requestMemoryDump request:
Traceback (most recent call last):
  File "/b/s/w/ir/third_party/catapult/telemetry/telemetry/internal/backends/chrome_inspector/tracing_backend.py", line 204, in DumpMemory
    response = self._inspector_websocket.SyncRequest(request, timeout)
  File "/b/s/w/ir/third_party/catapult/telemetry/telemetry/internal/backends/chrome_inspector/inspector_websocket.py", line 110, in SyncRequest
    res = self._Receive(timeout)
  File "/b/s/w/ir/third_party/catapult/telemetry/telemetry/internal/backends/chrome_inspector/inspector_websocket.py", line 149, in _Receive
    data = self._socket.recv()
  File "/b/s/w/ir/third_party/catapult/telemetry/third_party/websocket-client/websocket/_core.py", line 293, in recv
    opcode, data = self.recv_data()
  File "/b/s/w/ir/third_party/catapult/telemetry/third_party/websocket-client/websocket/_core.py", line 310, in recv_data
    opcode, frame = self.recv_data_frame(control_frame)
  File "/b/s/w/ir/third_party/catapult/telemetry/third_party/websocket-client/websocket/_core.py", line 323, in recv_data_frame
    frame = self.recv_frame()
  File "/b/s/w/ir/third_party/catapult/telemetry/third_party/websocket-client/websocket/_core.py", line 357, in recv_frame
    return self.frame_buffer.recv_frame()
  File "/b/s/w/ir/third_party/catapult/telemetry/third_party/websocket-client/websocket/_abnf.py", line 336, in recv_frame
    self.recv_header()
  File "/b/s/w/ir/third_party/catapult/telemetry/third_party/websocket-client/websocket/_abnf.py", line 286, in recv_header
    header = self.recv_strict(2)
  File "/b/s/w/ir/third_party/catapult/telemetry/third_party/websocket-client/websocket/_abnf.py", line 371, in recv_strict
    bytes_ = self.recv(min(16384, shortage))
  File "/b/s/w/ir/third_party/catapult/telemetry/third_party/websocket-client/websocket/_core.py", line 427, in _recv
    return recv(self.sock, bufsize)
  File "/b/s/w/ir/third_party/catapult/telemetry/third_party/websocket-client/websocket/_socket.py", line 83, in recv
    raise WebSocketTimeoutException(message)
WebSocketTimeoutException: timed out

Cc: perezju@chromium.org
Juan: the WebSocketTimeoutException exception when collecting memory trace in #12 looks oddly familiar. I remembered we was having them a lot on system health stories on a certain platform?
As of https://luci-milo.appspot.com/buildbot/chromium.perf/Linux%20Perf/611 I don't see this test failing anymore.

Maybe something in http://test-results.appspot.com/revision_range?start=467621&end=467729 fixed it?

On the specific question about WebSocketTimeoutException, we've seen it in the past, but not at a high rate as far as I can remember.
I think Randy's revert probably fixed it: https://codereview.chromium.org/2850143003/
Project Member

Comment 16 by bugdroid1@chromium.org, May 3 2017

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/408c39567722c70872f8d9e07a3d1f697df2a213

commit 408c39567722c70872f8d9e07a3d1f697df2a213
Author: ulan <ulan@chromium.org>
Date: Wed May 03 14:23:13 2017

Re-enable v8.infinite_scroll_tbmv2 on Linux and disable it on Android.

Android runs v8.mobile_infinite_scroll_tbmv2.

BUG= chromium:715716 , chromium:716435 

Review-Url: https://codereview.chromium.org/2857923003
Cr-Commit-Position: refs/heads/master@{#468960}

[modify] https://crrev.com/408c39567722c70872f8d9e07a3d1f697df2a213/testing/buildbot/chromium.perf.fyi.json
[modify] https://crrev.com/408c39567722c70872f8d9e07a3d1f697df2a213/tools/perf/benchmarks/v8.py

Comment 17 by u...@chromium.org, Aug 21 2017

Status: Fixed (was: Assigned)

Sign in to add a comment