Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8994010365414049424

Also failing on https://build.chromium.org/p/chromium.perf/builders/Android%20One%20Perf%20%283%29?numbuilds=200

Also failing on https://build.chromium.org/p/chromium.perf/builders/Android%20Nexus7v2%20Perf%20(3)

Bisect failed: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus9_perf_bisect/builds/2306
Failure reason: the build has failed due to infrastructure failure.

Looks like the bisect hit the hour timeout running the test. Is this the failure we were trying to catch? If not, should we just re-run?

That is not the expected failure, I will restart.

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8993996362305223360

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8993996038001611472

The test appears to take 3 hours. If the bisect timeout is 1 hour, it won't work :P

First failing build was:
https://build.chromium.org/p/chromium.perf/builders/Android%20Nexus9%20Perf%20%283%29/builds/3691


+ksakamoto, +nednguyen There's a CL related to this benchmark in the build's regression range. I'm going to revert.

But I also suspect it's not actually this CL, but rather maybe enabling traffic shaping increases the runtime so much that it hits some kind of timeout in the Android test runner code? It doesn't always stop at the same page, so I suspect some kind of timeout. The logging output says the failing test is "PrintStep" which is in local_device_perf_test_run.py.


https://codereview.chromium.org/2493063002
Enable traffic shaping in loading.mobile benchmark

After this patch, loading.mobile benchmark will run with two traffic
configurations:

- No traffic shaping (as before)
- "Regular 3G" (Down 750kbps, Up 250kbps, 100ms RTT)

Randy pointed me to the place where the timeout is, so instead of a revert, here's a CL to increase the timeout.
https://codereview.chromium.org/2556753003

For this benchmark we should only bisect per story.

Bisect failed: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus7_perf_bisect/builds/3537
Failure reason: the build has failed due to infrastructure failure.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/b08e4a84471a430cf1d76e38cf00685c997334e1

commit b08e4a84471a430cf1d76e38cf00685c997334e1
Author: dtu <dtu@chromium.org>
Date: Tue Dec 06 21:50:30 2016

Increase Android perf test timeout from 3 to 5 hours.

Now that loading.mobile is being run twice, both without and with traffic
shaping, it takes over 3 hours.

BUG= 671640 

Review-Url: https://codereview.chromium.org/2556753003
Cr-Commit-Position: refs/heads/master@{#436750}

[modify] https://crrev.com/b08e4a84471a430cf1d76e38cf00685c997334e1/build/android/pylib/local/device/local_device_perf_test_run.py

Bisect failed: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus9_perf_bisect/builds/2308
Failure reason: the build has failed due to infrastructure failure.

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8982628385813574896

Bisect failed: None
Failure reason: incorrect bisect configuation.

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8982580610908531664