All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=596100

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICg4KHqogoM


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

win-zenbook

Kicked off a bisect in https://chromeperf.appspot.com/buildbucket_job_status/9017185460022268096

 Issue 596065  has been merged into this issue.

This also bisected down to gavinp@'s change.

Issue 596052 has been merged into this issue.

For reference, here's the change: https://codereview.chromium.org/1790483003 , this is discussed in results pasted into  issue 596065 .

I don't expect any change in performance from this; but I am not an expert on what exactly these finch controls do. asvitkine, can you comment?

The CL enabled that field trial group to be tested by the perf bots and browser tests. It means that the field trial group that was enabled is responsible for the regression (assuming it's clear from bisect that it's that CL).

Aha, then we would expect that. In Canary and Dev, windows simple cache does have problems. E.G. see that we're proceeding to beta only in non-windows platforms.

Thanks asvitkine.

So I guess this is expected.

In that case, I think the right path is reverting the Windows testing config part of the change to restore the performance here on Windows - which shouldn't affect your trial plan for non-Windows platforms.

asvitkine: even though we continue to have a canary trial in Windows? OK. I am heading out very shortly, but I can write that revert up later tonight.

Yep, that should be OK. The testing config is required for experimenting on beta and stable.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/688b2ddd12857e078b629d3d7be1574efdb712e6

commit 688b2ddd12857e078b629d3d7be1574efdb712e6
Author: gavinp <gavinp@chromium.org>
Date: Wed Mar 30 16:11:09 2016

Revert SimpleCache fieldtrial testing information on windows.

This is not needed for the beta push, and the current state of
SimpleCache on windows makes this a regression.

BUG= 596100 ,490029
TBR=asvitkine

Review URL: https://codereview.chromium.org/1843633005

Cr-Commit-Position: refs/heads/master@{#383994}

[modify] https://crrev.com/688b2ddd12857e078b629d3d7be1574efdb712e6/testing/variations/fieldtrial_testing_config_win.json

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/688b2ddd12857e078b629d3d7be1574efdb712e6

commit 688b2ddd12857e078b629d3d7be1574efdb712e6
Author: gavinp <gavinp@chromium.org>
Date: Wed Mar 30 16:11:09 2016

Revert SimpleCache fieldtrial testing information on windows.

This is not needed for the beta push, and the current state of
SimpleCache on windows makes this a regression.

BUG= 596100 ,490029
TBR=asvitkine

Review URL: https://codereview.chromium.org/1843633005

Cr-Commit-Position: refs/heads/master@{#383994}

[modify] https://crrev.com/688b2ddd12857e078b629d3d7be1574efdb712e6/testing/variations/fieldtrial_testing_config_win.json

Graphs have recovered.