All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=690226

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgIDg6NSj8wsM


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

android-nexus6

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8988182096167732400

Bisect failed: http://build.chromium.org/p/tryserver.chromium.perf/builders/android_nexus6_perf_bisect/builds/2973
Failure reason: the build has failed due to infrastructure failure.

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8988006269758270336

=== Auto-CCing suspected CL author bcwhite@chromium.org ===

Hi bcwhite@chromium.org, the bisect results pointed to your CL, please take a look at the
results.


=== BISECT JOB RESULTS ===
Perf regression found with culprit

Suspected Commit
  Author : bcwhite
  Commit : 6ebe16e737195e35e4cb250dce734817bd3db339
  Date   : Mon Feb 06 16:10:28 2017
  Subject: Make LocalMemory (aka InMemory) the default for persistent histograms.

Bisect Details
  Configuration: android_nexus6_perf_bisect
  Benchmark    : storage.indexeddb_endure
  Metric       : vm_private_dirty_final_browser/vm_private_dirty_final_browser
  Change       : 1.38% | 25707.7575758 -> 26063.2727273

Revision             Result                  N
chromium@448235      25707.8 +- 55.0298      6      good
chromium@448255      25694.1 +- 31.6997      6      good
chromium@448260      25706.2 +- 46.1001      6      good
chromium@448263      25706.7 +- 46.5616      6      good
chromium@448264      25710.7 +- 48.3914      6      good
chromium@448265      26063.4 +- 59.9011      6      bad       <--
chromium@448275      26063.2 +- 54.0405      6      bad
chromium@448314      26063.3 +- 67.2029      6      bad

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 storage.indexeddb_endure

Debug Info
  https://chromeperf.appspot.com/buildbucket_job_status/8988006269758270336

Is this bisect wrong?
  https://chromeperf.appspot.com/bad_bisect?try_job_id=6179403844812800


| 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 believe that this is due to the addition of 2MB of memory shared between the browser and each renderer.  It's only realized when used so isn't wasting anything but may be counted double (once against the renderer and once against the browser) because of the sharing.

The actual memory use is pretty consistent with not using persistent histograms at all.

Does that explain the result?

Ping?

+test owner.

Looking at the graphs, looks like this recovered in 451072 - 451165:

https://chromeperf.appspot.com/group_report?bug_id=690226

Which includes r451093:
commit	8cf7e79b5bb95b8108861d8d7b910afdbc1db94d	
author	bcwhite <bcwhite@chromium.org>	Thu Feb 16 20:44:17 2017
Move to MappedFile as default for persistent histograms.

So seems fine to close.