All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=663172

Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICg48fG_wgM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgo-SrpQoM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgo6WmrgkM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgk7aL7QgM,agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICgk8v3tQoM


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

android-nexus5
android-nexus7v2
android-nexus9
linux-release

Started bisect job https://chromeperf.appspot.com/buildbucket_job_status/8996605834415574304

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

Hi reillyg@chromium.org, the bisect results pointed to your CL below as possibly
causing a regression. Please have a look at this info and see whether
your CL be related.


===== BISECT JOB RESULTS =====
Status: completed


===== SUSPECTED CL(s) =====
Subject : Port messages sent by WebIDBDatabaseImpl to Mojo.
Author  : reillyg
Commit description:
  
This is part two of the series of patches that converts the IPC messages
sent by WebIDBFactoryImpl, WebIDBDatabaseImpl and WebIDBCursorImpl to
Mojo messages.

This is a more mechanical change because the conversion of messages sent
by WebIDBFactoryImpl added most of the necessary infrastructure.

BUG= 627484 

Review-Url: https://codereview.chromium.org/2449953008
Cr-Commit-Position: refs/heads/master@{#430110}
Commit  : 963ec6df64089bef6867ccdc608101d5fc06d5c0
Date    : Sat Nov 05 01:13:13 2016


===== TESTED REVISIONS =====
Revision         Mean     Std Dev  N  Good?
chromium@430085  32963.4  917.086  5  good
chromium@430100  32403.6  802.43   5  good
chromium@430107  32236.8  261.709  5  good
chromium@430109  32321.2  585.359  5  good
chromium@430110  35946.2  509.502  5  bad    <--
chromium@430111  36887.2  864.5    5  bad
chromium@430114  36154.6  760.34   5  bad
chromium@430142  36612.0  865.585  5  bad

Bisect job ran on: linux_perf_bisect
Bug ID: 663172

Test Command: src/tools/perf/run_benchmark -v --browser=release --output-format=chartjson --upload-results --pageset-repeat=1 --also-run-disabled-tests storage.indexeddb_endure
Test Metric: vm_private_dirty_final_browser/vm_private_dirty_final_browser
Relative Change: 11.07%
Score: 99.9

Buildbot stdio: http://build.chromium.org/p/tryserver.chromium.perf/builders/linux_perf_bisect/builds/6824
Job details: https://chromeperf.appspot.com/buildbucket_job_status/8996605834415574304


Not what you expected? We'll investigate and get back to you!
  https://chromeperf.appspot.com/bad_bisect?try_job_id=5209742891286528

| 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 Tests>AutoBisect.  Thank you!

@reillyg: The regression was pretty significant and multiple bisects (see also issue 663161) have pointed to the patch in #3.

Was a regression expected? Should we revert?

Some regressions were expected and I am actively investigating this issue now.

Looks like there were a number of wins too:
https://chromeperf.appspot.com/group_report?rev=430110

Any movement here? This still hasn't recovered in some cases.

I investigated this and was unable to pinpoint any single cause for the increased memory usage because the data were simply too noisy.

These metrics have been deprecated.