Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(322)

Issue 2144073004: [system-health] Unify SH stories interface (Closed)

Created:
4 years, 5 months ago by petrcermak
Modified:
4 years, 5 months ago
Reviewers:
nednguyen, ulan
CC:
chromium-reviews, telemetry-reviews_chromium.org
Base URL:
https://chromium.googlesource.com/chromium/src.git@master
Target Ref:
refs/pending/heads/master
Project:
chromium
Visibility:
Public.

Description

[system-health] Unify SH story set interface This patch modifies SH stories so that all of their users instantiate them via page_sets.SystemHealthStorySet, which has the following parameters: platform: 'desktop' or 'mobile' (mandatory) case: 'load', 'browse', ... (default: None) take_memory_measurement: True or False (default: False) These parameters will be used by existing benchmarks as follows: take_ Benchmark name platform case memory_ measurement ---------------------------------------------------------------------- system_health.memory_desktop 'desktop' 'load' True system_health.memory_mobile 'mobile' 'load' True battor.system_health_loading_desktop 'desktop' 'load' False battor.system_health_loading_mobile 'mobile' 'load' False v8.browsing_desktop 'desktop' 'browse' False v8.browsing_mobile 'mobile' 'browse' False The case of the memory system health benchmarks will be changed to None in a subsequent patch, i.e. they will run all SH stories. As a result, all SH stories will also run in the SH smoke tests in the Chromium CQ. BUG=589726 CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.perf:android_s5_perf_cq;master.tryserver.chromium.perf:mac_retina_perf_cq Committed: https://crrev.com/fc79a085aec3b72004cbc99864291801e2b1ceca Cr-Commit-Position: refs/heads/master@{#405753}

Patch Set 1 #

Total comments: 5

Patch Set 2 : Restrict story selector #

Total comments: 1
Unified diffs Side-by-side diffs Delta from patch set Stats (+104 lines, -135 lines) Patch
M tools/perf/benchmarks/battor.py View 1 2 chunks +2 lines, -2 lines 0 comments Download
M tools/perf/benchmarks/system_health.py View 1 4 chunks +8 lines, -3 lines 0 comments Download
M tools/perf/benchmarks/v8_browsing.py View 1 3 chunks +6 lines, -3 lines 0 comments Download
D tools/perf/page_sets/data/browsing_desktop.json View 1 chunk +0 lines, -20 lines 0 comments Download
D tools/perf/page_sets/data/browsing_mobile.json View 1 chunk +0 lines, -29 lines 0 comments Download
M tools/perf/page_sets/data/memory_system_health_desktop.json View 1 chunk +16 lines, -1 line 0 comments Download
M tools/perf/page_sets/data/memory_system_health_mobile.json View 1 chunk +25 lines, -1 line 1 comment Download
M tools/perf/page_sets/system_health/browsing_stories.py View 4 chunks +2 lines, -39 lines 0 comments Download
M tools/perf/page_sets/system_health/loading_stories.py View 2 chunks +1 line, -8 lines 0 comments Download
M tools/perf/page_sets/system_health/system_health_stories.py View 1 1 chunk +24 lines, -18 lines 0 comments Download
M tools/perf/page_sets/system_health/system_health_story.py View 2 chunks +20 lines, -11 lines 0 comments Download

Messages

Total messages: 25 (10 generated)
petrcermak
PTAL. Thanks, Petr
4 years, 5 months ago (2016-07-14 15:53:24 UTC) #6
nednguyen
https://codereview.chromium.org/2144073004/diff/40001/tools/perf/page_sets/system_health/system_health_stories.py File tools/perf/page_sets/system_health/system_health_stories.py (right): https://codereview.chromium.org/2144073004/diff/40001/tools/perf/page_sets/system_health/system_health_stories.py#newcode20 tools/perf/page_sets/system_health/system_health_stories.py:20: def __init__(self, platform, story_name_regex=r'.*', this story_name_regex is a bit ...
4 years, 5 months ago (2016-07-14 20:01:49 UTC) #7
petrcermak
Thanks for your comments. PTAL. Petr https://codereview.chromium.org/2144073004/diff/40001/tools/perf/page_sets/system_health/system_health_stories.py File tools/perf/page_sets/system_health/system_health_stories.py (right): https://codereview.chromium.org/2144073004/diff/40001/tools/perf/page_sets/system_health/system_health_stories.py#newcode20 tools/perf/page_sets/system_health/system_health_stories.py:20: def __init__(self, platform, ...
4 years, 5 months ago (2016-07-15 09:01:43 UTC) #9
nednguyen
https://codereview.chromium.org/2144073004/diff/40001/tools/perf/page_sets/system_health/system_health_stories.py File tools/perf/page_sets/system_health/system_health_stories.py (right): https://codereview.chromium.org/2144073004/diff/40001/tools/perf/page_sets/system_health/system_health_stories.py#newcode20 tools/perf/page_sets/system_health/system_health_stories.py:20: def __init__(self, platform, story_name_regex=r'.*', On 2016/07/15 09:01:43, petrcermak wrote: ...
4 years, 5 months ago (2016-07-15 10:48:29 UTC) #10
petrcermak
On 2016/07/15 10:48:29, nednguyen wrote: > https://codereview.chromium.org/2144073004/diff/40001/tools/perf/page_sets/system_health/system_health_stories.py > File tools/perf/page_sets/system_health/system_health_stories.py (right): > > https://codereview.chromium.org/2144073004/diff/40001/tools/perf/page_sets/system_health/system_health_stories.py#newcode20 > ...
4 years, 5 months ago (2016-07-15 11:11:10 UTC) #11
nednguyen
lgtm For some reason I thought I did, sorry :P
4 years, 5 months ago (2016-07-15 11:45:33 UTC) #12
commit-bot: I haz the power
CQ is trying da patch. Follow status at https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.org/2144073004/60001
4 years, 5 months ago (2016-07-15 11:56:08 UTC) #14
commit-bot: I haz the power
Committed patchset #2 (id:60001)
4 years, 5 months ago (2016-07-15 13:42:57 UTC) #16
commit-bot: I haz the power
CQ bit was unchecked.
4 years, 5 months ago (2016-07-15 13:42:59 UTC) #17
commit-bot: I haz the power
Patchset 2 (id:??) landed as https://crrev.com/fc79a085aec3b72004cbc99864291801e2b1ceca Cr-Commit-Position: refs/heads/master@{#405753}
4 years, 5 months ago (2016-07-15 13:44:31 UTC) #19
ulan
https://codereview.chromium.org/2144073004/diff/60001/tools/perf/page_sets/data/memory_system_health_mobile.json File tools/perf/page_sets/data/memory_system_health_mobile.json (right): https://codereview.chromium.org/2144073004/diff/60001/tools/perf/page_sets/data/memory_system_health_mobile.json#newcode68 tools/perf/page_sets/data/memory_system_health_mobile.json:68: "browsing_mobile_000.wpr": [ Looks like this broke record_wpr: Traceback (most ...
4 years, 5 months ago (2016-07-18 11:10:47 UTC) #21
nednguyen
On 2016/07/18 11:10:47, ulan wrote: > https://codereview.chromium.org/2144073004/diff/60001/tools/perf/page_sets/data/memory_system_health_mobile.json > File tools/perf/page_sets/data/memory_system_health_mobile.json (right): > > https://codereview.chromium.org/2144073004/diff/60001/tools/perf/page_sets/data/memory_system_health_mobile.json#newcode68 > ...
4 years, 5 months ago (2016-07-18 11:18:26 UTC) #22
ulan
On 2016/07/18 11:18:26, nednguyen wrote: > On 2016/07/18 11:10:47, ulan wrote: > > > https://codereview.chromium.org/2144073004/diff/60001/tools/perf/page_sets/data/memory_system_health_mobile.json ...
4 years, 5 months ago (2016-07-18 13:10:25 UTC) #23
nednguyen
On 2016/07/18 13:10:25, ulan wrote: > On 2016/07/18 11:18:26, nednguyen wrote: > > On 2016/07/18 ...
4 years, 5 months ago (2016-07-18 13:14:23 UTC) #24
ulan
4 years, 5 months ago (2016-07-18 13:47:45 UTC) #25
Message was sent while issue was closed.
On 2016/07/18 13:14:23, nednguyen wrote:
> On 2016/07/18 13:10:25, ulan wrote:
> > On 2016/07/18 11:18:26, nednguyen wrote:
> > > On 2016/07/18 11:10:47, ulan wrote:
> > > >
> > >
> >
>
https://codereview.chromium.org/2144073004/diff/60001/tools/perf/page_sets/da...
> > > > File tools/perf/page_sets/data/memory_system_health_mobile.json (right):
> > > > 
> > > >
> > >
> >
>
https://codereview.chromium.org/2144073004/diff/60001/tools/perf/page_sets/da...
> > > > tools/perf/page_sets/data/memory_system_health_mobile.json:68:
> > > > "browsing_mobile_000.wpr": [
> > > > Looks like this broke record_wpr:
> > > > 
> > > > Traceback (most recent call last):
> > > >   <module> at
> > /usr/local/google/home/ulan/chrome/src/tools/perf/record_wpr:20
> > > >    
> sys.exit(record_wpr.Main(environment=chromium_config.ChromiumConfig()))
> > > >   Main at
> > > >
> > >
> >
>
/usr/local/google/home/ulan/chrome/src/third_party/catapult/telemetry/telemetry/record_wpr.py:282
> > > >     wpr_recorder.HandleResults(results, args.upload)
> > > >   HandleResults at
> > > >
> > >
> >
>
/usr/local/google/home/ulan/chrome/src/third_party/catapult/telemetry/telemetry/record_wpr.py:236
> > > >     upload_to_cloud_storage)
> > > >   AddRecordedStories at
> > > >
> > >
> >
>
/usr/local/google/home/ulan/chrome/src/third_party/catapult/telemetry/telemetry/wpr/archive_info.py:123
> > > >     (target_wpr_file, target_wpr_file_path) = self._NextWprFileName()
> > > >   _NextWprFileName at
> > > >
> > >
> >
>
/usr/local/google/home/ulan/chrome/src/third_party/catapult/telemetry/telemetry/wpr/archive_info.py:202
> > > >     ', doesn\'t begin with ' + base)
> > > > Exception: Illegal wpr file name browsing_desktop_001.wpr, doesn't begin
> > with
> > > > memory_system_health_desktop
> > > > 
> > > > 
> > > > Would renaming browsing_desktop_001.wpr to
> > > memory_system_health_desktop_NNN.wpr
> > > > and re-uploading work?
> > > > Or maybe change archive_info.py to skip the check?
> > > > 
> > > > Let me know if that doesn't work and I will re-record the story set.
> > > 
> > > I think renaming should work. Let keep the archive_info behavior for now.
> > 
> > On the second thought, if we rename the archive files and upload them using
> the
> > same SHA, then this will probably break bots?
> > Since they will have the old json files until the CL with new json files
> lands.
> 
> As long as we just create new copies and not delete the old ones, I think that
> shouldn't cause a problem?

I uploaded https://codereview.chromium.org/2161583003/

Powered by Google App Engine
This is Rietveld 408576698