media: Add UMA to report media pipeline start result

media: Add UMA to report media pipeline start result

Recently we see some changes in the pipeline start success rate.
However, we can only get this statistics indirectly through some other
UMAs. This CL adds a UMA directly for this.

Note that if the media pipeline is destroyed during the starting
process, this UMA will NOT be reported. This should be relatively rare.
Also, in this case, the result won't affect user experience anyways.

TEST=Manually tested to make sure the metric is reported.

Committed: https://crrev.com/1e865beae078d9f966d2775208d5429fc4469939
Cr-Commit-Position: refs/heads/master@{#422623}

[xhwang, 2016-09-30 03:35:54]

The CQ bit was checked by xhwang@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-09-30 03:36:13]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-09-30 05:01:41]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-09-30 05:01:42]

Dry run: This issue passed the CQ dry run.

[xhwang, 2016-09-30 06:00:42]

Description was changed from

==========
media: Add UMA to report media pipeline start result

Recently we see some changes in the pipeline start success rate.
However, we can only get this statistics indirectly through some other
UMAs. This CL adds a UMA directly for this.

Note that if the media pipeline is destroyed during the starting
process, this UMA will NOT be reported. This should be relatively rare.
Also, in this case, the result won't affect user experience anyways.

TEST=Manually tested to make sure the metric is reported.
BUG=
==========

to

==========
media: Add UMA to report media pipeline start result

Recently we see some changes in the pipeline start success rate.
However, we can only get this statistics indirectly through some other
UMAs. This CL adds a UMA directly for this.

Note that if the media pipeline is destroyed during the starting
process, this UMA will NOT be reported. This should be relatively rare.
Also, in this case, the result won't affect user experience anyways.

TEST=Manually tested to make sure the metric is reported.
==========

[xhwang, 2016-09-30 06:00:42]

xhwang@chromium.org changed reviewers:
+ sandersd@chromium.org

[xhwang, 2016-09-30 06:01:14]

xhwang@chromium.org changed reviewers:
+ dalecurtis@chromium.org

[xhwang, 2016-09-30 06:01:15]

sandersd: PTAL
dalecurtis: FYI

[DaleCurtis, 2016-09-30 16:44:18]

What failures end up reported via this path? Does this end up reporting
initialization failures?

[xhwang, 2016-09-30 17:45:41]

On 2016/09/30 16:44:18, DaleCurtis_OOO_Until_Oct_18 wrote:
> What failures end up reported via this path? Does this end up reporting
> initialization failures?

Good question.

Since we are using the serial runner, this will report any failure from
InitializeDemuxer(), ReportMetadata() and InitializeRenderer(). Since
ReportMetadata() is a closure and will not fail, so this is really just about
InitializeDemuxer() and InitializeRenderer().

https://cs.chromium.org/chromium/src/media/base/pipeline_impl.cc?rcl=0&l=242

I did a quick search, it seems InitializeDemuxer() will only return "DEMUXER"
errors (with one exception in FFmpegDemuxer [1]). InitializeRenderer() will
return other RENDERER/DECODER init errors I suppose (didn't check closely
though).

[1] It could return PIPELINE_ERROR_ABORT in some rare cases:
https://cs.chromium.org/chromium/src/media/filters/ffmpeg_demuxer.cc?rcl=0&l=...

[xhwang, 2016-09-30 17:45:44]

On 2016/09/30 16:44:18, DaleCurtis_OOO_Until_Oct_18 wrote:
> What failures end up reported via this path? Does this end up reporting
> initialization failures?

Good question.

Since we are using the serial runner, this will report any failure from
InitializeDemuxer(), ReportMetadata() and InitializeRenderer(). Since
ReportMetadata() is a closure and will not fail, so this is really just about
InitializeDemuxer() and InitializeRenderer().

https://cs.chromium.org/chromium/src/media/base/pipeline_impl.cc?rcl=0&l=242

I did a quick search, it seems InitializeDemuxer() will only return "DEMUXER"
errors (with one exception in FFmpegDemuxer [1]). InitializeRenderer() will
return other RENDERER/DECODER init errors I suppose (didn't check closely
though).

[1] It could return PIPELINE_ERROR_ABORT in some rare cases:
https://cs.chromium.org/chromium/src/media/filters/ffmpeg_demuxer.cc?rcl=0&l=...

[DaleCurtis, 2016-09-30 17:51:37]

Thanks for clarifying. Want to add some of those details to the histogram.xml
info? lgtm

[xhwang, 2016-09-30 18:20:19]

xhwang@chromium.org changed reviewers:
+ isherman@chromium.org

[xhwang, 2016-09-30 18:20:20]

isherman: Please OWNERS review tools/metrics/histograms/histograms.xml

[xhwang, 2016-09-30 18:20:33]

On 2016/09/30 17:51:37, DaleCurtis_OOO_Until_Oct_18 wrote:
> Thanks for clarifying. Want to add some of those details to the histogram.xml
> info?

Done.

[Ilya Sherman, 2016-10-03 21:13:04]

metrics lgtm

[xhwang, 2016-10-03 21:22:05]

The CQ bit was checked by xhwang@chromium.org

[xhwang, 2016-10-03 21:22:05]

The patchset sent to the CQ was uploaded after l-g-t-m from
dalecurtis@chromium.org
Link to the patchset: https://codereview.chromium.org/2383493004/#ps20001
(title: "update histogram comments")

[commit-bot: I haz the power, 2016-10-03 21:23:02]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-10-04 00:24:06]

Description was changed from

==========
media: Add UMA to report media pipeline start result

Recently we see some changes in the pipeline start success rate.
However, we can only get this statistics indirectly through some other
UMAs. This CL adds a UMA directly for this.

Note that if the media pipeline is destroyed during the starting
process, this UMA will NOT be reported. This should be relatively rare.
Also, in this case, the result won't affect user experience anyways.

TEST=Manually tested to make sure the metric is reported.
==========

to

==========
media: Add UMA to report media pipeline start result

Recently we see some changes in the pipeline start success rate.
However, we can only get this statistics indirectly through some other
UMAs. This CL adds a UMA directly for this.

Note that if the media pipeline is destroyed during the starting
process, this UMA will NOT be reported. This should be relatively rare.
Also, in this case, the result won't affect user experience anyways.

TEST=Manually tested to make sure the metric is reported.
==========

[commit-bot: I haz the power, 2016-10-04 00:24:09]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2016-10-04 00:26:49]

Description was changed from

==========
media: Add UMA to report media pipeline start result

Recently we see some changes in the pipeline start success rate.
However, we can only get this statistics indirectly through some other
UMAs. This CL adds a UMA directly for this.

Note that if the media pipeline is destroyed during the starting
process, this UMA will NOT be reported. This should be relatively rare.
Also, in this case, the result won't affect user experience anyways.

TEST=Manually tested to make sure the metric is reported.
==========

to

==========
media: Add UMA to report media pipeline start result

Recently we see some changes in the pipeline start success rate.
However, we can only get this statistics indirectly through some other
UMAs. This CL adds a UMA directly for this.

Note that if the media pipeline is destroyed during the starting
process, this UMA will NOT be reported. This should be relatively rare.
Also, in this case, the result won't affect user experience anyways.

TEST=Manually tested to make sure the metric is reported.

Committed: https://crrev.com/1e865beae078d9f966d2775208d5429fc4469939
Cr-Commit-Position: refs/heads/master@{#422623}
==========

[commit-bot: I haz the power, 2016-10-04 00:26:50]

Patchset 2 (id:??) landed as
https://crrev.com/1e865beae078d9f966d2775208d5429fc4469939
Cr-Commit-Position: refs/heads/master@{#422623}