Tagging with canary milestone.

Are we concerned with scroll latency, or just main thread busyness? If we just care about how busy the main thread is (and the potential impact to input latency), the Expected Queueing Time metric might be best.

Main thread busyness, yes. Should we just try it out or is there anything else we should know? Does the page have to scroll for expected queuing time?

Jank is often associated with scrolling, but I don't think that's a good example in this case.  Think about an application such as Hangouts where scrolling isn't possible, but animations, video rendering etc is pretty busy and smooth rendering is very important.

Expected Queueing Time measures what input latency would be on average, if input were to occur.

It doesn't require actual input.

Scrolling in particular is a bit tricky, because it sometimes happens on the compositor thread, in which case it's unaffected by main thread busyness.

Based on what you describe, I think EQT is probably the right thing to watch.

Changing back the component as per offline discussion.

Bringing this back to life...
How can we monitor the EQT?

EQT is defined here:
https://cs.chromium.org/chromium/src/third_party/catapult/tracing/tracing/metrics/system_health/expected_queueing_time_metric.html

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/e82554d88243ad84ee3329e7e17ddc6c22e8c250

commit e82554d88243ad84ee3329e7e17ddc6c22e8c250
Author: ehmaldonado <ehmaldonado@chromium.org>
Date: Wed Mar 15 15:17:09 2017

WebRTC: Add a TBMv2 version of the webrtc.webrtc_smoothness benchmark.

This adds the benchmark for the TBMv2 metric created at https://codereview.chromium.org/2711623002/

BUG= chromium:632295 , chromium:671121 

Review-Url: https://codereview.chromium.org/2742293002
Cr-Commit-Position: refs/heads/master@{#457085}

[modify] https://crrev.com/e82554d88243ad84ee3329e7e17ddc6c22e8c250/testing/buildbot/chromium.perf.fyi.json
[modify] https://crrev.com/e82554d88243ad84ee3329e7e17ddc6c22e8c250/testing/buildbot/chromium.perf.json
[modify] https://crrev.com/e82554d88243ad84ee3329e7e17ddc6c22e8c250/tools/perf/benchmarks/webrtc.py

Who should own this metric?

Let's hold off on this metric until the plan here is done: https://codereview.chromium.org/2742293002/#msg19

This issue has been Available for over a year. If it's no longer important or seems unlikely to be fixed, please consider closing it out. If it is important, please re-triage the issue.

Sorry for the inconvenience if the bug really should have been left as Available.

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot

Archiving due to inactivity. Please reopen if this is incorrect.