All graphs for this bug:
  https://chromeperf.appspot.com/group_report?bug_id=794889

(For debugging:) Original alerts at time of bug-filing:
  https://chromeperf.appspot.com/group_report?sid=2e8b265ec5c2fcbe495fe1b081092e0689054e42d84ed93468d8bee871e7c9b6


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

chromium-rel-win8-dual

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/16cb642a040000

📍 Couldn't reproduce a difference.
https://pinpoint-dot-chromeperf.appspot.com/job/16cb642a040000

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/11addee8840000

📍 Pinpoint job started.
https://pinpoint-dot-chromeperf.appspot.com/job/12dd58e4840000

The bisect and many of the graphs look bimodal. It looks like they all recovered. The noise here is pretty concerning though. +tdresser, any ideas?

Kicked off a few more bisects.

Are there specific cases that look noisy?

Of the random sample of 5 graphs I looked at, one became noisy fairly recently, and the rest seemed quite reasonable to me.

😿 Pinpoint job stopped with an error.
https://pinpoint-dot-chromeperf.appspot.com/job/12dd58e4840000

If you look at the pinpoint job in #3, it looks like the values were bimodal:
https://pinpoint-dot-chromeperf.appspot.com/job/16cb642a040000

Then I looked at some graphs:

This one seems to go back and forth between being bimodal and not, especially if you look at the history: https://pinpoint-dot-
chromeperf.appspot.com/job/16cb642a040000

This one too: https://chromeperf.appspot.com/report?sid=771fdfc409ff93eada78707bfc6ad058fa4c8866a7755dff0b3a94f27ba3031c&start_rev=519195&end_rev=526248

Other ones like this seem to go up for a while, then back down, without being bisectable: https://chromeperf.appspot.com/report?sid=0a4700b3292d75104b5e8bdf2a7a980f9fa9af56e0f868f23d8f8e0af9c60b45&start_rev=519195&end_rev=526248

😿 Pinpoint job stopped with an error.
https://pinpoint-dot-chromeperf.appspot.com/job/11addee8840000

dtu, are you looking at these errors?

This doesn't bisect.

The error was https://github.com/catapult-project/catapult/issues/4182, which we fixed 3 weeks ago. I'll kick off another bisect.

📍 Pinpoint job started.
https://chromeperf.appspot.com/job/1489299f840000

📍 Found a significant difference after 1 commit.
https://chromeperf.appspot.com/job/1489299f840000

Update RendererSideResourceScheduler testing config by yhirano@chromium.org
https://chromium.googlesource.com/chromium/src/+/47b8f2451e900fe2c5fc4d94046925f40fea2fc7

Understanding performance regressions:
  http://g.co/ChromePerformanceRegressions

It's already been reverted.