Analyze flakiness impact of a change on entire test corpus

Gabriel Charette

unread,

Sep 23, 2021, 2:45:45 PM9/23/21

to infra-dev, Dana Jansens, Daniel Cheng

Hi Infra,

I have two ideas for changes in //base that could reduce flakiness across all unit tests in Chromium. Is there a way to tell what the impact of such a change is (either experimentally or retroactively after landing)?

Thanks,

Gab

Dirk Pranke

unread,

Sep 23, 2021, 4:04:59 PM9/23/21

to Gabriel Charette, chrome-flakiness, infra-dev, Dana Jansens, Daniel Cheng

Hi Gab,

infra-dev@ is probably not the right list to handle this; let's try chrome-flakiness@ instead :).

-- Dirk

--
You received this message because you are subscribed to the Google Groups "infra-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to infra-dev+...@chromium.org.
To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/infra-dev/CAJTZ7LKcwU0-2UdzjUicJBZxyqbD5ju_4XFByff41Yary4kGsg%40mail.gmail.com.

Bruce Dawson

unread,

Sep 23, 2021, 6:43:20 PM9/23/21

to Dirk Pranke, David Bienvenu, Gabriel Charette, chrome-flakiness, infra-dev, Dana Jansens, Daniel Cheng

David Bienvenu is also a mandatory resource for this question since he has been doing a lot of work on reducing test flakiness, which has included monitoring test flakiness (and working with infra on improving the monitoring)

To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/infra-dev/CAEoffTCDD9G3VMfPjfbjx4Z31cqAWanVdqdPzuzQ-ruKEv_VuA%40mail.gmail.com.