Win Builder (dbg) builds on luci.chromium.ci are failing |
|||||||||||||||||
Issue descriptionhttps://ci.chromium.org/p/chromium/g/main/console The builds in chromium.win/debug/builder/32 are failing about 50% of the time. It looks like the successful builds are running on buildbot. (e.g. https://ci.chromium.org/buildbot/chromium.win/Win%20Builder%20(dbg)/40533) The non successful builds are on luci.chromium.ci (e.g. https://ci.chromium.org/p/chromium/builders/luci.chromium.ci/Win%20Builder%20(dbg)/133) Please check what is happening here.
,
Dec 18 2017
,
Dec 20 2017
Where did you the LUCI builds shown https://ci.chromium.org/p/chromium/g/main/console ? (screenshot would be nice) To my best knowledge it still shows only buildbot ones, but it's possible we've messed something up.
,
Dec 20 2017
+jchinlee@ who is currently in early stages of migrating builder to LUCI.
,
Dec 20 2017
IIUC this is just in the console header. The builder is not fully migrated yet; perhaps it'd be less confusing to remove the builder from the header. I'll put something together for that.
,
Dec 21 2017
+nodir@ Oh, I didn't think about the header. Good point - I think header should use the same logic as console column to display latest prod build from whichever builders are configured for it. Pri1 because this is blocking migrations.
,
Dec 21 2017
the problem seems to be stemming from the fact that we are using def.Builders instead of def.Def.Builders in https://cs.chromium.org/chromium/infra/go/src/go.chromium.org/luci/milo/buildsource/console.go?q=console.go+&sq=package:chromium&l=167 and thus flatten the list of builder ids (we should not) I believe +iannucci has most context
,
Dec 22 2017
Removing from sheriff-queue, as I don't think there is anything for sheriffs to do here.
,
Dec 22 2017
,
Jan 3 2018
When I filed the bug, both luci and buildbot builds were shown in the chromium.win/debug/builder/32 column, which was quite confusing as successful and non successful builds alternated. Now only the header seems to be affected.
,
Jan 25 2018
,
Jan 25 2018
,
Jan 25 2018
Marked as dup
,
Jan 30 2018
,
Jan 30 2018
,
Feb 1 2018
Not a dupe! This is specificially about the console headers, not the console view (which we don't have enough info for yet).
,
Feb 1 2018
,
Feb 1 2018
,
Feb 1 2018
So what's happening is that the console header-rendering code operates differently from the console rendering code. It should be pretty easy to make them behave identically.
,
Feb 13 2018
We have been looking at this actively and have seen this recently. dullweber@chromium.org if you see this again, please let us know. |
|||||||||||||||||
►
Sign in to add a comment |
|||||||||||||||||
Comment 1 by dullweber@chromium.org
, Dec 18 2017