New issue
Advanced search Search tips
Note: Color blocks (like or ) mean that a user may not be available. Tooltip shows the reason.

Issue 591995 link

Starred by 1 user

Issue metadata

Status: Fixed
Owner:
Last visit > 30 days ago
Closed: Aug 2016
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 2
Type: Bug

Blocked on:
issue 591536

Blocking:
issue 589965



Sign in to add a comment

Angle bisects not finishing.

Project Member Reported by sullivan@chromium.org, Mar 4 2016

Issue description

See https://chromeperf.appspot.com/group_report?bug_id=589965

Most of them don't descend into the angle roll, but this one did and only seemed to have tested one CL: http://build.chromium.org/p/tryserver.chromium.perf/builders/winx64nvidia_perf_bisect/builds/1546

Why didn't they finish? The results aren't clear from the quicklog.

Assigning to robertocn, since he is working on bug triage rotation today.
 
Blocking: 589965
Owner: qyears...@chromium.org
Status: Assigned (was: Unconfirmed)
Summary: Angle bisects not finishing. (was: Angle bisects not finishing)
Blockedon: 591536
Logs are lost, but from the buildbot page I notice that I build timedout after 24 hrs. Bisect posted a build request and was waiting for the archive for 5hrs.
I assume build might have failed on the builder or was in queue to get its turn.

In these kind of situation, bisector should constantly poll the builder for build status and move on with other revisions in the range if build fails or takes more time.
Labels: triaged
Cc: qyears...@chromium.org
Owner: ----
Status: Available (was: Assigned)
Cc: kbr@chromium.org
Is anyone working on this? Sorry I don't understand, how do we prevent these jobs from failing?
Owner: robert...@chromium.org
Roberto: ping, since you're on bug triage rotation again today.
The check for failed builds finally landed today, launching new jobs to see if we get better results.
Status: Fixed (was: Available)
Components: Speed>Bisection

Sign in to add a comment