New issue
Advanced search Search tips

Issue 901981 link

Starred by 1 user

Issue metadata

Status: Fixed
Owner:
Closed: Nov 5
Components:
EstimatedDays: ----
NextAction: ----
OS: Chrome
Pri: 1
Type: Bug



Sign in to add a comment

CQ HWTest failures: suite timeout

Project Member Reported by pprabhu@chromium.org, Nov 5

Issue description

A slew of paladins failed with suite timeout in the CQ run starting 11/5 08:26 AM PST

https://cros-goldeneye.corp.google.com/chromeos/healthmonitoring/buildDetails?buildbucketId=8930681237042883008
https://cros-goldeneye.corp.google.com/chromeos/healthmonitoring/buildDetails?buildbucketId=8930681235430619056
https://cros-goldeneye.corp.google.com/chromeos/healthmonitoring/buildDetails?buildbucketId=8930681226242570048
https://cros-goldeneye.corp.google.com/chromeos/healthmonitoring/buildDetails?buildbucketId=8930681191134995968
https://cros-goldeneye.corp.google.com/chromeos/healthmonitoring/buildDetails?buildbucketId=8930681155166688864
...

digging into one:

https://cros-goldeneye.corp.google.com/chromeos/healthmonitoring/buildDetails?buildbucketId=8930681237042883008

Failed suite logs: https://stainless.corp.google.com/browse/chromeos-autotest-results/255476296-chromeos-test/hostless/

The test that was "aborted" is not linked to directly in any way, but searching on http://cautotest-prod by the job name yields: 
http://cautotest-prod/afe/#tab_id=view_job&object_id=255476356
https://stainless.corp.google.com/browse/chromeos-autotest-results/255476356-chromeos-test/chromeos4-row8-rack3-host3/

Which completed just fine (according to status.log).

The timing of this failure is suspect. I think that the suite job failed to insert tko_job_keyvals on test completion, and then reported it poorly as a suite failure.
The failed tests are in the small window where there was a tko_job_keyvals outage due to crbug/890970#c9
 
Status: Assigned (was: Untriaged)
I verified one more builder for the same symptom. Gotta chalk it up to the failed migration. I don't expect the failures to continue.

But I did learn that we can expect these failures the next time we run the migration later this week.... 

bug OPEN --> CLOSE_WAIT
Status: Fixed (was: Assigned)
CLOSE_WAIT --> CLOSE

Spot checked that some of the currently running paladins have passed the HWTest stages. So this particular bug is fixed (it affects *all* hwtests)

Sign in to add a comment