Au test automated re-run did not happen |
|||||||
Issue descriptionChrome OS version: 8350.21.1/ 52.0.2743.32 Automated re-runs seem to not have kicked off and failed/ missing tests had to be manually re-run. Most of them subsequently passed. Below is a link to a failed test for Monroe, as a sample. This may have a real issue as several re-runs did not pass the test. https://uberchromeos-server38.corp.google.com/new_tko/#tab_id=spreadsheet_view&row=job_name%252Ctest_name&column=status&show_incomplete=true&show_only_latest=false&show_invalid=true&condition=job_name%2520LIKE%2520%27monroe-release/R52-8350.21.1/paygen%255C_au%255C_dev/%2525%27%2520AND%2520test_name%2520%253C%253E%2520%27SERVER_JOB%27%2520AND%2520test_name%2520NOT%2520LIKE%2520%27CLIENT%255C_JOB%2525%27
,
Jun 10 2016
In the first link you sent me, I am only seeing passing tests I think (also, not sure where you got that link from -- it is a tko query link? I'm not used to looking at those; also, it seems to be a link directly to an autotest shard server rather than to the master). Can you pull up the exact build, or the exact suite that was launched by it, for me to examine?
,
Jun 10 2016
Here's a daisy build that failed au and later passed after manual re-run. https://uberchromegw.corp.google.com/i/chromeos_release/builders/daisy-release%20release-R52-8350.B/builds/21
,
Jun 10 2016
Also there's something that's not clear to me: in GE, this page shows that most bvt-cq passed (https://cros-goldeneye.corp.google.com/console/listBuild?milestone=52#/details) but when I created the release record, there were a ton of failed bvts that I had to re-run (both cq and inline). I'll check with GE folks.
,
Jun 11 2016
Moving this nonessential bug to the next milestone. For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
,
Jun 14 2016
Re #3, do you happen to have a link to your re-run as well?
,
Jun 14 2016
For posterity here is the failing au test suite from #3 http://cautotest.corp.google.com/afe/#tab_id=view_job&object_id=66319797
,
Jun 14 2016
akeshet@: So can the retries be enabled for all failing tests so that we don't have to kick these off manually?
,
Jun 14 2016
This bug was opened after the re-runs were already done so there wasn't much to show for failed tests. But for bvt inline, bvt cq and au tests, I had to do at least 4 re-runs for about 29 failing boards in each test category. It would be a great help if these flaky tests could be re-run automatically because 99% of them end up passing after 3-4 re-runs, which always delays releases as we can't pass them to test unless everything is green.
,
Jun 14 2016
,
Jun 14 2016
RE #8 and 9 yes, that is what i'm trying to resolve. Would help to have the logs of both the failure (already found) and the subsequent retry.
,
Jun 17 2016
I'll update the bug early next week and include logs for before and after re-runs
,
Jul 6 2016
Moving this nonessential bug to the next milestone. For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
,
Nov 14 2016
Is this still relevant?
,
Nov 15 2017
This issue has been Available for over a year. If it's no longer important or seems unlikely to be fixed, please consider closing it out. If it is important, please re-triage the issue. Sorry for the inconvenience if the bug really should have been left as Available. If you change it back, also remove the "Hotlist-Recharge-Cold" label. For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
,
May 14 2018
|
|||||||
►
Sign in to add a comment |
|||||||
Comment 1 by gkihumba@chromium.org
, Jun 10 2016