2 Errors in vdsm storage tests

David Caro dcaroest at redhat.com
Fri Jun 6 07:43:30 UTC 2014


On Thu 05 Jun 2014 11:31:16 PM CEST, Nir Soffer wrote:
> Hi all,
>
> I found that vdsm storage tests failed [1], and checked console output. Turns out that
> the storage tests were fine, but the test infrastructure is broken:
>
> 21:59:17 StorageTest
> 21:59:17     testCreatePoolErrors                                        OK
> 21:59:23     testStorage(backendType='localfs', domVersion='0')          OK
> 22:00:40     testStorage(backendType='localfs', domVersion='3')          OK
> 22:00:50
> ...
> 22:00:52 [vdsm_master_storage_functional_tests_localfs_gerrit] $ /bin/sh -xe /tmp/hudson2560576493759199190.sh
> 22:00:52 + sudo cp /var/log/vdsm/vdsm.log
> 22:00:52 cp: missing destination file operand after '/var/log/vdsm/vdsm.log'
> 22:00:52 Try 'cp --help' for more information.
> 22:00:54 POST BUILD TASK : FAILURE
> 22:00:54 END OF POST BUILD TASK : 0
> 22:00:54 Archiving artifacts
> 22:00:55 ERROR: No artifacts found that match the file pattern "*vdsm.log". Configuration error?
> 22:00:55 ERROR: ?*vdsm.log? doesn?t match anything
> 22:00:55 Build step 'Archive the artifacts' changed build result to FAILURE
> 22:00:55 Recording test results
> 22:00:56 Build step 'Groovy Postbuild' marked build as failure
> 22:00:57 Finished: FAILURE
>
> This bogus failure showed on gerrit as:
>
> http://jenkins.ovirt.org/job/vdsm_master_storage_functional_tests_localfs_gerrit/1100/ : FAILURE
>
> There are 3 errors here:
>
> 1. cp command is missing an argument
>    I fixed this in the job configuration
>
> 2. Error in artifacts archiving - probably because the previous step in the script failed,
>    but nobody checks for error codes.
>    Maybe the script should run with "set -e" ?

Not really, The postbuild tasks shell scripts have the -e flag set by 
default, so that's already happening, that's why it failed.

>
> 3. Error in the tests environment should result in "ERROR" output on gerrit, not "FAILURE".

Can you please specify a way to distinguish env/infra failures from 
test failures from the logs? All I can do is filter the logs for 
strings matching a pattern and set one message or the other accordingly.
Usually looking for the test failure string[s] to set the test failure 
state is the best approach.


Can you please open a track ticket 
(https://fedorahosted.org/ovirt/newticket) for any future issue?

Thanks!

>
> Hopefully someone can fix these errors.
>
> [1] http://jenkins.ovirt.org/job/vdsm_master_storage_functional_tests_localfs_gerrit/1100/
>
> Thanks,
> Nir
> _______________________________________________
> Infra mailing list
> Infra at ovirt.org
> http://lists.ovirt.org/mailman/listinfo/infra



--
David Caro

Red Hat S.L.
Continuous Integration Engineer - EMEA ENG Virtualization R&D

Email: dcaro at redhat.com
Web: www.redhat.com
RHT Global #: 82-62605

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 473 bytes
Desc: OpenPGP digital signature
URL: <http://lists.ovirt.org/pipermail/infra/attachments/20140606/a41e0168/attachment.sig>


More information about the Infra mailing list