[9.2] [Tempest] Volume backup is in ERROR status

[Yury Tregubov]

Several tests Cinder API in Tempest suites are failed after about snapshot #518.

test_backup_create_attached_volume[compute,id-07af8f6d-80af-44c9-a5dc-c8427b1b62e6]

Traceback (most recent call last):

  File "tempest/test.py", line 100, in wrapper

    return f(self, *func_args, **func_kwargs)

  File "tempest/api/volume/test_volumes_backup.py", line 105, in test_backup_create_attached_volume

    backup['id'], 'available')

  File "tempest/common/waiters.py", line 224, in wait_for_backup_status

    raise lib_exc.VolumeBackupException(backup_id=backup_id)

tempest.lib.exceptions.VolumeBackupException: Volume backup 717029e6-bc04-4bc2-be78-331ef0194a6b failed and is in ERROR status

Diagnostic snapshot is available here: 172.18.162.134:/home/jenkins/snapshots/fuel-snapshot-2016-11-21_18-22-14.tar

Env where the problem usually reporduces is available on 172.18.162.134
Env name is MOS_CI_MirantisOpenStack-9.0.iso_ironic_cinder. Fuel master ip is 10.109.2.2.

[Michael Dovgal (mdovgal)]

In cinder-backup service's log (node-3) we have such a traceback.
http://paste.openstack.org/show/590199/
Cinder seems to have been configured without access to create backups in swift storage.

[Nikita Karpin (mkarpin)]

I cannot reproduce this manually, can you please provide commands for reproducing this on the environment? Also cinder config looks as usual if compare with MOS 9.0 .

[Nikita Karpin (mkarpin)]

I was able to create backup with command:

 cinder -d backup-create --force --name attached_backup1 42e0f507-85ba-4bca-990f-3c67f505eea8 --container volumes_backup

under admin/admin

output:

http://paste.openstack.org/show/590303/

[Ivan Udovichenko (iudovichenko)]

Please provide more information.

[Oleksiy Butenko (obutenko)]

tempest use role "Member" for users, but in swift config:
/etc/swift/proxy-server.conf:operator_roles = admin, SwiftOperator, _member_

Fix was added to tempest.conf for using role "_member_".

{0} tempest.api.volume.test_volumes_backup.VolumesBackupsV2Test.test_backup_create_attached_volume [119.323999s] ... ok
{1} tempest.api.volume.test_volumes_backup.VolumesBackupsV1Test.test_backup_create_attached_volume [120.664173s] ... ok
{0} tempest.api.volume.test_volumes_backup.VolumesBackupsV2Test.test_volume_backup_create_get_detailed_list_restore_delete [91.814946s] ... ok
{1} tempest.api.volume.test_volumes_backup.VolumesBackupsV1Test.test_volume_backup_create_get_detailed_list_restore_delete [94.073447s] ... ok

======
Totals
======
Ran: 4 tests in 229.0000 sec.
 - Passed: 4
 - Skipped: 0
 - Expected Fail: 0
 - Unexpected Success: 0
 - Failed: 0
Sum of execute time for each test: 425.8766 sec.

We need wait for result next nightly job.

[Ekaterina Shutova (eshutova)]

Verified on 9.2 snapshot #574. Isuue is fixed, tests ate passed.

[Yury Tregubov (ytregubov)]

Verified on 10.0 iso #1304. The problem is not reproduced on Tempest tests.