Problem Creating Volume Backups

asked 2017-07-13 23:36:03 -0600

CDOT gravatar image

I am trying to create a backup of a root volume and whose instance is in running state using :

 cinder backup-create --name BackupTest de3e9f5b-a628-4fa4-812c-b35ee1a9e29e --force

The output is :

+-----------+--------------------------------------+
| Property  | Value                                |
+-----------+--------------------------------------+
| id        | d9c49ba8-6614-42b6-b591-07ac378260a1 |
| name      | BackupTest                           |
| volume_id | de3e9f5b-a628-4fa4-812c-b35ee1a9e29e |
+-----------+--------------------------------------+

Then I run this repeatedly :

 cinder backup-list | grep de3e9f5b-a628-4fa4-812c-b35ee1a9e29e

The output is at first :

| d9c49ba8-6614-42b6-b591-07ac378260a1 | de3e9f5b-a628-4fa4-812c-b35ee1a9e29e | creating | BackupTest | 10 | 0 | volumebackups |

After sometime the output changes to :

| d9c49ba8-6614-42b6-b591-07ac378260a1 | de3e9f5b-a628-4fa4-812c-b35ee1a9e29e | error | BackupTest | 10 | 0 | volumebackups |

On running cinder backup-show d9c49ba8-6614-42b6-b591-07ac378260a1, the output is :

+-----------------------+-------------------------------------------------------------------+
| Property              | Value                                                             |
+-----------------------+-------------------------------------------------------------------+
| availability_zone     | nova                                                              |
| container             | volumebackups                                                     |
| created_at            | 2017-07-14T03:26:54.000000                                        |
| data_timestamp        | 2017-07-14T03:26:54.000000                                        |
| description           | None                                                              |
| fail_reason           | Timed out waiting for a reply to message ID 3eb06b612efc4e8399a689|
                          1a3d6f78f7                                                        |
| has_dependent_backups | False                                                             |
| id                    | d9c49ba8-6614-42b6-b591-07ac378260a1                              |
| is_incremental        | False                                                             |
| name                  | BackupTest                                                        |
| object_count          | 0                                                                 |
| size                  | 10                                                                |
| snapshot_id           | None                                                              |
| status                | error                                                             |
| updated_at            | 2017-07-14T03:27:55.000000                                        |
| volume_id             | de3e9f5b-a628-4fa4-812c-b35ee1a9e29e                              |
+-----------------------+-------------------------------------------------------------------+

I cannot guess the origin of this problem as no log is created in controller, cinder or nova nodes. Please suggest a way so that I can create a valid backup of my instance.

edit retag flag offensive close merge delete

Comments

Does openstack volume service list display the backup service?

Is there anything in the backup log file (probably /var/log/cinder/backup.log, but might be configured to be at a different location)?

Bernd Bausch gravatar imageBernd Bausch ( 2017-07-14 02:41:51 -0600 )edit

The cinder-backup service is enabled on each Cinder node. Nothing related to this event is recorded in /var/log/cinder/cinder-backup.log

CDOT gravatar imageCDOT ( 2017-07-14 03:14:59 -0600 )edit

Then my guess is either a glitch or cinder-api (or cinder-volume? I don't know too well how backups are orchestrated) is unable to reach cinder-backup. In my opinion, the latter is more likely.

Find the message ID in the Cinder log files and look at the context of the error message.

Bernd Bausch gravatar imageBernd Bausch ( 2017-07-16 00:46:45 -0600 )edit

No. there is nothing in cinder-api logs and cinder-volume logs files regarding this error.

CDOT gravatar imageCDOT ( 2017-07-17 23:29:53 -0600 )edit

Brute force then. On all your nodes, run

# grep -r 3eb06b612efc4e8399a689 /var/log

where 3eb06b612efc4e8399a689 is the message ID.

Bernd Bausch gravatar imageBernd Bausch ( 2017-07-18 00:34:28 -0600 )edit