cinder backup api can't make a back-up with attached volume

[mylq]

I try to use cinder backup api to backup a volume which is attached at an instance,but failed.It is ok for my environment to create a back-up with an available volume,which means that it isn't in use.My OS is ubuntu 14.04,and I installed openstack with swift component by devstack.The version of them is mitaka.
Here is the api link I used:
http://developer.openstack.org/api-ref/block-storage/v3/?expanded=list-all-back-end-storage-pools-detail,show-all-back-end-capabilities-detail,create-a-backup-detail
There is one snapshot below which could explain the bug I meet.
Here is the command I use to create a back-up of in-use volume:

stack@lq:~$ curl -s -H "X-Auth-Token: $OS_TOKEN" -H "Content-Type: application/json" -X POST http://127.0.0.1:8776/v3/$OS_TENANT_ID/backups -d '{ "backup":{"container":"vol-9-26","description":"backuptest-description3","name":"lq9-27","force":true,"volume_id":"c2ed4817-738a-47ea-92b9-3ec4da6873f0", "incremental": false} }' | python -m json.tool
{
    "backup": {
        "id": "1f3f8b03-6130-47a5-906b-39df15d5fa98",
        "links": [
            {
                "href": "http://127.0.0.1:8776/v3/539b278df88e4b70adac5479434ab01c/backups/1f3f8b03-6130-47a5-906b-39df15d5fa98",
                "rel": "self"
            },
            {
                "href": "http://127.0.0.1:8776/539b278df88e4b70adac5479434ab01c/backups/1f3f8b03-6130-47a5-906b-39df15d5fa98",
                "rel": "bookmark"
            }
        ],
        "name": "lq9-27"
    }
}

stack@lq:~$ curl -s -H "X-Auth-Token: $OS_TOKEN" http://127.0.0.1:8776/v3/$OS_TENANT_ID/backups/1f3f8b03-6130-47a5-906b-39df15d5fa98 | python -m json.tool
{
    "backup": {
        "availability_zone": "nova",
        "container": "vol-9-26",
        "created_at": "2016-07-13T11:58:00.000000",
        "data_timestamp": "2016-07-13T11:58:00.000000",
        "description": "backuptest-description3",
        "fail_reason": "Timed out waiting for a reply to message ID 11b441e1d4834c7ea9fe47a648c429ed",
        "has_dependent_backups": false,
        "id": "1f3f8b03-6130-47a5-906b-39df15d5fa98",
        "is_incremental": false,
        "links": [
            {
                "href": "http://127.0.0.1:8776/v3/539b278df88e4b70adac5479434ab01c/backups/1f3f8b03-6130-47a5-906b-39df15d5fa98",
                "rel": "self"
            },
            {
                "href": "http://127.0.0.1:8776/539b278df88e4b70adac5479434ab01c/backups/1f3f8b03-6130-47a5-906b-39df15d5fa98",
                "rel": "bookmark"
            }
        ],
        "name": "lq9-27",
        "object_count": 0,
        "size": 1,
        "snapshot_id": null,
        "status": "error",
        "updated_at": "2016-07-13T11:59:24.000000",
        "volume_id": "c2ed4817-738a-47ea-92b9-3ec4da6873f0"
    }
}

[Sean McGinnis (sean-mcginnis)]

What release of Cinder are you using?

[mylq (leibnizlq)]

the version of cinder I use is mitaka

[Ivan Kolodyazhny (e0ne)]

In the latest version of Cinder API we have --force flag for it

[Janice Chen (chenshujuan)]

I have met the same problem in mitaka version recently, even if use --force flag,it does not work for 'in-use' volume.it seems stucks while _transfer_data.But it works fine with the 'available' volume with the same configuration.

[Sean McGinnis (sean-mcginnis)]

Almost a year old, so this may be asking for a lot. Please include:

* Volume backend driver being used
* Backup driver being used
* Cinder logs capturing the operation failure