Cinder Volume : LVM
Cinder-Backup : CEPH
즉. Volmue 이 CEPH 백엔드에 저장되어있지 않고 LVM으로 운영하고 backup만 CEPH로 운영
볼륨이 사용중일 경우는 --force 옵션을 사용해서 Volume-detach 하지 않고 백업진행
# cinder backup-create --name <NAME> <Volume_ID> --incremental --force
# ceph -s
cluster 75e29f5f-7219-466b-bebd-2c989981d6ef
health HEALTH_OK
monmap e2: 3 mons at {open-mon0=172.16.11.200:6789/0,open-mon1=172.16.11.201:6789/0,open-mon2=172.16.11.202:6789/0}
election epoch 16, quorum 0,1,2 open-mon0,open-mon1,open-mon2
mdsmap e5: 1/1/1 up {0=open-mds=up:active}
osdmap e54: 3 osds: 3 up, 3 in
flags sortbitwise
pgmap v6004: 512 pgs, 5 pools, 28908 MB data, 7049 objects
57840 MB used, 50228 GB / 50285 GB avail
512 active+clean
client io 628 B/s rd, 41905 kB/s wr, 21 op/s
===>>> Volume 데이타가 있는 여부 상관없이 전체 백업 (full backup)
20G 약 8분정도 소요
4개 백업하고나서의 사용량...
증분백업되지 않고 계속해서 풀백업으로 진행
pgmap v6542: 512 pgs, 5 pools, 83604 MB data, 20730 objects
163 GB used, 50122 GB / 50285 GB avail
512 active+clean
# cinder backup-show <Volume_ID>
+-----------------------+--------------------------------------+
| Property | Value |
+-----------------------+--------------------------------------+
| availability_zone | nova |
| container | backups |
| created_at | 2016-03-30T02:43:48.000000 |
| description | None |
| fail_reason | None |
| has_dependent_backups | False |
| id | 22de2456-d134-40fe-9ba1-e1d762adb29a |
| is_incremental | True |
| name | <NAME> |
| object_count | 0 |
| size | 20 |
| status | available |
| volume_id | 1fc4a829-759d-429e-8d90-3238cf0e7e1d |
+-----------------------+--------------------------------------+
# cinder backup-list
+--------------------------------------+--------------------------------------+-----------+----------------------+------+--------------+-----------+
| ID | Volume ID | Status | Name | Size | Object Count | Container |
+--------------------------------------+--------------------------------------+-----------+----------------------+------+--------------+-----------+
| 12e205fd-802d-4f30-a610-0642d3b503fc | 1fc4a829-759d-429e-8d90-3238cf0e7e1d | available | 2016_0330_1122_incre | 20 | 0 | backups |
| 22de2456-d134-40fe-9ba1-e1d762adb29a | 1fc4a829-759d-429e-8d90-3238cf0e7e1d | available | 2016_0330_1143_incre | 20 | 0 | backups |
| 2ef9a9ad-f3ff-4ddb-a39c-4f2f63e1d002 | 1fc4a829-759d-429e-8d90-3238cf0e7e1d | available | 2016_0330_1132_incre | 20 | 0 | backups |
| 5a8d828f-3bb6-46f2-9d5f-fb4b14046944 | 1fc4a829-759d-429e-8d90-3238cf0e7e1d | available | 2016_0330_1111_incre | 20 | 0 | backups |
# rbd list backups
volume-1fc4a829-759d-429e-8d90-3238cf0e7e1d.backup.12e205fd-802d-4f30-a610-0642d3b503fc
volume-1fc4a829-759d-429e-8d90-3238cf0e7e1d.backup.22de2456-d134-40fe-9ba1-e1d762adb29a
volume-1fc4a829-759d-429e-8d90-3238cf0e7e1d.backup.2ef9a9ad-f3ff-4ddb-a39c-4f2f63e1d002
volume-1fc4a829-759d-429e-8d90-3238cf0e7e1d.backup.5a8d828f-3bb6-46f2-9d5f-fb4b14046944
# rbd info backups/volume-1fc4a829-759d-429e-8d90-3238cf0e7e1d.backup.22de2456-d134-40fe-9ba1-e1d762adb29a
rbd image 'volume-1fc4a829-759d-429e-8d90-3238cf0e7e1d.backup.22de2456-d134-40fe-9ba1-e1d762adb29a':
size 20480 MB in 5120 objects
order 22 (4096 kB objects)
block_name_prefix: rbd_data.9d067792204e
format: 2
features: layering, striping
flags:
stripe unit: 4096 kB
stripe count: 1
※ Volume back-end 가 동일한 ceph 저장소가 아닌 LVM 형태로 운영되어있다면...--incremental 옵션하고는 무관하게 백업은 항상 전체백업으로 진행
단. 백업된 Volume 을 삭제할경우에는 incremental 적용된것 처럼 나오기에 백업된 순서 역순으로 삭제해야 된다.
순서상관없이 삭제하게 될경우 아래와 같은 메세지 발생
# cinder backup-delete 12e205fd-802d-4f30-a610-0642d3b503fc
ERROR: Invalid backup: Incremental backups exist for this backup. (HTTP 400) (Request-ID: req-f2e2ad62-ae59-4256-906c-c25eedb0360f)
