Performing snapshot rollback
The following procedure requires root access (UNIX) or Administrator privilege (Windows).
To perform snapshot rollback (UNIX)
- Start the Backup, Archive, and Restore interface.
/usr/openv/netbackup/bin/jbpSA &
- Click the Restore Files tab.
- Click Actions > Specify NetBackup Machines to specify the server, source client, policy type, and destination client.
- For the Restore Type, select Point in Time Rollback.
The Browse directory field is grayed out, with root (/) as default.
Instant Recovery backups are displayed in the Backup History window, for all dates (you cannot set a range).
- Select an image from the list and click OK.
The image contents are displayed in the Directory Structure pane of the Restore Files tab.
You can select root level or mount points (file systems or volumes), but not folders or files at a lower level.
- In the Directory Structure list, click the check box next to the root node or a mount point beneath root.
You can select a file system or volume, but not lower-level components.
- Click the Restore option.
If the volume that you have selected, belongs to a storage array consistency group or volume set, you can see a list of volumes that are restored along with the selected volume, as they belong to the same array volume set or consistency group. In the open dialog,
To recover all the listed volumes, click OK.
To cancel, and select another restore method for single volume recovery, click Cancel.
For more details refer to NetBackup status code: 4311, in the NetBackup Status Codes Reference Guide
The only available destination option is Restore everything to its original location.
- For file systems, you can choose to skip file verification by placing a check in the Skip verification and force rollback option.
Warning:
Click Skip verification and force rollback only if you are sure that you want to replace all the files in the original location with the snapshot. Rollback deletes all files that were created after the creation-date of the snapshot that you restore.
If Skip verification and force rollback is not selected, NetBackup performs checks on the file system.
See Notes on rollback.
If the checks do not pass, the rollback aborts and the Task Progress tab states that rollback could not be performed because file verification failed.
- For volumes belonging to consistency group or volume set, select Force rollback even if it destroys the consistency group's state on the storage array option, to ignore any mismatch in consistency group configuration on the storage arrays during the restore. NetBackup completes the restore with warning messages.
Note:
Select this option only when it is safe to rollback all the devices on the concerned storage array consistency group.
To understand the concept of forced rollback, refer to the following examples.
The above figure shows a simple application server where the application data (/mnt1 and /mnt2) resides on five storage array devices (D1, D2, D3, D4 and D5), which are a part of storage array consistency group. Consider that your NetBackup policy is configured such that you are using VSO snapshot method and backup selection leads to /mnt1 and /mnt2. Once the snapshot is conducted, it will be conducted at consistency group level. So, now at the time of PIT Rollback Restore if you choose only /mnt1 for such a restore, you will get a warning stating partial selection. This is because of the fact that the snapshot being referred for restore has information about group snapshot consisting of devices D1, D2, D3, D4 and D5 where /mnt1 resides on D1, D2 and D3 only; and reverting back at consistency group will affect all devices in this group.
The above figure shows a simple application server where the application data (/mnt1 and /mnt2) resides on three devices (D1, D2, D3) which are part of a storage array consistency group. The consistency group is comprised of four devices (D1, D2, D3 and D4). Consider that your NetBackup policy is configured such that you are using VSO snapshot method and backup selection leads to /mnt1 and /mnt2. Once the snapshot is conducted, it will be conducted at consistency group level. So, now at the time of PIT Rollback Restore if you select both /mnt1 and /mnt2 for such a restore, the restore will fail due to mis-match of the devices. Because, the group consists of four devices whereas the devices being asked for restore are only three. Reverting back at consistency group will affect all devices in this group.
The rest of the procedure is identical to a normal restore as explained in the NetBackup Backup, Archive, and Restore Getting Started Guide and help.