Recover from repository lost

[flow90]

Hello,

I have two Windows Server ReFS repository dedicated for NAS backup and I'm using the v12 VBR version.

The NAS backup job is setup with this settings :
- backup the file shares once a day
- keep files versions for 365 days
- make a secondary copy to the other ReFS repository (offsite) with same retention as the first backup.

NAS backup are doing forever incrementals by default... and I would avoid to run a new full under no circumstances.

In the following scenarios, how can we recover from loosing one repository, for exemple, after 200 days of successfull daily backups ?

- If we completely loose our first repository (fire for exemple) :
Once we have a new repository deployed, we have to use "Copy Backup" feature from the second repo to the new one, and remap the NAS file share job source to the new repository. Is it exact ?

- From my understanding, things get more complicated if we loose the second repository :
Once we have a new repository deployed, we have two option :

1) Use Copy Backup to get back all the past 200 daily backups and because it's not possible to remap a "File share backup copy" we will have to select the new repository from the secondary target job setting but he will run a new full with only the last version of files (tested...).

2) Only change the secondary target into the job but as I say above, we will only have the last file version in the file share backup copy.
If it's exact, my question is : How can we avoid to run a new full and consume the storage twice ? (one for the 199 past version) and a new full mapped with the job with only the last version.

We really want to keep two copies with the full retention of the backups...
PS : With vSphere VM Backup, the "backup copy" copies all the previous versions... but that doesn't seem to be the case with NAS backup jobs...

Thank's for your informations !

[Mildur]

Hello Flow

- If we completely loose our first repository (fire for exemple) :
Once we have a new repository deployed, we have to use "Copy Backup" feature from the second repo to the new one, and remap the NAS file share job source to the new repository. Is it exact ?

Unfortunately you cannot map a backup job to backup files created by the "secondary location"-copy job. A Backup Job can only be mapped to backup files created by a backup job.

If it's exact, my question is : How can we avoid to run a new full and consume the storage twice ? (one for the 199 past version) and a new full mapped with the job with only the last version.

This is the only solution for now. I don't see any other trick you could try.

PS : With vSphere VM Backup, the "backup copy" copies all the previous versions... but that doesn't seem to be the case with NAS backup jobs...

Immediate backup copy jobs for VMs only copies the latest backup in the initial copy session. In subsequent sessions, all new restore points will be copied. It won't copy "everything" from the source location. But there is a registry key (BackupCopyMirrorAll) which allows you to copy all in the first run of a new backup copy jobs.

Best,
Fabian

[flow90]

Thank you for your answer.

I notice that I need to manage to have some free space to run a full again in case of disaster on the primary repository.

Is it possible to use "BackupCopyMirrorAll" registry key for a NAS job ?
Best regards

[flow90]

Hi,

I have tried the BackupCopyMirrorAll with a NAS job, Not working... I only get the last restore point.
Is this expected to be supported ?

Best regards