Best Practice for Backing up Backup Respositry

[dankingdon]

Dear all,

Looking for some guidance on the best route to go down with our backup infrastructure. We run vmware for everything, around 40 vm's and have two physical backup servers in alternative locations.

We are using standard backup jobs to store incremental backups for up to 999 recovery points. This gives us the necessary long term data retention. The problem is how do we make sure those long term backups are safe? If i create a copy job to the second backup server, the copy job data is different form the main backup job, so if anything happens to our main repository I will not be able to restore it from the second server. If the second server fails then i will not be able to restore it from the primary server?

I have read a lot of discussions about people attempting to use robo copy and other utilities to maintain a copy of the veeam repository in two separate locations but this seems woefully backwards in todays day and age.

Is there no way of telling veeam to make two copies of a repository, saving it in two distinct location? What is the usual recommendation in this situation?

Our over all back up aims are to protect against disaster recovery and to provide a much long term file data retention as possible whilst ensuring that long term data is also safe.

[foggy]

Daniel, the recommended approach is to use SureBackup to verify that your primary backups are recoverable and then backup copy jobs to copy data to the secondary location. Backup copy job copies the same VM state that is stored in the latest backup job restore point and has built-in integrity check to ensure data is bit-identical, so you always can restore the required VM state from another location.

[dankingdon]

Hi Foggy,

Thanks for the reply, that was my initial gut instinct. However, what happens if the primary backup repositry fails? I have no way of restoring the repositry. I can still restore from the backup copy but the original primary backups will have to start again from scratch.

I'm currently leaning towards something like synctool to maintain a proper replica of the primary backup repositry.

[foggy]

You can use secondary backups to seed the primary one. What is the exact use case you're trying to address? Am I understanding right that the ability to restore VMs from secondary location (in case primary is down when such restore is needed) is not enough for you and you just want to have an exact copy of the primary repository on site?

[dankingdon]

Yeah pretty much. The main goal is as follows.

Repositry A contains all backups going back say 300 days.
Repositry B contains a copy job of the same data.
Repositry A and B's files are different, despite containing identical vm's and retention.
if A or B is lost there is no way of rebuilding it from the other?
As a result if we lose both within our retention period we lose temporarily our historical restore points?

[foggy]

For that matter some third copy (tape?) of your backups is required.