Skip to main content
  • EN

Hello,

Regarding Forever Forward backups, is it necessary to create period full backups (either Active or Synthentic), or can you just go with the initial full backup and everything else is incremental?

 

(I know it is probably at least RECOMMENDED to do periodic full backups, but let’s not worry about that for purposes of this post/question)

 

If you were to go with just the initial full backup then leave the rest as incrementals, what happens if you have a corrupt/failed incremental backup in the chain, but need to do a restore of a file?  Is Veeam able (and smart enough) to work around it? (i.e., if it has a failed or corrupt incremental backup, is Veeam smart enough to start the next backup based on the last good incremental and basically ignore the failed/corrupt one?)

 

Thanks! :-)

This would be dependant on the number of restore points selected for the job. If small it would be fine (30 or less) but if a bigger number more risk.

If you are using ReFS or XFS you can do synthetic as it will be quicker and ensure a safe backup chain.

Regardless you should consider one or the other.

Yes, you can do FFwd without any Fulls. I do this in my environment. But, I don’t recommend having chains longer than say 1 ½ - 2mos. To have Backups for longer periods, configure GFS settings. See here from the Guide on how to do so.

If you do this as I do, to help mitigate against corruption, Veeam provides health checks you can configure within your Backup Job > Storage area > Advanced button > Maintenance tab as shown here.

And yes..you are correct, if you don’t have all backups in a Retention Period chain, you won’t be able to do restores, as noted in the Guide here. To help mitigate this possibility, along with the Maintenance setting you can configure as I mentioned above, it is recommended to do occasional Fulls, beit Active or Synthentic; which, as Chris noted above, if you use ReFS (Windows) or XFS (Linux) filesystems, the time it takes to do those is negligible.

Hope we clarified things for you.

Thank your for your responses.  They have helped a lot.

Sure...glad to help. 

Thank your for your responses.  They have helped a lot.

Always glad to help. Be sure to mark an answer for your question to help others.

1 piece of advice that isn’t often acknowledged with FF incremental without active fulls. 

I use this in my environment and have for years and it works great. 

If you are using REFS even if you are doing synthetic fulls, it will fragment over time. If you have slow storage, 6 months down the road you may notice your restores, testing and other things not quit as quick as they were to start. 

You can fix that by either using SSD/Flash storage as your primary, or running the defrag/compact on your jobs. 

Running SurebBackup every so often is also a good way to verify your backups are good. Even if you do an active full I suggest backup testing to make sure you have valid data.

 

 

 

Hi @SilkBC -

I’m just following back up on your FFwd post. Did any of the provided comments help answer your question? If so, we ask you select which comment helped you as ‘Best Answer’ so others with a similar question who come across your post may benefit.

If you do still have further questions, don’t hesitate to ask.

Thank you!

Comment


Terms & Conditions