Cloud copy retention

[collinp]

I want to build a scale out repository with local and azure storage using cloud copy. I would like to store most recent 30 days locally and most recent 90 days in azure. How do I configure the 2 retentions?

[sfirmes]

This is an example where you can use a combination of the Move Mode and Copy Mode when sending files to Object Storage. You would set your retention policy in the backup job to be 90 days. Use the Copy Mode to send the backup files immediately to Object Storage, then you can set the Move Mode to prune the files off of the local storage after 30 days. This will keep the latest 30 days locally and retain all the data in object storage for 90 days.

[dalbertson]

What Steve said is one way. Another possibility would be to set your Scale-out repo to enable move mode at 30 days. Then when you create a backup job you would set your retention to 90 days. This would accomplish 30 days on prem and then they would be tiered to object until the 90 day retention was hit. You would need to make sure to schedule weekly or monthly fulls to make sure that the data would be tiered. You would also need to think about a secondary copy.

The strategy that Steve talked about above where you enable copy mode would create a copy of a backup in cloud storage as soon as the backup job is finished. This would help meet the 3-2-1 rule

[collinp]

Are these scaleout and repository settings correct for the approach that Steve suggested? https://photos.app.goo.gl/V7uD5vuSrAf8vQKH9

Also, is there any issue with using Azure "cool" blob storage?

[dalbertson]

Azure cool is supported

https://helpcenter.veeam.com/docs/vbo36 ... tml?ver=40

Those settings are good for your SOBR

Once you set your job settings post those.

[sfirmes]

@collinp The settings look good.

[collinp]

My job setting are picture 1 in the link showing 90 days retention and the job is pointing to the sobr. Any other job setting I need to configure?

[dalbertson]

Sorry the guest processing page and the schedule.

[collinp]

Ok. I updated the pictures in the link

[sfirmes]

Thanks for the updated pictures. The current settings appear to be fine.

[dalbertson]

Thanks @collinp

What i was looking at was your retention schedule and your backup mode.

Reason for the retention schedule was that if you wanted 90 retore points you would need to have your job running daily. Which you do, so thats great.

The second thing was checking that you have either fulls or synthetic fulls going so that it will prune that data from on prem when your 30 days are met.

as steve said with these settings the way you have them, you are good to achieve what you wanted to achieve.

[dariusz.tyka]

Hi,

how SOBR/Job settings should be configured if I want to keep 1 month of data onprem and keep weekly backups for 2 months and monthly backups for 1 year? We want to switch from tapes to S3 completely and also use S3 immutability. For now we are using reverse forver incremental backups in our jobs to limit disk space. I suppose with current backup scheme it will not be possible. Should I go for BCJs or use the GFS "Keep certain full backups longer ...." option within the backup job?

[Gostev]

Hi,

The latter is the recommended approach, as it saves you from managing an extra Backup Copy job.

If you use ReFS or XFS for your backup repository, then periodic synthetic full backups will not require additional on-prem disk space compared to reversed incremental backups.

Thanks!

[dariusz.tyka]

Hi Gostev,

thanks for your comment. We are still using Windows Server 2012R2 on our backup server. Don't know if it is advised to use REFS on this OS? For now backup repository is formatted with NTFS.

[Gostev]

ReFS integration requires 2016 or later.

[dariusz.tyka]

OK. Would then be an option for us to reconfigure our backup jobs as follow:

1. Retention policy for backup job - XX days
2. Keep certain full backups longer....
4 weekly, 12 monthly, 10 yearly
3. Enable on SOBR - to copy new backups to object storage as soon as they are created
4. Enable on SOBR - move backup to object storage older than 14 days
5. Schedule full backup once a week.
What should be the job retention configured for backup job to have 2 full backup chains available locally and also all older backups offloaded to S3 to save space? Sth more than 14 days?

What do you think?

[Gostev]

That works, except this will require disk space for up to 3 full backups locally. This is just how incremental backup works: the oldest incremental backup chain simply cannot be removed in its entirety until the restore point in its last increment is still under the retention policy, as that last point is obviously dependent on the entire chain.

So if you want to guarantee there's no more than 2 full backup chains locally at any given time, you will want your Move policy at 7 days. In this case, the number of your local restore points will fluctuate between 7 and 13, depending on the day of the week.

[dariusz.tyka]

Thanks Gostev.