Scale out or not

[simcfc73]

Hi,

I'm a bit confused about the way to go regarding using the cloud as a secondary backup.

I've tried using rotating disks but it didn't work, I've also tried different backup jobs per disk but it's confusing so thought I would try scale-out.

I have an existing backup for my servers backing up to internal hard drives but want to utilize the cloud.

Simply, hold a weekly backup in the cloud for 6 months. 4 weekly, 6 monthly ones.

Is scale-out the way to go? I've set up my cloud repository using Backblaze as documented.

[HannesK]

Hello,
to use capacity tier (S3, Azure Blob...) you need to use scale out repositories. It's a functionality that only exists there. https://helpcenter.veeam.com/docs/backu ... ml?ver=100

"rotated hard drives" are not supported for scale out repositories.

I assume, that you like to keep a copy of your backups in Backblaze: https://helpcenter.veeam.com/docs/backu ... ml?ver=100 (check the "copy..." checkbox)

I also recommend reading the FAQ: post338749.html#p338749

Best regards,
Hannes

[simcfc73]

Ok thanks..

I'm using multiple disks as rotating disks didn't work for me.

When I try and create a scale out I get to the prompt. This scale out backup repository already contains data, should we copy it.. when I click All or Latest I get an error saying failed to save and then a path d:\backups\backup disk 2a which doesn't exist.

[simcfc73]

I had to disable the disk backups to get the scale out to configure.. now I have 99 backups trying to upload to Backblaze.

I really just want weekly backups up there.

[veremin]

If I get your requirements correctly, you need to:

* have a dedicated Scale-Out Backup Repository with object storage repository attached as a Capacity Tier
* have a Capacity Tier with either move (single copy of GFS backups stored in cloud) or copy (two copies of GFS backups stored locally and in cloud) policy enabled
* have a backup copy job with GFS retention enabled (weekly, monthly)
* point the backup copy job to the Scale-Out Backup Repository

Otherwise, all restore points from all jobs pointed to this Scale-Out Backup Repository will be a subject of Capacity Tier policies.

Thanks!

[simcfc73]

I thought if I left the 'copy as soon as they are created' unticked the move function selected for 14 days it would just move backups over 14 days old into the cloud. ( I have GFS setup on my backup job)

I'm not sure if I can use it anyway, the upload speed is abysmal and my 1.8Tb file will take over a year to upload.

1 year 10 months 22 days 11 hours 24 minutes 6 seconds

[veremin]

Correct, with the move policy configured backup files that belong to inactive backup chain and fall out of the specified operation restore window (14 days) will be moved to the object storage. All backup files from all backup jobs pointed to the Scale-Out Backup Repository, not only GFS backup files of particular job - this must explain why in your case the number of restore points offloaded was 99.

Anyway, with the currently available bandwidth (I think this is your major bottleneck) it does not seem that you can make any reasonable use of public object storage.

Thanks!