Backing up Azure VMs

[wormdoc]

Hi There,I have a cloud only customer everything in Azure, .

They will have a VBR server in azure with managed disks sounds to backup to

its going to be expensive as they have 15TB of data they want backed up…

They want to keep

daily backups retained for 2 weeks (14 restore points)
weekly backups retained for 5 weeks
monthly backups retained for 6 months

I was going to have two jobs

One backup job for the daily’s and retain locally on the azure managed disk (30tb disks)
Then a backup copy job to create copy the weekly and monthlys.
The backup copy job was going to copy to a cloud connect partner.

They’ve advised this will be expensive, which is true so have asked for alternative methods.. like sending to azure blob..


I have a few questions

If we use a backup copy to cloud connector provider do they support REFS on the remote repository?

If we use our standard Backup repository + azure cool blob (SOBR) would this be cheaper?
I keep hearing about the new archive tear allow native support for blobs.

Is my understanding correct that We create a SOBR which consists of our standard repository + the blob storage
The standard repository is the perf tier and the blob storage is the archive tier? is this understanding correct?
Would we still use back-up copy jobs for this? Its quite confusing

[HannesK]

Hello,

If we use a backup copy to cloud connector provider do they support REFS on the remote repository?

yes, the cloud connect provider can use REFS on its side and you can save disk space with REFS

SOBR which consists of our standard repository + the blob storage

SOBR can have one or multiple standard repositories and also include blob storage.

The standard repository is the perf tier and the blob storage is the archive tier?

correct

I keep hearing about the new archive tear allow native support for blobs.

it's to early to comment on details, but yes, we are working on significant improvements


Which option is cheaper for you is hard to say, but for your design, just keep the following in mind: today we always move old data to blob (tiering). The data moved to blob must be an "inactive" (no changes anymore) backup chain.

Would we still use back-up copy jobs for this?

in most cases yes. theoretical you could create a 180 days (6 months) reverse incremental chain, but if you loose the full backup, then everything is lost. So this is not recommended.

Best regards,
Hannes

[jmmarton]

The red flag I see is that if the source to be backed up is Azure VMs and the destination is VBR including repo in Azure, then to do BCJ to a Veeam Cloud Connect provider that's a whole bunch of potential egress charges.

Another option would be to create two repos in Azure. The primary repo is for the 14 RPs. Then use BCJ to another repo in Azure configured as SOBR for weekly/monthly retention. Then enable Capacity Tier to move those to Azure Blob, and that could be configured to move after 0 days which means within hours of archive retention point creation move it to Azure Blob. That would prevent egress charges.

Joe