Backup best practices for two sites

[filippoAdb]

We have two sites, each with a vmware cluster.

Site A
Veeam B&R server
HPE Storeonce A as backup repository, 2 catalyst stores, one for windows and one for Linux

Site B
Veeam gateway
HPE Storeonce B as backup repository, 2 catalyst stores, one for windows and one for Linux

The backup jobs use the local storeonce of the sites where the vm resides. We want to add a remote copy.
For this purpose we created for test a backup copy job of type "Storage copy", the one for storeonce repositories, from Linux repository on site B to a new dedicated repository on site A.
Documentation warns to avoid loops, so I cannot set Linux on site A as target repository because that way I cannot remote copy my site A Linux VM on site B. Is it correct ?
If I want to make crossed "Storage copy", A -> B and B -> A, I suppose it is necessary to create new catalyst stores dedicated to the remote data in order to avoid loops. Is there another way ?

"Storage copy" configuration is very simple, it is all or nothing, I wonder why I cannot choose a single VM. That way I could add new VMs avoiding an initial overload.

By the way I launched the test backup copy an hour ago and it printed 4 rows like the following:
10/09/2024 15:18:53 :: Processing Storage Copy Task one-vm.vm-34078D2024-08-25T190420_F379.vbk
Every data is at zero, processing rate, data read and transferred, etc, I suspect it will last days and can interfere with normal backup jobs.

Which are the best practices to follow in this scenario ?

[mjr.epicfail]

I don't understand the statement "storage copy".
The best practice (for me) would be to make a backup job to backup your workload, and a backup copy job to make a backup from your backups created in the original backupjob.
You can do this by creating a secondary backup repository int eh opposite site of your primary backup repository.
Create a backupjob with as source your backupjobs, target the secondary repository and you are good to go.

The other option would be using Veeam Cloud Connect if there is no (private) network connectivity possible.

More information about Best Practices can be found here: https://bp.veeam.com/vbr/

[filippoAdb]

I don't understand the statement "storage copy".

In a storeonce system you create some catalyst stores which are the target repositories for one or more backup jobs.
Storage copy is a special case of backup copy job dedicated to storeonce systems, according to my limited understanding it keeps a local and a remote catalyst stores synchronized in a smart way. If storage copy is available for and dedicated to storeonce systems I think it should be the preferred choice over traditional backup copy jobs linked to backup jobs.

[mjr.epicfail]

ahh, so it has nothing the do with Veeam.

Yes you can copy everything that way, but to tell Veeam all the files exist somewhere else can be a pain.
My suggestion would be to leverage Veeam software to create a BackupCopy of the original Backups in site B and copy those site A.
You need a new repository ofcourse and network access. It will be different If a LAN access is not possible. You will need VCC for that.
But my assumption is that there is LAN access between the sites.

BP backupcopy: https://bp.veeam.com/vbr/4_Operations/O ... y_job.html
Helpcenter: https://helpcenter.veeam.com/docs/backu ... ml?ver=120

Hope this helps

[filippoAdb]

ahh, so it has nothing the do with Veeam.

Yes you can copy everything that way, but to tell Veeam all the files exist somewhere else can be a pain.

@mjr.epicfail

The operation is done from Veeam gui console, it is a special case of its backup copy job, and Veeam is perfectly conscious that backup file are available locally and remotely.

To be more specific in my question, in order to avoid loop, I think I'm forced to create new repositories dedicated to the remote backups on both sites. Is it correct ? Is threre another possibility I 'm not aware ?

[mjr.epicfail]

Hi,

Sorry, I cannot answer your question about the storeonce integration, mainly because I have never used it.
tbh I only rely on the backupjob / backupcopyjob because this gives me the posibility to change retention, add immutability, maybe even use a CloudConnect target to get the 3-2-1.