Storage Snapshot and Backup Job design

[ralphmurray]

Was wondering what most people are doing when using integrated storage snapshots to backup VM's when you have multiple Volumes.
Do you create multiple backup jobs, one for each volume (datastore) and then have them backup one after the other ?
How does Veeam handle this if you were to backup all VM's in one job that are on multiple volumes, would you end up with multiple snapshots presented to the Proxy all at once ?

The site i'm currently looking at has 14 volumes with around 60 VM's, running on NFS Netapp.
Currently I have created 1 job for each volume (Datastore) as I don't think I want 14 Volumes mounted to the proxy all at once.
Problem is when you move a vm from one datastore to another you have to do a full backup again etc..

or does Veeam handle this by only doing one volume at a time ?

Any recommendations/help around this would be appreciated.

Cheers

Ralph...

[Vitaliy S.]

Hi Ralph,

How does Veeam handle this if you were to backup all VM's in one job that are on multiple volumes, would you end up with multiple snapshots presented to the Proxy all at once ?

You will have storage snapshot created for every set of VMs residing in the same volume. The number of concurrently processed VMs will depend on the max allowed tasks in the proxy server configuration.

The site i'm currently looking at has 14 volumes with around 60 VM's, running on NFS Netapp.
Currently I have created 1 job for each volume (Datastore) as I don't think I want 14 Volumes mounted to the proxy all at once.
Problem is when you move a vm from one datastore to another you have to do a full backup again etc..

Moving VM from one datastore to another should not cause a full VM backup. What have you specified as source for the backup job?

or does Veeam handle this by only doing one volume at a time ?

Any recommendations/help around this would be appreciated.

Since you're using NFS storage Veeam backup server will create 14 volumes on your storage and will mount them based on the VMs processing list. Once your job is complete, all storage snapshots will be removed.

Thanks!

[ralphmurray]

Thnkyou for the quick reply.

>>>Since you're using NFS storage Veeam backup server will create 14 volumes on your storage and will mount them based on the VMs processing list. Once your job is complete, all storage snapshots will be removed.

Would Veeam present all 14 to the proxy at the same time (start of the job) or just as it needs them.

>>>Moving VM from one datastore to another should not cause a full VM backup. What have you specified as source for the backup job?

In the current test setup we have a seperate backup job for each nfs volume so we can control how many nfs volumes are presented to the proxy at once, moving a VM from one datastore to another means I would need to move it from one backup job to another. the source is the VM's from Vcenter.

Doing it this way also allows us to have different retention settings for VM's by putting less important servers on the same datastore etc..


Ralph...

[Vitaliy S.]

Ralph,

Would Veeam present all 14 to the proxy at the same time (start of the job) or just as it needs them.

The second. The number of storage snapshots mounted to the proxy server will depend on the number of concurrently processed VMs (and the number of VMs located on each LUN). In other words if you proxy server allows backing up VMs from two LUNs (or more), these two LUNs will be mounted to the proxy.

In the current test setup we have a seperate backup job for each nfs volume so we can control how many nfs volumes are presented to the proxy at once, moving a VM from one datastore to another means I would need to move it from one backup job to another. the source is the VM's from Vcenter.

Doing it this way also allows us to have different retention settings for VM's by putting less important servers on the same datastore etc..

Ok, got it. You can combine VMs from different LUNs to the same job, and then control how many LUNs will be mounted to the proxy via maximum allowed tasks counter. On top of that, if you don't use storage snapshots, and still move VMs between the LUNs, your job will still process this VM and create and incremental restore point, but via regular backup mode (SAN, hotadd or NBD).

Thanks!