Use NetApp snapshots as source for separate jobs

[dbr]

Hi guys,

I'm currently thinking about the possibility to back up vms from NetApp snapshots that were created by a backup job with NetApp snapshot as destination repository. Guess we trigger application consistent NetApp snapshots via an ordinary backup job every hour and want to backup (or copy) vms from the latest application consistent snapshot to a separate Veeam repository once a day without (vmware-)snapshot the vm at all.

Long story short: Is it possible to use application consistent snapshots created by snapshot-only jobs as source for backup or backup copy jobs?

Thanks for any replies.

[DGrinev]

Hi Daniel,

You can't use snapshot created by the Snapshot-only job as a source for the backup or copy job.
You can stay with the ordinary backup job from the snapshot by Netapp and set up secondary destination job.
Please keep in mind that VMware snapshot will be taken anyway, even in a Snapshot-only job but for a very short duration. Thanks!

[dbr]

Thanks for clarification. I've just checked the creation of a VMware Snapshot and want to add that VMware Snapshots are created only if application processing is enable for a vm.

[DGrinev]

Thanks for follow up.
You're right if VMware quiescence or Veeam Application-Aware processing checkbox is enabled, then the VMware snapshot will be created otherwise isn't.

[dbr]

I have one more question regarding this matter. Assuming I have a application consistent NetApp snapshot (all containing vms with application aware processing). Does this snapshot contain my vms including a vmware snapshot? If yes and in case of disaster, would it be possible to revert to the NetApp snapshot or would I additionally have to revert to the vm's snapshot afterwards in order to get the application consistent state? Or is a desaster recovery not the intended use of NetApp snapshots created by Veeam?

[larry]

For DR I use Veeam created netapp snapshots. You can recover any point in time you have a snapshot for. The RT is just as long as it takes to boot the VM, I stopped using Veeam replicas off site as NetApp snap shots created by Veeam take up way less space with dedupe on NetApp. Think of each NetApp snap shot created by Veeam as a App consistent replica all ways ready to boot, and will boot at SAN time. I have Veeam create Netapp snapshots every 30 minutes for two days, daily snap shots for 3 days and weekly's for 2, Veeam controls of the retention policies. I would guess 95 percent of my FLR is from snapshots this way. I let NetApp replicate to DR high priority data but have Veeam tell NetApp to replicate at night low priority snap shots.

Sorry for long post just the Veeam/ NetApp remote snapshots combo is the best DR tool I have ever used.

[dbr]

Really nice. I will consider this design when we plan our Exchange DR. But back to my original question. Just booting from a reverted NetApp snapshot will use the crash consistent state unless I use instant recovery or revert to vmware snapshot that should be within the NetApp snap as well. Is this correct? Sure, it's possible just to boot the vm and check if it boots correctly but knowingly booting an application consistent state makes me feel better. Can anyone confirm or deny that?

[foggy]

Not sure what VMware snapshot are you referring to. Veeam B&R doesn't backup existing VM snapshots. Temporary VM snapshot created during backup is removed right after creating storage snapshot. Storage snapshot will contain application-consistent state of the VM.

[dbr]

I'm referring to the VMware snapshot which is created when running a veeam job with with application aware processing activated and NetApp snapshot as target. Maybe I just think too complicated. What I try to explain is that if Veeam creates the necessary VMware snapshot, then create a NetApp snapshot while the VMware snapshot exists and delete the vm snapshot afterwards you would end in a NetApp snapshot with a VMware snapshot in it. Or more precisely, the NetApp snapshot will contain a vmdk file (read-only) plus a vmdk delta file to which all disk writes are redirected. So my question is: If I revert to the NetApp snapshot with built-in commands and then boot the vm (which consists of one vmdk and one vmdk-delta file) it will boot from crash consistent state with delta file, won't it? This is not case if I use Veeam to recover from restore point out of NetApp snapshot because Veeam uses the vmdk, more specifically its configuration (like it would do with instant recovery directly out of the backup file), independently of existing delta file, right? I know that I will definitely get the application consistent state with instant recovery, full vm restore, file level recovery etc... Perhaps this is more a VMware- / NetApp-related and less a Veeam-related question... but maybe somebody else can jump in.

[foggy]

Got it now, yes, your understanding is correct. Unless you revert the VM snapshot after reverting the NetApp snapshot manually, the VM will boot to the state it had at the moment of storage snapshot creation (crash-consistent).