Veeam & NetApp Integration Questions

[jsc8041]

We currently utilize NetApp Snapmirror for DR to an off-site environment.

In v8, we should be able to do instant VM recovery from the SAN snapshots on the DR side, correct? Is this considered a viable alternative to Veeam Replication jobs?

Also, in our environment with existing utilization of Snapmirror, is there an advantage to the Backup from Snapshots feature in the Enterprise Plus version?

What are the pros and cons of doing Veeam incremental backups v. Veeam backups from SAN snapshots? Is it just performance? When using backups from SAN snapshots, are all the VM's on a single volume typically run in one job, since the SAN is acting at the volume level?

Before NetApp worked with Veeam, we were set on choosing one over the other, but now with these options, I'd love to leverage both, and would appreciate anyone's thoughts or experiences using both.

thanks,
Claude

[Gostev]

In v8, we should be able to do instant VM recovery from the SAN snapshots on the DR side, correct? Is this considered a viable alternative to Veeam Replication jobs?

It can be, but for certain disaster scenarios only (such as a single VM outage due to in-guest issue) because:
1. With SnapMirror, data remains in the same fault domain. SnapMirror will mirror any corruption just like it mirrors "good" data.
2. Instant VM Recovery from SnapMirror does not provide advanced functionality such as Failover Plans, re-IP etc.

That said, Veeam replication is not always a replacement for SnapMirror either. Depending on your RPO, environment size and load, crash-consistent SnapMirror may be the only choice. Veeam is best for replicating fewer VMs over WAN (and can land replicas to any storage), SnapMirror is best for replicating large amounts of VMs over thick pipe (and requires another FAS).

Also, in our environment with existing utilization of Snapmirror, is there an advantage to the Backup from Snapshots feature in the Enterprise Plus version?

Yes, as this will provide true, read-only backups stored outside of that single fault domain. Remember the "2" in the "3-2-1" rule, which stands for two different medias. You cannot possibly consider identical storage devices staying in sync with each other (2 FAS boxes in SnapMirror relationship) as two different medias, as effectively it is the same media - just geographically spanned. Sure, this deployment will protect you from fire or flood in your primary data center - but how often do those events happen comparing to a data corruption? In my 8 years at Veeam, I've seen data protection strategies based solely on storage replication fail big times on multiple occasions.

What are the pros and cons of doing Veeam incremental backups v. Veeam backups from SAN snapshots? Is it just performance?

VM spends much less time running off snapshot comparing to classic VADP backup, which reduces impact on production storage and actual workload they are running. There are no cons.

When using backups from SAN snapshots, are all the VM's on a single volume typically run in one job, since the SAN is acting at the volume level?

Depends solely on how you choose to populated your jobs - there are no restrictions or limitations on our side. You can organize your jobs per datastore, or in any other way.

[jsc8041]

Thanks Anton for your input. I really appreciate it.