Hyper-V Cluster with CSV Storage used as Backup Repository?

[jaredplaisted]

Using on host backup mode, would it be possible to use a repository located on a CSV within the same Hyper- cluster?
I was thinking of defining the repository using the Cluster Name/IP address instead of an individual Hyper-V host.
I understand the network stack would still be used, but at least would be faster for VM's that

In our environment it would give these benefits:
-Backup repository would be highly available in the case of master node failing, a new Hyper-V node would become cluster master.
-Backup traffic over the network would stay on our management network (which is separate to vm traffic).

Would this be possible, any potential issues?
Would there be any conflicts due to the defined Backup Repository server being defined twice (repository and hyper-v node).

In any case it would be a great feature to have a CSV aware Backup Repository, so that on host backups could always target local CSV for SAN speed backups, kind of opposite to the Offhost Proxy scenario.

[Vitaliy S.]

Hi Jared,

In order to point your backup jobs to repository, you should it add it to the backup console first. Since you can only add Windows hosts and not Failover Clusters, then you will have to choose what Hyper-V node to add. Once you do this, you will be able to point your repository to any folder Windows Explorer can see on the host.

Given all this, I can say it should be possible to use CSV as a backup repository, however if you lose connection to the Windows host added to the backup console, Veeam backup server will not push traffic through the second Hyper-V node to access a CSV node.

If you want to have HA connection to the repository, I would suggest using HA SMB share (hosted on a CSV volume) as a target for your backups.

Hope this helps!

[jaredplaisted]

Hi there thanks for the feedback.

Yes i understand that Windows hosts can be added, since the Failover Cluster IP address is always assigned to the master node you would in fact be addressing a Windows host using the cluster IP.
The cluster IP is a secondary address and can move between nodes during failures etc.
I've tested and it appears that the cluster IP works when adding a storage repository.

Not much info on using CSV in this mannor, probably best to steer clear.

[Vitaliy S.]

Even though it worked now, I still doubt that you will have redundancy for your repository. In other words, we install our repository services on the Windows host, and if it dies, then services will not automatically be deployed to the secondary host, unless these services are configured to work in HA mode via Failover Cluster settings.

[Cragdoo]

Even though it worked now, I still doubt that you will have redundancy for your repository. In other words, we install our repository services on the Windows host, and if it dies, then services will not automatically be deployed to the secondary host, unless these services are configured to work in HA mode via Failover Cluster settings.

Has the HA mode ever been successfully applied and tested ?

Would you happen to have any information regarding this ?

[Gostev]

Craig, check out this backup repository HA solution (see in the middle of #2 section). Thanks!

[Cragdoo]

Thanks Anton

I have managed to successfully create a HA Role for the Veeam Backup Proxy Service on a 2 node Windows Failover Cluster , backed by a Shared Volume running on Windows Storage Spaces, and have tested manual/automatic failover. I have tested failing over the HA Role to the stand by node while a job was running, and the job successfully completed on the re-try. No need for SMB/CIF shares

[d.lansinklesscher]

Is it possible to create a backup repository directly on the CSV on a hyper-v host? Or is this not supported/recommended for some reason?

[foggy]

Dennis, this is possible/supported, however if you're after an HA repository solution, then see above for the recommended approach.