HPE StoreOnce Catalyst Store and Veeam Cloud Connect

[tzcorreia]

Hello all.
On my company we are implementing a new Veeam cloud connect infrastructure with HPE StoreOnce. I know that the support of catalyst store as cloud connect repositories in on road map but for when?
Case # 07217825.

[Andreas Neufert]

We are working with HPE on implementing longer backup chain compatibility (HPE read from parent will enable us to work with more backup files within a chain in parallel). After release, we will collect feedback from larger customers to balance the load between multiple Backup and Replication servers accessing single StoreOnce Catalyst Store with this feature.

Can you please share a bit your scenario. How many customers would work with the same StoreOnce?
How many would work with the same Repository (same StoreOnce Catalyst store)?
What model are you planning to use?

Background: StoreOnce Deduplication is not global and work only within same Catalyst Store but the Catalyst Store has in total connection limitations.

[tzcorreia]

Hi.
This is our scenario: VCC server on 3 different locations/sites, clients are distributed for this 3 locations/sites and each client should have is own catalyst store ideally.

At this moment we have only 15 clients already in production on the old infrastructure and plan to move them to this one soon. We want to add new clients already in the new infrastructure but the use of NAS shares is a big concern.

[Andreas Neufert]

For the NAS shares, use NFS it should work a bit faster compared with SMB.

How many VMs do these 15 customers backup?

[tzcorreia]

Hi.

On March the total was almost 400.

[tzcorreia]

But how do I add StoreOnce NFS Share as repository? I will have to mount that on a Linux Server first, and this will put another machine on the middle.

[Andreas Neufert]

No, Veeam Backup & Replication can work as well from Windows Gateway Servers with NFS shares since some versions.

[tzcorreia]

Hi.

We have now 2 tenants using SMB and 2 tenants using NFS. Tenants with NFS are not able to perform backups because they receive the error "Failed to call RPC function 'NfsGetDirectorySize': Operation is not supported. NFS status code: 10004". We have already case #07306893 opened but we still not solve this issue. One of the tenants using SMB is having a hugh issue, syntethic fulls take more that 48h to complete (before they tooke less than 12h) and so they are loosing restore points on the SP side. We have a case opened for this issue too case #07303117, also not solved. Do you have any ideas?

[foggy]

Hi Teresa, thanks for opening support cases, our engineers will be able to address the technical issues, if there are any. However, please keep in mind that lower performance of synthetic operations is expected in the case of using StoreOnce without Catalyst integration.

[oczkov]

It has been almost a year since it was mentioned that the HPE StoreOnce will be supported as a cloud repository. Do you have any updates on this?

[veremin]

The API we currently use to work with StoreOnce has a limit on the number of concurrently opened files, and the maximum length of backup chains on HPE StoreOnce is also limited based on the specific storage model. Due to these limitations, we initially blocked the usage of StoreOnce in Cloud Connect.

However, in one of the upcoming releases, we are removing this limitation by switching to another API implemented by the HPE team. With this change, it will be time for us to remove the restriction in Cloud Connect that prevented using StoreOnce as a cloud repository for tenants.

Thanks!

[oczkov]

Do you have a raw estimation *when* this is going to be available (GA)?

[veremin]

In one of the updates for version 13. Unfortunately, at the moment, we cannot provide more specific timeframes. Thanks!