Scale-Out Repo as Cloud Connect SP

[Refruit]

Hello again,

we have a storage backbone with 4 identical Hardware Nodes which are forming a global storage pool. Each Node can be accessed via NFS but the storage in the backend is the same. Each customer has it's own simple repository on the storage pool

//<vip01>/nfsshare/customer1
//<vip02>/nfsshare/customer2
//<vip03>/nfsshare/customer3
//<vip04>/nfsshare/customer4

Now i would like to know if this can be accomplished using a Scale-Out Repo when each Node represent a single extent. It would look like this:

//<vip01>/nfsshare/Directory1
//<vip02>/nfsshare/Directory2
//<vip03>/nfsshare/Directory3
//<vip04>/nfsshare/Directory4

Keep in mind that even if the directory name is different, the content of each directory is identical so that regardless of which extent is polled, the backups for each customer is available on each extent.
The question for me is: How would i add and assign a subfolder for each customer on the single Scale-Out Repo with 4 extents. Thats, what i'd like to accomplish.

Regards,
Refruit

[Mildur]

Hi Refruit

In a SOBR, it‘s expected that a single backup file is available only once on all extends. The metadata file will be stored on all extends. When the backup server requires access to the backup data for retention or job run, it needs to open the metadata file on each extend and write new information to the file. I see issues in your plan because of this. Even if it‘s the same metadata file, it has to be opened by each SOBR extend with read/write permission.

Multiple extends can use the same server, but they must have their own folder structure on the server and not use virtualized paths which all link to the same storage folder with the same content.

Thanks
Fabian

[Refruit]

Hi Fabian,

maybe i wrote the scenario a bit confusing.
Yes it is correct the backup file is available only once on all extents and it's a a different metafile for each extent as we have a single NFS share, but with 4 subfolder which are representing 4 different repositories.
What i meant is, that theoretically every repository (subfolder) can be accessed through every vip/node.

We are using that scenario already in our managed infrastructure (a seperate veeam installation) where all customer backups are shared through the same one SOBR. That is no issue here as those customers don't ever have access to those backups, because it is fully managed by us. But in the case of Cloud Connect, each customer has access to the repository so we cannot use the same configuration as described for the managed infrastructure.

I hope that made things a bit clearer
That brings me to the question, if we can add a sub repository/folder to the single SOBR when adding it as cloud repository for the customer, so that it remains unique or would we need to create a seperate SOBR for each customer.

Regards,
Stefan

[Mildur]

Hi Stefan

In case you have a SOBR as a cloud connect repo, esch customer will have his own folder on each SOBR extend.
Customer 1 will never see the data from Customer 2 on a Veeam Cloud Connect SOBR.
The folder on the SOBR is automatically created by veeam when you assign the tenant a new repository on that sobr.

Thanks
Fabian