[PVE-User] Single BPS for multiple PVE lead to namespace conflict

[Jan Brand]

Hi Martin,

you can create multiple datastores on one zpool, just create an 
additional zfs dataset and configure the second pbs datastore on it.
This way all data is stored on the same RAIDZ array, if this is your 
intention.
If you want use the deduplication feature of pbs across backups of 
multiple clusters, I have to disappoint you. Afaik the deduplication is 
per datastore.

Another "solution" would be to manually avoid overlapping VMIDs, but 
this would be error-prone and means a lot of work in an existing 
environment.
I would create one backup datastore per cluster and call it a day.

Best regards,

Jan

Am 09.01.2021 um 22:11 schrieb Konold, Martin:
>
> Hi there,
>
> I am pretty new to Proxmox and deeply impressed by the quality of many 
> aspects of its design and implementation.
>
> In my testing I observed that in case I have multiple PVE Clusters and 
> perform backups to a single datastore on a single PBS I experience a 
> lack of namespaces.
>
> Why a single datastore on the PBS for multiple PVE Clusters?
>
> For reasons of efficiency and avoidance of fragmentation I would like 
> to use a single RAIDZ2 as a target. The problem now arises that both 
> PVE clusters see the same "vm/100/{dateTime}".
>
> What about prefixing the backups with the Cluster-Name e.g. 
> "pve1/vm/100{dateTime}"?
>
> Is there something I overlooked sofar?
>
> Regards
> ppa. Martin Konold
>
> -- 
> Martin Konold - Prokurist, CTO
> KONSEC GmbH -⁠ make things real
> Amtsgericht Stuttgart, HRB 23690
> Geschäftsführer: Andreas Mack
> Im Köller 3, 70794 Filderstadt, Germany
>
> _______________________________________________
> pve-user mailing list
> pve-user at lists.proxmox.com
> https://lists.proxmox.com/cgi-bin/mailman/listinfo/pve-user