Backup data storage pool

[SpinningRust]

Hello everyone,

I have a question regarding backups, that doesn't seem to be too common. I use proxmox at home on my own little server / in a 2-3 node cluster.
Since the server is a home server, it houses quiet some files that i would like to back up one way or another. These files are stored in their own storage pool, separated from VM/CT data, and in their own dataset. That dataset is then forwarded to a container with a mount point.

To my actual question: Can one backup this storage pool/dataset natively with proxmox to another node or to another target? Or would it be better to just have a rsync server on another node that then backs up the data to another storage pool?

VM and CT data is already being backed up, so that's not a concern of mine.


Thanks in advance for any suggestions




P.S.: is it possible to add drives lateron to a RAIDZ1 to expand capacity? I couldn't find an up to date answer to that

 

[wolfgang]

Hi,

if you use ZFS you can use pve-zsync to backup the data offside.
Or use pvesr in the cluster but this is no backup.

Expending a ZFS is only possible with new dev. This means you always add an extend to the other vdevs.

1.) https://pve.proxmox.com/wiki/PVE-zsync
2.) https://pve.proxmox.com/wiki/Storage_Replication

 

[SpinningRust]

Hi,

if you use ZFS you can use pve-zsync to backup the data offside.
Or use pvesr in the cluster but this is no backup.

Expending a ZFS is only possible with new dev. This means you always add an extend to the other vdevs.

1.) https://pve.proxmox.com/wiki/PVE-zsync
2.) https://pve.proxmox.com/wiki/Storage_Replication

Hello Wolfgang,

thank you for your reply.
Right now i don't have a backup server yet so i can not test this; therefore one more question:

would the command

pve-zsync create --source 192.168.1.1:/HQ/sharing --dest BKP/drop --verbose --maxsnap 4 --name replica_01 --limit 512 --skip

then create a recurring job that backs up the dataset "sharing" in the pool HQ to the pool BKP/drop located at the specified IP?


many thanks,

JohnTanner


Edit:

I thought so that a RAIDZ was not expandable after creation but i wanted to be sure. Thank you for clarifying

 

[wolfgang]

then create a recurring job that backs up the dataset "sharing" in the pool HQ to the pool BKP/drop located at the specified IP?

Yes default it runs every 15 minutes but it is just a cron entry so you can set every time interval you like.

I thought so that a RAIDZ was not expandable after creation but i wanted to be sure.

The vdev (RaidZ) itself is not extendable, but you can add a new vdev to the pool and expend the pool.