I'm running proxmox backup server in a container but I think I may have over complicated my re-install scenario which caused me to lose my data store accidently.
[ rpool - zfs 180GB ]
proxmox ve
container: 900 proxmox backup server - 16GB
[tank0]
- vm1
- vm2
[tank1]
- container 900: mount_point_1 data store - 1TB (set not to backup)
Re-install proxmox on rpool wipes out the backup server container. Understandable however, I did have a vzdump of the main pbs server container.
But when I copied back the lxc directory, the /etc/pve/lxc/900.conf now created the container info and when I went to restore the 900-pbs-vzdump it already thought 900 existed - rightfully so...
This is where I made the silly mistake of using the gui to delete the 900 container... That's when I lost my [ tank1 1TB mount point - datastore ]
I believe when I copied the /etc/pve/lxc directory for the pbs this was my big mistake. I think if I did not copy the /etc/pve/lxc/900.conf the restore of the vzdump would have been fine. Or at least I should have restored to a new 901 then somehow pointed the right container to it.
Of course if I just put the pbs server on [tank0] to begin with none of this would have been an issue. I'm going to move it after this lesson....
Anyway, just wanted to show my mistake if it helps anyone re-think their pbs container solution and pve setup on re-installing.
I've made new backups and all is well.
EDIT: Also looking into a 2nd remote PBS sync can also help in these situations.
[ rpool - zfs 180GB ]
proxmox ve
container: 900 proxmox backup server - 16GB
[tank0]
- vm1
- vm2
[tank1]
- container 900: mount_point_1 data store - 1TB (set not to backup)
Re-install proxmox on rpool wipes out the backup server container. Understandable however, I did have a vzdump of the main pbs server container.
But when I copied back the lxc directory, the /etc/pve/lxc/900.conf now created the container info and when I went to restore the 900-pbs-vzdump it already thought 900 existed - rightfully so...
This is where I made the silly mistake of using the gui to delete the 900 container... That's when I lost my [ tank1 1TB mount point - datastore ]
I believe when I copied the /etc/pve/lxc directory for the pbs this was my big mistake. I think if I did not copy the /etc/pve/lxc/900.conf the restore of the vzdump would have been fine. Or at least I should have restored to a new 901 then somehow pointed the right container to it.
Of course if I just put the pbs server on [tank0] to begin with none of this would have been an issue. I'm going to move it after this lesson....
Anyway, just wanted to show my mistake if it helps anyone re-think their pbs container solution and pve setup on re-installing.
I've made new backups and all is well.
EDIT: Also looking into a 2nd remote PBS sync can also help in these situations.
Last edited:
