proxmox-stuff
proxmox-stuff copied to clipboard
DerDanilo
/etc/pve is completely empty
I run your script and I am not getting the point: the folder /etc/pve is completely empty in the file: proxmoxetc.2023-03-10.19.30.04.tar
I followed your wiki and while all other folders are ok, this one - quite important I believe - is empty what am I doing wrong?
but when I enter the below commands manually:
cd /
tar cfz etc.tar.gz etc
I get the file with everything from etc, including /etc/pve as well
I don't use the script anymore since we implemented PBS based backup of the PVE cluster config. Still waiting for the proxmox team to actually integrate PBS backup for PVE cluster config/database as they announced.
Did you run the script with proper permission? /etc/pve is only accessible with root.
Same problem here. Also no symlinks for the files in pve.
lrwxr-xr-x 1 root www-data 0 Jan 1 1970 qemu-server -> nodes/atlas/qemu-server
lrwxr-xr-x 1 root www-data 0 Jan 1 1970 openvz -> nodes/atlas/openvz
lrwxr-xr-x 1 root www-data 0 Jan 1 1970 lxc -> nodes/atlas/lxc
lrwxr-xr-x 1 root www-data 0 Jan 1 1970 local -> nodes/atlas
nodes folder is missing.
@DerDanilo said he is not maintaining this anymore ;-) but to be honest he should remove it from github or announce that it is broken for the current PVE version - some people could rely on it and then... you know - shit happens ;)
remove --one-file-system from https://github.com/DerDanilo/proxmox-stuff/blob/master/prox_config_backup.sh#L128 and https://github.com/DerDanilo/proxmox-stuff/blob/master/prox_config_backup.sh#L130
not sure L130 is needed but L128 for sure
‘--one-file-system’ Prevents tar from crossing file system boundaries when archiving. Use in conjunction with any write operation.
Proxmox uses cluster file system for /etc/pve
@Fma965 Excellent! I wonder if this script ever worked (without your amendments) to copy /etc/pve. Would seem an oversight by the original @DerDanilo . I wonder why he added --one-file-system at all. He must have had a good reason, I imagine.
Have never restored (yet!). Would like to try someday - to see if it really works!
The script worked at some point. And yes, it actually copied the files in /etc/pve. It was intended to backup single node setups. If I remember correctly (and I may be wrong) single node systems don't have a mounted filesystem at /etc/pve.
This is correct although it's still odd that that parameter was added, especially if you never used a cluster 🤔
I don't even remember the time I wrote the script. I must have had a reason for it. :)
I don't even remember the time I wrote the script. I must have had a reason for it. :)
Haha, I know that feeling! 😉
I don't even remember the time I wrote the script. I must have had a reason for it. :)
Happens to me all the time - if there is a time-gap when relooking at any script! Anyway thanks again.