sinuz

Hi all, In my backup strategy i use tree different types of backup. Local, USB (offline) and remote (online). I'm slowly moving all over to unraid and got the local (cache->array) backup working with the rsnapshot docker. Now i'm trying to get the the second step working. However: i think more of this could be automated. In my ideal world the following would happen: I plug in the USB drive Unraid checks if this is the backup drive The backup runs The usb drive unmounts I get a notification that the backup is finished and it is safe to remove the drive I remove the USB drive and store it safely To accomplish this I made a second rsnapshot docker and connected the USB drive with the Unassigned Devices plugin. Making a backup works but requires CLI-intervention. Any help in further automating this would be greatly appreciated.

Thanks! I moved my data one folder deeper. I'll report back if it works Update: after about a week and multiple tries to reproduce the problem everything is still working, moving the data one level deeper solved this issue

Both appdata and the data share are on my pool which is indeed formatted with zfs (Unraid 6.12.6).

Every night i use the appdata-backup to stop, backup, check for updates and restart my nextcloud-ffmpeg container. Most of the mornings after i am greeted with an error explaining the data directory is readable by other users suggesting to change the folder permissions to 0770. The folder permissions as visible from within the container are drwxrwx---. The only thing notices that has changed is the owner of the data folder(volume) in the docker container, when i change it back with chown www-data:www-data data (from within the container cli) the error goes away. Perhaps relevant: the data volume is mapped to a different share ("nextcloud"). Any suggestions on debugging this issue?