I’m using Autorestic, a wrapper for Restic that lets you specify everything in a config file. It can fire hooks before/after backups so I’ve added it to my healthchecks instance to know if backups were completed successfully.
One caveat with Restic: it relies on hostnames to work optimally (for incremental backups) so if you’re using Autorestic in a container, set the host: option in the config file. My backups took a few hours each night until I fixed this - now they’re less than 30 minutes.
Like some others, I have separate storage and compute servers.
The data directory is an NFS share on my storage server and I run Nextcloud in docker on my compute server.
I have the NFS share defined as a volume of type nfs in the docker compose, mounted to /var/www/html/data. Nextcloud itself just treats it like a local directory.