This is very specific to me honestly but someone else might find it useful
I use mergerfs and a local directory with my media in it. i also run websites on my seedbox and generally want to back up as many directories and configurations on my seedbox as possible.
i used to do this with cron jobs and shell scripts, but this is a lot easier.
https://github.com/dannymichel/rbaker
This does a few things:
- moves your media files to your remote
- backs up general directories to your remote
- backs up all your mysql databases and moves it to your remote
1 Like
Nice work on this script 
One thing that improved reliability on my side was splitting transfer logs from verification logs, so it is easier to spot throttling patterns by provider over time.
For seedbox backups, I also run a periodic integrity pass on a sample set before pruning old data.
Another thing that helped on long seedbox copy jobs was keeping a small manifest of the last verified state, so retries do not force you to re-check everything blindly after a transient provider issue.
If you ever extend the script, a simple summary for transferred / verified / skipped items plus total elapsed time makes it much easier to compare runs over a few weeks.
One extra safeguard that paid off for me was treating database dumps separately from the media copy step. I keep a tiny post-run check that compares dump size and age against the previous successful run, because media transfers can look fine while the SQL dump quietly fails.
For large libraries, it also helps to decide explicitly where checksum verification is worth the extra read cost and where size/date checks are good enough. That kept my backup window much more predictable.