Hmm, the required drive space is estimated by a dry run and then add drive block size once for each file. If this is then larger than free space on target (minus existing backup in case), then the warning appears.
If the target SDcard really has 16G free space, then no chance that the block size addition can be the reason, unless you have billions of tiny byte-sized files.
There was one other case where this appeared due to a single corrupted very large file: https://dietpi.com/forum/t/dietpi-backup/2867/1
The file size there has wrong value due to fs corruption. The problem is that the backup indeed writes this size with meaningless data to the target until it’s full. I had this once myself on SDcard, as well within the kernel modules directory, where one file would have been several Exabyte size. Perhaps a certain RPi kernel update shipped a broken file or the update method caused file corruption in some cases, not sure.
Above the warning prompt there should have been a size estimation print in terminal. Can you see and copy&paste this to see how large DietPi-Backup expects the required disk space?
You can try to ignore. But if it takes more than e.g. 5 minutes, please cancel/interrupt and check /var/log/dietpi-backup.log. The last line should then contain the file that is too large due to corruption or at least the directory where this file is inside. So we can investigate and see if it’s safe to simply remove it.
Just one question:
- Why you place the backup onto an SDcard? This is no recommended, as SDcards to not survive long when regularly writing much data on them.
- If you already have an external HDD/SDD then I would recommend to use this instead.