Statistics unreliable?

[Ted_Clark]

One full backup and one incremental for Fran Davis and the stats show like two full backups.
For LC one full backup (a small folder for testing) and multiple incremental (adding additional folders of larger amounts) the stats look realistic.

To be truly useful statistics should only be actual disk usage, no hard-links. Let us know space usage for each client and free space without having to look elsewhere. The pie chart should be entire drive space, not just space used. There should also be an option to assign clients to alternate drives instead of assigning drives to folders. Not everyone wants to run an expandable raid solution.

[silversword]

That is correct, it’s unreliable all over the place because of multiple reasons:

Statistics seem to wrong/ Possible issues with quota?

I’m using Urbackup Server 2.033 on a Debian Jessie Host using BTRFS as storage filesystem. So one of my clients is a database server, which I backup incrementally once every hour. Just the MyISAM table files. If I look at the statistics of the urbackup webui, it says that the total usage of files is 2.3TB Now I’m guessing it’s not taking into account things like compression and deduplication when counting the statistics. Since if I look at the available space of the server directly via comman…

No two backup sizes are the same

doesn’t include . files, you could use du -hs .* separately , or ls -la to get a listing But for what you want, better install ncdu instead The missing folder should be .directory_pool, it’s for fully unchanged folders that get symlinked

[aj_potc]

I can’t add anything useful to this. I just want to say that I heartily support your suggestions and can confirm that, yes, the statistics are highly unreliable in my case as well.

UrBackup is a great piece of software and could certainly benefit by having a better overview of storage used, especially for those of us who manage a lot of systems and want a good dashboard/overview of our backup space usage.