Help understanding stats against the full repository

Getting Help
1

I’m trying to understand why my restore-size is higher than the raw-size when I run stats against all snapshots across all hosts.

For example running will show me how much storage is used for the whole repository:

restic stats --mode raw-data --json

which gives

{"total_size":32789877906,"total_file_count":0,"total_blob_count":299351}

But if I run with restore-size, what does this mean exactly?

restic stats --mode restore-size --json

which gives

{"total_size":30487544688,"total_file_count":5002615}

and 30487544688 < 32789877906 so my total restore size is less than the total size in the repo (after deduplication).

Have I understood --mode restore-size correctly? How does it work when there are multiple hosts in the repository and I don’t pass latest to stats?

Is there a way to get an idea of deduplication % for all snapshots across all hosts? Or is this something you can only calculate on a snapshot by snapshot basis?

2

The restore size calculation is currently buggy: Fix restore size calculation for multiple snapshots by MichaelEischer · Pull Request #3740 · restic/restic · GitHub . It should report the size on disk when restoring all snapshots (restore each snapshot and sum up all restored snapshots).

1 Like
3

Thanks, I ran stats against the latest build from master with your PR and it makes much more sense.

1 Like