I have restic 0.15.1 backing up to an AWS S3 bucket. That part works great. What I then try to do is attempt a restore. The first thing I do is check the restore size before I waste time restoring to a volume that can’t fit the restore. Sadly, it took me more than four hours only for that.
$ export RESTIC_REPOSITORY="s3:s3.amazonaws.com/foo/backups/gitlab" $ export AWS_ACCESS_KEY_ID=ABCD $ export AWS_SECRET_ACCESS_KEY=1234 $ export RESTIC_PASSWORD="XXXXX" $ alias restic="restic --no-cache" $ time restic stats 3e09b991 repository fbc2cbd6 opened (version 2, compression level auto) scanning... Stats in restore-size mode: Snapshots processed: 1 Total File Count: 210397 Total Size: 46.788 GiB real 250m49.565s user 3m26.763s sys 1m8.307s
As the backup is done on Kubernetes and the cache is on an EFS (NFS) volume, I have no access to it when I restore to my laptop, unless of course I copy the cache locally.
Is it expected to take this long without cache? The total repo is about 3TB. Would cache have made it any quicker? It all feels extremely slow.
Any advice would be welcome.