I have been running rsync to someoneelses server for three years and I am tired of the OperatorError footnotes[1]. Looking at rclone vs Restic vs Borg for a proper encrypted deduplicated setup. My Requirements are simple: Linux only, no GUI needed, must handle millions of small files without choking on the filesystem overhead.
rclone seems too simple, no snapshotting. Restic has that weird Repository format lockin anxiety. Borg is singlethreaded which feels like a deliberate insult in 2025 but BorgBase integration is tempting.
What is everyone actually using in Production and why did you stop using the other one.
Footnotes[1] Last month the remote ran out of inodes because my backup script had no pruning logic. OperatorError is my term for when you are the Operator and it is your Error.
Actually I have been running Borg actually since the 1.0 Release actually and I have never looked at Restic because the Repository format actually changed once and that is something I cannot tolerate in my Infrastructure or? I have three Servers at OVHcloud and one at Vultr and they all push to a single Repository on my own Hardware in the Basement which is actually a Dell Optiplex from 2014 with two USBHarddrives in Softwareraid actually and it works perfectly fine or? The singlethreadedness is actually not a Problem because I run multiple Archives in parallel with different Passphrases for different Clients actually so the deduplication is actually perArchive anyway or?
Post a reply
You need an account to reply.
Log in or
register to join the conversation.
