Edit: interestingly the restic repo's Readme is better in that regard. It's the demo video but in text form with explanations: https://github.com/restic/restic#quick-start
Edit: interestingly the restic repo's Readme is better in that regard. It's the demo video but in text form with explanations: https://github.com/restic/restic#quick-start
repo=/mnt/mybackupdrive/somefolder
restic init -r $repo
restic backup /my/files -r $repo
restic mount /mnt/backedup -r $repo
cp -r /mnt/backedup/snapshots/latest/accidentally-removed-directory ~/restored
Where the repository can also be sftp or various other things.There are a ton of other commands and options, but to literally just get started, this is all you need for making and restoring data from a backup.
This quickstart assumes some context that not every person starting to use Restic has. It could be improved by offering some more context for each line:
# RESTIC QUICKSTART FOR LINUX
# Choose were you want to save your backups
MYREPO=/mnt/mybackupdrive/somefolder
# Initialize a new restic repo at your chosen backup location
restic init -r $MYREPO
# Backup your files to the newly created repo
restic backup /my/files -r $MYREPO
# To restore a backup, first mount the repo
restic mount /mnt/backedup -r $MYREPO
# Browse the latest backup at /mnt/backedup/snapshots/latest
ls -la /mnt/backedup/snapshots/latest
# Copy the files you want to restore out of the repo
cp -r /mnt/backedup/snapshots/latest/directory-to-restore ~/restoredQuick start != guide, to me. A guide will guide you through thing at a slow pace, a quick start is the quickest way to get something running/started. And the commands seem fairly self-explanatory, but I could be suffering from the so-called curse of knowledge.
Fwiw, I've used restic for a few years, and (still, if perhaps less) agree the docs are not great.
Mostly I just use restic --help or restic subcommand --help anyhow, rather than the docs. Perhaps I'm just not doing fancy enough things with it and that's why I didn't run into edge cases yet where it lacks documentation?
Like...maybe it's time to stop working on new features, and focus on a release.
Duplicati has a similar problem. They're endlessly tinkering with new features instead of squashing bugs and meeting user expectations. I think it still chokes and permanently corrupts backup archives if you interrupt it during the initial backup.
Like: guys. People expect backup software to be able to handle interruptions and disconnects. It's actually one of the things I liked the most about Crashplan, and it could handle being interrupted without issue...nearly a decade ago.