Kopia: A Backup Tool That’s Quietly Smarter Than Most
Most backup software either tries too hard or not hard enough. Kopia is one of the rare ones that gets it just right — doesn’t talk much, doesn’t lock you into anything, and doesn’t mess up your system with daemons or half-working services.
It just works. You run it, point it at a folder, and it starts creating snapshots — encrypted, deduplicated, and fast. No server needed. No GUI dependency (though there is one if you want it). It feels like something built by people who actually needed to back things up at scale — without turning every setup into a research project.
What It’s Actually Like to Use
You install it (literally one command), set up a repository — could be local disk, SFTP, S3, whatever — and start snapping directories. Doesn’t matter if it’s your home folder or /etc, it figures out what changed and only stores the differences.
Each snapshot is like a moment frozen in time, but it’s not just a full copy — it’s smart. It hashes every file, deduplicates on the fly, encrypts everything with AES by default. And you can mount it like a filesystem if you want to look inside.
No weird formats, no lock-in.
Typical Use Case (Real One)
Let’s say someone has a bunch of workstations, and wants to back up configs, home dirs, and some app data to a remote NAS over SSH.
Instead of setting up an rsync script with a dozen flags, you do:
kopia repository create sftp –host=nas.local –user=backup –path=/backups
kopia snapshot create /home
kopia snapshot create /etc
And that’s it. Every time you run it again, it’ll only store what’s new. No overwriting, no bloat, no drama.
You can even set a policy:
kopia policy set –keep-last 5
So it keeps just five snapshots per target. Simple, predictable, and surprisingly fast.
Key Characteristics of Kopia
| Feature | Description |
| Platform Support | Windows, Linux, macOS |
| Storage Backends | Local filesystem, SFTP/SSH, Amazon S3, Google Cloud, WebDAV, B2, and more |
| Encryption | Built-in AES-GCM encryption per snapshot |
| Deduplication | Yes — block-level, content-based |
| Snapshots | Immutable, mountable, and easy to list or restore |
| User Interface | Command-line by default, optional Web UI via embedded server |
| Rootless Operation | Fully supports running as non-root |
| Retention Policies | Per-path or global settings: keep last N, keep hourly/daily/weekly, etc. |
| Automation Friendly | Scriptable CLI, ideal for cron jobs or CI/CD tasks |
| Installation | Single binary; install via script or manual download |
Why It’s Good (And Sometimes Great)
– Doesn’t need root. That alone makes it great for devs and shared systems
– Works over anything — local disk, cloud buckets, SSH, even Google Drive
– Snapshots are encrypted out of the box
– CLI feels clean. Nothing cryptic, nothing hidden behind a wizard
– Web UI is optional, and you can run it with one command (kopia server start)
It’s also one of the few backup tools that actually plays well with scripting. You can cron it. You can CI it. It doesn’t complain when it runs headless. Basically, you control it — not the other way around.
Gotchas & Gaps
– It’s not real-time — it runs on snapshots, so it’s not tracking changes constantly
– First-time setup might feel weird — especially policies and retention
– It’s not a system image tool. File-level only
– GUI is getting better, but still not as polished as enterprise stuff
But honestly, for what it does, it’s one of the best open-source backup tools around right now — especially for people who want clean automation without noise.
