Plakar

Plakar creates snapshots of your data, stores them encrypted and deduplicated, and lets you inspect, verify, and restore them later. What it removes is the opaque backup: an archive you cannot look inside, cannot check, and only learn the truth about the day a restore fails.

Snapshots live in a Kloset, an immutable data store, and one or more of them can be exported as a ptar archive that stays self-contained, deduplicated, compressed, and encrypted. Encryption covers metadata as well as data, and the cryptography implementation has been audited. Plakar is driven from the command line or through its built-in web interface.

What sets it apart comes down to a few properties.

• Browsable snapshots: inspect contents, diff two snapshots, or restore a single file without touching the rest.
• Independent verification: every snapshot can be verified without restoring anything first.
• Cheap retention: deduplication and compression mean keeping many snapshots does not multiply storage costs.
• Integrations: a package manager adds sources and backends including PostgreSQL, MySQL, etcd, Kubernetes, and S3-compatible object stores.
• Distributed stores: Kloset stores synchronize across locations for 3-2-1 strategies and for push, pull, and sync workflows.

Building from source needs Go 1.23.3 or higher and a single go install, and prebuilt binaries are published for the usual platforms. Creating a repository, backing up a directory, listing snapshots, restoring, and opening the web interface are five short commands, which makes it a reasonable first backup tool for an engineer who wants no vendor lock-in.