Thanks to visit codestin.com
Credit goes to github.com

Skip to content

plakar is a backup solution powered by Kloset and ptar

License

Notifications You must be signed in to change notification settings

tracepanic/plakar

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6,265 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Plakar Backup & Restore Solution

plakar - Effortless backup & more

Join our Discord community Subscribe on YouTube Join our Subreddit

Deutsch | Español | français | 日本語 | 한국어 | Português | Русский | 中文

🔄 Latest Release

Join Plakar v1.1.0 Beta

V1.1.0-beta.1 - Beta Release: Performance, UI, and Architecture (January 2026)

  • New Terminal UI: Completely reworked terminal output with a new tui renderer for better visibility during long-running operations, alongside the classic stdio renderer for verbose output.
  • Dramatic Performance Improvements:
    • Restore operations: ~95% faster (from ~60 minutes to ~3 minutes for 1M items)
    • Backup operations: up to 33% faster with optimizations
    • Sync operations: ~20% faster
  • Significant RAM Reduction:
    • Backup: -43% (from ~3.0 GiB to ~1.3 GiB)
    • Restore: -66% (from ~2.3 GiB to ~800 MiB)
    • Sync: -30% (from ~3.6 GiB to ~2.5 GiB)
    • Check: -40% (from ~1.3 GiB to ~800 MiB)
  • Reduced Cache Footprint: -55% on-disk cache usage (from 4 GiB to 1.8 GiB for 1M items) by removing VFS cache and trading bandwidth for disk space.
  • Architecture Redesign: Replaced the agent with cached, a lightweight process dedicated exclusively to cache maintenance and locking. Commands now execute directly in the CLI.
  • Multi-directory Support: Back up multiple directories in a single snapshot (e.g., plakar backup /etc /home).
  • Improved FUSE Support: Completely rewritten for better reliability on both Linux and macOS, including support for FUSE-T. New capabilities to mount specific snapshots, directories, or serve them over HTTP.
  • New Package Manager: Brand new package manager with simpler interface and support for integration updates.
  • Redesigned Integration Interfaces: Simpler and more explicit importer, exporter, and storage interfaces, lowering the barrier for third-party integrations.

📝 Release article

V1.0.6 - Bugfix Release: State Synchronization and Memory Fixes (November 2025)

  • Critical Fix: Resolved state-synchronization bug that could cause snapshots to appear correct on the backup machine but not on others. Introduced two-stage commit to guarantee remote state updates before local visibility.
  • New Repair Tool: Added plakar repair command to detect and fix state inconsistencies. Recommended for all users to run once after upgrading.
  • Memory Leak Fixes: Fixed storage API memory leak in go-kloset-sdk affecting all third-party integrations (SFTP, S3, etc.) during list, check, and restore operations.
  • Improved Memory Usage: Resolved large buffer retention issue during restore and check operations with external integrations, significantly reducing RAM usage for large snapshots on S3 and SFTP backends.
  • Integration Updates: Users should reinstall integrations (plakar pkg rm/plakar pkg add) to benefit from the corrected go-kloset-sdk.

📝 Release article

🧭 Introduction

plakar provides an intuitive, powerful, and scalable backup solution.

Plakar goes beyond file-level backups. It captures application data with its full context.

Data and context are stored using Kloset, an open-source, immutable data store that enables the implementation of advanced data protection scenarios.

Plakar's main strengths:

  • Effortless: Easy to use, clean default. Check out our quick start guide.
  • Secure: Provide audited end-to-end encryption for data and metadata. See our latest crypto audit report.
  • Reliable: Backups are stored in Kloset, an open-source immutable data store. Learn more about Kloset.
  • Vertically scalable: Backup and restore very large datasets with limited RAM usage.
  • Horizontally scalable: Support high concurrency and multiple backups type in a single Kloset.
  • Browsable: Browse, sort, search, and compare backups using the Plakar UI.
  • Fast: backup, check, sync and restore are operations optimized for large-scale data.
  • Efficient: more restore points, less storage, thanks to Kloset's unmatched deduplication and compression.
  • Open Source and actively maintained: open source forever and now maintained by Plakar Korp

Simplicity and efficiency are plakar's main priorities.

Our mission is to set a new standard for effortless secure data protection.

🖥️ Plakar UI

Plakar includes a built-in web-based user interface to monitor, browse, and restore your backups with ease.

🚀 Launch the UI

You can start the interface from any machine with access to your backups:

$ plakar ui

📂 Snapshot Overview

Quickly list all available snapshots and explore them:

Snapshot browser

🔍 Granular Browsing

Navigate the contents of each snapshot to inspect, compare, or selectively restore files:

Snapshot browser

📦 Installing the CLI

From binaries

Visit https://www.plakar.io/download/

From source

plakar requires Go 1.23.3 or higher, it may work on older versions but hasn't been tested.

go install github.com/PlakarKorp/plakar@latest

🚀 Quickstart

plakar quickstart: https://www.plakar.io/docs/v1.0.6/quickstart/

A taste of plakar (please follow the quickstart to begin):

$ plakar at /var/backups create                             # Create a repository
$ plakar at /var/backups backup /private/etc                # Backup /private/etc
$ plakar at /var/backups ls                                 # List all repository backup
$ plakar at /var/backups restore -to /tmp/restore 9abc3294  # Restore a backup to /tmp/restore
$ plakar at /var/backups ui                                 # Start the UI
$ plakar at /var/backups sync to @s3                        # Synchronise a backup repository to S3

🧠 Notable Capabilities

  • Instant recovery: Instantly mount large backups on any devices without full restoration.
  • Distributed backup: Kloset can be easily distributed to implement 3,2,1 rule or advanced strategies (push, pull, sync) across heterogeneous environments.
  • Granular restore: Restore a complete snapshot or only a subset of your data.
  • Cross-storage restore: Back up from one storage type (e.g., S3-compatible object store) and restore to another (e.g., file system)..
  • Production safe-guarding: Automatically adjusts backup speed to avoid impacting production workloads.
  • Lock-free maintenance: Perform garbage collection without interrupting backup or restore operations.
  • Integrations: back up and restore from and to any source (file systems, object stores, SaaS applications...) with the right integration.

🗄️ Plakar archive format : ptar

ptar is Plakar’s lightweight, high-performance archive format for secure and efficient backup snapshots.

Kapsul is a companion tool that lets you run most plakar sub-commands directly on a .ptar archive without extracting it. It mounts the archive in memory as a read-only Plakar repository, enabling transparent and efficient inspection, restoration, and diffing of snapshots.

For installation, usage examples, and full documentation, see the Kapsul repository.

📚 Documentation

For the latest information, you can read the documentation available at https://www.plakar.io/docs/v1.0.6/

💬 Community

About

plakar is a backup solution powered by Kloset and ptar

Resources

License

Code of conduct

Contributing

Security policy

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Go 91.1%
  • Roff 8.6%
  • Other 0.3%