Over the past two years, the AI agent space has gone through wave after wave of "engineering paradigm shifts". From Prompt Engineering to Context Engineering, and on to Harness Engineering, each iteration has focused on making individual agents smarter and more reliable.
However, complex tasks in the real world are never completed by a single person.
For example, multi-disciplinary medical consultations require experts from various fields working in tandem.
And yet, AI industry is still trying to build an "all-powerful" single agent rather than a well-coordinated team of agents.
This is precisely the problem openJiuwen aims to address.
What is openJiuwen?
openJiuwen is an open-source AI agent platform jointly built by Huawei Laboratories, Huawei Cloud, and the Terminal Xiaoyi team, released under the Apache 2.0 license. Its goal is to help developers build production-grade AI agents and enable large-scale multi-agent collaboration.
In May 2026, the openJiuwen community released a major component: JiuwenSwarm. It marks the first engineering implementation of "Coordination Engineering", moving the focus from "making a single agent smarter" to "making multiple agents collaborate efficiently."
In collaboration with Huawei's Terminal Tablet & PC Product Line, Software Department, and other teams, openJiuwen launched the HarmonyOS PC version of JiuwenSwarm. This means JiuwenSwarm shares the exact same technical DNA with the HarmonyOS ecosystem right from the architectural foundation.
What makes JiuwenSwarm different?
At the core of JiuwenSwarm is the Agent Swarm mechanism, which enables multiple AI agents to autonomously divide labor, dynamically negotiate, and collaborate efficiently.
Below are its key features.
Heterogeneous model routing for different tasks
JiuwenSwarm supports routing members to different models, providing appropriate model capabilities for different roles, thus allocating heavy-reasoning models to complex logic and lightweight models to simple tasks. This reduces system load and improves overall execution quality. Developers simply declare the role-to-model mapping in a configuration file, and the system automatically manages scheduling.
JiuwenSwarm has already achieved seamless integration with platforms like HarmonyOS Xiaoyi and Lark (Feishu). Developers can build JiuwenSwarm-mode agents that are invokable directly via HarmonyOS devices.
Layered memory architecture
Paired with a layered persistent memory system, JiuwenSwarm retains full-dimensional records across sessions, including operation history, enterprise profiles, and policy guidelines. Through a cross-Swarm experience inheritance mechanism, the experience accumulated by one Swarm can be directly inherited by another.
Distributed deployment
JiuwenSwarm supports distributed Agent Swarms. The Leader Agent and swarm members can run across different processes, different nodes, or even entirely different servers. This capability is especially critical for tasks requiring access to isolated environments, such as intranet databases.
This architecture naturally adapts to the multi-device distributed scenarios of HarmonyOS PC. In the HarmonyOS PC "Agent PC" environment, the PC acts as the home base for the Leader Agent, while HarmonyOS devices like phones and tablets act as Teammate Agent nodes, working together without needing extra adaptation.
With the Channels mechanism supporting platform integrations like Xiaoyi out of the box, developers can build multi-agent applications for the HarmonyOS PC ecosystem straight away.
Observability and self-evolution
JiuwenSwarm provides an intuitive interactive interface, helping users track Swarm running states, task progress, and member workloads in real time.
Simultaneously, its experience closed-loop mechanism captures task breakdown paths, role allocation strategies, tool invocation sequences, and result evaluations, converting them into reusable "task templates." The next time a similar task arises, the Swarm can skip the planning phase and jump straight into execution.
In enterprise office scenarios, OfficeClaw has already leveraged JiuwenSwarm to achieve automated collaboration across content generation, document processing, and knowledge search.
Meanwhile, the Swarm Skills Hub has attracted contributions from numerous developers, fostering a virtuous ecosystem loop.
Where it's headed
In openJiuwen's commercialization roadmap, the native HarmonyOS agent system represents one of the largest and strategically most significant deployment cases.
With JiuwenSwarm now open source, developers can directly build multi-agent applications for the HarmonyOS PC ecosystem using this framework.
There's no shortage of free backup tools on Linux. The problem is choosing which one to use.
The choice of backup tool depends on what you're actually doing. Someone backing up a home folder on a GNOME desktop has nothing in common (backup tool wise) with someone running five servers and a NAS.
In other words, different types of Linux users would have different need of a backup software.
Let me share you a variety of backup tools you can use in a variety of situations in Linux.
Back up tools for Linux users
At a glance, for our busy readers:
Tool
Best For
Backup Type
Cloud Support
Free Tier Limits
Déjà Dup
Linux desktop users (GNOME)
File-level (Restic backend)
Yes – Google Drive (via GUI)
None – fully free and open source
MSP360 Free Backup
Cross-platform users wanting cloud backup
File-level (Linux)
Yes – BYOS (S3, B2, Wasabi)
Personal use only, 5TB cap, no image backup on Linux
Kopia
Porfessionals and commercial use, off-site backup
File-level (snapshot-based)
Yes – BYOS (S3, B2, Azure, SFTP)
None – fully free and open source
BorgBackup
Linux servers, homelabs, terminal-native users
File-level (snapshot-based, dedup)
Local/SSH only – cloud needs extra tooling
None – fully free and open source
Restic
Developers, scripting, automation
File-level (snapshot-based)
Yes – BYOS (S3, B2, SFTP, and more)
None – fully free and open source
UrBackup
Multiple machines, small office fleets
File + image (mainly Windows clients)
No – local/network storage only
None – fully free, but requires server setup
Desktop Linux users running GNOME will get the smoothest experience from Déjà Dup. It's already in most GNOME-based distros and asks almost nothing of you. Connect it to Google or One drive or some network server for offsite backups.
Individual professionals and small teams need commercial-use licensing as much as features. Kopia is the strongest fit here. Open source, no restriction on business use. MSP360 Free covers the cloud-backup workflow well but is licensed for personal use only, so it's a better match for a side project than client data.
Sysadmins and homelabbers managing more than one servers should look at BorgBackup, Restic, or UrBackup. The choice really comes down to whether you want a CLI tool on each machine or one server watching all of them.
1. Déjà Dup: Best For Desktop Linux Users
Take Ubuntu, Fedora Workstation, basically any GNOME-based distro, odds are Déjà Dup is sitting in your app menu. Just search for "Backups." in the GNOME Activity. Pick what to back up, pick where it goes, set a schedule. That's the whole process.
It moved to a Restic backend a while back, so under the hood you're getting all the goodness of Restic, proper incremental snapshots, not just a folder copy. And all this without ever opening a terminal, Restic is CLI after all.
You can choose to store the backup in Google Drive, OneDrive, local folder (but what's the point), network server (NAS) or use RClone to any cloud storage of your choice.
There is one thing you should note. Déjà Dup backs up what your user account can access. Your home folder, basically. It's not a system backup tool, and it won't help you recover a broken OS install. For protecting personal files on a single desktop, though, it's close to ideal. For system backups, try Timeshift.
✋
Non-FOSS Warning! MSP360 offers 5 TB of cloud storage for personal use. However, the backup tool is not open source.
2. MSP360 Free Backup: Best For Personal-Use Cloud Backup Option
MSP360 Free Backup is one of the few free tools that bring real cloud backup to Linux without asking for a subscription. On Linux, MSP360 is more of CLI-first tool. The free edition is described as a command-line tool that automates file-level backups and schedules through cron. And that puts it closer to how Restic or Borg operate. There's a GUI installer if you want it, but MSP360's documents point Linux users toward cron-based automation as the main path.
You get flexible choice for storing the backup files. Local storage, network shares, or your own cloud account through Backblaze B2, Wasabi, or any S3-compatible provider.
The free edition is capped at 5TB locally and 5TB in the cloud. Incremental backup support is included, and automation runs through cron rather than a built-in GUI scheduler.
The free tier is licensed for personal use and won't run on domain-joined machines (if you don't know what that is, don't worry). Some advanced security and management features, including centralized management, are reserved for paid editions.
If you want free cloud backup on Linux and don't mind a cron job standing in for a GUI scheduler, this could be worth a try. Kopia is the nearest comparison to MSP360.
3. Kopia: Open Source and Cloud-Ready
Kopia is an open source cross platform backup tool for Linux, Windows and macOS. This doesn't mean that it is strictly a GUI tool. KopiaUI sits on top of the CLI, and scheduling, deduplication, and end-to-end encryption all works from there.
On the storage side, it covers a lot of groundyou have all the standard options. S3-compatible buckets, Backblaze B2, Azure Blob, Google Cloud Storage, SFTP, WebDAV, plus the usual local and network drives. All of it runs through your own storage accounts. Kopia doesn't provide storage, just the backup tool.
There is some learning curve here. You'll need to understand repositories and policies first. Not difficult, exactly, but not plug-and-play either. And because Kopia has less of a track record than Restic, testing your restores before you depend on them is worth the extra ten minutes.
4. BorgBackup: The Terminal-Native Choice for Linux Servers
BorgBackup feels like it was built specifically for the hardcore sysadmins. SSH, cron, systemd timers, no GUI required unless you bolt one on separately like Vykar. It's a strong fit for home servers, NAS boxes, and homelab setups where a GUI is a secondary need.
Borg builds deduplication, compression, and authenticated encryption straight into how it stores data, so both storage costs and bandwidth stay down over the long run. The feature that I find awesome is that you can mount a Borg backup and poke around it like a normal filesystem, instead of pulling an entire archive just to grab one file.
Cloud object storage isn't native here. Borg expects local drives or SSH-accessible remote servers. Getting it to talk to S3 or Backblaze B2 usually means a hosted Borg-compatible service or extra tooling. If you're already comfortable with SSH and server management, none of that is a real obstacle.
5. Restic – Maximum Control, Minimum Hand-Holding
Restic asks for more from the user than anything else on this list, and gives back the most control in return. There's no GUI on Linux, no built-in scheduler. You bring your own cron job or systemd timer. What you get is a fast, snapshot-based backup tool where every run after the first only stores what changed, and every snapshot is independently restorable.
Encryption is handled automatically using a key derived from your password. Storage destination span local drives, SFTP servers, and S3-compatible storage, Backblaze B2, Azure, Google Cloud.
This is a tool for serious sysadmins and terminal junkies. If that's you, it's hard to find a more capable free option. If it's not, Kopia gets you most of the same capability with considerably less setup.
6. UrBackup: When You're Backing Up More Than One Machine
Every othre tool here is built around backing up one machine. UrBackup is what you could use when you manage a fleet of servers or computers. It's a central server with a web dashboard, pulling backups from all client machines you point it at, deduplicating across all of them to keep storage cost in check.
The process involves setting up a machine ((Linux, Windows, NAS etc) ) that runs the server and installing a lightweight client agent on every machine you want covered. After that, scheduling, monitoring, and restores all happen from the web interface, no need to log into each machine separately.
One thing worth noting is that UrBackup's image-based backup feature was built primarily for Windows clients. Newer versions do support extX and XFS image backup for Linux clients, but it's a less "battle-tested" path than the Windows side. That's why file-level backup is still the safer bet for Linux machines. Worth planning around if you're running a mixed fleet of Windows and Linux endpoints off one server.
Few Things to Keep in Mind
Backup type matter. True image-based, bootable disk backup, the kind EaseUS or Veeam offer on Windows, barely exists for free on Linux. Clonezilla is there but I can never make it work. Most tools in Linux do file-level, snapshot-based backup instead which is fast, space-efficient, but not a one-click bare-metal restore.
Automation varies more than you'd expect. Déjà Dup, Kopia, and UrBackup all include scheduling as part of the product experience. MSP360, Restic, and BorgBackup are the exceptions. They expect you to wire up cron or systemd timers yourself, which is standard practice for Linux users but worth knowing if you're coming from a Windows background where scheduling is baked into the GUI by default.
Multi-machine management is where UrBackup earns its place on this list. None of the other five tools here are built to backup a fleet of machines from one dashboard.
FAQs
Do you need GNOME to use Déjà Dup? No, but you'll get the smoothest experience there since it's designed around GNOME conventions. It runs on other desktop environments too, just without the same level of integration.
Isn't rsync good enough for backup? Rsync syncs files. It doesn't keep historical versions by default. Delete or corrupt a file, then sync, and now your "backup" has the corrupted version too. Restic, Kopia, and Borg all keep point-in-time snapshots for exactly this reason, so you can go back to before the mistake happened.
Can any of these handle a remote VPS or headless server? Yes. Restic and BorgBackup are both built around exactly that scenario, run entirely from the command line with no desktop required. UrBackup can also back up headless machines if you install just the client agent.
What's the real difference between free and open source here? MSP360 Free is a limited free tier of a commercial product personal use only, some features are reserved for paid plans. Déjà Dup, Kopia, Restic, Borg, and UrBackup are all open source, broader usage rights, no features locked behind a paywall.
Should you combine more than one of these backup tools? Yes. See what fits your need.
So, Which Backup Tool to Use on Linux?
If you just want the shortcut:
GNOME/Linux desktop, zero setup: Déjà Dup
Free cloud backup, don't mind a cron job or non-open source: MSP360
Want a GUI with open source: Kopia
Live in the terminal, want full control: Borg or Restic
Managing more than one machine: UrBackup
And let me end this with excellent advice on backups: