What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
The safest OpenClaw backup is a verified archive created by OpenClaw itself, restored first into an empty staging directory. Use openclaw backup create for a complete machine move or rollback; use the SQLite, Git, or continuous-replication options when database recovery must be more frequent or granular. Never copy OpenClaw’s live database files by hand.
Antihost offers VPS plans with one-click OpenClaw, NVMe storage and a 14-day free trial.
What an OpenClaw backup contains
OpenClaw’s native openclaw backup command group handles archive creation, verification, restoration, SQLite snapshots, and Git-backed database history. A full archive is intended to reproduce an OpenClaw installation’s operational state rather than just save one preferences file.
As an Amazon Associate I earn from qualifying purchases.
| Data | Included in a full archive | Why it matters during recovery |
|---|---|---|
| Configuration | Active OpenClaw configuration | Restores settings needed for the Gateway and agents |
| Authentication profiles | Saved auth profiles | Preserves sign-in and provider access configuration |
| Provider and channel credentials | Credential records used by configured providers and channels | Allows restored integrations to reconnect without rebuilding every credential record |
| Sessions | Session history and associated state | Retains conversational and operational context |
| Databases | OpenClaw database state, captured through the backup mechanisms | Preserves structured application data without unsafe live-file copying |
| Workspaces | Included unless you disable them | Protects workspace files alongside application state |
Because archives can contain credentials and session history, treat every archive as secret material. A local archive created by OpenClaw is a plaintext .tar.gz file. Encrypting an upload at the storage destination does not encrypt that local file before it is uploaded.
Choose the right backup method
The method depends on what you need to recover and how often the state changes.
#1 Best Overall
- Easily store and access 2TB to content on the go with the Seagate Portable Drive, a USB external hard drive
- Designed to work with Windows or Mac computers, this external hard drive makes backup a snap just drag and drop
- To get set up, connect the portable hard drive to a computer for automatic recognition no software required
- This USB drive provides plug and play simplicity with the included 18 inch USB 3.0 cable
- The available storage capacity may vary.
| Method | Best use | Main trade-off |
|---|---|---|
| Full archive | One-off protection before updates, resets, uninstalls, or a machine move | Consumes more storage and re-uploads the complete copy |
| SQLite snapshot | Compact, verified recovery of database state | Does not cover every workspace file or other non-database asset |
| Git backup | Versioned, incremental database history | Requires a private repository and a deliberate policy for secrets |
Litestream or sqlite3_rsync |
Frequent or near-continuous database protection | Adds operational components and requires a tested restore plan |
For a simple move to another server, start with a full archive. For a busy installation where losing a day of database changes is unacceptable, add a database-focused or continuous method rather than trying to make full archives run constantly.
Prepare a safe backup destination
Protect the destination itself
- Use a directory with permissions limited to the OpenClaw operator and backup administrators.
- Keep at least one copy separate from the running machine. An external SSD that is disconnected between backup runs provides physical separation for local archives.
- For offsite copies, use an encrypted and access-restricted destination. Named offsite locations are supported by OpenClaw’s managed backup workflows.
- Plan credential rotation before storing an archive in any location you do not fully control. A leaked archive may expose provider credentials, channel credentials, auth profiles, and session history.
Check space and scope
Workspaces can be large. Decide whether the archive needs them before running a long upload. A configuration-only archive is useful for repairing a malformed configuration, but it is not a substitute for a full recovery copy. If you omit workspaces, document where those files are protected by another system.
Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallOutdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchCreate and verify a full OpenClaw archive
1. Preview the archive
Use a dry run to see what OpenClaw plans to capture without writing the archive:
openclaw backup create --dry-run --jsonThe JSON output is useful for logging the planned scope in an automated job. Review whether workspaces and the expected state components are present before committing the backup.
2. Create a local archive
Write the archive to a protected backup directory:
openclaw backup create --output ~/BackupsThis creates a complete archive containing configuration, auth profiles, provider and channel credentials, sessions, databases, and workspaces unless workspace inclusion is disabled. Keep the resulting .tar.gz file protected even if you later copy it to encrypted storage.
3. Verify during creation
Add --verify when you want the command to validate the archive as part of the run:
The Tool Desk
Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →openclaw backup create --output ~/Backups --verifyA backup job that reports success without verification only proves that an output file was written. Verification gives you an additional check before you rely on the archive for a rollback or migration.
4. Exclude workspaces when appropriate
To protect OpenClaw state without copying workspace files, use:
Rank #2
- Easily store and access 5TB of content on the go with the Seagate portable drive, a USB external hard Drive
- Designed to work with Windows or Mac computers, this external hard drive makes backup a snap just drag and drop
- To get set up, connect the portable hard drive to a computer for automatic recognition software required
- This USB drive provides plug and play simplicity with the included 18 inch USB 3.0 cable
- The available storage capacity may vary.
openclaw backup create --no-include-workspaceUse this only when workspace content is protected elsewhere or is intentionally disposable. Restoring such an archive will not recreate those omitted files.
5. Send an archive to an offsite location
OpenClaw can apply retention settings while sending archives to a configured destination:
openclaw backup create --to offsite --keep-daily 7 --keep-weekly 4 --keep-monthly 12These options retain seven daily copies, four weekly copies, and twelve monthly copies according to the configured offsite workflow. Confirm that the destination is private, encrypted, and accessible during a restore; retention is not a substitute for access control.
Make a configuration-only safety copy
If the immediate problem is a malformed configuration, save the active JSON before editing or repairing it:
openclaw backup create --only-config --output ~/Backups/openclaw --verifyThis operation saves only the active configuration file. It does not parse or include files referenced through $include, so preserve those dependencies separately if the configuration relies on them. After the repair succeeds, create a new full archive rather than treating the configuration-only file as a complete backup.
Back up OpenClaw databases without corrupting them
OpenClaw explicitly warns against copying live database files. Do not copy files ending in .sqlite, -wal, -shm, or -journal while OpenClaw is running. Those files can represent an in-progress transaction and may not form a consistent database when copied independently.
Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Repair Windows errors before they cause bigger problemsFix Now →SQLite snapshots
Use OpenClaw’s snapshot command so the database is captured and checked as a database:
openclaw backup sqlite create --global --repository ~/Backups/openclaw-sqliteFor an individual agent, specify its name instead:
openclaw backup sqlite create --agent main --repository ~/Backups/openclaw-sqliteSnapshots are compact and suited to database-focused recovery. They do not automatically preserve every workspace file, configuration artifact, or credential outside the selected database scope, so pair them with full archives when you need a complete machine image.
Rank #3
- Easily store and access 1TB to content on the go with the Seagate Portable Drive, a USB external hard drive.Specific uses: Personal
- Designed to work with Windows or Mac computers, this external hard drive makes backup a snap just drag and drop. Reformatting may be required for Mac
- To get set up, connect the portable hard drive to a computer for automatic recognition no software required
- This USB drive provides plug and play simplicity with the included 18 inch USB 3.0 cable
- The available storage capacity may vary.
Git-backed database history
A private Git repository can hold versioned database snapshots. Initialize a repository with access restricted to trusted operators, then use OpenClaw’s Git backup commands:
openclaw backup git createopenclaw backup git logopenclaw backup git verifyopenclaw backup git restoreGit history is useful when you need to identify or restore a particular database version. Its incremental nature reduces repeated full uploads, but repository privacy and secret handling become part of the recovery design.
Do these 3 things before closing this tab:
1Clear out junk files and repair common Windows errors2Fix the driver behind crashes, sound loss and screen glitches3Repair Windows errors before they cause bigger problemsContinuous and incremental protection
For installations where a daily archive leaves too much recovery-point loss, OpenClaw documents Litestream and sqlite3_rsync as alternatives for higher-frequency protection. These tools add components to operate, monitor, and restore. Test the complete failover procedure, not just the replication process, before relying on it during an outage.
Schedule backups and understand secret handling
OpenClaw’s managed backup jobs support archive and Git modes, with a documented default interval of 24 hours. A schedule is useful only if its destination remains reachable and its output is monitored; configure alerts for failed jobs and periodically perform a test restore.
Pushed schedules redact credential-bearing tables and machine-state rows whose names begin with the secret prefix by default. This reduces the impact of exposing a remote history, but it also means a redacted remote copy may not be sufficient for a full-fidelity recovery. The --include-secrets option enables full-fidelity remote history only when the remote is private and trusted. Local non-push schedules retain full fidelity.
Restore a full archive safely
1. Start with a trusted archive
Use an archive you created or one whose provenance you can verify. If credentials may have leaked, rotate them before bringing the restored instance online; restoring a compromised secret does not make it safe.
2. Extract into a new, empty staging directory
Run the restore with a target that does not contain live OpenClaw data:
openclaw backup restore ARCHIVE.tar.gz --target ./restored-openclawThe target must be outside the active state directory and outside configured agent directories. OpenClaw refuses a non-empty target to prevent accidental overwrites. The restore process checks archive structure, manifests, links, and database snapshots before writing. If extraction fails, incomplete output is cleaned up.
Rank #4
- Easily store and access 4TB of content on the go with the Seagate Portable Drive, a USB external hard drive.Specific uses: Personal
- Designed to work with Windows or Mac computers, this external hard drive makes backup a snap just drag and drop
- To get set up, connect the portable hard drive to a computer for automatic recognition no software required
- This USB drive provides plug and play simplicity with the included 18 inch USB 3.0 cable
- The available storage capacity may vary.
3. Preserve agent paths
Custom agent roots are a common migration trap. Either recreate the same paths on the new machine or update each configured agentDir to the restored location before starting the Gateway. A successful archive extraction does not guarantee that an agent whose files were stored outside the default root will be found automatically.
4. Put the staged state into service
After inspecting the staged files, move the restored state into the intended state location, or point OpenClaw at it with OPENCLAW_STATE_DIR. Do not delete the old live state until the staged copy has passed checks and you have a separate rollback option.
5. Run the doctor before restarting
openclaw doctorRun the doctor command against the restored state before restarting the Gateway. Address path, configuration, permission, or database findings first; then restart OpenClaw and confirm that agents, channels, providers, and sessions behave as expected.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Move OpenClaw to a new server
- Create and verify a full archive on the old server. Include workspaces unless they are protected independently.
- Copy the archive to the new server through a protected transfer or retrieve it from the secured offsite destination.
- Install the compatible OpenClaw environment and create an empty staging directory on the new host.
- Restore with
openclaw backup restore; never extract over an existing state directory. - Recreate or update custom
agentDirpaths, then setOPENCLAW_STATE_DIRif the restored state will live outside the default location. - Run
openclaw doctor, correct its findings, and only then start the Gateway. - Validate provider and channel connectivity, session visibility, database access, and workspace permissions before switching production traffic.
The old server should remain untouched until the new instance has passed this validation. That gives you a fallback if a path, permission, or integration does not transfer as expected.
Restore database snapshots and Git history
Database restore commands also require a fresh target. Keep the target separate from live state and let OpenClaw perform its integrity checks, including foreign-key checks, schema validation, and hash verification. A database that opens is not necessarily a database that is internally consistent; do not bypass these checks by replacing files manually.
For Git-backed recovery, identify the required revision with openclaw backup git log, verify it, and then invoke openclaw backup git restore into a clean restore target. Treat the repository as sensitive even when its history is incremental, because an earlier commit may contain data that a later policy removed.
Troubleshoot common backup and restore failures
| Symptom | Likely cause | Safe response |
|---|---|---|
| Restore refuses to run because the target is non-empty | The target contains existing files or is the live state directory | Create a new empty directory outside live state and configured agent roots, then retry |
| Agents start but cannot find their files | A custom agent root was not recreated or its path changed | Restore the files at the original location or update the affected agentDir settings |
| Archive is present but verification fails | The archive may be incomplete, damaged, or not the file you intended to restore | Keep the failed copy for diagnosis, obtain another verified archive, and do not start production from the unverified file |
| Configuration repair did not restore included settings | The configuration-only command does not parse $include dependencies |
Restore the referenced files separately, repair the active JSON, then create a full archive |
| Remote history lacks credentials needed for recovery | Pushed schedules redact credential-bearing data by default | Use a trusted private remote with --include-secrets when full fidelity is required, or keep a protected local full-fidelity copy |
| Database behaves inconsistently after a manual copy | Live SQLite, WAL, SHM, or journal files were copied independently | Discard that copy and create a SQLite snapshot or full archive through OpenClaw |
openclaw doctor reports permission or path errors |
The new host has different users, directory permissions, or filesystem paths | Fix ownership, permissions, and configured paths before restarting the Gateway |
A practical backup plan
- Before a risky change: create a verified full archive, for example with
openclaw backup create --verify. - Every day: run a managed archive job or another scheduled method and monitor its result. The documented managed-job default is 24 hours.
- For frequent database changes: add SQLite snapshots, private Git history, Litestream, or
sqlite3_rsyncaccording to your recovery-point target. - For offsite protection: retain encrypted copies with restricted access and a documented secret policy.
- At regular intervals: perform a staging restore, run
openclaw doctor, and verify an agent, a session, a provider, a channel, and any required workspace. - After a suspected leak: revoke or rotate credentials represented in the archive before reconnecting the restored installation.
This approach separates complete machine recovery from high-frequency database recovery, while ensuring that every restore is tested in a disposable location before it can affect a live Gateway.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
