Self-hosted beta · GNU GPLv3
BackupSheep is the open-source control plane for websites, databases, servers, volumes, and cloud infrastructure. One place to schedule the work, keep copies where you choose, and recover with confidence.
The recovery control plane
Cloud snapshots in one console. Database dumps in another. Website archives somewhere else. BackupSheep gives the whole recovery surface one operating model—without taking ownership away from you.
Read the usage guideBring cloud accounts, Linux hosts, databases, WordPress, and Basecamp into one inventory.
Use cron, intervals, or one-time schedules. Add retention and protected-copy requirements.
Fan one backup out to local disk, object storage, cloud drives, or several destinations at once.
Restore files and databases from the console; rebuild native cloud snapshots through the provider.
What it protects
BackupSheep uses the right mechanism for the source: incremental file archives, native database dumps, or provider snapshots. Nothing is flattened into a lowest-common-denominator backup.
INCREMENTAL FILE ENGINE
After the first mirror, only new and changed files cross the network. Every run still becomes a complete, standalone zip—never a fragile restore chain.
NATIVE DATABASE ENGINE
Version-matched native clients produce portable dumps over direct or SSH-tunnelled connections, with table-level selection when a full database is too much.
PROVIDER-NATIVE SNAPSHOTS
Discover servers, volumes, and supported managed databases from connected providers. BackupSheep schedules native snapshots while they remain in your cloud account.
APPLICATION DATA
Back up WordPress and Basecamp alongside infrastructure so application data is visible in the same inventory, schedule view, and audit trail.
Your data, your topology
Fan every archive out to several destinations in the same run. Keep it on your own disk, use the object storage you already buy, or add a separate copy beyond the blast radius.
Amazon S3 · Backblaze B2 · Wasabi · Cloudflare R2 · DigitalOcean Spaces · Google Cloud Storage · Azure Blob · IDrive e2 · IBM COS · Oracle · Scaleway · Linode · Vultr · UpCloud · Exoscale · Filebase · IONOS · Leviia · RackCorp · Tencent COS · Alibaba OSS
Dropbox · Google Drive · Microsoft OneDrive · pCloud
Local Storage · bind-mounted disk · NFS path
26 includes Local Storage. Hetzner Object Storage connects through its S3-compatible endpoint. Amazon S3 Object Lock controls apply specifically to Amazon S3 destinations.
Recovery is the feature
Open a historical backup, choose the stored copy, and start the restore from the same console that created it. Live status and redacted logs keep the recovery visible.
WebsitesReverse-mirror files, with optional exact-mirror deletion.
DatabasesImport native dumps and recreate missing databases.
Cloud snapshotsRebuild through the provider’s native recovery workflow.
Aug 12, 2026 · 02:58684 MB · Cloudflare R2
✓The hard parts stay visible
BackupSheep is designed to fail loudly, resume deliberately, and keep sensitive details out of the logs your team needs to read.
Transfer and native command exit status is checked. One failed file fails the run instead of archiving a gap.
Interrupted transfers continue, retries reuse the backup record, and concurrent runs for one node are serialized.
Database engines check available space against the expected dump before doing expensive work.
Secrets are encrypted at rest, kept out of process arguments, and redacted from backup and restore logs.
A schedule can require an air-gapped destination to validate successfully before the backup begins.
Amazon S3 destinations can apply Object Lock governance or compliance retention to every new archive.
More than a scheduler
See upcoming runs and attach keep-last retention to each protected source.
Route success, failure, restore, and storage-validation events to the right people.
Invite users, group them, scope nodes, and separate backup, schedule, integration, and storage rights.
Track actions, errors, acting users, and successful or failed sign-ins from one log.
Automate the same workflows your operators use in the UI.
Scale cloud, database, file, storage, and log work without starving the web console.
Open source today
Run the complete stack on a fresh Ubuntu 22.04+ or Debian 12+ server. The installer adds Docker and Compose, generates secrets, builds every service, and waits for a healthy app.
# Run on a fresh Ubuntu or Debian server
$ curl -fsSL https://raw.githubusercontent.com/
bilal414/backupsheep/main/install.sh | sudo bash
✓ Docker + Compose ready
✓ Application secrets generated
✓ Web, workers, scheduler, DB and broker healthy
Onboarding URL is ready →The self-hosted edition is the product available now. A managed edition may follow later; this release keeps the code, data path, and operating decisions in your hands.
Before you install
Yes. The self-hosted application is licensed under GNU GPLv3. It is currently marked beta, so evaluate and operate it with the care you would give any backup system.
The open-source edition runs in your environment. File and database archives go from your BackupSheep workers to the destinations you configure; provider-native snapshots stay with the cloud provider.
No. Websites and databases have console-driven restore workflows. Cloud servers and volumes are restored through the provider’s native snapshot workflow.
The quickest route is a fresh Ubuntu 22.04+ or Debian 12+ server. The installer provisions the Docker Compose stack, including PostgreSQL and RabbitMQ. Add HTTPS before public exposure.
Not currently. The product available today is the self-hosted open-source edition; a managed edition may be offered later.