Rclown
Friendly, self-hosted web UI for backing up object storage buckets, powered by rclone.

Canonical docs:
- https://rclown.com
- https://github.com/marckohlbrugge/rclown
- https://github.com/marckohlbrugge/rclown/blob/main/docs/docker-deployment.md
- https://github.com/marckohlbrugge/rclown/blob/main/docs/kamal-deployment.md
- https://github.com/marckohlbrugge/rclown/blob/main/docs/dokku-deployment.md
- https://github.com/marckohlbrugge/rclown/blob/main/docs/backup-retention.md

Primary use case:
- Mirror-style bucket-to-bucket backups between Cloudflare R2, Backblaze B2, Amazon S3, and other S3-compatible providers.
- Scheduled (daily/weekly) or manual runs, with live logs, run history, and email notifications on failure.

Quick setup (self-hosted):
1) Deploy Rclown with Docker (image includes Ruby, SQLite, and rclone):
   docker run -v rclown:/rails/storage \
     -e SECRET_KEY_BASE=$(openssl rand -hex 64) \
     -e HTTP_AUTH_USERNAME=admin \
     -e HTTP_AUTH_PASSWORD=your-password \
     -p 80:80 ghcr.io/marckohlbrugge/rclown:main
2) Add provider credentials (Cloudflare R2, Backblaze B2, or Amazon S3) in the web UI.
3) Let Rclown discover and import your buckets.
4) Create a backup: pick a source and destination bucket (optionally a sub-path), choose a daily or weekly schedule.
5) Watch live logs during runs; configure email notifications for failures.

Important implementation notes:
- Backups are mirror-style (rclone): the destination mirrors the source; no proprietary format, restore with any S3-compatible tool.
- Comparison modes: default (size + mtime), size-only, or checksum.
- Files deleted from the source are retained in the destination for a configurable period before cleanup.
- Data transfers run through your own server via rclone; credentials and files never touch third-party services.
- Single-user, single-server design protected by HTTP basic auth.