Features
HashBackup is designed to be easy to use, because if backup isn’t easy, it won’t happen at all!
Easy
-
Installs as a single static executable without dependencies or system config issues
-
Backup data is deduped, compressed, then encrypted locally with the AES encryption standard
-
Each backup is a complete snapshot for "point in time" restores
-
Full filesystem attribute support for accurate restores
-
Backs up to local and/or remote storage. Automatically synchronizes destinations & migrates backup data to new destinations
-
Easy, flexible file retention policies like last 7 days + 1 per month
-
Mount backup as a FUSE filesystem for easy, direct access to all versions. Mountable backup means no proprietary file-format lock in
-
Correctly handles deleted files
-
Backup admin can remove files & directories from the backup
-
Runs as root or regular userid
Secure
-
Backup key is created and stored locally only
-
Secure remove option removes confidential data
-
Backup admin can restrict commands with a password
-
Unalterable audit logs
-
Public key encryption option for write-only backups (readkey)
Reliable
-
Incremental selftest verifies all backup data over time
-
Sampled selftest verifies random blocks now or over time
-
Layers of checksums ensure data integrity:
-
SHA1 hash on each block
-
SHA1 hash on each file
-
Service-specific hash on file uploads & downloads
-
Capable
-
Scales from backing up /etc to multi-TB backups of millions of files. The largest reported backup is 53M files and 121TB, requiring 8GB of RAM. Sharding enables parallel backups of huge filesystems. Dedup is memory-efficient and usable even with multi-TB backups
-
Supports: sftp, scp, ftp & ftps, rsync, WebDAV, Amazon S3 and compatibles, Backblaze B2, USB thumb drives & removable hard drives, mounted remote storage (NFS, Samba/CIFS, sshfs, WebDAV, "cloud" drives)
-
Scriptable shell destination for storage not supported natively
-
Rclone destination for access to anything Rclone supports: Google Drive, Microsoft One Drive, Dropbox, etc.
Efficient
-
Backups are incremental for fast + small backups. Periodic full backups are an option
-
Incremental restore uses local file data to reduce downloads
-
Quick verify of remote data without downloading (dest verify)
-
Utilizes multiple CPU cores for fast backups and restores. All destination uploads and downloads are multithreaded
-
Backup to multiple destinations with one filesystem scan
-
Selective download retreives only the blocks needed (partial backup file download)
-
Extremely space-efficient incremental backups of virtual machine disk images: .vmdk, .hdd, .qcow, etc.
-
Backup and restore devices and LVM snapshots directly
-
Fifo backup directly saves program output and database dumps
-
Smart cache enables restoring large backups with minimal disk space
-
Efficiently handles "sparse" (thin-provisioned) files with OS support
-
Checkpoints allow restarting interrupted backups & restores