Description
The system still boots but is slow enough that you are thinking about a reinstall — or there are thirty machines in a lab that all need the same image. Doing it by hand costs a day; the commercial Ghost-style tools charge per seat. Clonezilla is the open-source answer: put it on a USB stick, boot into it, and take a whole disk or a single partition into an image you restore from the same stick when something breaks.
The clever part is that it copies **only used blocks**. Free space is skipped, so a 1 TB drive holding 200 GB images in the time and space of 200 GB, not 1 TB. Images can land on a local disk, an external drive, or straight onto SSH, Samba, NFS, WebDAV or S3. For mass deployment the SE edition multicasts one image to dozens of machines at once.
Whole disk or single partitions: savedisk and saveparts capture a drive or chosen partitions, restoredisk and restoreparts put them back — all picked from a menu rather than memorised as commands.
Used-block copying: it reads allocation for ext2/3/4, XFS, Btrfs, NTFS, FAT and HFS+ and skips free blocks; an unrecognised filesystem falls back to a sector-by-sector copy rather than refusing.
Images go anywhere: local and external drives plus SSH servers, Samba shares, NFS, WebDAV, AWS S3 and OpenStack Swift, so no local scratch space is required first.
Boots from USB or CD: shipped as ISO and ZIP — burn or PXE-boot the ISO, unzip the other onto a stick — so whether the machine's own OS still boots is irrelevant.
Multicast deployment: Clonezilla SE with DRBL pushes one image to dozens of machines simultaneously, which is the difference between an afternoon and a week in a computer lab.
Encryption and verification: images can be encrypted with ecryptfs and checked for integrity before a restore, instead of discovering a bad image halfway through.
Unattended runs: every choice can be preset as a boot parameter, so the stick boots and completes a backup or restore on its own — which is what makes a fixed recovery stick practical.
The clever part is that it copies **only used blocks**. Free space is skipped, so a 1 TB drive holding 200 GB images in the time and space of 200 GB, not 1 TB. Images can land on a local disk, an external drive, or straight onto SSH, Samba, NFS, WebDAV or S3. For mass deployment the SE edition multicasts one image to dozens of machines at once.
Features
Whole disk or single partitions: savedisk and saveparts capture a drive or chosen partitions, restoredisk and restoreparts put them back — all picked from a menu rather than memorised as commands.
Used-block copying: it reads allocation for ext2/3/4, XFS, Btrfs, NTFS, FAT and HFS+ and skips free blocks; an unrecognised filesystem falls back to a sector-by-sector copy rather than refusing.
Images go anywhere: local and external drives plus SSH servers, Samba shares, NFS, WebDAV, AWS S3 and OpenStack Swift, so no local scratch space is required first.
Boots from USB or CD: shipped as ISO and ZIP — burn or PXE-boot the ISO, unzip the other onto a stick — so whether the machine's own OS still boots is irrelevant.
Multicast deployment: Clonezilla SE with DRBL pushes one image to dozens of machines simultaneously, which is the difference between an afternoon and a week in a computer lab.
Encryption and verification: images can be encrypted with ecryptfs and checked for integrity before a restore, instead of discovering a bad image halfway through.
Unattended runs: every choice can be preset as a boot parameter, so the stick boots and completes a backup or restore on its own — which is what makes a fixed recovery stick practical.
