I rebuilt everything from scratch using scripts: ISO remastering in a chroot environment, a kernel driver built through DKMS, Linux kernel 7.0, a hybrid BIOS + UEFI ISO, and automated boot tests.
HDDLive is a live Linux system that boots from a USB drive, runs in RAM, and leaves the client’s PC untouched.
Here’s what it includes and what each tool does:
- Cloning failing drives: OpenSuperClone clones in phases, accesses the controller directly, and creates a virtual disk that reads only the sectors requested by the recovery tool. ddrescue copies data while skipping read errors and saves a map so the process can resume later.
- Recovering files and partitions: TestDisk rebuilds lost partitions; PhotoRec recovers files by signature, even after formatting; DMDE, R-Studio, R-Linux, and UFS Explorer reconstruct file systems and RAID arrays.
- Diagnostics: smartmontools, nvme-cli, hdparm, and Hard Disk Sentinel reveal the drive’s actual health before any decisions are made.
- RAID and encryption: mdadm, LVM, cryptsetup (LUKS), and dislocker (BitLocker) open volumes in read-only mode.
- Forensics: Sleuth Kit and bulk_extractor scan entire disk images for deleted files, emails, and metadata.

The screenshot, shows pt-br interface, but dispiste the screen, the program is 100% english

Source: https://github.com/hawkinf/hddlive-osc
Download: https://sourceforge.net/projects/hawkinf/

thank you

That's a really good idea. Awesome.

The RAM-only approach makes sense for cloning failing drives, since you're not writing to the suspect disk at all. Curious about the OpenSuperClone phased cloning: is the controller-level access something that needs specific hardware support, or does it work through standard tools on most boards? Also, how large does the ISO end up with Sleuth Kit, bulk_extractor, and the diagnostic suite all bundled? Keeping the pt-br interface alongside English seems like a practical choice for wider use.

Curious about the choices here, especially OpenSuperClone being used for phased cloning. On failing drives I'd worry that a live-in-RAM system plus USB I/O adds overhead; does the ramdisk setup leave enough headroom for imaging large disks? Also, with TestDisk and PhotoRec bundled alongside Sleuth Kit and bulk_extractor, is there a recommended order so you don't overwrite evidence before imaging? The pt-br interface looks clean from the screenshot. How did you verify the ISO remastering in the chroot was reproducible?

The detail about running cloning in phases sounds sensible: a drive that is already struggling would not cope well with a single continuous pass. I also like that you included nvme-cli and dislocker; far too many recovery images stop at SATA-era tools. Since the interface screenshot is in pt-br, I am curious whether the scripts handle locale and keyboard setup cleanly when booting on different hardware, because that often causes problems in remastered ISOs. Running entirely in RAM is good for working on a failing drive without mounting anything.