Data recovery is the process of salvaging data from damaged, failed, corrupted, or inaccessible secondary storage media when it cannot be accessed normally. Often the data are being salvaged from storage media such as hard disk drives, storage tapes, CDs, DVDs, RAID, and other electronics. Recovery may be required due to physical damage to the storage device or logical damage to the file system that prevents it from being mounted by the host operating system.

The most common “data recovery” issue involves an operating system (OS) failure (typically on a single-disk, single-partition, single-OS system), where the goal is to simply copy all wanted files to another disk. This can be easily accomplished with a Live CD, most of which provide a means to 1) mount the system drive, 2) mount and backup disk or media drives, and 3) move the files from the system to the backup with a file manager or optical disc authoring software. Further, such cases can be mitigated by disk partitioning and consistently moving valuable data files to a different partition from the replaceable OS system files.

The second type involves a disk-level failure such as a compromised file system, disk partition, or a hard disk failure —in each of which the data cannot be easily read. Depending on the case, solutions involve repairing the file system, partition table or MBR, or hard disk recovery techniques ranging from software-based recovery of corrupted data to hardware replacement on a physically damaged disk. These last two typically indicate the permanent failure of the disk, thus “recovery” means sufficient repair for a one-time recovery of files.

A third type involves the process of retrieving files that have been “deleted” from a storage media, since the files are usually not erased in any way but are merely deleted from the directory listings.

Although there is some confusion as to the term, the term “data recovery” may be used to refer to such cases in the context of forensic purposes or spying.

Recovering data after physical damage

A wide variety of failures can cause physical damage to storage media. CD-ROMs can have their metallic substrate or dye layer scratched off; hard disks can suffer any of several mechanical failures, such as head crashes and failed motors; tapes can simply break. Physical damage always causes at least some data loss, and in many cases the logical structures of the file system are damaged as well. This causes logical damage that must be dealt with before any files can be salvaged from the failed media.

Most physical damage cannot be repaired by end users. For example, opening a hard disk in a normal environment can allow airborne dust to settle on the platter and become caught between the platter and the read/write head, causing new head crashes that further damage the platter and thus compromise the recovery process. Furthermore, end users generally do not have the hardware or technical expertise required to make these repairs. Consequently, costly data recovery companies are often employed to salvage important data. These firms often use “Class 100” / ISO-5 cleanroom facilities to protect the media while repairs are being made. (Any data recovery firm without a pass certificate of ISO-5 or better will not be accepted by hard drive manufacturers for warranty purposes

Recovery techniques

Recovering data from physically-damaged hardware can involve multiple techniques. Some damage can be repaired by replacing parts in the hard disk. This alone may make the disk usable, but there may still be logical damage. A specialized disk-imaging procedure is used to recover every readable bit from the surface. Once this image is acquired and saved on a reliable medium, the image can be safely analysed for logical damage and will possibly allow for much of the original file system to be reconstructed.

Hardware repair

Examples of physical recovery procedures are: removing a damaged PCB (printed circuit board) and replacing it with a matching PCB from a healthy drive, performing a live PCB swap (in which the System Area of the HDD is damaged on the target drive which is then instead read from the donor drive, the PCB then disconnected while still under power and transferred to the target drive), read/write head assembly with matching parts from a healthy drive, removing the hard disk platters from the original damaged drive and installing them into a healthy drive, and often a combination of all of these procedures. Some data recovery companies have procedures that are highly technical in nature and are not recommended for an untrained individual. Any of them will almost certainly void the manufacturer’s warranty.

Disk imaging

The extracted raw image can be used to reconstruct usable data after any logical damage has been repaired. Once that is complete, the files may be in usable form although recovery is often incomplete.

Open source tools such as DCFLdd or DOS tools such as HDClone can usually recover data from all but the physically-damaged sectors. Studies have shown that DCFLdd v1.3.4-1 installed on a Linux 2.4 Kernel system produces extra “bad sectors” when executed with certain parameters, resulting in the loss of information that is actually available. These studies state that when installed on a FreeBSD Kernel system, only the bad sectors are lost. DC3dd, a tool that has superseded DCFLdd, and ddrescue resolve this issue by accessing the hardware directly. Another tool that can correctly image damaged media is ILook IXImager.

Typically, Hard Disk Drive data recovery imaging has the following abilities: (1) Communicating with the hard drive by bypassing the BIOS and operating system which are very limited in their abilities to deal with drives that have “bad sectors” or take a long time to read. (2) Reading data from “bad sectors” rather than skipping them (by using various read commands and ECC to recreate damaged data). (3) Handling issues caused by unstable drives, such as resetting/repowering the drive when it stops responding or skipping sectors that take too long to read (read instability can be caused by minute mechanical wear and other issues). and (4) Pre-configuring drives by disabling certain features, such as SMART and G-List re-mapping, to minimize imaging time and the possibility of further drive degradation.