Data Recovery Glossary
Technical terms for storage, data recovery, security and systems, explained directly by HD Doctor engineers.
Active Directory (AD)
Active Directory (AD) is Microsoft's directory service that centralizes authentication, authorization and configuration of users, computers and resources in a corporate Windows network.
See definitionAES-256
AES-256 (Advanced Encryption Standard with 256-bit key) is the symmetric encryption standard adopted by NIST in 2001, used in virtually every modern security solution.
See definitionBad Blocks
Bad blocks are disk sectors that cannot be reliably read or written. They can be physical (hardware damage) or logical (data corruption or ECC failure).
See definitionClass 100 Cleanroom (ISO 5)
Controlled environment with at most 100 particles (β₯0.5 ΞΌm) per cubic foot of air, standardized as ISO 5. Minimum technical requirement to safely open an HDD without contaminating the platters.
See definitionext4
ext4 (fourth extended filesystem) is the default filesystem of most Linux distributions since 2008. Successor to ext3, adds extents, delayed allocation, journaling and supports volumes up to 1 exabyte.
See definitionGPT (GUID Partition Table)
GPT (GUID Partition Table) is the modern partition table standard (part of UEFI), replacing MBR. Supports disks larger than 2 TB, 128 partitions per disk and keeps a redundant header at both start and end of the disk.
See definitionHDD Firmware
HDD firmware is the embedded software inside the disk itself that controls the motor, head positioning, ECC, sector mapping and translation of logical address (LBA) to physical (CHS).
See definitionImmutable Backup
Immutable backup is a copy of data that cannot be altered, deleted or encrypted after writing, not even by administrators with maximum privilege, for a defined period.
See definitionInode
Inode (index node) is the metadata structure that represents an individual file in Unix-like filesystems (ext, XFS, JFS, ZFS, APFS).
See definitionIOPS
IOPS (Input/Output Operations Per Second) is the metric measuring how many read/write operations a storage device can execute per second.
See definitionJournaling (Filesystem)
Journaling is the technique used by modern filesystems to record (in a log called journal) intended changes before applying them to data blocks.
See definitionMBR (Master Boot Record)
MBR (Master Boot Record) is the legacy partition table occupying sector 0 of the disk. Holds boot code (446 bytes), up to 4 primary partition entries (64 bytes) and signature 0x55AA.
See definitionMTBF / AFR
MTBF (Mean Time Between Failures) is the statistical estimate of hours until a disk fails. AFR (Annualized Failure Rate) is the percent probability of failure in 1 year of continuous operation.
See definitionNAND Flash
NAND Flash is the non-volatile memory used in SSDs, USB drives, SD cards and embedded storage. Each cell stores 1 to 4 bits and has finite write cycles.
See definitionNTFS
NTFS (New Technology File System) is the default Microsoft Windows filesystem since 1993 (NT 3.1). Supports journaling, detailed ACLs, compression, EFS encryption and sparse files.
See definitionNVMe
NVMe (Non-Volatile Memory Express) is a storage protocol designed specifically for flash over the PCIe bus, with reduced latency and up to 65,535 parallel command queues.
See definitionPCB (controller board)
PCB (Printed Circuit Board) is the electronic board attached to the bottom of the HDD that contains the main CPU, DDR cache memory, spindle motor chip, head preamplifier and SATA/SAS interface.
See definitionRAID Controller
A RAID controller is the component (hardware or software) that manages data distribution (stripe), parity and mirroring across multiple physical disks to present a single volume to the operating system.
See definitionRansomware
Ransomware is a malware category that encrypts the victim's files and demands cryptocurrency payment to provide the decryption key.
See definitionRead-write head
The read-write head is the magnetic transducer that reads and writes data on HDD platters. It flies approximately 10 nanometers above the platter surface, sustained aerodynamically.
See definitionS.M.A.R.T.
SMART (Self-Monitoring, Analysis and Reporting Technology) is an internal disk telemetry standard reporting health, temperature, cycles and reallocated sectors.
See definitionService Area (HDD)
Service Area is a region reserved at the start or end of HDD platters, outside user-visible capacity, where the disk stores firmware modules, LBA translation tables, G-List and P-List.
See definitionSHA-256 Hash
SHA-256 (Secure Hash Algorithm 256-bit) is a cryptographic hash function that produces a 64-hex-character (256-bit) digest from any input, with computationally infeasible collision.
See definitionSnapshot
A snapshot is a point-in-time image of a volume, filesystem or VM, created via copy-on-write or redirect-on-write without duplicating unchanged data.
See definitionTranslator (HDD)
The Translator is the HDD firmware module that converts the logical address (LBA, Logical Block Address) used by the operating system to the actual physical address (PBA: cylinder, head, sector) where data lives on the platters.
See definitionTRIM
TRIM is an ATA command that tells the SSD controller which logical blocks are no longer used by the filesystem, allowing the controller to physically erase those cells in background.
See definitionVHDX (Hyper-V)
VHDX is Microsoft Hyper-V's virtual disk format (Windows Server 2012+), replacement for the legacy VHD. Supports up to 64 TB, 4KB blocks and internal protection log against power-loss corruption.
See definitionVMDK (Virtual Machine Disk)
VMDK (Virtual Machine Disk) is VMware's virtual disk format, used in vSphere ESXi, Workstation and Fusion. Can be thick, thin, sparse or descriptor + flat.
See definitionVolume Shadow Copy (VSS)
Volume Shadow Copy Service (VSS) is the native Windows infrastructure to create point-in-time snapshots of NTFS volumes without stopping running applications.
See definitionWear Leveling
Wear leveling is the SSD controller algorithm distributing writes across all NAND cells to prevent premature wear of specific blocks.
See definition