Glossary
48 IT terms, explained clearly.
We explain the terms you will come across when dealing with data recovery, PC service or business IT. From HDD to ZFS, from BitLocker to NIS2 — in plain English, no marketing.
17 terms
Drives and storage
- HDD Hard Disk Drive
- A magnetic hard drive with spinning platters and moving read heads. The most common storage in personal computers until 2015, and still used in businesses today for high-capacity archives. Mechanically vulnerable to shocks, with a typical lifespan of 3–7 years in continuous operation. More →
- SSD Solid State Drive
- A drive with no mechanical parts, based on flash memory (NAND). Orders of magnitude faster than an HDD for random reads and writes. Lifespan is measured by the number of terabytes written (TBW). Resistant to shocks, but if the memory chips or controller fail, data recovery becomes difficult. More →
- NVMe Non-Volatile Memory Express
- A modern protocol for communicating with an SSD over the PCIe bus. 5–10× faster than SATA (for example PCIe 4.0 x4 = 7 GB/s). Now standard in laptops and servers. Available in M.2 or U.2 form factors. More →
- M.2
- A physical form factor — a small card for internal storage in laptops. Sized 22×30 / 42 / 60 / 80 / 110 mm. It can carry either the SATA or the NVMe protocol — confusing the two is a common mistake during an upgrade.
- SATA Serial ATA
- An interface for connecting drives (both HDD and SSD) — today typically SATA III at 600 MB/s. Gradually being replaced by NVMe in laptops, but still standard for large 2.5"/3.5" HDDs.
- CMR Conventional Magnetic Recording
- The traditional way of writing to an HDD platter — the tracks do not overlap. Robust, with predictable write performance. Preferred for RAID, NAS and archiving. Seagate IronWolf, WD Red Plus/Pro, Toshiba N300. More →
- SMR Shingled Magnetic Recording
- A newer way of writing to an HDD platter, where the tracks overlap like roof shingles. Higher recording density, but slow rewrites — a disaster in RAID arrays and NAS units with frequent writes. Risky for data recovery (more sensitive to a repeated write attempt when errors occur). More →
- S.M.A.R.T. Self-Monitoring, Analysis and Reporting Technology
- Built-in drive diagnostics that continuously report status (reallocated sector count, temperature, read errors, power-on hours). The attributes can be read with CrystalDiskInfo or `smartctl`. Warning values mean one thing: stop using the drive and back it up immediately.
- TRIM
- A command by which the operating system tells the SSD: "these LBA blocks are free, you may erase them at the flash level." It extends lifespan and maintains speed. After formatting, an SSD usually retains only metadata — which makes data recovery very demanding.
- Head crash
- Physical contact between the read head and the HDD platter — typically after a drop or shocks. It damages both the platter (a spot with no magnetisation) and the head. A clicking drive is the classic audible symptom. Every further start-up attempt makes things worse — the rule is "switch it off at once". More →
- Clean room ISO 14644-1
- A space with controlled air filtration (class 100 = < 100 particles ≥ 0.5 μm per cubic foot). Essential for opening an HDD — even the tiniest dust particle on a platter causes damage within a few rotations. At ITHOPE we operate our own clean room. More →
- Head comb
- A specialist tool used to safely remove the heads from an HDD. Without it, the heads collide with one another or with the platter. Each drive family needs a different one — Seagate F3, WD Enterprise and Toshiba MG all differ.
- Firmware
- A microprogram stored on the drive (outside the user-accessible area) that governs its behaviour — head calibration, sector remapping, ECC, power management. Faulty firmware means the drive does not respond normally, and recovery requires specialist PC-3000 hardware.
- Platter
- A glass or aluminium disc coated with a magnetic layer — the physical medium on which HDD data is stored. It spins at 5,400 or 7,200 rpm in consumer drives, and 10,000–15,000 rpm in enterprise drives.
- NAND
- The type of flash memory from which SSDs and USB flash drives are built. Variants: SLC (1 bit/cell, the most reliable, expensive), MLC, TLC, QLC (4 bits/cell, the cheapest, the most prone to wear).
- Chip-off recovery
- A recovery method in which the flash chips are physically desoldered from the board (USB flash drives, phones, damaged SSDs) and read with a specialist programmer. The data is in raw form and must be reconstructed using knowledge of the FTL (Flash Translation Layer).
- JTAG Joint Test Action Group
- A debug interface on microcontrollers and SoCs. In data recovery it is used to read the memory of smart devices without disassembly — typically older mobile phones and network devices.
7 terms
RAID and NAS
- RAID 0
- Striping data across 2+ drives — it increases performance, but if one drive fails you lose everything. There is no redundancy. Used for video editing, scratch disks and cache. Needlessly risky in the home. More →
- RAID 1
- Mirroring — the same data on 2 drives. Survives the failure of 1 drive. Capacity = 1 drive. The safest option for small servers and home NAS units with 2 bays. More →
- RAID 5
- Striping with parity — needs 3+ drives, survives the failure of 1 drive, capacity = (N-1) × drive size. Popular in SME environments. A rebuild after a failure takes hours to days, and a further failure during the rebuild means data loss. More →
- RAID 6
- Striping with double parity — 4+ drives, survives the simultaneous failure of 2 drives. Today the recommended choice for arrays over 8 TB. The rebuild is slower than RAID 5 but safer. More →
- RAID 10
- A stripe across mirror pairs — 4+ drives, the fastest RAID with redundancy. Survives the failure of 1 drive in each pair. Used for databases and virtualisation. More →
- NAS Network-Attached Storage
- A standalone network storage device (Synology, QNAP, Asustor, TerraMaster, WD My Cloud). It holds one or more drives in a RAID and exports them over SMB/NFS/iSCSI. It often uses proprietary RAID formats (Synology SHR, QNAP QTS) — data recovery requires knowledge of the specifics. More →
- SHR Synology Hybrid RAID
- Synology's proprietary RAID algorithm — it automatically makes use of space even on drives of different sizes. Under the hood it is a combination of mdraid + LVM + btrfs/ext4. A rebuild requires all 3 layers to be correctly assembled.
4 terms
Encryption
- BitLocker
- The native Windows tool for full-disk encryption — AES 128/256 XTS. The key is stored in the TPM chip and/or protected by a password. The recovery key (48 digits) is the only backup — without it, the encrypted data is mathematically inaccessible. More →
- FileVault
- Native encryption in macOS (since 10.7 Lion). XTS-AES-128 with a 256-bit key. The recovery key is a 24-character code. On macOS with a T2/M1/M2 chip, FileVault is always on.
- LUKS Linux Unified Key Setup
- The standard format for encrypted partitions in Linux (on top of dm-crypt). Supports multiple slots for passwords/keys. Without the passphrase or key file the data cannot be recovered — but the key can be stored on a USB token.
- TPM Trusted Platform Module
- A hardware chip in modern PCs and laptops (or firmware TPM — fTPM) that keeps encryption keys in secure storage. Windows 11 requires it. Replacing the motherboard invalidates the keys bound to the TPM — which is why the recovery key matters.
10 terms
Software and filesystems
- NTFS
- The Windows file system — journaled, with ACL permissions, compression, encryption (EFS), and support for large files and volumes. Standard for Windows partitions. Data recovery from NTFS is well supported thanks to the available documentation.
- APFS
- Apple File System (since macOS 10.13) — copy-on-write, snapshots, natively encrypted, optimised for SSDs. Recovery is more complex than HFS+ because of the split into containers and volumes.
- ext4
- The standard file system of most Linux distributions — journaled, supporting large volumes (up to 1 EiB) and extents. Stable and well documented.
- btrfs
- A modern Linux filesystem with copy-on-write, snapshots, native RAID 1/5/6 and compression. Used by Synology (DSM 6+) and Fedora. Recovery after metadata corruption requires specific tools.
- ZFS
- An enterprise-grade filesystem and volume manager in one — checksums on every block, self-healing on read, snapshots, compression and RAIDZ. TrueNAS, FreeBSD, Proxmox and Linux (ZFS on Linux).
- Bit rot
- The silent degradation of data on disks or tapes — individual bits flip without warning. Modern filesystems (ZFS, btrfs) detect it using checksums. After years of archiving without checks, some of the data may be invalid even though the drive reports "OK".
- Data scrubbing
- Regularly reading all data in a RAID/ZFS array and checking its consistency. It uncovers bit rot and bad sectors before they cause loss. Recommended once a month.
- Active Directory AD
- Microsoft's central directory service — it manages users, computers, groups and Group Policy within a domain. The backbone of corporate Windows environments. The cloud equivalent: Microsoft Entra ID (formerly Azure AD).
- Microsoft 365 M365
- Microsoft's cloud platform (Exchange Online, SharePoint, Teams, OneDrive, Office desktop/web). The default choice for businesses with 5–500 employees. It has no native long-term archiving — this must be added with a third party (Altaro, Acronis, Veeam).
- BIOS / UEFI
- The motherboard firmware — it initialises the hardware and loads the bootloader. UEFI (the newer one) supports GPT, Secure Boot and a graphical menu. A BIOS lock (supervisor password) complicates service — resetting it requires a specialist procedure or replacing a component.
5 terms
Networks and security
- VLAN Virtual LAN
- Logical separation of a network on the switch at layer L2. It allows traffic to be segmented (guest Wi-Fi, VoIP, IoT) without separate hardware. Necessary for NIS2 compliance in businesses.
- VPN Virtual Private Network
- An encrypted tunnel between networks, or between a client and a network. Uses: remote access (WireGuard, OpenVPN), linking branch offices (IPsec). The freely available WireGuard is now replacing older solutions.
- NIS2
- An EU cyber-security directive in force since 10/2024. It extends NIS1 to more sectors — it applies to companies with 50+ employees in the listed sectors. It requires organisational and technical measures (MFA, logging, incident response, audits). More →
- MFA Multi-Factor Authentication
- Sign-in verified by two or more factors (password + token/TOTP/biometrics). Today the minimum for corporate accounts — MFA stops 99% of attacks on compromised accounts.
- EDR Endpoint Detection and Response
- Advanced endpoint protection — telemetry, behavioural analysis, response playbooks. It replaces the classic "antivirus". Bitdefender GravityZone, Microsoft Defender for Endpoint, SentinelOne.
5 terms
General terms
- 3-2-1 backup
- The standard rule: 3 copies of your data, on 2 different types of media, with 1 copy off-site. It protects against hardware loss, ransomware and physical disasters. More →
- RPO Recovery Point Objective
- The maximum acceptable data loss over time — "how much data can we afford to lose?". It depends on how often you back up. RPO of 1 h = you back up every hour.
- RTO Recovery Time Objective
- The maximum acceptable downtime — "how quickly must we be back up?". It determines how much to invest in redundancy and a disaster recovery plan.
- DoD 5220.22-M
- The US standard for secure data erasure — a multi-pass overwrite. At ITHOPE we use it when returning rented laptops and after handing over recovered data.
- No recovery, no fee
- The billing principle in data recovery — if the data cannot be recovered, the client pays nothing (apart from any shipping). We are bound to it by one rule: there is no charge for diagnostics.
Missing something?
Have a specific technical question?
We are happy to advise — a free phone consultation or an email. We are here in person in Brno, and serve the whole country by courier.