Glossary

ACCESS TIME:

Of a hard disk, the combined time for seek, head switch, and rotational latency in a read/write operation. Of an optical disk, the time it takes to access a data track and begin transferring data. In an optical jukebox, it's the time it takes to locate a specific disk, insert it in an optical drive, and begin transferring data to the host system. If the disk is already in the drive, then access time is determined by seek time. Otherwise , it's determined by disk swap time, spin-up time, and seek time.

ACCESS:

Refers to the process of obtaining data from, or placing data into a disk storage device, register, or RAM. (i.e., accessing a memory location).

ACCESS FREQUENCY BASED DATA MIGRATION:

A strategy for migrating data between storage platforms based on frequency of access rather than by the date that the data was last modified. This strategy is superior to HIERARCHICAL STORAGE MANAGEMENT (HSM) from a storage capacity optimization perspective.

ACTUATOR ARM:

A mechanical device used to carry all read/write heads in a multi-platter disk drive.

ACTUATOR:

An electro-mechanical device that moves an object, such as the robotic arm that moves an optical disk within the jukebox, or the device that controls the read/write head on a disk drive.

ADAPTIVE CACHING:

Technology that allows the drive to tune the cache (number of segments and segment size) to best suit the system's needs.

ADDRESS MARK:

Two byte address at the beginning of both the ID field and the data field of the track format. The first byte is the "A1" data pattern, the second byte is used to specify either an ID field or a data field.

ADDRESS:

(physical) A specific location in memory where a unit record, or sector, of data is stored. To return to the same area on the disk, each area is given a unique address consisting of three components : cylinder, sector, and head. CYLINDER ADDRESSING is accomplished by assigning numbers to the disk's surface concentric circles ( cylinders ). The cylinder number specifies the radial address component of the data area. SECTOR ADDRESSING is accomplished by numbering the data records (sectors) from an index that defines the reference angular position of the disks. Index records are then counted by reading their ADDRESS MARKS . Finally, HEAD ADDRESSING is accomplished by vertically numbering the disk surfaces, usually starting with the bottom-most disk data surface. For example, the controller might send the binary equivalent of the decimal number 610150 to instruct the drive to access data at cylinder 610, sector 15, and head 0.

ADJUSTABLE INTERLEAVE :

Interleaving permits access to more than one memory module, e.g., if one memory module contains odd-numbered address and another even-numbered address, they can both be accessed simultaneously for storage. If the interleave is adjustable, the user may select which ranges or areas are to be accessed each time.

ADVANCED INTELLIGENT TAPE ( AIT ) :

A helical scan technology developed by Sony for tape backup/archive of networks and servers, specifically addressing midrange to high-end backup requirements.

AFC (Anti-Ferromagnetic-Coupled) MEDIA:

See [PIXIE DUST]

ALLOCATION UNIT:

A group of sectors on the disk reserved for specified information. On hard drives for small computer systems, the allocation unit is usually in the form of a block, cluster, or sector.

See also [CLUSTER]

See also [BLOCK]

See also [SECTOR]

ALLOCATION:

The process of assigning particular areas of the disk to specific data or instructions.

See also [ALLOCATION UNIT]

AMERICAN NATIONAL STANDARDS INSTITUTE ( ANSI ) :

A standard-setting, independent organization that develops and publishes manufacturing and design standards for the United States.

AMERICAN STANDARD FOR CODED INFORMATION INTERCHANGE:

ASCII.

ANSI:

See [AMERICAN NATIONAL STANDARDS INSTITUTE]

APPLICATION PROGRAM:

A sequence of programmed instructions that tell the computer how to perform an end use task (i.e., accounting, word processing, or other work for the computer system user). To use a program, it must first be loaded into MAIN MEMORY from some AUXILIARY MEMORY such as a floppy diskette or hard disk.

ARCHIVAL MANAGEMENT:

A storage management solution for cataloging files and moving them to long- term storage, where they can be stored and accessed inexpensively.

ARCHIVE:

A copy of reference data or document images that are stored on optical disks, floppies, tape, paper, or microfiche. Typically refers to long-term storage of data for later possible access.

AREAL DENSITY:

Bit density (bits per inch, or BPI) multiplied by track density (tracks per inch, or TPI), or bits per square inch of the disk surface. Bit density is measured around a track (circumferential on the disk), and track density is radially measured.

ARRAY:

A group of disk drives which have been combined into a common array and appear as a single LSU (Logical Storage Unit).

See also [DISK ARRAY]

ASCII:

See [AMERICAN STANDARD FOR CODED INFORMATION INTERCHANGE]

ASYNCHRONOUS DATA:

Data sent usually in parallel mode without a clock pulse. Time intervals between transmitted bits may be of unequal lengths.

ATA (AT Attachment):

This term defines the signal and logical protocol described in X3.221 for IDE (Integrated Drive Electronics) peripherals.

See also [INTERFACE]

ATAPI:

See [AT ATTACHMENT PACKET INTERFACE]

AT ATTACHMENT PACKET INTERFACE ( ATAPI ) :

A command protocol used for accessing ATA (IDE) peripheral devices. Widely used on CD-ROM and Tape Backup units attached to ATA bus.

AT BUS:

An acronym representing Advanced Technology bus. The standard PC compatable peripheral bus to which video cards, I/O cards, internal modem cards, and sound cards are added. Also called the ISA bus, it runs at a maximum of 8.33 MHz and has a 16-bit wide data path .

ATOMIC RESOLUTION STORAGE:

A technology derived from atomic probe microscopy in which a storage medium moves along under tiny probe tips that emit lasers no wider than one atom. To read already-recorded patches of medium, the probe tips produce a weaker, even more sensitive, laser.

AUTOLOADER:

A single-drive, tape-based backup device that houses a number of tape cartridges. An autoloader is designed to support routine, automatic backup procedures, using a mechanical arm to sequentially load a new tape for daily backup.

AUXILIARY MEMORY:

Memory other than main memory; generally a mass storage subsystem, it can include disk drives, backup tape drives, controllers, and buffer memory. Typically, AUXILIARY MEMORY is non-volatile.

AUXILIARY STORAGE DEVICE:

Devices, generally magnetic tape and magnetic disk, on which data can be stored for use by computer programs. Also known as secondary (in a two tier storage configuration) or tertiary (in a three tier storage configuration) storage.

AVERAGE ACCESS TIME:

The average track access time, calculated from the end of the CONTROLLER commands to access a drive, to drive "seek complete" time averaged over all possible track locations at the start of ACCESS, and over all possible data track ADDRESSES .

AVERAGE LATENCY:

The average time required for any byte of data stored on a disk to rotate under the disk drive's read/write head. Equal to one half the time required for a single rotation of the platter.

AVERAGE SEEK TIME:

The average time it takes for the read/write head to move to a specific location. Calculated by dividing the time it takes to complete a large number of random seeks by the number of seeks performed.

BACKUP DEVICE:

Disk or tape drive used with a fixed Winchester disk drive to make copies of files or other data for off line storage, distribution, or protection against accidental data deletion from the Winchester drive, or against drive failure.

BACKUP FILE:

File copies made on another removable media device (disk, tape, or sometimes a remote hard disk system) and kept to ensure recovery of data lost due to equipment failure, human errors, updates, and disasters.

BACKUP:

1. A duplicate copy of a program, disk, or data files. 2. A procedure for duplicating key data files, often automatically, and storing them in a safe place for the purpose of file recovery.

BACKWARD COMPATIBILITY:

A design standard that assures that new software, hardware, devices, and media will be compatible with earlier versions.

BAD BLOCK:

A block (usually the size of a sector) that cannot reliably hold data due to a physical flaw or damaged format markings .

BAD TRACK TABLE:

A list affixed to the casing of a hard disk drive that states which tracks are flawed and cannot hold data. This list of bad tracks is entered into the low-level formatting program when the drive is formatted at the factory.

BANDWIDTH:

The amount of data that can be transmitted via a given communications channel (e.g., between a hard drive and the host PC) in a given unit of time.

BASE CASTING:

The rigid structure which holds the mechanical sub-assemblies of a hard disk drive. Together with the top cover, creates an airtight, extremely clean enclosure.

BI-DIRECTIONAL BUS:

A bus that may carry information in either direction but not in both simultaneously.

BINARY:

A number system like the decimal numbers, but using 2 as its base and having only the two digits 0 (zero) and 1 (one). It is used in computers because digital logic can only determine one of two states ”"OFF" and "ON." Digital data is equivalent to a binary number.

BIOS (BASIC INPUT OUTPUT SYSTEM):

A program permanently stored in the memory of the computer and is available without an operating system disk. For example, it performs the internal self test of the computer and searches for the operating system on the disk drive.

BIT CELL LENGTH:

Physical dimension of the bit cell in direction of recording along the disk circumference of a track.

BIT CELL TIME:

The time required to pass one bit of information between the controller and the drive. Cell time is the inverse of the drive's data rate.

BIT DENSITY:

Expressed as "BPI" (for bits per inch), bit density defines how many bits can be written onto one inch of a track on a disk surface. It is usually specified for "worst case," which is the inner track. Data is the densest in the inner tracks where track circumferences are the smallest.

BIT JITTER:

The time difference between the leading edge of read and the center of the data window.

BIT SHIFT:

A data recording effect, which results when adjacent 1's written on magnetic disks repel each other. The "worst case" is at the inner cylinder where bits are closest together. BIT SHIFT is also called pulse crowding.

BIT:

An abbreviation for a binary digit which can be either 0 or 1. A bit is the basic data unit of all digital computers. It is usually part of a data byte or word, but bits may be used singly to control or read logic ON/OFF functions. A bit is a single digit in a binary number. Bits are the basic unit of information capacity or a computer storage device. Eight bits equal one byte.

BIT:

The smallest unit of data. Consists of a single binary digit that can take the value of 0 or 1.

BLOCK:

A group of BYTES handled, stored, and accessed as a logical data unit, such as an individual file record. Typically, one block of data is stored as one physical sector of data on a disk drive. In UNIX workstation environments, the smallest contiguous area that can be allocated for the storage of data. UNIX blocks are generally 8 KB (16 sectors) in size. In DOS environments, the block is referred to as a cluster. (Note: This usage of the term block at the operating system level is different from its meaning in relation to the physical configuration of the hard drive.)

See also [CLUSTER]

See also [SECTOR]

BLOCK ERROR RATE:

The average number of errors that occur (or can occur) while writing or transmitting a block of data.

BOOT:

Short for bootstrap. Transfer of a disk operating system program from storage on diskette or hard disk drive to computer's working memory.

BPI (Bits per inch):

A measure of how densely information is packed on a storage medium.

See also [FCI]

BPSI (Bits per square inch):

A measure of areal density calculated by multiplying bits per inch (BPI) by tracks per inch (TPI).

BUFFER:

A temporary data storage area that compensates for a difference in data transfer rates and/or data processing rates between sender and receiver.

BURST MODE:

A temporary, high-speed data transfer mode that can transfer data at significantly higher rates than would normally be achieved with non-burst technology; the maximum throughput a device is capable of transferring data.

BUS:

A length of parallel conductors that forms a major interconnection route between the computer system CPU and its peripheral subsystems. Depending on its design, a bus may carry data to and from peripheral's addresses, power, and other related signals.

BUS MASTERING:

A method of data transfer which allows data to be moved between a peripheral controller and system memory without interaction with the host CPU or a third party DMA controller. This technique allows the peripheral controller to take control of the system bus, and in the case of EISA, to move data at up to 33MB/s.

BYTE:

A sequence of adjacent BINARY digits or BITS considered as a unit, 8 bits in length. One byte is sufficient to define all the alphanumeric characters . There are 8 BITS in 1 BYTE. The storage capacity of a disk drive is commonly measured in MEGABYTES, which is the total number of bits storable , divided by eight million.

CACHE:

A temporary storage location, usually Random Access Memory (RAM), for data. In input/output operations, the organization of the cache is important because it enables time-saving functions such as read-ahead.

CACHE HIT:

This occurs when the data requested is already in the cache. A cache hit saves the time of getting the data from the rotating disk; the seek, latency, and read times.

CACHE MEMORY:

A portion of RAM allocated for storing frequently accessed information from a storage device.

CAPACITY:

Amount of memory (measured in megabytes) that can be stored in a disk drive. Usually given as formatted capacity.

See also [FORMAT OPERATION]

See also [FORMATTED CAPACITY]

CAPACITY ALLOCATION EFFICIENCY:

The automated provisioning of storage capacity to applications based on thresholds and policies.

CAPACITY UTILIZATION EFFICIENCY:

The automated provisioning of data to storage platforms based on data access and retention characteristics and platform cost.

CARRIAGE ASSEMBLY:

Assembly in a hard disk drive which holds read/write heads and roller bearings. It is used to position the heads radially by the actuator, in order to access a track of data.

CD-ROM (Compact Disk Read Only Memory):

A read only storage device which retrieves up to 660 Mbytes of information from a removable laser disk similar to an audio compact disk.

CENTRAL PROCESSOR UNIT ( CPU ) :

The heart of the computer system that executes programmed instructions. It includes the arithmetic logic unit (ALU) for performing all math and logic operations, a control section for interpreting and executing instructions, fast main memory for temporary (VOLATILE) storage of an application program and its data.

CHANNEL:

In regard to disk drives, a channel is an electrical path for the transfer of data and control information between a disk and a disk controller. The Primary and Secondary Hard Drive Port Addresses are an example to two channels.

CHARACTER:

An information symbol used to denote a number, letter, symbol or punctuation mark stored by a computer. In a computer a character can be represented in one (1) byte or eight (8) bits of data. There are 256 different one-byte binary numbers, sufficient for 26 lower case alphas, 26 upper case alphas, 10 decimal digits, control codes, and error checks.

CHIP:

An integrated circuit fabricated on a chip of silicon or other semiconductor material, e.g., a CHIP is an integrated circuit, a microprocessor, memory device, or a digital logic device.

CIFS:

See [COMMON INTERNET FILE SYSTEM]

CLIENT:

Typically, a desktop computer hooked up to a network, and designed to work with a more powerful server that runs applications and stores data.

CLIENT/SERVER:

An environment that allows interactions between " clients " (typically desktop computers) and "servers" (computers that store data and run software programs). In client-server environments, data may be stored on a remote server rather than a computer's hard disk; applications may be stored on a server and delivered to individual desktop computers as needed. The server acts as a gateway to the network, running administrative software controls, and providing access to the network and its resources.

CLOCK RATE:

The rate at which bits or words are transferred between internal elements of a computer or to another computer.

CLOSED LOOP:

A control system consisting of one or more feedback control loops in which functions of the controlled signals are combined with functions of the command to maintain prescribed relationships between the commands and the controlled signals. This control technique allows the head actuator system to detect and correct off-track errors. The actual head position is monitored and compared to the ideal track position, by reference information either recorded on a dedicated servo surface, or embedded in the inter-sector gaps. A position error is used to produce a correction signal (FEEDBACK) to the actuator to correct the error.

See also [TRACK FOLLOWING SERVO]

CLUSTER SIZE:

Purely an operating system function or term describing the number of sectors that the operating system allocates each time disk space is needed. In DOS environments, the smallest contiguous area that can be allocated for the storage of data. DOS clusters are usually 2 KB (4 sectors) in size.

CLUSTERED SERVERS:

The concept of combining multiple host computers together through a private communication line, such as Ethernet backbone, to form a ring of host computers; this ring of host computers act as a single entity, capable of performing multiple complex instructions by distributing the workload across all members of the ring.

CLUSTERED STORAGE:

The concept of combining multiple storage servers or intelligent storage devices together to form a redundant ring of storage devices; clustered storage systems typically perform multiple read and write requests through parallel access lines to the requesting computer.

CODE:

A set of unambiguous rules specifying the way which digital data is represented physically, as magnetized bits, on a disk drive. One of the objectives of coding is to add timing data for use in data reading.

See also [DATA SEPARATOR]

See also [MFM]

See also [RLL]

COERCIVITY:

A measurement in units of orsteads of the amount of magnetic energy to switch or "coerce" the flux change (di-pole) in the magnetic recording media.

COMMAND:

(1) An instruction sent by the central processor unit (CPU) to a controller for execution. (2) English-like commands entered by users to select computer programs or functions. (3) A CPU command, which is a single instruction such as "add two binary numbers" or "output a byte to the display screen."

COMMAND DESCRIPTOR BLOCK ( CDB ) :

SCSI commands are issued from an initiator by transferring a Command Descriptor Block to the target device. For some commands, a parameter list sent during a Data Out phase accompanies the request. A CDB contains an opcode, logical unit number, set of command parameters, and control byte.

COMMAND OVERHEAD:

See [OVERHEAD]

COMMON ACCESS METHOD ( CAM ) :

Defines a set of software and hardware interfaces which attempt to standardize an operating system's access to peripheral devices.

COMMON INTERNET FILE SYSTEM ( CIFS ) :

A proposed standard protocol that lets programs make requests for files and services on remote computers on the Internet or across any TCP/IP network. CIFS uses the client/server programming model. A client program makes a request of a server program (usually in another computer) for access to a file or to pass a message to a program that runs in the server computer. The server takes the requested action and returns a response. CIFS is a public or open variation of the Server Message Block (SMB) protocol developed and used by Microsoft. The SMB protocol is widely used in today's local area networks for server file access and printing.

COMPACT DISK-READ ONLY MEMORY ( CD-ROM ) :

An optical disk recording format in which the optical disk carries pre-recorded data, music or software. Users cannot add or delete data to a CD-ROM.

COMPACT DISK-RECORDABLE ( CD-R ) :

An optical disk recording format that allows data to be written to optical disks. The disks can be recorded just once, but played virtually without limit.

COMPACT DISK-REWRITABLE ( CD-RW ) :

An optical disk recording format that allows disks to be recorded and re-recorded, much like floppy disks or audio tapes. The disks can be rewritten up to 10,000 times and played virtually without limit.

COMPUTER OUTPUT TO LASER DISK ( COLD ) :

An optical storage technology for transferring computer-based information to an optical disk for near-online storage. Typically used as an alternative to paper or microfiche-based storage of computer-generated reports .

CONSOLE (also called CRT or Terminal):

A device from which a computer can be operated; often includes a monitor and keyboard.

CONTROLLER:

A controller is a printed circuit board required to interpret data access commands from host computer (via a BUS), and send track seeking, read/write, and other control signals to a disk drive. The computer is free to perform other tasks until the controller signals DATA READY for transfer via the CPU BUS.

COST OF OWNERSHIP:

The purchase price of equipment plus the cost of operating this equipment over its projected life span.

CPU:

See [CENTRAL PROCESSOR UNIT]

CRASH:

A malfunction in the computer hardware or software, usually causing loss of data.

See also [HEAD CRASH]

CRC:

See [CYCLIC-REDUNDANCY-CHECK]

CYCLIC-REDUNDANCY-CHECK ( CRC ) :

Used to verify data block integrity. In a typical scheme, 2 CRC bytes are added to each user data block. The 2 bytes are computed from the user data, by digital logical chips. The mathematical model is polynomials with binary coefficients. When reading back data, the CRC bytes are read and compared to new CRC bytes computed from the read back block to detect a read error. The read back error check process is mathematically equivalent to dividing the read block, including its CRC, by a binomial polynomial. If the division remainder is zero, the data is error free.

CYLINDER:

The cylindrical surface formed by identical track numbers on vertically stacked disks. At any location of the head positioning arm, all tracks under all heads are the cylinder. Cylinder number is one of the three address components required to find a specific ADDRESS , the other two being head number and sector number.

DAFS:

See [DIRECT ACCESS FILE SYSTEM]

DAISY CHAIN:

A way of connecting multiple drives to one controller. The controller drive select signal is routed serially through the drives, and is intercepted by the drive whose number matches.

DAT:

See [DIGITAL AUDIO TAPE]

DATA:

Information processed by a computer, stored in memory, or fed into a computer.

DATA ACCESS TIME:

See [ACCESS TIME]

DATA ACCESS:

When the controller has specified all three components of the sector address to the drive, the ID field of the sector brought under the head by the drive is read and compared with the address of the target sector. A match enables access to the data field of the sector.

DATA ADDRESS:

To return to the same area on the disk, each area is given a unique address consisting of the three components: cylinder, head, and sector. HORIZONTAL: accomplished by assigning numbers to the concentric circles (cylinders) mapped out by the heads as the positioning arm is stepped radially across the surface, starting with 0 for the outermost circle. By specifying the cylinder number the controller specifies a horizontal or radial address component of the data area. ROTATIONAL: once a head and cylinder have been addressed, the desired sector around the selected track of the selected surface is found by counting address marks from the index pulse of the track. Remember that each track starts with an index pulse and each sector starts with an address mark. VERTICAL: assume a disk pack with six surfaces, each with its own read/write head, vertical addressing is accomplished by assigning the numbers 00 through XX to the heads, in consecutive order. By specifying the head number, the controller specifies the vertical address component of the data area.

DATA BASE MANAGEMENT SYSTEM ( DBMS ) :

Application program used to manage, access, and update files in a data base.

DATA COMPRESSION:

An automatic utility that reduces the size of a data file by removing redundant bits of information. An algorithm built into the hardware, firmware, or software handles compression and decompression .

DATA ENCODING:

To use a code such as GCR, MFM, RLL, NZR, etc. to represent characters for memory storage.

DATA FIELD:

The portion of a sector used to store the user's DIGITAL data. Other fields in each sector include ID, SYNC, and CRC, which are used to locate the correct data field.

DATA MIGRATION:

See [HIERARCHICAL STORAGE MANAGEMENT (HSM)]

DATA SEPARATOR:

The circuit that extracts data from timing information on drives that store a combined data and clock signal.

DATA TRACK:

Any of the circular tracks magnetized by the recording head during data storage.

DATA TRANSFER:

The movement of data from one point to another within a computer system, for example, from an optical disk to a computer's hard disk.

DATA TRANSFER RATE ( DTR ) :

Speed at which bits are sent: In a disk storage system, the communication is between CPU and controller, plus controller and the disk drive. Typical units are bits per second (BPS), or bytes per second. I/O transfer rate is the data rate between the drive and the CPU. Internal transfer rate is the rate data is written to/from the disk.

See also [TRANSFER RATE]

DATA WAREHOUSE:

A large centralized database designed to hold and manage a company's information over a long period of time. Data ware houses are often used to mine key data for reference, for example, to detect trends, spot new market opportunities, and monitor business results.

DECREASE THE FLYING HEIGHT:

A method for increasing areal density. Since the head core is closer to the media surface, the lines of flux magnetize a smaller area. Thus, more bits can be recorded in a given distance, and higher BPI (bits per inch) is achievable.

DEFAULT:

A particular value of a variable that is used by a computer unless specifically changed, usually via an entry made through a software program.

DENSITY:

Generally, bit recording density.

See also [AREAL]

See also [BIT]

See also [STORAGE DENSITY]

DFHSM (Data Facility Hierarchical Storage Manager):

IBM's hierarchical storage management system, part of DFSMS.

DFSMS (Data Facility Storage Management System):

Data management, backup, and HSM software from IBM for MVS and OS/390 mainframes. Introduced in 1993, it combines separate backup, copy, HSM, and device driver routines into one package, which provides all the I/O management for the operating system.

DIGITAL:

Any system that processes digital binary signals having only the values of a 1 or 0. An example of a non-digital signal is an analog signal which continuously varies, e.g., TV or audio.

DIGITAL AUDIO TAPE ( DAT ) :

A storage technology that uses 4mm tape to record data. DAT is similar to an audio tape, but instead of recording data linearly along the length of the tape, data is recorded at an angle. This recording format, called DDS, is the industry standard for all DAT devices.

DIGITAL DATA STORAGE ( DDS ) :

A recording format used by all major DAT drive and media manufacturers, and the only recognized industry standard for DAT systems. A number frequently follows the DDS designation to indicate the generation of the standard: for example, DDS-3 represents a third-generation product.

DIGITAL LINEAR TAPE ( DLT ) :

A serpentine technology first introduced by Digital Equipment Corporation and later developed by Quantum for tape backup/archive of networks and servers; DLT technology addresses midrange to high-end tape backup requirements.

DIGITAL VIDEO DISK ( DVD ) :

A disk that closely resembles a standard CD in size, color , and physical format, but holds about seven times as much data. A typical CD holds about 650 MB of data, whereas today's DVDs hold 4.7 GB, with a targe capacity of about 17 GB in the future. A two- hour feature-length movie can fit on a DVD, making it an attractive medium for the entertainment industry as well as PC makers . The current state of DVD technology, allowing play-back but not recording on DVDs. Multiple DVD-RAM and DVD-Rewritable standards exist to support both play-back and recording on DVDs. At this time, no standards have been adopted industry-wide for recordable/rewritable DVD.

DIRECT ACCESS:

Generally refers to an AUXILIARY MEMORY device, having all data on-line. A tape drive without a tape mounted is not direct access, but a WINCHESTER DRIVE is direct access device.

DIRECT ACCESS FILE SYSTEM ( DAFS ) :

A high-performance file sharing protocol based on the VI memory-to-memory architecture. DAFS provides bulk data transfer directly between the application buffers of two machines without having to packetize the data. It also allows applications to access hardware without operating system intervention.

DIRECTORY:

A special disk storage area (usually cylinder zero) that is read by a computer operating system to determine the ADDRESSES of the data records that form a DISK FILE .

DIRECT MEMORY ACCESS ( DMA ) :

A means of data transfer between peripheral and host memory without processor intervention. DMA improves speed and efficiency by allowing the system to continue processing even while it is retrieving new data from the drive.

DIRTY CACHE:

A cache page in which data has been written or modified but which has not yet been copied to the storage device. Once the data has been copied to disk, the page is said to be clean.

DISK ARRAY (or ARRAY):

A linked group of small, independent hard disk drives used to replace larger, single disk drive systems. The most common disk arrays implement RAID (redundant array of independent disks) technology.

See also [RAID]

DISK CONTROLLER:

The chip or circuit that controls the transfer of data between the disk and buffer.

See also [DISK DRIVE CONTROLLER]

See also [INTERFACE CONTROLLER]

DISK DRIVE CONTROLLER:

The hard disk drive controller electronics, which include the disk controller and the interface controller.

See also [DISK CONTROLLER]

See also [INTERFACE CONTROLLER]

DISK FILE:

A file of user data, e.g., the company employee list, with all NAMEs and information. The data in the file is stored in a set of disk SECTORS (records).

DISK OPERATING SYSTEM ( DOS ) :

A computer program which continuously runs and mediates between the computer user and the APPLICATION PROGRAM , and allows access to disk data by DISK FILE NAMEs .

DISK OVERHEAD:

See [OVERHEAD]

DISK PACK:

A number of metal disks packaged in a canister for removal from the disk drive. WINCHESTER DRIVES do not have disk packs .

DISK STORAGE:

Auxiliary memory system containing disk drives.

DISK SWAP:

1. The act of swapping one optical disk for another. To complete a swap, a jukebox autochanger mechanism must remove a disk from the drive, put it away, retrieve a new disk, and insert it in the drive. The drive then spins -up the new disk and the operation is complete. 2. Changing out a defective or malfunctioning hard disk drive.

DISK PLATTER:

For rigid disks, a flat, circular aluminum disk substrate, coated on both sides with a magnetic substance (iron oxide or thin film metal media) for non-VOLATILE data storage. The substrate may consist of metal, plastic, or even glass. Surfaces of disks are usually lubricated to minimize wear during drive start-up or power down.

DISK-TO-DISK ( D2D ) :

A two-tier storage architecture.

DISK-TO-DISK TO TAPE ( DDT ) :

A three-tier storage architecture.

DISKETTE:

A floppy disk. A plastic (mylar) substrate, coated with magnetic iron oxide, enclosed in a protective jacket.

DISTRIBUTED NETWORK:

A network that divides data processing, storage, and other functions into separate units rather than having them all handled by a single computer.

DMA:

See [DIRECT MEMORY ACCESS]

DOCUMENT IMAGE MANAGEMENT ( DIM ) :

A storage management solution for converting paper documents, photos, and receipts into an electronic format that can be accessed from a computer.

DOS:

See [DISK OPERATING SYSTEM]

DRIVE GEOMETRY:

The functional dimensions of a drive, including the number of heads, cylinders, and sectors per track.

DRIVE:

A computer memory device with moving storage MEDIA (disk or tape).

DRIVER:

A software component or set of file commands that allow an application to communicate with another application, driver, or hardware device. A driver receives I/O requests from higher levels within the operating system and converts those requests to the protocol required by a specific hardware device.

ECC:

See [ERROR CORRECTION CODE]

EFFECTIVE ACCESS TIME:

The actual time it takes to access data. In an optical jukebox, it involves variables such as disk swap time, disk spin-up time, seek time, and transfer rates of the host computer and software application.

EISA:

See [EXTENDED INDUSTRY STANDARD ARCHITECTURE]

ELECTROMAGNETIC INTERFERENCE ( EMI ) :

Interference resulting from the presence of electromagnetic fields from electrical and/or electronic devices.

ELECTRO-STATIC DISCHARGE ( ESD ) :

A cause of integrated circuit (CHIP) failure. Since the circuitry of CHIPs are microscopic in size, they can be damaged or destroyed by small static discharges. People handling electronic equipment should always ground themselves before touching the equipment. Electronic equipment should always be handled by the chassis or frame. Components, printed circuit board edge connectors should never be touched.

ELEVATOR SORTING:

A method of sorting records or cache pages by physical location on disk so that the information may be written to disk with less seek and rotational latency.

EMBEDDED SERVO SYSTEM:

Servo data is embedded or superimposed along with data on every cylinder. Timing and positioning signals are interspersed in data tracks. These signals provide the information the actuator needs to fine-tune the position of the read/write heads. ( See also DEDICATED SERVO .)

ENCODING:

The conversion of data into a pattern of On/Off or 1/0 signals prior to writing them to the disk surface.

See also [MFM]

See also [RLL]

ENCRYPTION:

A security method in which electronic data is scrambled and decoded using a software algorithm.

ENTERPRISE NETWORK:

A system of network connections that links all of a company's LANs, allowing enterprises to communicate across many geographic locations and sites.

ENTERPRISE STORAGE NETWORK ( ESN ) :

According to EMC Corporation, an integrated suite of products and services designed to maximize heterogeneous connectivity and management of enterprise storage devices and servers. ESN, generically, constitutes a dedicated, high-speed network connected to the enterprise's storage systems, enabling files and data to be transferred between storage devices and client mainframes and servers.

EPROM:

See ERASABLE PROGRAMMABLE READ ONLY MEMORY .

ERASE:

To remove previously recorded data from magnetic storage media.

ERROR:

See also [HARD ERROR]

See also [SOFT ERROR]

ERROR CORRECTION CODE ( ECC ) :

An embedded code that allows detection of a mismatch between transmitted and received data in a communications system, or between stored and retrieved data in a storage system. The ECC can correct errors, but within limits.

ERROR DETECTION:

A software or firmware algorithm that looks for inconsistencies or errors in a data file as it is being stored. More advanced levels of error detection will not only detect problems, but also correct errors or inconsistencies automatically.

ERROR RATE:

The ratio of data that is incorrectly recorded relative to the entire amount of data written.

ESDI:

See ENHANCED SMALL DEVICE INTERFACE .

ETHERNET:

A local area network standard for hardware, communication, and cabling.

EUROPEAN COMPUTER MANUFACTURERS ASSOCIATION ( ECMA ) :

An international organization founded in 1961 and dedicated to the standardization of information and communication systems.

EXECUTE:

To perform a data processing operation described by an instruction or a program in a computer.

EXTENDED INDUSTRY STANDARD ARCHITECTURE ( EISA ) :

An enhanced AT bus architecture designed by nine manufacturers of PC compatibles and announced in September 1988. EISA provides backwards compatibility with existing 8 and 16-bit hardware cards. In addition, EISA supports 32-bit data paths, 33 Mbytes/sec data transfers from Bus Mastering peripheral cards, automatic configuration, and a more sophisticated I/O addressing scheme.

See also [AT]

See also [ISA]

EXTERNAL DRIVE:

A hard disk drive mounted in an enclosure separate from the computer system enclosure. An external drive has its own power supply and fan and is connected to the system by a cable.

FABRIC SWITCH:

A type of storage area network (SAN) switch characterized by the fact that any port on any switch can provide (subject to bandwidth availability) full speed access to any other port on the network. The network consists of a fabric of linked switches.

FAILOVER:

The transfer of operation from a failed component (e.g., controller, disk drive) to a similar, redundant component to ensure uninterrupted data flow and operability.

FAST ATA:

Fast ATA is the market identity given to disk drives that support the high-speed data transfers resulting from implementing the industry standard protocols: Programmed input/output (PIO) mode 3, Multiword direct memory access, read/write multiple sectors per interrupt.

FAST SCSI:

The original SCSI specification defined synchronous data transmission speeds of up to 5MHz. By assuming transceivers that provide tighter timing margins, the SCSI-2 standard allows synchronous transfers of up to 10MHz. Devices that utilize these faster timings are called Fast SCSI devices.

FAT (File allocation table):

A data table stored on the outer edge of the disk and used by the operating system to determine which sectors are allocated to each file and in what order.

FAULT TOLERANCE:

The ability of a system to cope with internal hardware problems (e.g., a disk drive failure) and still continue to operate with minimal impact, such as by bringing a backup system online.

FC-AL:

See [FIBRE CHANNEL ARBITRATED LOOP (FC-AL)]

FCI:

See [FLUX CHANGES PER INCH]

FCIP:

See [FIBRE CHANNEL OVER IP]

FEEDBACK:

A closed-loop control system, using the head-to-track positioning signal (from the servo head) to modify the HEAD POSITIONER signal (to correctly position the head on the track).

FETCH:

A CPU read operation from MAIN MEMORY and its related data transfer operations.

FIBRE CHANNEL:

A high-speed, serial, storage/networking interface that offers higher performance, greater capacity and cabling distance, increased system configuration flexibility and scalability, and simplified cabling. One can view Fibre Channel simply as a transport vehicle for the supported command set (usually SCSI commands). In fact, Fibre Channel is unaware of the content of the information being transported. It simply packs data in frames , transports them to the appropriate devices, and provides error checking.

FIBRE CHANNEL ARBITRATED LOOP ( FC-AL ) :

One of the possible physical topologies of Fibre Channel. In this topology, the Fibre Channel is connected in a loop with devices all connecting to the loop. It can be thought of as a similar structure to a token ring network. Up to 126 nodes can be connected to the loop.

FIBRE CHANNEL CLASS 1 SERVICE:

This service level guarantees bandwidth and ordering of packets. It also returns confirmations of transmission.

FIBRE CHANNEL CLASS 2 SERVICE:

This service level is connectionless and can deliver packets out-of-order. Delivery of packets is however guaranteed and confirmations are sent.

FIBRE CHANNEL CLASS 3 SERVICE:

This is the lowest service level and does not guarantee either ordering or delivery.

FIBRE CHANNEL FABRIC:

One of the physical topologies of Fibre Channel. In this topology, the addressing of ports on a network of Fibre Channel nodes is made independently of the physical location or address of the target port. Switches are responsible for passing Fibre Channel packets to the target port regardless of which Fibre Channel loop or switch where the port physically resides.

FIBRE CHANNEL LOOP COMMUNITY ( FCLC ) :

An international non-profit organization whose members include manufacturers of servers, disk drives, RAID storage systems, switches, hubs, adapter cards, test equipment, cables, and connectors, and software solutions.

FIBRE CHANNEL OVER IP ( FCIP ) :

Fibre Channel over IP (FCIP or FC/IP, also known as Fibre Channel tunneling or storage tunneling) is an Internet Protocol (IP)-based storage networking technology developed by the Internet Engineering Task Force (IETF). FCIP mechanisms enable the transmission of Fibre Channel (FC) information by tunneling data between Fibre Channel fabrics using IP networks. While this capacity facilitates the joining of FC fabric "islands" over a geographically distributed enterprise, it does so at a cost of local SAN identity: all FC fabrics become part of one large fabric.

FIBRE CHANNEL POINT-TO-POINT:

One of the physical topologies of Fibre Channel. This topology provides a simple, direct connection between just two nodes, approximating traditional SCSI topology.

FIBRE CHANNEL PORTS:

Fibre Channel ports come in a number of types depending on the topology of the Fibre Channel network. N_Ports are simple equipment node ports in a point-to-point connection topology. NL_Ports are node ports connected to an Arbitrated loop. F_Ports are point-to-point ports connected to a fabric. Generally this means that the F_Port is a port on a switch. FL_Ports are ports connecting from one loop to a switch and thus to a fabric. E_Ports are expansion ports used to interconnect switches together. G_Ports are classified by some switch companies as ports that can be either E_Ports or F_Ports depending on usage.

FIBRE CHANNEL PROTOCOL OVER INTERNET or INTERNET FIBRE CHANNEL PROTOCOL ( iFCP ) :

iFCP is a mapping of FCP to the Internet Protocol family for use in extending Fibre Channel storage networks across the Internet. iFCP provides a means of passing data to and from Fibre Channel storage devices in a local storage area network (SAN) or on the Internet using TCP/IP. TCP provides congestion control as well as error detection and recovery services. It can either replace or be used in conjunction with existing Fibre Channel protocols, such as FCIP (Fibre Channel over IP). Unlike FCIP, which is a tunneling protocol that encapsulates Fibre Channel data and forwards it over a TCP/IP network as an extension of the existing Fibre Channel network, iFCP operates the Fibre Channel Protocol as an application across an IP network. iFCP gateways can either replace or complement existing Fibre Channel fabrics, and therefore can be used to facilitate migration from a Fibre Channel SAN to an IP SAN or a hybrid network.

FIBRE CHANNEL SAN:

An oxymoronic reference to a storage deployment topology more appropriately termed "switched, server-attached storage" or a "Fibre Channel Fabric."

FIBRE CHANNEL SPECIFICATION:

A specification for the Fibre Channel serial interconnect, which may be used as the infrastructure for a SAN. FC-0 is the Physical layer of the Fibre Channel protocol stack. This layer includes the definition of all physical components used in Fibre Channel. FC-1 is the Encode/Decode layer in Fibre Channel specification. It covers the byte encoding and character-level error control. FC-2 is called the Framing Protocol Layer. It incorporates the management of frames, flow control, and CRC generation. It also manages sequences of frames comprising a transmission, and exchanges between nodes on the Fibre Channel to accomplish commands akin to the SCSI I/O sequence. This layer also provides the management of the three service classes: Class 1, Class 2, and Class 3. FC-3 is called the Common Services Layer and is currently not used. FC-4 is the Protocol Mappings Layer and is the layer that maps protocols such as SCSI and IP to the underlying layer protocols.

FIBRE CHANNEL TOPOLOGY:

A number of possible topologies, or methods for interconnecting devices or nodes, have been specified for Fibre Channel. These include point-to-point, Arbitrated Loop, and Fabric topologies.

See also [FIBRE CHANNEL POINT-TO-POINT]

See also [FIBRE CHANNEL ARBITRATED LOOP]

See also [FIBRE CHANNEL FABRIC]

FILE ALLOCATION TABLE ( FAT ) :

What the operating systems uses to keep track of which clusters are allocated to which files and which are available for use. FAT is usually stored on Track-0.

FILE RECOVERY:

The process of using backup files to replace lost files after a power failure, facility damage, virus infection, system crash, or human error.

FIRMWARE:

Permanent instructions and data programmed directly into the circuitry of read-only memory for controlling the operation of the computer. Distinct from software, which can be altered by programmers.

FIXED DISK:

A disk drive with disks that cannot be removed from the drive by the user.

FLEXTURE:

An extremely pliable plastic circuit which connects the accuator assembly electronics to the base casting. The drive may have access times that move the heads from the outer diameter to the inner as low as 20 milliseconds and this flexture rides along.

FLOPPY DISK:

A flexible plastic disk coated with magnetic media and packaged in a stiff envelope. Comes in 8-inch, 5-1/4-inch, and various sub-4 inch sizes. FLOPPY DISKS generally exhibit slow ACCESS TIME and smaller CAPACITY compared to WINCHESTER DRIVES , but feature removable diskettes.

FLUX CHANGE:

Location on the data track, where the direction of magnetization reverses in order to define a 1 or 0 bit.

FLUX CHANGES PER INCH ( FCI ) :

Linear recording density defined as the number of flux changes per inch of data track. The number of magnetic field patterns that can be stored on a given area of disk surface, used as a measure of data density. Also known as flux density. Synonymous with FRPI (flux reversals per inch). In MFM recording 1 FCI equals 1 BPI (bit per inch). In RLL encoding schemes, 1 FCI generally equals 1.5 BPI.

See also [BPI]

FLUX DENSITY:

See [FLUX CHANGES PER INCH]

FLYING HEIGHT:

The distance between the read/write head and the disk surface, created by the cushion of air that results from the velocity of the disk rotation, which keeps the two objects from touching. Smaller flying heights permit denser data storage but require more precise mechanical designs. Also known as fly height.

FOOTPRINT:

The amount of floor space that a piece of equipment (e.g., a rackmount enclosure) occupies.

FORMAT:

To write a magnetic track pattern onto a disk surface, specifying the locations of the tracks and sectors. This information must exist on a disk before it can store data. The purpose of a format is to record "header" data that organize the tracks into sequential sectors on the disk surfaces. This information is never altered during normal read/write operations. Header information identifies the sector number and also contains the head and cylinder ADDRESS in order to detect an ADDRESS ACCESS error.

FORMATTED CAPACITY:

The amount of space left to store data on a disk after writing the sector headers, boundary definitions, and timing information during a format operation. Actual capacity available to store user data.

FORM FACTOR:

The industry standard that defines the physical, external dimensions of a particular device.

FRICTION:

Resistance to relative motion between two bodies in contact; e.g., there is sliding friction between head and disk during drive power up/down.

FULL HEIGHT DRIVE:

Winchester 5-1/4" drive which fits in the same space as full height mini-floppy drive (called the full-height form factor), which is 3.25 inches high.

G:

A G is a unit of force applied to a body at rest equal to the force exerted on it by gravity. Hard disk drive shock specifications are usually called out in Gs. A shock specification of 40 Gs non-operating means that a drive will not suffer any permanent damage if subjected to a 40 G shock . This is roughly equivalent to a drop of the drive to a hard surface from a distance of 1 inch.

GAP:

(1) Part of the disk format. Allows mechanical compensations (e.g., spindle motor rotational speed variations) without the last sector on a track overwriting the first sector. (2) An interruption in the permeable material of a read/write head, usually a glass bonding material with high permeability , allowing the flux fields to exit the head structure to write or read data bits in the form of flux changes on the recording media.

GAP LENGTH:

Narrowing the head gap length achieves higher bit density because the lines of force magnetize a smaller area where writing data in the form of flux changes on the recording media.

GAP WIDTH:

The narrower the gap width, the closer the tracks can be placed. Closer track placement results in higher TPI.

GIGABYTE ( GB ) :

1 Gigabyte = 1,073,741,824 bytes (or approximately one thousand million bytes).

GUARD BAND :

1. Non-recorded band between adjacent data tracks, 2. For closed loop servo drives, extra servo tracks outside the data band preventing the Carriage Assembly from running into the crash stop.

HALF HEIGHT:

A standard drive height of 1.6 inches ”the equivalent to half the vertical space of a 5.25-inch drive.

See also [LOW PROFILE]

HARD DISK DRIVE:

Sometimes called rigid disk drives, or Winchester disk drives. An electromechanical device that can read rigid disks. Though similar to floppy disk drives, the hard disks have higher bit density and multiple read/write surfaces.

HARD ERROR:

A data error that persists when the disk is reread, usually caused by defects in the physical surface. Hard errors are caused by imperfections in the disk surface, called media defects. When formatting hard disk drives, hard error locations, if known, should be spared out so that data is not written to these locations. Most drives come with a hard error map listing the locations of any hard errors by head, cylinder and BFI (bytes from index ”or how many bytes from the beginning of the cylinder).

HARD SECTOR MODE:

A hardware-controlled convention defining a fixed number of sectors per track in any specified zone.

HARD SECTORED:

A term describing a hard drive that determines the starting location of each sector from information in the embedded servo. This method is the most common and is newer and more precise than soft sectored techniques.

See also [SOFT SECTORED]

HARD ZONING:

A method of SAN zoning that uses only WWNs to specify each device. Traffic across the switch is regulated passing requests through the switch's route table. If two ports are not authorized to communicate with each other, the route table for those ports is disabled, and the communication between those ports is blocked.

HARDWARE:

Computer equipment (as opposed to the computer programs and software).

HARDWARE ARRAY:

A group of disk drives which are all members of the same array and share the same logical name or unit number.

HDA:

See [HEAD/DISK ASSEMBLY]

HDA INTERCONNECT:

Connects the drive electronics to the mechanical assembly.

HEAD:

An electromagnetic device that can write (record), read (playback), or erase data on magnetic media.

HEAD CRASH:

A head landing occurs when the disk drive is turned on or off. This function normally does not damage the disk as the disk has a very thin lubricant on it. A head crash occurs when the head and disk damage each other during landing, handling, or because a contaminant particle gets between them. Head crash is a catastrophic failure condition and causes permanent damage and loss of data.

HEAD/DISK ASSEMBLY ( HDA ) :

A sealed Winchester assembly including disks, heads, filter, and actuator assembly.

HEAD LANDING AND TAKEOFF:

In Winchester drives, the head is in contact with the platter when the drive is not powered . During the power up cycle, the disk begins rotation and an "air bearing" is established as the disk spins up to full RPM ( rotations per minute). This air bearing prevents any mechanical contact between head and disk.

HEAD LANDING ZONE:

An area of the disk set aside for takeoff and landing of the Winchester heads when the drive is turned on and off.

HEAD POSITIONER:

Also known as the ACTUATOR , a mechanism that moves the CARRIAGE ASSEMBLY to the cylinder being accessed.

HEAD SKEW:

See TRACK SKEW .

HEAD SLAP:

Similar to a head crash but occurs while the drive is turned off. It usually occurs during mishandling or shipping. Head slap can cause permanent damage to a hard disk drive.

See also [HEAD CRASH]

HELICAL SCAN RECORDING ( HSR ) :

Used widely in VHS formats for video recording, the HSR tape recording method writes at an angle across the width of a tape, allowing higher storage densities on half-inch tape.

HELICAL SCAN:

A DAT recording method whereby heads record data at an angle rather than in a straight line (linear), which uses the entire width of the tape.

HEXADECIMAL ( HEX ) :

A number system based on sixteen, using digits 0 through 9 and letters A through F to represent each digit of the number. (A = 10, B = 11, C = 12, D = 13, E = 14, F = 15).

HIERARCHICAL FILE SYSTEM ( HFS ) :

The file management system in which directories have sub-directories and sub-subdirectories. In MS Windows and Macintosh operating systems, the directories and sub-directories are represented as folders nested within other folders.

HIERARCHICAL STORAGE MANAGEMENT ( HSM ) :

A storage system in which new, frequently modified data is stored on the fastest , most accessible (and generally more expensive) media (e.g., RAID) and older, less frequently used data is stored on slower (less expensive) media (e.g., tape).

HIGH-LEVEL FORMATTING:

Formatting performed by the operating system to create the root directory, file allocation tables, and other basic configurations.

See also [LOW-LEVEL FORMATTING]

See also [FAT]

HOME PAGE:

The main page on a Web site that serves as the primary point of entry to related pages within the site and may have links to other sites as well.

HOME:

The reference track of a hard disk, usually the outermost track (track 0), used for recalibration of the actuator.

HOLOGRAPHIC STORAGE:

An optical technology that records data as holograms. The hologram is created by two lasers targeted at the media. One laser is beamed into the material through a matrix of LCD shutters, called a "spatial light modulator ." The shutters are opened or closed based on the binary pattern of the page of data being stored. For example, using a matrix of 1,024 pixels on each side, the page could hold a million bits. A second reference laser is angled into and intersects with the data laser at the storage unit. If the angle and/or frequency is changed, another hologram can be created overlapping and filling the same space as the first hologram. In fact, 10,000 holograms (pages) can overlap each other. The page is read by directing just the reference laser back into the hologram. The light is diffracted into an original copy of the data, which is sensed by a matrix of CCD sensors. The technology has been in development since the 1960s.

HOST ADAPTER:

A plug-in board that acts as the interface between a computer system bus and the disk drive.

HOST BUS ADAPTER ( HBA ) :

A hardware card that resides on the PC bus and provides an interface connection between a SCSI device (such as a hard drive) and the host PC.

HOT SPARE:

A spare disk drive which, upon failure of a member of a redundant disk array, will automatically be used to replace the failed disk drive.

HOT SWAP:

The operation of removing a failed disk drive, which is a member of a redundant array, and replacing it with a good drive.

HUB:

A device that splits one network cable into a set of separate cables, each connecting to a different computer; used in a local area network to create a small-scale network by connecting several computers together.

I/O PROCESSOR:

Intelligent processor or controller that handles the input/output operations of a computer.

ID FIELD:

The address portion of a sector on a formatted hard disk. The ID field is written during the Format operation. It includes the cylinder, head, and sector number of the current sector. This address information is compared by the disk controller with the desired head, cylinder, and sector number before a read or write operation is allowed.

IDE:

See [INTEGRATED DRIVE ELECTRONICS]

IFCP:

See FIBRE CHANNEL OVER INTERNET .

IMAGE-BACKUP MODE:

Used with streaming tape, image-backup mode records an exact copy of the disk, including unused sectors and bad tracks.

INDEX (PULSE):

The Index Pulse is the starting point for each disk track. The index pulse provides initial synchronization for sector addressing on each individual track.

INDEX TIME:

The time interval between similar edges of the index pulse, which measures the time for the disk to make one revolution. This information is used by a disk drive to verify correct rotational speed of the media.

INDUSTRY STANDARDS:

Rules or guidelines, established by independent consortia, to control the development and manufacture of products and devices in the electronics industry. Industry standards for audio CDs, for example, are what assure consumers that any audio CD will work in any CD player.

INDUSTRY STANDARD ARCHITECTURE ( ISA ) :

The standard 16-bit AT bus designed by IBM for the PC/AT system. ISA was the only industry standard bus for PCs until the release of IBM's MCA (MicroChannel Architecture) and EISA (Extended Industry Standard Architecture).

See also [EISA]

INFRASTRUCTURE:

The physical equipment (computers, cases, racks, cabling, etc.) that comprises a computer system.

INITIALIZATION:

See [LOW-LEVEL FORMATTING]

INITIATOR:

A SCSI device that requests another SCSI device (a target) to perform an operation; usually a host computer acts as an initiator and a peripheral device acts as a target.

INPUT:

Data entered into the computer to be processed, or User commands or queries.

INPUT/OUTPUT ( I/O ) :

The process of entering data into or removing data from a computer system. The reception (read) or transmission (write) of computer signals; the entire connection path between the CPU bus and the disk drives.

INSTITUTE OF ELECTRICAL AND ELECTRONICS ENGINEERS ( IEEE ) :

The largest technical society in the world, consisting of engineers, scientists, and students. Articulates standards for computers and communications.

INSTRUCTION:

The most basic task that a computer performs. A single instruction involves a single, simple calculation processed by the computer. When working on a computer, many instructions are strung together one after another to complete the larger tasks that the operating system or software directs.

See also [MIPS]

INTEGRATED DRIVE ELECTRONICS ( IDE ) :

A disk drive interface that incorporates the drive controller into the drive electronics. IDE drives are used in IBM-compatible computers. Also known as ATA.

INTELLIGENT PERIPHERAL:

A peripheral device that contains a processor or microprocessor to enable it to interpret and execute commands, thus relieving the computer for other tasks.

INTERCHANGEABILITY:

The ability to use one brand or type of storage media in a variety of drives. For example, manufacturers of audio tapes and tape players support industry standards for interchangeability, so any tape will work in any player.

INTERFACE:

A connection between hardware devices, applications, or different sections of a computer network. A hardware or software protocol (contained in the electronics of the disk controller and drive) that manages the exchange of data between the drive and the computer. The most common interfaces for small computer systems are ATA (also known as IDE) and SCSI.

See also [AT]

See also [SCSI]

INTERFACE CONNECTOR:

Attachment point for the interface ribbon cable: 40 pins for ATA; 50 pins for SCSI Narrow, etc.

INTERFACE CONTROLLER:

The chip or circuit that translates computer data and commands into a form suitable for use by the hard drive and controls the transfer of data between the buffer and the host.

See also [DISK CONTROLLER]

See also [DISK DRIVE CONTROLLER]

INTERFACE STANDARD:

The interface specifications agreed to by various manufacturers to promote industry-wide interchangeability of products such as disk drives and controllers. An interface standard generally reduces product costs, allows buyers to purchase from more than one source, and allows faster market acceptance of new products.

See also [SCSI]

See also [ESDI]

INTERLEAVE:

The arrangement of sectors on a track. The interleave value tells the controller where the next logical sector is located in relation to the current sector. For example, an interleave value of one (1) specifies that the next logical sector is physically the next sector on the track. Interleave of two (2) specifies every other physical sector, three (3) every third sector, and so on. Interleaving is used to improve the system throughout based on overhead time of the host software, the disk drive and the controller. For example, if an APPLICATION PROGRAM is processing sequential logical records of a DISK FILE in a CPU time of more than one second but less than two, then an interleave factor of 3 will prevent wasting an entire disk revolution between ACCESSES .

See also [INTERLEAVE FACTOR]

INTERLEAVE FACTOR:

The ratio of physical disk sectors skipped for every sector actually written. The number of sectors that pass beneath the read/write heads before the next sector arrives. For example, a 3:1 interleave factor means that the heads read a sector, then let two pass by before reading another, requiring three full revolutions of the disk to access the complete data track.

INTERNAL DRIVE:

A disk drive mounted inside a computer either in a drive bay or on a card installed in an expansion slot.

INTERNATIONAL ORGANIZATION FOR STANDARDIZATION ( ISO ) :

A worldwide organization that develops, publishes, and promotes international industrial and technical standards. The term ISO is not an acronym, but a derivative of the Greek word isos , which means "equal."

INTERNET:

A worldwide system of linked computer networks.

INTERNET FIBRE CHANNEL PROTOCOL:

See [FIBRE CHANNEL PROTOCOL OVER INTERNET]

INTERNET SCSI PROTOCOL:

See [SCSI OVER INTERNET]

INTEROPERABILITY:

The ability of one computer system to control another, even though the two systems are made by different manufacturers.

INTERRUPT:

A signal, usually from a peripheral device to a CPU, to signify that a commanded operation has been completed or cannot be completed.

INTRANET:

A computer network, based on Internet technology, that is designed to meet the internal needs for sharing information within a single organization or company.

IOPS (I/Os PER SECOND):

A measure of performance for a host-attached storage device or RAID controller.

IP Storage:

Storage platforms accessed via IP-based protocols such as iSCSI or iFCP.

ISA:

See [INDUSTRY STANDARD ARCHITECTURE]

ISCSI:

See [SCSI OVER INTERNET]

JBOD:

See [JUST A BUNCH OF DISKS]

JUKEBOX:

Also called an optical disk library, a jukebox is a standalone cabinet that holds multiple optical disk drives and cartridges for high-speed, high-capacity storage. It includes a robotic arm to pick an optical cartridge from its storage slot, move it to one of several drives, then return it to the slot when it is no longer needed.

JUST A BUNCH OF DISKS ( JBOD ) :

A non-RAID disk array.

KERNEL:

The core of an operating system such as Windows 98, Windows NT, Mac OS, or Unix; provides basic services for the other parts of the operating system, making it possible for it to run several programs at once (multitasking), read and write files and connect to networks and peripherals.

KILOBYTE ( KB ) :

A unit of measure consisting of 1,024 bytes.

LAN:

See [LOCAL AREA NETWORK]

LANDING ZONE:

A non-data position on the disk's inner cylinder where the heads can land when the power is off. In a portable computer, the heads also move to the landing zone after a period of inactivity to save power and extend battery life.

LATENCY, ROTATIONAL:

The time for the disk to rotate the accessed sector under the head for read or write. On the average, latency is the time for half of a disk revolution.

LEGACY:

A computer, system, or software, that was created for a specific purpose but is now outdated ; anything left over from a previous version of the hardware or software.

LIBRARY:

See [JUKEBOX]

LIGHT INTENSITY MODULATION DIRECT OVERWRITE ( LIMDOW ) :

An optical recording technology that allows write procedures to be accomplished in one pass of the optical read/write head over the media. LIMDOW accelerates optical performance, but requires somewhat complex and expensive optical media, and involves a more complicated write procedure that leaves less margin for error compared with other optical technologies.

LINEAR TAPE OPEN ( LTO ) :

A tape format developed by HP, IBM, and Seagate Technology.

LOCAL AREA NETWORK ( LAN ) :

A communications network used to connect computers and other electronic devices within a confined geographical area. For example, a LAN can connect users within a single site, allowing them to share data, exchange e-mail, and share peripherals.

LOGIC:

Electronic circuitry that switches on and off ("1" and "0") to perform digital operations.

LOGICAL BLOCK:

See [SECTOR]

LOGICAL FORMAT:

Refers to low-level formatting. In relation to DOS-specific format requirements, refers to the translations accomplished by the controller in situations where the hard drive data configurations do not match DOS format limitations.

LOGICAL UNIT NUMBER ( LUN ) :

An addressing scheme used to define SCSI devices on a single SCSI bus.

LOOK AHEAD:

The process of anticipating events in order to speed up computer operations. For example, a disk drive can use look ahead caching to speed subsequent sequential data requests. When the drive receives a request for data, it reads not only the data requested into the cache buffer but also "looks ahead" and reads data immediately following the requested data into cache until the cache buffer is filled. Once loaded into cache, the information can be accessed almost instantaneously, without the need for additional read operations.

LOOKUP:

The action of obtaining and displaying data in a file.

LOW LEVEL FORMAT:

The first step in preparing a drive to store information after physical installation is complete. The process sets up the "handshake" between the drive and the controller. In an XT system, the low level format is usually done using DOS's debug utility. In an AT system, AT advanced diagnostics is typically used. Other third party software may also be used to do low level format on both XTs and ATs.

LOW PROFILE:

A standard drive height of 1 inch.

See also [HALF HEIGHT]

LUN:

See [LOGICAL UNIT NUMBER]

LUN MASKING:

A method of masking multiple LUNs behind a single port. By using World Wide Port Names (WWPNs) of server host bus adapters, LUN masking is configured at the RAID-array level. LUN masking also allows disk storage resource sharing across multiple independent servers. A single large RAID device can be sub-divided to serve a number of different hosts that are attached to the RAID through the SAN fabric with LUN masking. So that only one or a limited number of servers can see that LUN (e.g., disk slice, portion, unit), each LUN inside the RAID device can be limited. LUN masking can be done either at the RAID device (behind the RAID port) or at the server HBA. It is more secure to mask LUNs at the RAID device, but not all RAID devices have LUN masking capability. Therefore, in order to mask LUNs, some HBA vendors allow persistent binding at the driver-level.

MAGAZINE:

A removable chamber that holds multiple optical disk cartridges or magnetic tapes, often used for high-volume automated backup.

MAGNETIC MEDIA:

A disk or tape with a surface layer containing particles of metal, or metallic oxides that can be magnetized in different directions to represent bits of data, sounds or other information.

MAGNETIC RECORDING:

The use of a head, recording head, recording media (tape or disk), and associated electronic circuitry for storing data or sound or video.

MAGNETO-OPTICAL ( MO ) :

An optical disk storage technology that uses a combination of magnet and laser to alter the magnetic flux directions on a disk's recording surface, much like a magnetic hard disk. The laser heats a small point on the disks' recording surface to 150 degrees Celsius. At this temperature, the polarity of the disk's recording surface can be altered with a magnet , causing a change in reflectivity that is detected during reads. MO is the dominant technology for rewritable optical disks, and meets ANSI, ISO, and ECMA industry standards for interchangeable optical disk cartridges.

MAGNETO-RESISTIVE HEAD ( MRH ) :

A special read head technology designed to support data acquisition from media with very high recording densities. Based on materials with special magneto-resistive properties, e.g., whose electrical resistance changes in the presence of a magnetic field, read heads incorporating MR material are more sensitive to changes in the strength of magnetic fields, enabling them to read more densely-packed bits stored on tape or disk media. This mechanism cannot be used for writing, so a conventional thin film inductive write head element is deposited alongside the MR stripe.

MAIN MEMORY:

Random-access memory used by the CPU for storing program instructions and data currently being processed by those instructions. ( See RANDOM-ACCESS MEMORY .)

MAINFRAME COMPUTER:

A large computer generally found in data processing centers.

See also [MINICOMPUTER]

See also [MICROCOMPUTER]

MANAGED HUB:

This is a technique for providing statistics information about the traffic on a hub. Typically, no actual management of the hub is possible using this interface, but information and notification of failures can be achieved. This interface often uses Simple Network Management Protocol (SNMP) Management Information Bases (MIBs) as a standard protocol for providing this information.

MEAN SWAPS BETWEEN FAILURE ( MSBF ) :

A measure of reliability specific to optical jukeboxes, usually determined in benchmark testing. MSBF refers to the average number of disk "swaps" a jukebox and its internal mechanisms can be expected to deliver before maintenance is required. A statistical calculation used to predict the average usefulness of a robotic device (e.g., a tape library) with any interruption of service.

MEAN TIME BETWEEN FAILURE ( MTBF ) :

The average time before a failure will occur. This is not a warranty measurement. MTBF is a calculation taking into consideration the MTBF of each component in a system and is the statistical average operation time between the start of a unit's lifetime and its time of a failure. After a product has been in the field for a few years , the MTBF can become a field proven statistic. The higher the MTBF, the more reliable the equipment.

MEAN TIME TO REPAIR ( MTTR ) :

A measure of the complexity of design in electronic equipment. Highly modular designs ”i.e., those that use interchangeable, hot-swappable components ”typically have a low MTTR since failed components can be replaced with functioning components.

MEAN TIME until DATA LOSS ( MTDL ) :

The average time from startup until a component failure causes a permanent loss of user data in a disk array. The concept is similar to MTBF, but takes into account the possibility that RAID redundancy can protect against loss due to single component failures.

MEDIA:

A physical storage medium. Includes optical disks, CDs, magnetic tapes, hard disks, and other technologies used to store computer-based information. The magnetic layers of a disk or tape. ( See DISK/PLATTER .)

MEGABYTE ( MB ) :

A megabyte is 106 or 1,000,000 bytes. One megabyte can store more than one million characters.

MEMORY:

Any device or storage system capable of storing and retrieving information.

MESSAGE DIGEST HEADER:

A data header containing the results of an algorithm applied to the contents of the dataset. The algorithm produces a unique " fingerprint " for the dataset that may be referenced to determine dataset authenticity and integrity. Several versions of digest algorithms exist, the latest being MD5. MDx hashes may be a useful component of a scheme for ensuring data integrity, authenticity, and self-description -based management.

MFM (Multiple frequency modulation):

A method of encoding analog signals into magnetic pulses or bits.

MICROCOMPUTER:

A computer whose central processor unit (CPU) is manufactured as a chip or a small number of chips. Personal computers are examples of microcomputers.

MICROINCH:

One-millionth of an inch.

MICROSECOND:

One-millionth of a second.

MILLISECOND:

One-thousandth of a second.

MICROSOFT WINDOWS:

An operating system with several variants developed by Microsoft for high-performance processors and networked systems.

MINICOMPUTER:

A computer midway in size and processing power between a MICROCOMPUTER and a MAINFRAME COMPUTER .

MIPS (MILLIONS OF INSTRUCTIONS PER SECOND):

A measure of the speed at which a CPU operates.

See also [INSTRUCTION]

MIRRORING:

A method of storage in which data from one disk is duplicated on another disk so that both drives contain the same information, thus providing data redundancy. A popular term for RAID-1.

MISSION CRITICAL:

Any computer process that cannot fail during normal business hours; some computer processes (e.g., telephone systems) must run all day long and require 100 percent uptime.

MISSION CRITICAL DATA:

Data or information considered to be so important that its loss would cause grave difficulty to all or part of a business. For example, customer account information at a bank, or patient information at a hospital.

MNEMONIC:

A shortened code for a longer term.

MODIFIED FREQUENCY MODULATION ( MFM ) :

A method of recording digital data, using a particular CODE to get the flux reversal times from the data pattern. MFM recording is self-clocking because the CODE guarantees timing information for the playback process. The controller is thus able to synchronize directly from the data. This method has a maximum of one bit of data with each flux reversal.

See also [NRZ]

See also [RLL]

MTBF:

See [MEAN TIME BETWEEN FAILURE]

MTTR:

See [MEAN TIME TO REPAIR]

MULTIMEDIA:

The combination of several media formats used for the delivery of information. Many commercial CD-ROMs use a multimedia format, combining text, photos, audio, animation, and video on a single disk.

MULTIPLATFORM:

The ability of a product or network to support a variety of computer platforms (e.g., IBM, Sun, Macintosh); also referred to as cross-platform.

MULTIPLE SEGMENT CACHING:

A technique enabling the division of the cache into segments so that different blocks of data can be cached simultaneously and subsequent commands will have a better probability of a cache hit.

MULTIPROCESSOR:

A computer containing two or more processors.

MULTITASKING:

The ability of a computer system to execute more than one program or program task at a time.

MULTI-THREADED:

Having multiple concurrent or pseudo-concurrent execution sequences. Used to describe processes in computer systems. Multi-threaded processes are one means by which throughput intensive applications can make maximum use of a disk array to increase I/O performance.

MULTI-USER:

The ability of a computer system to execute programs for more than one user at a time.

NEAR-FIELD RECORDING ( NFR ) :

An optical disk storage technology that combines elements of hard disk and magneto-optical (MO) storage. Though still under development, the technology is expected to deliver hard disk-like performance with greater storage capacities and lower cost storage costs than current technologies.

NEAR-ONLINE STORAGE:

A cross between online and offline storage, usually consisting of data stored in optical jukeboxes. Near-online storage is less expensive, more durable, and takes only slightly longer to access than online storage kept on high-speed hard disks. It is significantly faster and easier to access than offline storage.

NETWORK-ATTACHED STORAGE ( NAS ) :

A disk array with a thin server operating system optimized for storage and IP network attachement. NAS devices typically present a networked file system in accordance with a protocol such as NFS or CIFS/SMB.

NETWORKING:

The ability to interconnect a number of PCs, workstations, servers, and peripherals for the purpose of sharing, sending, receiving and managing information, files, e-mail, and other data.

NETWORK FILE SYSTEM ( NFS ) :

A file system originated by Sun Microsystems that will mount remote file systems across homogenous and heterogenous systems. NFS is a client/server application. An NFS server can export local directories for remote NFS clients to use. Typically, NFS runs over IP using UDP, but there are NFS implementations that will work using TCP as the network transport service. NFS has been accepted by the Internet Engineering Task Force (IETF) in certain areas as a standard for file services on TCP/IP networks on the Internet.

NFS:

See [NETWORK FILE SYSTEM]

NODE (or NETWORK NODE):

Any device that is directly connected to the network, usually through Ethernet cable; nodes include file servers and shared peripherals.

NOISE:

Extraneous electronic signals that interfere with information signals (similar to radio static or TV interference). Sources of noise in computers can be power supplies , ground loops, radio interference, cable routing, etc.

NON-RETURN TO ZERO ( NRZ ) :

A method of magnetic recording of digital data in which a flux reversal denotes a one bit, and no flux reversal a zero bit. NRZ recording requires an accompanying synchronization clock to define each cell time, unlike MFM or RLL recording.

NONREPUDIATION:

A method for ensuring that a transferred message has been sent and received by the parties claiming to have sent and received the message. Nonrepudiation is necessary to guarantee that the sender of a message cannot later deny having sent the message and that the recipient cannot deny having received the message.

OFF-LINE:

Processing or peripheral operations performed while not connected to the system CPU via the system BUS. A collection of data that requires some manual intervention such as loading a disk or tape before it can be accessed. For example floppy disk, tape, CD-ROM.

OFFLINE STORAGE:

Infrequently accessed data that is stored offline in a tape archive or file cabinet. Offline storage is the least expensive and slowest storage method, consisting primarily of tape, microfiche and paper media. Restoring offline data to an online environment must be handled manually.

ON-LINE:

A collection of data, typically stored on hard disks or arrays, that is immediately accessible.

ONLINE STORAGE:

The fastest and most expensive storage alternative, consisting of frequently accessed files found a computer's hard disk.

OPEN SYSTEMS NETWORK:

A network comprised of equipment that conforms to industry standards of interoperability between different operating systems (e.g., Unix, Windows NT).

OPERATING SYSTEM:

An operating system is a program which acts as an interface between the user of a computer and the computer hardware. The purpose of the operating system is to provide an environment in which a user may run programs. The goal of the operating system is to enable the user to conveniently use the computer's resources such as the CPU, memory, storage devices, and printers.

OPTICAL DISK:

A storage medium that generally uses a laser to write and read data.

See also [WORM]

See also [CD-ROM]

See also [CD-R]

See also [CD-RW]

OPTICAL STORAGE:

A storage alternative to hard disks that provides random-access capability like hard disks. Compared to hard disk storage, optical storage offers higher reliability and a higher degree of removability and transportability. However, optical disk access times are two to four times slower than hard disks, due primarily to the weight of the optical head that reads and writes data. Most optical storage technologies use a read/write laser to store data: to write, the laser heats the recording surface of the media, causing a physical change that is detected during reads.

OPTICAL STORAGE TECHNOLOGY ASSOCIATION ( OSTA ) :

An international trade association founded in 1992 to promote the use of writable optical technologies and products.

OUTPUT:

Processing data being transferred out of the computer system to peripherals (i.e., disk, printer, etc.). This includes responses to user commands or queries.

OVERHEAD:

Overhead refers to the processing time required by the controller, host adapter, or drive prior to the execution of a command. Lower command over-head yields higher drive performance. Disk overhead refers to the space required for non-data information such as servo data. Disk overhead often accounts for about ten percent of drive capacity. Lesser disk overhead yields greater disk capacity.

See also [ZERO COMMAND OVERHEAD]

PARITY:

A data error checking method using an extra bit in which the total number of binary 1's (or 0's) in a byte is always odd or always even; thus, in an odd parity scheme, every byte has eight bits of data and one parity bit. If using odd parity and the number of 1 bits comprising the byte of data is not odd, the 9th or parity bit is set to 1 to create the odd parity. In this way, a byte of data can be checked for accurate transmission by simply counting the bits for an odd parity indication. If the count is ever even, an error is indicated.

PARITY DATA:

A block of information mathematically created from several blocks of user data to allow recovery of user data contained on a drive that has failed in an array; used in RAID levels 3 and 5.

PARKING:

Parking the disk drive heads means the recording heads are moved so that they are not over the platter's data area. Many drives have an auto-park feature where the heads are automatically parked when power to the drive is shut off. Other drives require the user to run some kind of parking software to park the heads.

PARTIAL RESPONSE MAXIMUM LIKELIHOOD ( PRML ) :

A method for detecting and sampling a signal generated from the reading of a hard disk or tape. Applied in conjunction with MAGNETO-RESISTIVE (MR) HEADs , PRML enables extremely high AREAL DENSITIES to be achieved on magnetic media.

PARTITION:

A portion of a hard drive dedicated to a particular operating system or application and accessed as a single logical volume.

PARTITIONING:

Method for dividing an area on disk drive for use by more than one disk operating system or for dividing large disk drives into areas which the File Allocation Table (FAT) can deal with when in use.

PATH:

The DOS term "path" has three definitions and each definition involves directories. A PATH may be defined as: (1) the NAMEs of the chain of directories leading to a file; (2) the complete file or directory NAME; (3) a DOS command.

PATTERNED MEDIA:

A set of technologies for future storage devices that approach the issue of areal density limitations imposed by superparamagnetism by writing data onto extremely small bit locations on media that is pre-etched with "mesas and valleys" in order to preclude cross-over and bit flipping.

PCB:

See [PRINTED CIRCUIT BOARD]

PEER-TO-PEER ARCHITECTURE:

A network of two or more computers using the same programs or types of programs to communicate and share data.

PERFORMANCE:

A measure of the speed of a hard drive during normal operation. Factors affecting performance are seek times, transfer rate, and command overhead.

PERIPHERAL EQUIPMENT:

Auxiliary memory, displays, printers, disk drives, and other equipment usually attached to computer systems' CPU by controllers and cables (they are often packaged together in a desktop computer).

PERPINDICULAR RECORDING:

A method of digital recording on a magnetic material in which the bits are arranged vertically rather than horizontally in order to reduce crossover effects of their electromagnetic fields. Perpendicular recording is one method for coping with the effects of superparamagnetism, which imposes a limit on areal density scaling in conventional magnetic media.

PERSISTENT BINDING:

Persistent binding is a host-centric method for enabling an operating system to assign specific and non-changing SCSI target IDs and LUNs to certain storage resources. Operating systems and upper-level applications (such as backup software) typically require a static or predictable SCSI target ID for their storage reliability.

PERSONAL COMPUTER INTERCONNECT ( PCI ) :

An industry-standard bus used in servers, workstations, and PCs.

PETABYTE:

1,024 terabytes.

PHASE-CHANGE TECHNOLOGY:

An optical disk storage technology that uses a plastic disk and metal recording layer to store data. Heat generated by the drive's laser changes the molecular structure of the metal, transforming it from an amorphous to highly reflective crystalline state. The changes in reflectivity are detected during reads. Used for CD-RW and DVD re-writable disks.

PIXIE DUST:

A term coined by IBM engineers to describe Anti-Ferromagnetic-Coupled coating for hard disk platters. It refers to a three atom thick layer of ruthenium sandwiched between two magnetic layers. The magnetic layers allow for thinner bits, and the bit above and below are polarized in opposite directions. This technology pushes back the superparamagnetic limit to between 150 and 200 Gb/in ² .

PLATED MEDIA:

Disk platters that are covered with a hard metal alloy instead of an iron-oxide compound. Plated disks can store greater amounts of data than their oxide-coated counterparts.

PLATED THIN FILM DISKS:

Magnetic disk memory media having its surface plated with a thin coating of a metallic alloy instead of being coated with oxide.

PLATFORM:

A hardware standard, such as IBM, Sun, or Macintosh.

PLATTER:

An actual metal (or other rigid material) disk that is mounted inside a fixed-disk drive. Many drives use more than one platter mounted on a single spindle (shaft) to provide more data storage surfaces in a small package.

PLUG AND PLAY:

An auto detect method used by the system BIOS to identify and configure peripheral devices. Identification numbers, interrupts and port addresses are set when the system boots.

POLLING:

A technique for allocating CPU cycles to specific peripheral devices and tasks.

PORTABILITY:

The ability to move storage media from one point to another. Tapes and optical disks are highly portable since they can be easily moved from a working environment to a different location for storage.

POSITIONER:

See [ACTUATOR]

PRECOMPENSATION:

A technique used with some oxide media drives to write data bits closer together on the disk in order to offset the repelling effect (bit shift) caused by magnetic recording.

PREVENTIVE MAINTENANCE:

A method of doing a scheduled routine observation or exchanging a part, prior to a breakdown of a piece of equipment.

PRIMARY STORAGE:

Online storage of electronic data, typically found on a computer's hard disk. This includes frequently used data, work in process, or data that is not frequently used but must be immediately available at all times.

PRINTED CIRCUIT BOARD ( PCB ) :

The circuit board with the chips attached to a drive.

PROCESSING (DATA PROCESSING):

The handling, manipulating, and modifying of data by a computer in accordance with software instructions.

PROGRAM:

A sequence of instructions stored in memory and executed by a processor or microprocessor. ( See also APPLICATION PROGRAMS .)

PROPRIETARY:

Privately developed and owned technology.

PROTOCOL:

A set of conventions governing the format of messages to be exchanged within a communications system.

PRML:

See [PARTIAL RESPONSE MAXIMUM LIKELIHOOD]

QUARTER-INCH CARTRIDGE ( QIC , pronounced "Quick"):

A tape media recording technology that uses a mini-cartridge for tape storage. There are many QIC standards, each defining a method for reading and writing data to tapes.

RACKMOUNT:

The cabinet that houses a server/storage workstation (also referred to as a server rack); to mount equipment into a cabinet.

RADIAL:

A way of connecting multiple drives to one controller. In radial operation, all output signals are active even if the drive is not selected.

See also [DAISY CHAIN]

RAID (Redundant Array of Inexpensive Disks):

A method of combining hard disks into one logical storage unit, which offers disk-fault tolerance and can operate at higher throughput levels than a single hard disk.

RAID ADVISORY BOARD ( RAB ) :

An organization of storage system manufacturers and integrators dedicated to advancing the use and awareness of RAID and associated storage technologies; started in 1992, RAB states its main goals as education, standardization, and certification.

RAM:

See [RANDOM ACCESS MEMORY]

RAM DISK:

A DOS operation, where part of the computer's random access memory is used to simulate a disk drive. The RAM disk and its contents will disappear if power is lost or DOS MAIN MEMORY is restarted. RAM is far faster (microseconds ACCESS TIME ) than disks (milliseconds), so APPLICATION PROGRAMS that access the disk run faster.

RANDOM ACCESS:

The ability to skip randomly from track to track on a storage medium. Optical disks, magnetic hard disks, and audio CDs allow random access to data tracks. Audio tapes, by comparison, allow only sequential access (i.e., fast-forward or reverse) to locate stored data.

RANDOM ACCESS MEMORY ( RAM ) :

An integrated circuit memory chip that allows information to be stored and retrieved by a microprocessor or controller. The information can be stored or accessed in any order, and all storage locations are equally accessible. Random access memory usually refers to volatile memory where the contents are lost when power is removed. The user addressable memory of a computer is random access memory.

RDMA:

See [REMOTE DIRECT MEMORY ACCESS]

RDMA OVER IP:

Remote Direct Memory Access Over IP is an implementation of RDMA that provides transmission of IP packets from the memory of one computer to the memory of another without involving the CPU.

See also [REMOTE DIRECT MEMORY ACCESS]

READ:

To access a storage location and obtain previously recorded data.

READ AFTER WRITE:

A mode of operation requiring that the system read each sector after data is written, checking that the data read back is the same as the data recorded. This operation lowers system speed but raises data reliability.

READ ONLY MEMORY ( ROM ) :

A chip that can be programmed once with bits of information. This chip retains this information even if the power is turned off. When this information is programmed into the ROM, it is called burning the ROM.

READ VERIFY:

A data accuracy check performed by having the disk read data to the controller, which then checks for errors but does not pass the data on to the system.

READ/WRITE HEAD:

See [HEAD]

REAL-TIME:

Immediate processing of input or notification of status.

RECALIBRATE:

Return to Track Zero. A common disk drive function in which the heads are returned to track 0 (outermost track).

RECORD:

A record is a single unit made up of logically related fields.

REDUCED INSTRUCTION SET COMPUTER ( RISC ) :

A computer processing architecture that requires fewer instructions to run applications, thus increasing processing speed.

REDUCED WRITE CURRENT:

A signal input (to some older drives) which decreases the amplitude of the write current at the actual drive head. Normally this signal is specified to be used during inner track write operations to lessen the effect of adjacent bit "crowding." Most drives today provide this internally and do not require controller intervention.

REMOTE DIRECT MEMORY ACCESS ( RDMA ) :

A communications protocol that provides transmission of data from the memory of one computer to the memory of another without involving the CPU. InfiniBand, Virtual Interface (VI), and RDMA Over IP are all forms of RDMA. Implemented in hardware on the network adapter (NIC), RDMA techniques have been developed to accommodate the ever-increasing network speeds.

RESOLUTION:

In regard to magnetic recording, the bandwidth (or frequency response) of the recording heads.

RESTORE:

The act of copying files or data from a backup storage device to their normal location on a computer's hard disk, often to replace files or data that were accidentally lost or deleted.

RLL:

See [RUN LENGTH LIMITED]

ROBOT:

A machine that can sense and react to input, and cause changes in its surroundings with some degree of intelligence, ideally with no human supervision.

ROBOTICS:

The internal components of an optical jukebox or automated tape library, usually consisting of a mechanical arm that automatically transports media inside the cabinet for use and storage.

ROM:

See [READ ONLY MEMORY]

ROTARY ACTUATOR:

See [ACTUATOR]

ROTATIONAL LATENCY:

See LATENCY, ROTATIONAL .

ROUTER:

An electronic device that connects two or more networks and routes incoming data packets to the appropriate network.

RUN LENGTH LIMITED ( RLL ) :

A method of recording digital data in which the combinations of flux reversals are coded/decoded to allow greater than one (1) bit of information per flux reversal. This compaction of information increases data capacity by approximately 50 percent.

SAN HUB:

This is a simple connectivity device that allows for devices to be connected to a Fibre Channel loop by being attached to a hub port. The advantage of this is that failures of a single device on the loop can be isolated from the other ports on the loop. The aggregate bandwidth of the hub is still that of a single Fibre Channel loop however.

SAS:

See [Serial Attached SCSI]

SATA:

See [SERIAL ATA]

SCALABLE:

The ability of a product or network to accommodate growth.

SCALEABLE LINEAR RECORDING:

A tape format from Tandberg Data ASA.

SCATTER/GATHER:

A feature which allows data to be transferred to or from multiple discontiguous areas of host computer memory with a single I/O command.

SCSI:

See SMALL COMPUTER SYSTEMS INTERFACE .

SCSI OVER INTERNET or INTERNET SCSI ( iSCSI ) :

iSCSI is an IETF standards-based protocol for operating SCSI as an application across a TCP/IP network. It provides an IP-based alternative to Fibre Channel for interconnecting servers and storage devices in a SAN and may also be used to connect devices to a Fibre Channel fabric via a hybrid iSCSI/FC Fabric switch.

SECTOR:

A sector is a section of a track whose size is determined by formatting. When used as an address component, sector and location refer to the sequence number of the sector around the track. Typically, one sector stores one user record of data. Drives typically are formatted from 17 to 26 sectors per track. Determining how many sectors per track to use depends on the system type, the controller capabilities, and the drive encoding method and interface. On Macintosh and UNIX drives, sectors are usually grouped into blocks or logical blocks that function as the smallest data unit permitted. Since these blocks are often defined as a single sector the terms block and sector are sometimes used interchangeably in this context. ( Note: The usage of the term block in connection with the physical configuration of the disk is different from its meaning at the system level.)

SECTOR-SLIP:

Sector-slip allows any sector with a defect to be mapped and bypassed. The next contiguous sector is given that sector address.

SEEK:

The radial movement of the heads to a specified track address.

SEEK TIME:

The time required to move between tracks when seeking data.

SELF-DESCRIBING DATA:

Data that carries with it a header or other mechanism providing descriptive information regarding its origins, criticality, stale date, and other information that might facilitate its lifecycle management.

SERIAL ATA ( SATA ) :

SATA is a serial implementation of the Parallel ATA physical storage interface. Serial ATA is a serial link ”a single cable with a minimum of four wires creates a point-to-point connection between devices. Transfer rates for Serial ATA begin at 150 MBps. One of the main design advantages of Serial ATA is that the thinner serial cables facilitate more efficient airflow inside a server cabinet and permits smaller chassis designs. In contrast, IDE cables used in parallel ATA systems are bulkier than Serial ATA cables and can only extend to 40 cm long, while Serial ATA cables can extend up to one meter.

SERIAL ATTACHED SCSI ( SAS ) :

SAS is a serialized implementation of the parallel SCSI bus. Features include smaller cables and connectors, full duplex connections, the capability of addressing thousands of devices per port, and an initial transfer rate of 3 Gbps. Smaller connectors enable dual-ported SAS drives, reducing the risk of a failing controller and extending the compact 2.5-inch format to enterprise-class drives, with obvious benefits in space and power requirements. SAS will have point-to-point connections; but by using expanders ” essentially , fan-out devices ”the new protocol makes it possible to address multiple devices from a single port. SAS and SATA devices may be connected to the same controller (but on different ports) enabling the new protocol to consolidate the best characteristics of SCSI, FC, and SATA into a single interface.

SERIAL STORAGE ARCHITECTURE ( SSA ) :

A high-speed method of connecting disk, tape, and CD-ROM drives, printers, scanners , and other devices to a computer.

SEQUENTIAL ACCESS:

Writing or reading data in a sequential order, such as reading data blocks stored one after the other on magnetic tape (the opposite of random access).

SERVER:

A computer that provides access to a network and its resources, runs administrative software controls, and provides services (such as file storage and retrieval) for desktop computers.

SERVER-ATTACHED STORAGE:

A storage system that is connected directly to the network server; also referred to as direct-attached storage.

SERVO DATA:

See EMBEDDED SERVO .

SERVO MOTOR:

A motor used to position the actuator arm and read/write heads on hard disk media.

SERVO SURFACE:

See DEDICATED SERVO .

SERVO TRACK:

A prerecorded reference track on the dedicated servo surface of a closed-loop disk drive. All data track positions are compared to their corresponding servo track to determine "off-track/on-track" position. Information written on the servo surface that the electronics of the drive uses to position the heads over the correct data track. This information is written on the drive by the servo track writer.

SETTLE TIME:

The interval between the arrival of the read/write head at a specific track and the lessening of the residual vibration to a level sufficient for reliable reading or writing.

SETUP:

Program used by at type computers to store configuration in CMOS. This program is sometimes found in the system bios and can be accessed from the keyboard. On other systems, the program is on a diskette.

SHOCK RATING:

A rating, expressed in terms of the force of gravity (Gs), of how much shock a disk drive can sustain without damage.

SILICON:

Semiconductor substrate material generally used to manufacture micro-processors and other integrated circuit chips.

SIMPLE NETWORK MANAGEMENT PROTOCOL ( SNMP ) :

A standard protocol that runs over an IP link to provide management of network type devices without performing continual polling.

SINGLE-ENDED:

An electrical signal protocol which transmits information through changes in voltage. Single-ended SCSI uses standard TTL signal-and-ground pairs to transmit information over the SCSI bus.

SKEWING:

Some low-level formatting routines may ask for a Head and/or Cylinder Skew value. The value will represent the number of sectors being skewed to compensate for head switching time of the drive and/or track-to-track seek time allowing continuous read/write operation without losing disk revolutions.

SMALL COMPUTER SYSTEM INTERFACE ( SCSI ) :

An industry standard for connecting peripherals such as printers, scanners, optical drives, and tape drives to a microprocessor. SCSI covers both hardware and software standards for allowing computers and peripherals to communicate with each other.

SMART:

Self-Monitoring, Analysis, and Reporting Technology can help prevent data loss and unscheduled computer downtime. It provides advanced warning of certain types of drive failures, allowing the user of data-management software to backup the data. This technology is not supported by all system manufacturers or hard drive distributors and requires third party software.

SMS:

See [SYSTEM MANAGED STORAGE]

SOFT ERROR:

A bit error during playback which can be corrected by repeated attempts to read, usually caused by power fluctuations or noise spikes.

SOFT SECTOR:

A convention, defined by software, of setting a variable numbers of sectors per track in direct relationship to the drive's FCI rating in regards to the area of media that passes beneath the head. This schema takes advantage of the fact that, in actual surface area, the outermost tracks are longer than the innermost.

SOFT SECTORED:

A term describing a hard drive that determines the starting location of each sector from information stored in data fields. This method is older and results in more overhead than hard sectored techniques.

See also [HARD SECTORED]

SOFT ZONING:

A method of defining and maintaining SAN zones though software. The zoning process uses the name server database located in the fibre-channel switch, which stores port numbers and world wide names (WWN) used to identify devices during the zoning process. The devices in the database receive Registered State Change Notification (RSCN) when a zone change is made. In order to change related communication paths, each device must correctly address the RSCN. Any device that does not correctly address the RSCN and continues to transfer data to a specific device after a zoning change, will be blocked from communicating with its targeted device.

SOFTWARE APPLICATION PROGRAMS:

Disk operating systems and other programs (as opposed to HARDWARE ). The instructions or programs, usually stored on floppy or hard disks, which are used to direct the operations of a computer, or other hardware.

SOFTWARE PATCH:

Software modification which allows or adds functions not otherwise available using the standard software program.

SPINDLE:

The drive's center shaft, on which the platters are mounted. A synchronized spindle is a shaft that allows two disks to spin simultaneously as a mirror image of each other, permitting redundant storage of data.

SPINDLE MOTOR:

The spindle motor is the electro-mechanical part of the disk drive that rotates the platters.

SPUTTER:

A special method of coating the disk that results in a hard, smooth surface capable of storing data at a high density.

SSA:

See [SERIAL STORAGE ARCHITECTURE]

ST-506/ST-412 INTERFACE:

One of several industry standard interfaces between a hard disk and hard disk controller. In the ST-506/ST-412 interface, the "intelligence" is on the controller rather than the drive.

See also [INTERFACE STANDARD]

See also [ESDI]

See also [SCSI]

STEP PULSE:

The pulse sent from the controller to the stepper motor on the step interface signal line to initiate a step operation.

STEP TIME:

The time required by the drive to step the heads from the current cylinder position to a target cylinder.

STEP:

An increment or decrement of the head positioning arm to move the heads in or out, respectively, one track from their current position. In buffered mode (open loop drives), the head motion is postponed until the last of a string of step pulses has been received.

STEPPER:

A type of motor that moves in discrete amounts with each electrical pulse. Steppers were originally the most common type of actuator engine, since they can be geared to advance a read/write head one track per step.

STEPPER MOTOR:

The stepper motor is the electro-mechanical part of the disk drive that positions the heads by step pulse on the tracks of the disk to read and write data.

STORAGE APPLIANCE:

Concept of an intelligent, network-attached, storage device.

STORAGE AREA NETWORK ( SAN ) :

A network comprising multiple hosts and storage peripherals, currently conceived as Fibre Channel/SCSI Command Set-based. However, any interconnect and any network protocol could be used, theoretically, to establish a SAN, provided that the strict latency and throughput requirements of storage are met.

STORAGE CAPACITY:

Amount of data that can be stored in a memory, usually specified in kilobytes (KB) for main memory and floppy disk drives and megabytes (MB) for hard disk and tape drives. The maximum amount of data that can be stored on a given media.

STORAGE DENSITY:

Usually refers to recording density (BPI, TPI, or their product, AREAL DENSITY ).

STORAGE LOCATION:

A memory location, identified by an ADDRESS , where information is to be read or written.

STORAGE MODULE DRIVE ( SMD ) :

Storage module drive interface. An interface, used in larger disk drives, e.g., 14" drives.

STRIPE:

A contiguous region of disk space. Stripes may be as small as one sector or may be composed of many contiguous sectors.

STRIPING:

A method of storage in which a unit of data is distributed and stored across several hard disks, which improves access speed but does not provide redundancy. Also called RAID-0.

SUBSTRATE:

In disk technology, the material underneath the magnetic coating of a platter. Common substrates include aluminum or magnesium alloys for hard drives, glass for optical disks, and mylar for floppy disks.

SUPERPARAMAGNETISM:

In storage technology, when the magnetic energy holding bits in their recorded state becomes equal to the thermal energy generated by the drive itself, random bit flipping can occur. This is referred to as the superparamagnetic effect or SPE. Superparamagnetism imposes a limit on the areal density improvements possible with contemporary magnetic disk technology.

SURFACE:

The top or bottom side of a platter, which is coated with the magnetic material for recording data. On some hard drives, one of the surfaces on one of the platters is reserved for servo data.

SURFACE MOUNTED DEVICE ( SMD ) :

A CHIP in a smaller integrated surface package, without connection leads.

SUSTAINED MODE:

The measured transfer rate of a given device during normal operation.

SWITCH:

A network traffic monitoring device that controls the flow of traffic between multiple network nodes.

SYNCHRONOUS DATA:

Data sent, usually in serial mode, with a clock pulse.

SYSTEM MANAGED STORAGE ( SMS ) :

Software used to routinely back up and archive files. Also abbreviation of DFSMS: enhanced data management software for MVS mainframes from IBM. Introduced in 1988, it provides functions such as automatically allocating data, which prevents most out-of-space errors when disk volumes become full. Also a moniker for Storage Management Services software from Novell that allows data to be stored and retrieved on NetWare servers independent of the file system the data are maintained in (DOS, OS/2, Mac, etc.). It is used to back up data from heterogeneous clients on the network. Various third-party backup products are SMS compliant.

SYSTEMS INTEGRATOR:

An individual or company that combines various components and programs into a functioning system, customized for a particular customer's needs.

TAPE DRIVE:

A sequential access memory device whose magnetic media is tape in a cassette, reel, or continuous loop.

TARGET:

A device that performs an operation requested by an initiator.

TELCO:

Abbreviation for "telecommunications company."

TERABYTE:

Terabyte equals 1,099,511,627,776 bytes or approximately one trillion bytes.

TERMINATION:

A method of matching the transmission impedance of a electrical bus so as to eliminate signal reflections from the physical ends of the bus.

THIN FILM:

A type of coating deposited on a flat surface through a photolithographic process. Thin film is used on disk platters and read/write heads.

THIN FILM HEADS:

A read/write head whose read/write element is deposited using integrated circuit techniques rather than being manually fabricated by grinding ferrite and hand winding coils.

THIN SERVER:

Name given to a network-attached device with an embedded micro-kernel operating system. Storage appliances are thin servers, as are many NAS devices.

THROUGHPUT:

A performance measurement indicating the volume and speed of data as it flows from one point to another through a data pipeline. High throughput indicates a system architecture that can carry high volumes of data at high speeds, resulting in high system performance.

TOP COVER:

Together with the base casting, creates an airtight, extremely clean environment. Any attempt to remove this top cover outside of a clean room will immediately contaminate and ruin a disk drive. Needless to say, any warranty is immediately voided if the top is removed.

TOPOLOGY:

Geometric arrangement of nodes and cable links in a local area network; may be either centralized, and decentralized.

TPI:

See [TRACKS PER INCH]

TRACK:

A track is the circular ring traced over the disk surface by a head as the disk rotates under the heads. Also known as a channel.

TRACK ACCESS TIME:

See [AVERAGE ACCESS TIME]

TRACK DENSITY:

See [TRACKS PER INCH]

TRACK FOLLOWING SERVO:

A closed-loop positioner control system that continuously corrects the position of the disk drive's heads by utilizing a reference track and a feedback loop in the head positioning system.

See also [CLOSED LOOP]

TRACK PITCH:

Distance from centerline to centerline of adjacent tracks (TPI divided into 1.0).

TRACKS PER INCH ( TPI ) :

Tracks per inch. The number of tracks written within each inch of the disk's surfaces, used as a measure of how closely the tracks are packed on a disk surface. Also known as track density.

TRACK-TO-TRACK SEEK TIME:

The time required for the read/write heads to move to an adjacent track.

TRACK WIDTH:

Width of data track. Also called core width of Read/Write Head.

TRACK ZERO:

Track zero is the outermost data track on a disk drive. In the ST 506 INTERFACE, the interface signal denotes that the heads are positioned at the outermost cylinder.

TRACK ZERO DETECTOR:

An obsolete technology that RECALIBRATES by sensing when infrared beams between a LED and infrared sensitive photo- transistor are blocked by the track zero interrupter (TZI).

TRANSFER RATE:

The rate of speed at which data travels through a bus or device, typically measured in bits, bytes, kilobytes or megabytes per second. The rate at which the disk sends and receives data from the controller. The sustained transfer rate includes the time required for system processing, head switches, and seeks and accurately reflects the drive's true performance. The burst mode transfer rate is a much higher figure that refers only to the movement of data directly into RAM.

TUNNEL ERASE:

An erase scheme where both sides of the recorded data is erased when writing data to eliminate track to track interference. This is primarily used on floppy disk drives.

TURBOCODE:

A type of channel coding that uses a convolutional code and a type of Viterbi decoder that outputs a continuous value rather than a 0 or 1. Convolutional coding adds patterns of redundancy to data in order to improve the signal to noise ratio for more accurate decoding at the receiving end. The Viterbi algorithm is used to decode a particular type of convolutional code.

TURN TIME:

The constant rate of rotation of a hard disk platter, typically measured in rotations per minute (RPM). Typical disks have rotation speeds of 4,500 to 7,200 RPM, but 10,000 RPM drives are beginning to enter the market.

TURNKEY:

A product or system that can be plugged in, turned on, and operated with little or no additional configuring.

ULTRA-SCSI:

A variant of SCSI. Doubles the bandwidth of SCSI Fast. It provides 8-bit (SCSI Narrow) data rates of 20 Mbytes per second and 16-bit (SCSI Wide) data rates of 40 Mbytes per second. Shorter cables may be required. Formerly known as Fast-20.

UNFORMATTED CAPACITY:

The total number of bytes on a disk, including the space that is required to record location, boundary definitions, and servo data.

See also [FORMATTED CAPACITY]

UNIX:

An operating system that supports multitasking and is ideally suited to multi-user applications (such as networks).

UPGRADE PATH:

Generally, with disk products, a family having multiple products with varying capacities such that the system storage capacity can increase with changing application requirements simply using a different disk drive within the product family.

VAR (Value-Added Reseller):

A business that repackages and improves hardware (or software) manufactured by an original equipment manufacturer (OEM).

VCX:

A tape technology from Ecrix Corporation.

VERIFICATION:

A process of re-reading data just written to disk to ensure the data was written correctly.

VOICE COIL:

A fast and reliable actuator motor that works like a loudspeaker, in which the force of a magnetic coil causes a proportionate movement of the head. Voice coil actuators are more durable than their stepper counterparts, since fewer parts are subject to daily stress and wear.

VOICE COIL MOTOR:

An electro-magnetic positioning motor in the rigid disk drive similar to that used in audio speakers . A wire coil is placed in a stationary magnetic field. When current is passed through the coil, the resultant flux causes the coil to move. In a disk drive, the CARRIAGE ASSEMBLY is attached to the voice coil motor. Either a straight line (linear) or circular (rotary) design may be employed to position the heads on the disk's surface.

VOLATILE:

Memory that will be erased if power is lost. Typically, MAIN MEMORY is volatile, while AUXILIARY MEMORY is non-volatile and can be used for permanent (but changeable at will) storage of programs and data.

WAN:

See [WIDE AREA NETWORK]

WEB CACHE:

A Web cache fills requests from the Web server, stores the requested information locally, and sends the information to the client. The next time the web cache gets a request for the same information, it returns the locally cached data instead of searching over the Internet, thus reducing Internet traffic and response time.

WEB SITE:

A location on the World Wide Web that is owned and managed by an individual, company, or organization; usually contains a home page and additional pages that include information provided by the site's owner, and may include links to other relevant sites.

WEDGE SERVO SYSTEM:

A certain part of each CYLINDER contains servo positioning data. Gap spacing between each sector contains servo data to maintain position on that cylinder.

WEDGE SERVO:

See EMBEDDED SERVO .

WIDE AREA NETWORK ( WAN ) :

A network that uses high-speed, long-distance communications technology (e.g., phone lines and satellites ) to connect computers over long distances.

WIDE SCSI:

The Wide SCSI interface provides a 16-bit wide SCSI bus, as compared to the narrow 8-bit SCSI bus. The wider 16-bit bus provides a transfer rate of 20 Mbytes per second compared to 10 Mbytes per second with Fast SCSI 8-bit interface.

WINCHESTER DISK:

Former code name for an early IBM hard disk model, sometimes still used to refer to the technology and design of most traditional hard drives.

WINCHESTER DRIVE:

A disk drive with a Winchester head and non-removable (fixed) disks sealed in a contaminant-free housing.

WORD:

Number of bits processed in parallel (in a single operation) by a CPU. Standard word lengths are 8, 16, 32, and 64 (1, 2, 4, or 8 bytes).

WORLD WIDE WEB ( WWW ) :

A global hypertext system operating on the Internet that enables electronic communication of text, graphics, audio, and video.

WORM:

See [WRITE ONCE READ MANY]

WRITE:

To access a storage location and store data on the magnetic surface.

WRITE CURRENT:

The optimum HEAD write current necessary to saturate the magnetic media in a cell location.

WRITE IMMEDIATE:

With host-controlled write immediate, status is returned when data is transferred to the drive buffer instead of waiting until the data is written to the media. The seek, latency, and write times are cut out of the total command completion as seen by the host.

WRITE ONCE READ MANY ( WORM ) :

An optical storage technology that burns pits into the recording layer of an optical disk, allowing the disks to be written just once but read without limit. WORM drives write directly to an optical disk from a host computer. Both the drives and disks include built-in safeguards to assure that data, once written, cannot be erased, overwritten or altered. Tape-based WORMs have just begun to enter the market.

WRITE PROTECT:

The use of various safeguards to prevent a computer system from overwriting a storage medium. Floppy disks have a sliding tab for "physical" write protect. Hard disks support "logical" write protect in software. Optical disks often use a combination of physical and logical write protect safeguards.

XOR ENGINE:

Process or set of instructions that calculates data bit relationships in a RAID subsystem.

ZERO COMMAND OVERHEAD:

To reduce command overhead to zero, processing traditionally done through software can be placed in the hardware, where it is completed almost instantaneously. Zero command overhead yields a substantial improvement in system performance.

ZERO LATENCY READ:

This reduces the delay in transferring data from the drive to the initiator due to rotational latency delays. Data is read out of order from the disk and transferred to the host where the requested order is restored.

ZONING:

This is the term used by some switch companies to denote the division of a storage area network or fabric into sub networks or fabrics that provide different levels of connectivity or accessibility between specific hosts and devices on the network. In effect routing tables are used to control access of hosts to devices. This zoning can be performed by cooperative consent of the hosts or can be enforced at the switch level. In the former case, hosts are responsible for communicating with the switch to determine if they have the right to access a device.