Storage Networking Fundamentals: An Introduction to Storage Devices, Subsystems, Applications, Management, and File Systems (Vol 1)

A component that is performing work or is waiting to perform work.

An architecture for high availability where all components are active and can assume extra work if another component fails.

An architecture for high availability where one or more components are idle, waiting to take over for another that fails.

A mechanism in disk drives that rotates and moves disk arms across disk platters.

The combination of storage address spaces, making a larger space from multiple smaller storage address spaces.

A file system process that selects the storage blocks to use in creating or updating a file.

A special-purpose server or device that is easy to install and use.

The method of determining which initiator or target gets access to the bus or network. Used with parallel SCSI and Fibre Channel loops.

A copy of data written to secondary storage for the purpose of making historical versions of files.

The ratio of storage capacity to area on magnetic storage media media.

An assembly that connects disk drive heads to the disk actuator.

A logical structure of same-sized storage address spaces configured as a single storage entity with a composite storage address space. Can be manipulated by virtualization techniques. Usually operated as a RAID array.

A method of I/O processing that allows unacknowledged I/Os.

Advanced Technology Attachment. An interface for desktop disk drives.

File system metadata that is used to manage data. Attributes indicate basic information about files, including file creation and access dates and times, size, access rights, and update indicators.

The average delay incurred in disk I/O waiting for heads to be relocated from one track to another.

The base layer of tape media, not magnetic.

The common rigid connecting circuit board that carries all I/O traffic in a subsystem or system.

The process of copying data to secondary storage for the purpose of restoring it when it is needed.

Backup management software and the system it runs on.

The amount of time in a 24-hour period in which to perform backup operations.

The adhesive used to adhere the magnetic layer to the backing in tape media.

A granular unit of storage capacity used by a file system or database.

A location of a block of storage within a storage address space.

Memory used in data transfers to temporarily store data.

Storage located in proximity to the main data center that allows high-performance copies of data to be made (either remote copy or mirroring). Bunker storage is typically used to "stage" data that will be transmitted to storage a longer distance away.

A business practice to prepare for the resumption of IT operations following a disaster.

Contiguous data locations within a file, where bytes are used as the granular units of storage.

Temporary storage of data to enhance performance.

Content Addressable Storage. A way to identify and manage data based on its contents as opposed to its name or metadata.

Command Descriptor Block. A SCSI command protocol data unit.

Cluster File System.

Common Internet File System. The evolution of Server Message Block in Windows technology.

A high-availability architecture where more than one system can provide application services.

A lubricating layer on tape media.

A backup process that runs while applications are not running.

A granular unit of work in storage operations.

A method of command processing that allows logical units to determine the order in which tasks are processed.

A method of reducing the number of bits needed to convey data.

Appending one storage address space to another to create a larger one.

Along with filing and storing, one of three main elements in storage networks having to do with buses or networks.

A complete recording of all related data on media as intended by the writing application and including layout reference data updates and metadata.

Adjoining and in order. Usually used with blocks, as in contiguous blocks.

A combination of hardware and software used to manage and control data transmissions.

A system function that manages storage transfers. Usually a hardware implementation with firmware.

A process that allows data management to be performed on files while they are being updated by applications.

Multiple tracks used together on parallel platters.

Disk-to-disk backup.

Disk-to-disk-to-tape backup.

Disk-to-tape backup.

Direct Attached Storage.

The process of relocating data from one system or subsystem to another one.

An initiator that performs third-party copy processes.

Multiple systems that can access common data.

A RAID array operating with a failed member.

A redundancy method where data can be recreated based on applying or subtracting change data to or from a known state.

A storage drive. A disk drive or tape drive.

A component of a SCSI logical unit that is responsible for receiving and responding to commands from initiators.

Distributed File System.

A backup operation that copies all data that has changed since the last full backup.

An extended-distance SCSI bus technology.

A method of processing updates in host systems that bypasses host buffer memory and transfers writes and updates directly to storage.

A file containing a listing of files and serving as an access point for those files within a name space.

A granular component of disk storage; a device with rotating media platters.

A rigid rotating medium coated with a magnetic surface and used in disk drives.

A limit placed on the amount of storage capacity available to a specific user or application.

In the direction away from an initiator.

A storage controller with both target and initiator functionality.

A redundancy method where data is duplicated on two independent components.

A file system process that adds block storage capacity to an storage address space without stopping system operations and halting application processing.

Virtual or real storage in a disk subsystem.

The inverse of failover. Changing work from a redundant component back to a reinstated or replaced component.

The process of changing work from a failed, active component to other active, redundant components. May involve the activation of a passive component.

The ratio of downstream to upstream entities in a network.

The logical segregation of resources and processes used in network operations.

A networking technology developed in the 1990s that was the first network used for SANs.

A data storage abstraction used by applications to store data by relative byte addressing.

A type of storage transfer that is used between applications and operating systems.

The server application that services client requests for file I/O in a network file system.

System software that determines the location of files in storage address spaces and provides access control for files and directories.

Along with connecting and storing, one of three main elements in storage networks having to do with locating data in storage address spaces and managing data objects.

Software that operates on circuit boards or adapters.

A method of governing data transfers in networks.

A controller used in remote copy systems to transmit data to a receiving controller.

The natural progression in a file system resulting in shorter contiguous available blocks.

Block addresses that are available to be allocated to files by a file system.

A filename appended to a complete directory path, beginning with the root directory.

A replicated, identical view of the file and directory structure of a cluster or distributed file system.

Host Bus Adapter. A system storage controller providing initiator functionality and acting as the storage interface between the system's I/O bus and a storage bus or network.

A microscopic electro-physical device responsible for recording and detecting data signals on media.

A low-latency message used in clusters to validate system availability.

An alignment used in tape equipment where the heads make diagonal tracks across tape media.

A characteristic describing equipment and implementations designed for minimal downtime.

A backup process that runs while applications are running.

A redundant component available to replace a failing or failed component.

Hierarchical Storage Management. A data management application that attempts to match the priority of data with the cost of storage.

Integrated Device Electronics. The same as ATA.

Information Life Cycle Management.

Unchanged; impossible to change.

A process that runs in a hardware or software component of the I/O path.

A backup operation that copies only data that has changed since the last backup.

A SCSI architectural component that creates an I/O command.

A command issued to discover storage resources accessible through a particular target address.

A file system that runs as a "guest" using the interfaces provided by the operating system.

A device connection method. Includes hardware, software, and cabling (bus or network).

Links between switches in a switched topology network.

A storage transfer.

A system bus used to connect adapters such as HBAs.

A sequence of hardware and software components traversed in the transmission of data between systems and storage.

A virtualization method for accessing data on a network file server. Runs in a client of a network file system.

A SCSI command to read or write data.

Used with storage virtualization to identify the processes used by an in-path virtual logical unit where storage commands are managed and received.

Just a Bunch of Disks. A disk subsystem that does not use RAID or mirroring.

A logging method in file systems that records all intended I/Os and their completion.

Operating system core processes that manage the execution of work in a system.

A mode of system operation that is typically reserved for system services and is not used for application processing.

A backup operation in a SAN where data is not transmitted over a LAN.

Delay in transmission or process.

The organization of files in a storage address space.

The method used to locate file data within a storage address space; in UNIX, these are the inodes or vnodes.

Logical Block Addressing. Used by disk drive controllers to virtualize storage locations inside disk drives.

A method of controlling the sequence of access and access permissions in a file system.

A file maintained by a database system that records all intended I/Os and their completion.

A ring topology typically implemented as a Fibre Channel device interconnect where disk drives and initiators share a single connecting path.

Logical Unit. A SCSI architectural component that receives and processes commands on behalf of a storage address space.

Logical Unit Number. A SCSI architectural component that provides an access point for a SCSI logical unit.

Low-Voltage Differential. The most recent implementation of the parallel SCSI bus with high data rates, cable lengths of 12 meters, and backward compatibility with single-ended SCSI.

The materials used for recording data.

A method of determining the usage of backup tapes.

Multiple media that are used for a common purpose, such as a pool of backup tapes.

The performance of transfers on storage media.

A storage address space functioning as a component in a RAID array.

Data used to describe data.

Cache memory belonging to a pair of storage controllers that are connected over a high-speed link. All updates and changes are reflected in both caches simultaneously.

A technique of creating duplicate copies of data on different storage address spaces.

Mean Time Between Failure. The statistical calculation of failure rates based on testing large numbers of products.

The use of redundant connections for high-availability access to a single storage logical unit.

A file system component that displays the logical organization of directories and files.

Network-Attached Storage.

A standalone network-attached storage computer system without integrated storage.

Network Data Management Protocol.

Secondary or tertiary storage that is not online but that can be quickly accessed through automated processes.

A relationship set of SCSI components made up of initiators, target, LUN, and LUs.

Network File System. The original network file system developed by Sun Microsystems.

Storage that cannot be immediately accessed.

Primary or secondary storage that is readily accessed.

A process that runs in a hardware or software component not in the I/O path.

The ability to process multiple SCSI commands on multiple devices connected to the same physical interconnect.

The ability to perform work in parallel.

A redundancy method where multiple pieces of data are used to generate a single piece of data. In RAID, the parity algorithm is XOR.

The process of re-creating data in a RAID array on a replacement member.

A logical structure establishing boundaries of a contiguous storage address space in a disk drive.

A component that is not used until it is needed to assume the work of a failed component.

The sequence of directory names, starting with the root directory, used to access a given directory.

Protocol Data Unit. Examples are frames, packets, and cells.

An I/O command that was issued and is waiting for a response.

A physical or virtual view of data that existed in the system at some previous point in time.

The re-creation of system data as it existed at a previous, particular time.

A line topology connecting two entities.

A physical connection to a network or bus.

Storage used to support application and system processing.

The process of defining LUs and LUNs in a subsystem, including interconnect, device, partitions, and RAID levels.

quality of service. A way to designate performance and service quality for systems or applications having designated priorities.

Redundant Array of Inexpensive Disks.

Parity RAID with a static parity location.

Parity RAID with a changing parity location.

A RAID array formed by striping data across mirrored pairs of storage address spaces.

An indication of RAID algorithms used in a RAID array.

A storage address space recognized by the operating system, but lacking file system management. Raw partitions are commonly associated with database storage where the database system provides its own layout reference system.

Data transfer from storage to system.

Data transmission electronics used inside a disk drive.

A controller used in remote copy systems to receive data from a forwarding controller and commit it to remote storage.

The use of multiple components to create higher data availability.

Used with storage virtualization to identify the process of creating one or more new storage commands from commands received and managed by a virtual logical unit.

A data management application that creates redundant data when a dual-mode storage controller forwards writes to a remote storage controller.

The process of establishing access to a network file system with an I/O redirector.

A command issued to check the status of the SCSI target, LUN, LU, or task.

The process of copying data back from secondary backup storage to an application system.

The top-level directory in a file system's hierarchy of directories.

The delay incurred in disk I/O while waiting for a sector to rotate beneath the heads.

The process of determining a new network path based on network algorithms.

An open-source software initiative to provide Windows server message block support on Linux systems.

Storage Area Network.

Serial Attached SCSI. The serial version of SCSI server interconnect technology.

Serial ATA. The serial version of the ATA desktop disk drive interconnect.

The ability to increase the size of a system; in storage, this usually means capacity.

Small Computer Systems Interface.

A SCSI standard separating physical and logical components, enabling the implementation of serial SCSI SANs.

Storage used to support data and storage management functions to make redundant copies of data.

A regularly sized subsection of a track on a disk platter.

Programmatic interfaces used by applications to access file system services.

A method of I/O processing allowing up to two pending write I/Os.

The implementation of SCSI protocols on independent communications networks.

A method of computing where multiple servers run the same applications and access the same data.

A backup operation in a SAN where data is not transmitted through a server system, instead using third-party copy.

A cluster storage design where all systems can access all storage.

A cluster storage design where storage is accessed through its corresponding controlling system.

A method of establishing a single partition on a disk drive to limit actuator movements and hence seek times.

The original bus technology used for parallel SCSI.

Server Message Block. A client/server protocol developed for Windows server systems in the early 1990s.

The process of creating a point-in-time copy of data.

A file system function that selects blocks in a storage address space for storing file contents.

A rotating shaft that spins disk platters.

Single Point of Failure. Describes single components that compromise high availability.

Storage Resource Management.

A mode of tape operation where tape is stopped and restarted.

A method of distributed processing where participating systems retain connection details following a communication interruption.

A method of distributed processing where participating systems do not retain connection details following a communication interruption.

Contiguous blocks used to store data. Managed as a granular, complete storage container by a single filing system.

Transmissions of storage data and commands.

Storage address spaces used for a common purpose, such as a pool of secondary storage or a pool of storage for disk backup.

A term used with virtualized disk storage. The aggregation and subdivision of storage address spaces to create storage address spaces.

A network device used to exchange storage transmissions across different types of storage connecting technologies, such as Fibre Channel and parallel SCSI.

Along with filing and connecting, one of the three main elements in storage networks having to do with the methods and operations of initiators, devices, and subsystems.

A mode of tape operation for sustained high performance.

A composite storage entity in RAID arrays formed from the set of aligned strips (all strips have the same relative block addresses in their corresponding members).

The amount of storage capacity (in blocks) corresponding to a defined RAID stripe: (the number of blocks per strip) * (the number of members in the array).

A granular unit of data written to a single member of a RAID array.

The process of writing storage I/Os to multiple storage address spaces. Usually done across multiple disk drives to reduce disk drive contention.

The segmentation of resources, making several smaller resources from a single larger resource. Partitioning a disk drive is one example.

Replacing the function of one resource with a different resource.

The rigid base layer of disk platters.

An assembly of storage devices, power distribution, cooling, and connection technology operating as a single integrated product.

A private storage area used by the file system to store configuration and operating status information.

The performance of transfer in and out of a disk drive.

A hub-and-spoke topology using network switches where end nodes have dedicated links and interswitch links are shared.

A method of I/O processing that requires a response for each write command before the next command is issued.

Memory used by an operating system to temporarily store data during I/O activity.

Files used by relational database systems to store user and application data.

The receiver of storage commands.

The component of a SCSI logical unit that is responsible for maintaining the order of the task (command) list.

Storage used to make third-generation copies of data.

A method of making data copies where a managing process uses an initiator located on another system or subsystem to issue read and write commands.

A storage management approach that attempts to match the cost of storage with the relative importance of the data stored on it.

Circular storage entities on disk platters having a unique address and many embedded sectors for storing data.

In the direction of an initiator.

A mode of system operation that is used for application processing.

The percentage of storage occupied by system and application data.

A storage address space made from a combination or subdivision of other storage address spaces. Typically accessed as a SCSI LUN in a SAN.

The creation of a storage address space from a subset or superset of other storage address spaces.

The operating system definition of a storage address space.

System software that manipulates storage address spaces to create RAID, mirroring, aggregation, and so on.

Write Once, Read Many. A technology that is used to store data that is inteneded to never be changed.

Data transfer from system to storage.

An I/O process that maintains the sequence of I/O write commands as issued by an initiator.

The process of reading old data and parity data before writing new data and parity in parity RAID 5 arrays.

World Wide Name. A unique 64-bit identifier used in Fibre Channel networking equipment.

World Wide Node Name. The WWN for a system or subsystem that may contain one or more ports, each with its own WWPN.

World Wide Port Name. The WWN for an individual network port.

An exclusive OR algorithm used in parity RAID.

A method of recording on disk drives that places more sectors in tracks as the radius of the track increases.