Skip navigation and jump to content
active/active array - A storage array whose multipathing solution allows efficient access to an array volume from more than one path at the same time.
active/passive array - A storage array whose multipathing solution relies upon only one path to an array volume being active at any one time.
array - A physical storage device (subsystem) with multiple hard drives and RAID functionality.
array adapter - An interface on an array controller that connects to a host or Fibre Channel network.
array cache - Memory inside a storage array used to store data until it can be written to disk, or to store a copy of data for efficient access. Caching policies and redundancy mechanisms are specific to the storage array vendor and model type.
array controller - A processor in a storage array that provides volume management and other storage services. A single storage array usually contains multiple controllers.
array volume - A SCSI logical unit presented by a storage array.
bound array volume - An array volume which has been assigned to management by the Incipient NSP Administrator.
cache - Memory used to store data until it can be written to disk, or to store a copy of data for efficient access.
clone - The target of a snapshot done using split-mirror technology, on which a complete copy of the data on the source at a specified point-in-time is made.
consistency group - A collection of Network-Volumes that are treated as a single unit for snapshot operations, ensuring referential data integrity across multiple volumes. Applications such as large databases or email servers may span multiple volumes, and the use of consistency groups guarantees that snapshots of all volumes represent a single, atomic point in time.
ControlPath - A component of the intelligent FC switch responsible for modifying metadata state, handling I/O errors, and managing coherency and synchronization with other software ControlPaths. There is one ControlPath for multiple FastPaths. The core Incipient NSP software is located on the ControlPath.
copy-on-write - A virtual copy method (Snapshot), which uses mapping to create a logical replica of a Network-Volume instantaneously, without immediately copying any data. Instead, data is only copied if it is imminently to be overwritten.
data mobility - A feature of Incipient NSP that allows the assignment of data to physical storage to be dynamically modified without host disruption. See also migration, virtualization.
Deep Discovery - An Incipient NSP Administrator operation which queries attached storage arrays for detailed information about the physical layout of array volumes, including their RAID properties and mappings to physical hard drives.
discovery - The process of obtaining information about the physical and logical devices connected to a SAN, either through in-band (fabric-based) or out-of-band (TCP/IP-based) queries.
encapsulated volume - A Network-Volume that maps directly to an underlying array volume. An array volume that contains data before virtualization is introduced may be encapsulated and presented as a Network-Volume in order to preserve the existing data and layout.
extent - A set of consecutively addressable disk blocks. An extent may be virtual or physical. Virtual extents are consecutively addressable disk blocks on a Network-Volume, while physical extents are consecutively addressable blocks on an array volume.
fabric - One or more interconnected Fibre Channel switches that can physically transmit data between any two N_Ports on any of the switches.
failover - The process of nondisruptively routing data to an alternate data path or device in the event of device or connection failure.
failure domain - The set of all components impacted by the failure of a single component. For example, a physical hard drive is contained within the failure domain of an entire array.
FastPath - The component of the intelligent FC switch responsible for routing almost all host I/O requests. This component is supplied by the switch vendors.
fault tolerance - A system's ability to continue to perform its function in response to a failure of some of its parts, so that data is not lost or corrupted and service is not disrupted.
Fibre Channel (FC) - Fibre Channel (FC) is a technology for networking computer devices. Fibre Channel is primarily used for attaching computer servers to shared storage devices and for interconnecting storage controllers and drives. See also storage area network (SAN).
high availability (HA) - The ability of a system to perform its function continuously, without interruption.
host - A computer that hosts applications and may initiate I/O commands to a target.
host bus adapter (HBA) - A host bus adapter (HBA) is the interface card in a host, which connects a server to a Fibre Channel network.
host cluster - A set of two or more hosts configured to cooperatively share SCSI volumes.
host-side port - Intelligent switch connection that can receive host I/O requests. For example, Incipient NSP target ports are considered host-side ports because they receive host I/O.
host-side multipathing - The use of multiple paths between a host (initiator) and a Network-Volume (target). Host-side multipathing requires the use of host software.
Incipient NSP Cluster - A set of two or more Incipient NSP nodes that act together as a single unit to provide storage services on the network.
Information Lifecycle Management (ILM) - A process which comprehensively and dynamically manages the assignment of storage resources to data, based on attributes such as the criticality of the data, its reference and update frequency, regulatory requirements, and retrieval costs.
initiator - A SCSI device that that sends I/O requests to another device (target).
intelligent FC switch - An intelligent FC switch is a Layer-3 switch that is bundled with a programmable virtual storage router to enable highly efficient network-based storage services.
Layer-2 (L2) switch - A switch that forwards Fibre Channel I/O frames using physical network addresses, to connect initiators (such as hosts) with targets (such as storage arrays).
Layer-3 (L3) switch - A switch bundled with a programmable virtual storage router, to enable network-based storage services such as volume management, data migration, and copy services.
logical fabric - One or more virtual Fibre Channel networks created by the shared configurations on interconnected switches. The Incipient NSP Administrator discovers logical fabrics by querying switches.
logical unit (LU) - A SCSI object that processes IO commands sent by a host. Both Network-Volumes and array volumes are logical units.
logical unit number (LUN) - The SCSI identifier of a logical unit within a target. Targets can present different LUNs to different initiator ports for the same logical unit.
LUN mapping - The process of assigning initiator-port-specific LU numbers to Network-Volume exposures.
LUN masking - The process of assigning initiator-port-specific access permissions to exposures.
metadata - In the Incipient NSP environment, metadata is essential information about Network-Volume structures and the array storage used to construct those Network-Volumes.
metadata repository - An array volume or set of array volumes used by Incipient NSP to store information about and manage access to metadata for a set of Network-Volumes. Multiple metadata repositories may exist in a single Incipient NSP environment.
migration - The process of moving data from one set of physical extents to another while the Network-Volume remains online.
mirror - A Network-Volume which maintains at least one complete, dynamic copy of the data stored within its logical extents.
mirror set - The set of all mirror sides associated within a Network-Volume.
mirror side - A single dynamic copy of a set of extents composing a Network-Volume. A single mirror always has at least two mirror sides.
mirroring (RAID-1) - A method for data redundancy that maintains two or more identical, dynamic, real-time copies of data on separate disks. Mirroring is transparent to the host, which perceives a single volume.
multipathing - Multipathing is the use of multiple paths between an initiator and a logical unit for the purpose of load balancing or fault tolerance. See also active/passive array, active/active array, preferred path array.
Network-Volume - A virtual SCSI logical unit created by the Incipient NSP, constructed from extents of one or more array volumes.
node - One instance of the Incipient NSP software residing in the network (e.g., on an intelligent FC switch processor). Nodes are always deployed in sets of two. Each node is allowed to be in only one cluster.
partition - A logical division of storage on a fixed disk. In Incipient NSP, a partition is a contiguous extent of blocks on an array volume that is used as a building block for a Network-Volume.
point-in-time copy - A virtual copy of a Network-Volume which contains a logical version of all of the data at a single reference point. See also snapshot, copy-on-write, split-mirror, clone
policy - An enforceable set of rules to restrict and ensure accountability, a particular sequence of actions, or correct allocation and use of storage resources.
port - The connection point of a device such as a host or switch to a bus or network. Incipient NSP ports are represented with a worldwide name (WWN) and can be virtual connection points. Fibre Channel switch ports have many different logical operating modes, such as N_Port, F_Port, E_Port.
preferred path array - A storage array whose multipathing solution allows access to an array volume down multiple paths simultaneously, but for which efficient access is through a single path.
provisioning - The process of assigning storage to data, to optimize performance, cost, and reliability requirements.
RAID - Redundant Array of Independent Disks (RAID) is a technology that uses multiple physical disks to enhance I/O performance, data redundancy, or both, transparent to the host operating system. Common RAID levels are listed below.
snapshot - A copy service that duplicates the contents of a source Network-Volume at a single point in time to a target Network-Volume.
split-mirror - A snapshot copy service that builds a full, real-time copy of a source Network-Volume and then breaks the mirroring relationship in order to provide a point-in-time copy. See also clone.
storage area network (SAN) - A dedicated network for storage operations, whose primary purpose is to transfer data between computer systems and storage devices. SANs provide the ability to share a single storage array across many servers or applications.
storage array - An intelligent storage device whose software provides functionality such as RAID protection, volume management, and replication. Also called a storage subsystem.
storage pool - A collection of bound array volumes that provide capacity that can be allocated for a specified class of Network-Volumes. For policy control, each user can be given access to one or more storage pools, and storage pools may be defined based on a number of different storage attributes.
storage-side port - Intelligent switch connection than can initiate I/O requests to storage. For example, Incipient NSP initiator ports are considered storage-side ports because they send I/O requests to storage arrays.
storage-side multipathing - The use of multiple paths between a storage-side port and an array volume.
striping (RAID-0) - A method for improving performance by distributing I/O load across disks. Striping spreads data alternately across two or more physical or virtual disks in small, equal-sized portions.
switch - A physical device that selects a path for sending a unit of data to its next destination.
synchronize snapshot - Update the relationship between a source Network-Volume and its target Network-Volume(s), bringing the target's copy of data up to the current point in time.
target - A SCSI device that that receives I/O requests from another device (initiator).
target Network-Volume - The destination disk used in a copy operation. A source Network-Volume can have more than one target Network-Volume.
tiered storage - The assignment of data to different types of storage media, based on cost, reliability, and performance requirements.
Universal LUN Support (ULS) - The ability of the Incipient NSP to bind to and use any LUN that is visible to a cluster, without discovering or managing the storage array that is presenting the LUN. It possible to use storage from arrays even if the Incipient NSP Administrator is not able to probe the array.
virtual target port - A logical port on an Incipient NSP node used to expose Network-Volumes to hosts. The Incipient NSP Administrator creates virtual target ports when you expose a Network-Volume to a host and configures them for use as SCSI targets.
virtualization - The insertion of an abstraction layer between hosts and storage which enables advanced, network-based functionality, such as: network-based volume management, pooling of storage from multiple SAN-attached storage devices from various vendors, and advanced network-based copy services.
virtualizer - The network component responsible for virtualization.
volume - A virtual disk (logical volume) created by a storage-, network-, or host-based volume manager. A volume may be composed of one or more physical extents, but to the host application, a volume appears as a single logical extent.
volume class - User-defined provisioning template that specify the RAID attributes, number of paths, min-max volume size, and the storage pools from which resources are selected.
VSAN - A logical implementation of a SAN. VSAN technology partitions a single physical SAN fabric into multiple virtual SANs. Each VSAN is a logically separate fabric with its own fabric services.