It has been suggested thatData at rest bemerged into this article. (Discuss) Proposed since March 2025.
For broader coverage of this topic, seeData storage.
1 GiB ofSDRAM mounted in acomputer. An example ofprimary storage.15 GBPATA hard disk drive (HDD) from 1999. When connected to a computer it serves assecondary storage.
Computer data storage ordigital data storage is the retention ofdigital data viatechnology consisting ofcomputer components andrecording media. Digital data storage is a core function and fundamental component of computers.[1]: 15–16
Generally, the faster and volatile storage components are referred to as "memory", while slower persistent components are referred to as "storage". This distinction was extended in theVon Neumann architecture, where thecentral processing unit (CPU) consists of two main parts: Thecontrol unit and thearithmetic logic unit (ALU). The former controls the flow of data between the CPU and memory, while the latter performs arithmetic andlogical operations on data. In practice, almost all computers use amemory hierarchy,[1]: 468–473 which puts memory close to the CPU and storage further away.
Text, numbers, pictures, audio, and nearly any other form of information can be converted into a string ofbits, or binary digits, each of which has a value of 0 or 1. The most common unit of storage is thebyte, equal to 8 bits.Digital data comprises the binary representation of a piece of information, often beingencoded by assigning a bit pattern to eachcharacter,digit, ormultimedia object. Many standards exist for encoding (e.g.character encodings likeASCII, image encodings likeJPEG, and video encodings likeMPEG-4).
Forsecurity reasons, certain types of data may beencrypted in storage to prevent the possibility of unauthorized information reconstruction from chunks of storage snapshots. Encryption in transit protects data as it is being transmitted.[4]
Data compression methods allow in many cases (such as a database) to represent a string of bits by a shorter bit string ("compress") and reconstruct the original string ("decompress") when needed. This utilizes substantially less storage (tens of percent) for many types of data at the cost of more computation (compress and decompress when needed). Analysis of the trade-off between storage cost saving and costs of related computations and possible delays in data availability is done before deciding whether to keep certain data compressed or not.
Impending failure onhard disk drives is estimable usingS.M.A.R.T. diagnostic data that includes thehours of operation and the count of spin-ups, though its reliability is disputed.[5] The health ofoptical media can be determined bymeasuring correctable minor errors, of which high counts signify deteriorating and/or low-quality media. Too many consecutive minor errors can lead to data corruption. Not all vendors and models ofoptical drives support error scanning.[6]
Without a significant amount of memory, a computer would only be able to perform fixed operations and immediately output the result, thus requiring hardwarereconfiguration for a new program to be run. This is often used in devices such as deskcalculators,digital signal processors, and other specialized devices.Von Neumann machines differ in having a memory in which operatinginstructions and data are stored,[1]: 20 such that they do not need to have their hardware reconfigured for each new program, but can simply bereprogrammed with new in-memory instructions. They also tend to be simpler todesign, in that a relatively simpleprocessor may keepstate between successive computations to build up complex procedural results. Most modern computers are von Neumann machines.
In contemporary usage, the term "storage" typically refers to a subset of computer data storage that comprises storage devices and their media not directly accessible by theCPU, that is,secondary ortertiary storage. Common forms of storage includehard disk drives,optical disc drives, andnon-volatile devices (i.e. devices that retain their contents when the computer is powered down).[7] On the other hand, the term "memory" is used to refer tosemiconductorread-write data storage, typicallydynamic random-access memory (DRAM).Dynamic random-access memory is a form of volatile memory that also requires the stored information to be periodically reread and rewritten, orrefreshed;static RAM (SRAM) is similar to DRAM, albeit it never needs to be refreshed as long as power is applied.
In contemporary usage, thememory hierarchy of primary storage and secondary storage in some uses refer to what was historically called, respectively,secondary storage andtertiary storage.[8]
Primary storage (also known asmain memory,internal memory, orprime memory), often referred to simply asmemory, is storage directly accessible to the CPU. The CPU continuously reads instructions stored there and executes them as required. Any data actively operated on is also stored there in a uniform manner. Historically,early computers useddelay lines,Williams tubes, or rotatingmagnetic drums as primary storage. By 1954, those unreliable methods were mostly replaced bymagnetic-core memory. Core memory remained dominant until the 1970s, when advances inintegrated circuit technology allowedsemiconductor memory to become economically competitive.
This led to modernrandom-access memory, which is small-sized, light, and relatively expensive. RAM used for primary storage isvolatile, meaning that it loses the information when not powered. Besides storing opened programs, it serves asdisk cache andwrite buffer to improve both reading and writing performance. Operating systems borrow RAM capacity for caching so long as it's not needed by running software.[9] Spare memory can be utilized asRAM drive for temporary high-speed data storage. Besides main large-capacity RAM, there are two more sub-layers of primary storage:
Processor registers are the fastest of all forms of data storage, being located inside the processor, with each register typically holding aword of data (often 32 or 64 bits). CPU instructions instruct thearithmetic logic unit to perform various calculations or other operations on this data.
Processor cache is an intermediate stage between faster registers and slower main memory, being faster than main memory but with much less capacity. Multi-levelhierarchical cache setup is also commonly used, such that primary cache is the smallest and fastest, while secondary cache is larger and slower.
Primary storage, includingROM,EEPROM,NOR flash, andRAM,[10] is usuallybyte-addressable. Such memory is directly or indirectly connected to the central processing unit via amemory bus, comprising anaddress bus and adata bus. The CPU firstly sends a number called thememory address through the address bus that indicates the desired location of data. Then it reads or writes the data in thememory cells using the data bus. Additionally, amemory management unit (MMU) is a small device between CPU and RAM recalculating the actual memory address. Memory management units allow formemory management; they may, for example, provide anabstraction ofvirtual memory or other tasks.
Non-volatile primary storage contains a small startup program (BIOS) is used tobootstrap the computer, that is, to read a larger program from non-volatile secondary storage to RAM and start to execute it. A non-volatile technology used for this purpose is calledread-only memory (ROM). Most types of "ROM" are not literally read only but are difficult and slow to write to. Someembedded systems run programs directly from ROM, because such programs are rarely changed. Standard computers largely do not store many programs in ROM, apart fromfirmware, and use large capacities of secondary storage.
Secondary storage (also known asexternal memory orauxiliary storage) differs from primary storage in that it is not directly accessible by the CPU. Computers useinput/output channels to access secondary storage and transfer the desired data to primary storage. Secondary storage is non-volatile, retaining data when its power is shut off. Modern computer systems typically have two orders of magnitude more secondary storage than primary storage because secondary storage is less expensive.
To reduce the seek time and rotational latency, secondary storage, includingHDD,ODD andSSD, are transferred to and fromdisks in large contiguous blocks. Secondary storage isaddressable by block; once thedisk read/write head on HDDs reaches the proper placement and the data, subsequent data on the track are very fast to access. Another way to reduce the I/O bottleneck is to use multiple disks in parallel to increase the bandwidth between primary and secondary memory, for example, usingRAID.[11]
Secondary storage is often formatted according to afile system format, which provides the abstraction necessary to organize data intofiles anddirectories, while also providingmetadata describing the owner of a certain file, the access time, the access permissions, and other information. Most computeroperating systems use the concept ofvirtual memory, allowing the utilization of more primary storage capacity than is physically available in the system. As the primary memory fills up, the system moves the least-used chunks (pages) to a swap file or page file on secondary storage, retrieving them later when needed.
A largetape library, with tape cartridges placed on shelves in the front, and a robotic arm moving in the back. The visible height of the library is about 180 cm.
Tertiary storage ortertiary memory typically involves arobotic arm which mounts and dismount removable mass storage media from a catalogdatabase into a storage device according to the system's demands. It is primarily used for archiving rarely accessed information, since it is much slower than secondary storage (e.g. 5–60 seconds vs. 1–10 milliseconds). This is primarily useful for extraordinarily large data stores, accessed without human operators. Typical examples includetape libraries,optical jukeboxes, and massive arrays of idle disks (MAID). Tertiary storage is also known asnearline storage because it is "near to online".[12]Hierarchical storage management is an archiving strategy involving automatically migrating long-unused files from fast hard disk storage to libraries or jukeboxes.
Offline storage is computer data storage on a medium or a device that is not under the control of aprocessing unit.[13] The medium is recorded, usually in a secondary or tertiary storage device, and then physically removed or disconnected. Unlike tertiary storage, it cannot be accessed without human interaction. It is used totransfer information since the detached medium can easily be physically transported. In modern personal computers, most secondary and tertiary storage media are also used for offline storage.
Storage area network (SAN) is a specialized network, that provides other computers with storage capacity. SAN is commonly associated withFibre Channel networks.
Cloud deployment models define the interactions between cloud providers and customers.[16]
Private clouds, for example, are used incloud security to mitigate the increasedattack surface area ofoutsourcing data storage.[17] A private cloud is cloud infrastructure operated solely for a single organization, whether managed internally or by a third party, or hosted internally or externally.[18]
Hybrid cloud storage are another cloud security solution, involving storage infrastructure that uses a combination of on-premises storage resources with cloud storage. The on-premises storage is usually managed by the organization, while the public cloud storage provider is responsible for the management and security of the data stored in the cloud.[19][20] Using a hybrid model allows data to be ingested in an encrypted format where the key is held within the on-premise infrastructure and can limit access to the use of on-premisecloud storage gateways, which may have options to encrypt the data prior to transfer.[21]
Cloud services are considered "public" when they are delivered over the public Internet.[22]
Avirtual private cloud (VPC) is a pool of shared resources within a public cloud that provides a certain level of isolation between the different users using the resources. VPCs achieve user isolation through the allocation of a private IPsubnet and a virtual communication construct (such as aVLAN or a set ofencrypted communication channels) between users as welll as the use of avirtual private network (VPN) per VPC user, securing, by means of authentication and encryption, the remote access of the organization to its VPC resources.[citation needed]
Block-level storage is a concept in cloud-hosteddata persistence wherecloud services emulate the behaviour of a traditionalblock device, such as a physicalhard drive,[25] where storage is organised asblocks. Block-level storage differs fromobject stores or 'bucket stores' or tocloud databases. These operate at a higher level of abstraction and are able to work with entities such as files, documents, images, videos or database records.[26] At one time, block-level storage was provided bySAN, and NAS provided file-level storage.[27] With the shift from on-premises hosting to cloud services, this distinction has shifted.[28]
Instance stores are a form of cloud-hosted block-level storage, being provided as part of a cloudinstance.[29] Unlike other forms of block storage, instance store data will be lost the cloud instance is stopped.[30]
Storage technologies at all levels of the storage hierarchy can be differentiated by evaluating certain core characteristics as well as measuring characteristics specific to a particular implementation. These core characteristics are:
Anuninterruptible power supply (UPS) can be used to give a computer a brief window of time to move information from primary volatile storage into non-volatile storage before the batteries are exhausted. Some systems, for exampleEMC Symmetrix, have integrated batteries that maintain volatile storage for several minutes.
Types of access includerandom access andsequential access. In random access, any location in storage can be accessed at any moment in approximately the same amount of time. In sequential access, the accessing of pieces of information will be in a serial order, one after the other; therefore the time to access a particular piece of information depends upon which piece of information was last accessed.
Storage performance metrics includelatency,throughput, granularity and reliability.
Energy
Low capacitysolid-state drives have no moving parts and consume less power than hard disks.[31][32][33] Also, memory may use more power than hard disks.[33] Large caches, which are used to avoid hitting thememory wall, may also consume a large amount of power.
In modern computers, primary storage almost exclusively consists of dynamic volatile semiconductorrandom-access memory (RAM), particularlydynamic random-access memory (DRAM). Since the turn of the century, a type of non-volatilefloating-gate semiconductor memory known asflash memory has steadily gained share as off-line storage for home computers. Non-volatile semiconductor memory is also used for secondary storage in various advanced electronic devices and specialized computers that are designed for them.
As early as 2006,notebook anddesktop computer manufacturers started using flash-basedsolid-state drives (SSDs) as default configuration options for the secondary storage either in addition to or instead of the more traditional HDD.[35][36][37][38][39]
Magnetic storage uses different patterns ofmagnetization on amagnetically coated surface to store information. Magnetic storage isnon-volatile. The information is accessed using one or more read/write heads which may contain one or more recording transducers. A read/write head only covers a part of the surface so that the head or medium or both must be moved relative to another in order to access data. In modern computers, magnetic storage will take these forms:
Magnetic-tape was often used for secondary storage;
Tertiary (e.g.NCR CRAM) or off line storage in the form of magnetic cards.
Magnetic storage does not have a definite limit of rewriting cycles like flash storage and re-writeable optical media, as altering magnetic fields causes no physical wear. Rather, their life span is limited by mechanical parts.[40][41]
Optical storage, the typicaloptical disc, stores information in deformities on the surface of a circular disc and reads this information by illuminating the surface with alaser diode and observing the reflection. Optical disc storage isnon-volatile. The deformities may be permanent (read only media), formed once (write once media) or reversible (recordable or read/write media). The following forms are in common use as of 2009[update]:[42]
CD,CD-ROM,DVD,BD-ROM: Read only storage, used for mass distribution of digital information (music, video, computer programs);
CD-R,DVD-R,DVD+R,BD-R: Write once storage, used for tertiary and off-line storage;
Ultra Density Optical or UDO is similar in capacity toBD-R or BD-RE and is slow write, fast read storage used for tertiary and off-line storage.
Magneto-optical disc storage is optical disc storage where the magnetic state on aferromagnetic surface stores information. The information is read optically and written by combining magnetic and optical methods. Magneto-optical disc storage isnon-volatile,sequential access, slow write, fast read storage used for tertiary and off-line storage.
Light induced magnetization melting in magnetic photoconductors has also been proposed for high-speed low-energy consumption magneto-optical storage.[43]
Paper data storage, typically in the form ofpaper tape orpunched cards, has long been used to store information for automatic processing, particularly before general-purpose computers existed. Information was recorded by punching holes into the paper or cardboard medium and was read mechanically (or later optically) to determine whether a particular location on the medium was solid or contained a hole.Barcodes make it possible for objects that are sold or transported to have some computer-readable information securely attached.
Relatively small amounts of digital data (compared to other digital data storage) may be backed up on paper as amatrix barcode for very long-term storage, as the longevity of paper typically exceeds even magnetic data storage.[44][45]
Optical tape is a medium for optical storage, generally consisting of a long and narrow strip of plastic, onto which patterns can be written and from which the patterns can be read back.
Holographic data storage stores information optically insidecrystals orphotopolymers, for example, in HVDs (Holographic Versatile Discs). Holographic storage can utilize the whole volume of the storage medium, unlike optical disc storage, which is limited to a small number of surface layers.
Magnetic photoconductors store magnetic information, which can be modified by low-light illumination.[43]
^"ISO/IEC 22123-2:2023(E) - Information technology — Cloud computing — Part 2: Concepts".International Organization for Standardization. September 2023.
^Mell, Peter; Timothy Grance (September 2011).The NIST Definition of Cloud Computing (Technical report). National Institute of Standards and Technology: U.S. Department of Commerce.doi:10.6028/NIST.SP.800-145. Special publication 800-145.
^Jones, Margaret (July 2019)."Hybrid Cloud Storage".SearchStorage.Archived from the original on 21 June 2022. Retrieved22 June 2022.
^Rouse, Margaret."What is public cloud?". Definition from Whatis.com.Archived from the original on 16 October 2014. Retrieved12 October 2014.
^Kolodner, Elliot K.; Tal, Sivan; Kyriazis, Dimosthenis; Naor, Dalit; Allalouf, Miriam; Bonelli, Lucia; Brand, Per; Eckert, Albert; Elmroth, Erik; Gogouvitis, Spyridon V.; Harnik, Danny; Hernandez, Francisco; Jaeger, Michael C.; Bayuh Lakew, Ewnetu; Manuel Lopez, Jose; Lorenz, Mirko; Messina, Alberto; Shulman-Peleg, Alexandra; Talyansky, Roman; Voulodimos, Athanasios; Wolfsthal, Yaron (2011). "A Cloud Environment for Data-intensive Storage Services".2011 IEEE Third International Conference on Cloud Computing Technology and Science. pp. 357–366.CiteSeerX10.1.1.302.151.doi:10.1109/CloudCom.2011.55.ISBN978-1-4673-0090-2.S2CID96939.
^S. Rhea, C. Wells, P. Eaton, D. Geels, B. Zhao, H. Weatherspoon, and J. Kubiatowicz, Maintenance-Free Global Data Storage. IEEE Internet Computing, Vol 5, No 5, September/October 2001, pp 40–49.[1]Archived 2012-03-29 at theWayback Machine[2]Archived 2011-06-23 at theWayback Machine
^Wittig, Andreas; Wittig, Michael (2015).Amazon web services in action. Manning press. pp. 204–206.ISBN978-1-61729-288-0.
