Cloud computing glossary of terms

Application migration occurs when an organization or client moves their computer programs to a new environment. A migration may move programs from internal servers to a computing cloud or from one computing cloud to another.

Cloud computing uses the internet to deliver computing applications and services to consumers. "Cloud" can also be used as a prefix when describing any service or activity that utilizes a cloud computing model.

• Cloud applications are software programs that combine locally hosted and cloud-based components.
• Cloud integrations combine multiple cloud-based resources into a unified computing suite.
• Cloud providers deliver cloud services and resources to clients, usually on a pay-as-you-go basis.

Cloud-native applications are computer programs that perform in cloud environments.

A container is a software unit that aligns an application's code with the configuration files and dependencies it requires to run properly. In cloud computing, containers help conserve resources by isolating applications from the host server's operating system.

A content delivery network (CDN) uses networked servers in various physical locations to deliver data-rich media to end users. CDN models typically use the server closest to the end user's geographic location to optimize efficiency and performance.

Data migration describes the process of relocating digital data from its original storage location, such as an in-house server, to a new location, such as a cloud network.

The migration process engages data governance, which describes the policies an organization uses to keep its data secure. When data can be moved between different applications or computing environments, it has data portability.

Edge computing is a type of distributed architecture in which external data-processing resources remain physically proximate to the device(s) using them. The model reduces response times, delivering smooth, efficient functionality.

Some cloud computing systems have the built-in ability to increase or decrease resource availability automatically and on demand. This feature is known as elasticity.

Grid computing uses a group of networked computers together to complete data-intensive or mass-scale tasks.

Hybrid cloud models may pair public and/or private clouds with on-site servers, edge computing networks, or both to run applications and process data.

Associated sub-terms include hybrid cloud architecture, hybrid cloud management, and hybrid cloud security. These terms respectively describe the design, administration, and information security strategies used in hybrid cloud systems.

IaaS models deliver information technology infrastructure over the internet via a computing cloud. This broad term may encompass servers, networking resources, and/or software applications.

The IoT consists of internet-connected appliances, devices, machines, or systems. It allows users to operate connected equipment remotely via an online interface. The IoT also gathers, processes, and analyzes data to improve the performance and responsiveness of connected devices or equipment.

Load balancing describes the practice of evenly spreading processing loads or network traffic across a pool of servers. When used in cloud computing, load balancing can improve resource efficiency and deliver a smoother experience for the end user.

Middleware is a class of computer software that facilitates data management and communications between different applications. It occupies a virtual space between an operating system and the computing resources running on that system, which inspired its name.

Some cloud computing models distribute service delivery across multiple providers. This practice, known as multicloud computing, can improve performance efficiency, reduce costs, and address the operational risks associated with relying on a single provider for all cloud services.

This cloud architecture model provides shared resource access to multiple end users (tenants). It can be used in both public and private clouds to reduce costs while preserving scalability.

When a particular tenant consumes a disproportionate share of the cloud's available resources, that tenant is known as a noisy neighbor.

A computing platform combines applications and services with an operating system to deliver a suite of functionalities. PaaS models virtualize platforms and deliver them to users over the internet via the cloud.

Private clouds deliver computing resources and services across a proprietary or privately operated internal network and only to a selected group of specially authorized clients or end users.

Public clouds are managed platforms owned and operated by third-party providers. These providers make cloud resources available to anyone who wishes to access them.

A relational database has an architecture model that uses defined relationships between data types and data points to improve searchability and access efficiency. Relational databases use a programming language known as Structured Query Language (SQL) to process queries and store resources.

Nonrelational databases, also known as NoSQL databases, do not use definitional schema to establish relationships between data points. They are generally preferred for applications involving large quantities of unstructured data subject to growth or change.

Replication is the practice of copying data and storing each copy in a different end destination. Cloud systems use replication as a data backup strategy and disaster recovery tool to improve the on-demand availability of stored information.

Scaling refers to decreasing or increasing access to computing resources on an as-needed basis. It has two subtypes: vertical and horizontal scaling.

Vertical scaling, also known as scaling up (or down), increases or decreases the amount of available computing resources. Horizontal scaling, also known as scaling out, uses a partitioning model known as sharding to turn a single large database into a series of smaller databases known as nodes.

SaaS models allow end users to access computer programs and store associated data remotely. Rather than physically installing software on a user's machine, SaaS applications run over the internet. SaaS models liberate users from the need to purchase or license software for personal or business use.

The "SPI model" collectively refers to the three most widely used cloud services: software (SaaS), platform (PaaS), and infrastructure (IaaS). Cloud clients can access these three services together in a single bundled package.

Virtualization involves creating simulated versions of computing resources such as operating systems, computer hardware, or data storage units. It creates virtual computers, which share the common resources of the host computer while retaining the independent ability to deploy different operating systems or software applications.

Virtual desktop infrastructure allows users to access a specific computer system remotely from any authorized internet-connected device. Virtual private networks, or VPNs, connect an end user's computer to a remote third-party server to mask the user's actual location and bypass geographic access restrictions.

A cloud computing model designed, configured, and optimized for use in a specific industry is known as a vertical cloud. Here, the term "vertical" refers to a particular business model or niche. This denotation differs from the way "vertical" is used in cloud computing terms like "vertical scaling."

Vertical clouds are also known as industry cloud platforms.

XaaS is the standard acronym for "anything as a service." It describes a generalized cloud-based service delivery configuration that may include the SPI model and any additional cloud-based services.