Cloud Infrastructure

Cloud infrastructure is best understood as a way of obtaining computing capacity. The US National Institute of Standards and Technology, or NIST, defines cloud computing around shared resources that customers can obtain and release rapidly over a network. Its framework emphasizes self-service, resource pooling, elasticity, and measured use. The definition is useful for comparing offerings without assuming that every product carrying the word 'cloud' has identical capabilities. s1

NIST also distinguishes infrastructure, platform, and software services. At the infrastructure layer, a customer can deploy operating systems and applications while the provider operates the underlying infrastructure. A platform service moves more of the execution environment into the provider's responsibility. These distinctions help readers understand why a virtual machine, an application platform, and a finished online application belong to different layers even when purchased from the same company. s1

The inspected provider catalogs show overlapping portfolios. ArvanCloud lists cloud servers, object storage, containers, content delivery, video services, and managed databases. Its catalog illustrates how one supplier can offer both resources used to run applications and services used to distribute their output. These listings establish advertised product categories, rather than independently measured capacity or reliability. s2

ParsPack advertises cloud servers alongside virtual and dedicated servers, cloud hosting, storage, private-cloud services, and application platforms. Its homepage distinguishes hosting locations in Iran and abroad and describes storage offerings in Iran and Europe. This makes geography an explicit part of the commercial offering. The presence of a location in a catalog does not establish that every product or configuration is available there. s3

ParsPack also links to cloud-server API documentation, object-management instructions, backup guidance, and application-platform configuration documentation. Those materials indicate that the service relationship extends beyond purchasing a server: customers must also manage credentials, stored data, configurations, and recovery arrangements. This page does not rank the providers or establish their shares of the Iranian market. s3

Storage and delivery solve different problems. ArvanCloud's object-storage page describes uses including bulk files, application data, logs, backups, and unstructured information. It also lists tools for transferring and managing objects and links storage with content-delivery services. The practical distinction is that keeping an object and delivering it to an audience are related operations, but they need not use the same service. s4

This distinction also clarifies the branches of this guide. Data-center services concern the facilities supporting equipment; cloud-computing infrastructure concerns resources exposed as services; cloud storage concerns persistent data; and content delivery concerns distribution. The categories can overlap in a deployed application. Their separation is an explanatory framework, rather than a claim that Iranian providers operate isolated markets for each function. s1s2s3s4

Iran's connectivity environment makes infrastructure measurement especially consequential. A research preprint by Ali Sadeghi Jahromi and Jason Jaskolka examines nationwide shutdowns in 2026 using internet routing data, host scans, and active connection tests. Its abstract reports that many network prefixes remained announced during disruptions even when traffic was discarded. A routing announcement therefore did not reliably demonstrate that a service could be reached. s5

The authors also identify misleading changes in apparent host counts and argue for interpreting several measurement methods together. The implication for reading infrastructure claims is specific: a visible network, an apparently responsive address, and successful access to an application are different observations. The paper is a preprint, and this account is limited to findings stated in its inspected abstract. It does not provide a provider-by-provider uptime assessment. s5

The available evidence supports a layered account of cloud infrastructure in Iran: providers advertise computing, storage, platforms, and distribution services, while independent connectivity research shows why reachability needs separate examination. Product variety is documented more clearly here than sector size or service outcomes. A fuller account requires dated infrastructure inventories, comparable performance measurements, and evidence about how services behaved during disruptions. s2s3s4s5