Public Cloud

Public cloud describes a deployment model, not a particular product or application architecture. A provider operates infrastructure and offers computing resources to a broad customer base through services such as virtual machines, object storage, databases, and serverless execution. Customers typically provision these resources through a console, command-line tools, or APIs, with usage-based pricing and optional pricing commitments.

The provider manages physical facilities, hardware, and foundational infrastructure. Customers retain responsibilities that depend on the services they choose. With virtual machines, customers generally manage guest operating systems, application patches, and workload configuration. With managed databases or serverless services, the provider handles more of the underlying operation, but customers still manage their data, access permissions, and application behavior. Moving to public cloud changes operational responsibilities; it does not eliminate them.

For example, a retailer can deploy an application on cloud compute resources, store product images in object storage, and use a managed database for orders. The retailer can configure scaling policies to add capacity during busy periods and remove it afterward. This avoids purchasing hardware sized for peak demand, although elasticity requires suitable application design, configuration, service quotas, and available capacity. Public cloud does not make every application automatically scalable.

For deployment-model questions, distinguish a cloud-based deployment from an on-premises deployment and a hybrid deployment. In a cloud-based deployment, the application’s components run in cloud infrastructure, whether migrated from an existing environment or built there. In an on-premises deployment, workloads run in the organization’s own facilities or other infrastructure it manages. A hybrid deployment integrates cloud and on-premises environments, such as a cloud-hosted application accessing a database retained in a company data center through private connectivity. Merely maintaining unrelated workloads in both places does not demonstrate meaningful hybrid integration.

Public cloud is also distinct from private cloud. A private cloud is provisioned for exclusive use by one organization and may be operated on or off premises. An ordinary on-premises data center is not necessarily a private cloud: it may lack cloud characteristics such as on-demand self-service and resource pooling. Likewise, a logically isolated virtual network within a public cloud does not turn the deployment into a private cloud.

The word public refers to the provider’s offering, not to the visibility of customer data or workloads. Public cloud resources can use private IP addresses, restricted access policies, encryption, and private network connections. Providers commonly pool infrastructure across customers while enforcing tenant isolation, and some services offer dedicated hardware options within the public cloud model.

Trade-offs include reduced upfront hardware investment and faster provisioning versus ongoing service charges, network dependencies, data-transfer costs, and provider-specific dependencies. Engineers must also evaluate data residency, service availability, recovery requirements, and compliance controls. Provider infrastructure does not automatically make an application highly available: teams must choose appropriate locations, design redundancy, test recovery, and control spending through monitoring and resource lifecycle management.