2 Cloud Deployment Models
Compare the four cloud deployment models by who uses and manages each environment, their common uses and trade-offs, and the factors that guide workload placement.
How deployment models differ
A describes who can use a cloud environment, who owns or manages it, and how it relates to other environments. The four models are public, private, community, and hybrid. They differ in control, sharing, cost, and operational complexity; none is automatically the safest or best choice for every workload.
Public and private clouds
A is provided for use by customers generally. The provider operates the underlying infrastructure, and multiple customers may share physical resources while their cloud environments are logically isolated. Common uses include websites and applications with changing traffic, development and testing, analytics, and workloads that benefit from services available on demand.
can be quick to start and scale without purchasing data-center hardware. Customers have less control over the underlying infrastructure and must configure and secure their own data, identities, and applications. Usage-based pricing requires monitoring to avoid unexpected costs. “Public” describes availability of the cloud service to customers; it does not mean customer data is publicly accessible.
A is provisioned for the exclusive use of one organization. It may be operated by that organization, a third party, or both, and located on or off the organization’s premises. It can suit workloads needing specialized configuration, close control, or integration with existing systems, including some regulated or legacy applications.
A private environment can offer more direct control, but the organization remains responsible for ensuring it meets security and compliance needs. Building and maintaining it may require substantial investment, staff expertise, and capacity planning. It is not automatically more secure simply because it is dedicated to one organization.
A is an isolated virtual network in a , not a private-cloud deployment by itself. Its underlying infrastructure remains part of the public-cloud provider’s environment.
Community clouds
A is provisioned for exclusive use by organizations that share concerns such as mission, security requirements, policy, or compliance needs. It may be managed by one or more participating organizations, a third party, or a combination. Organizations in the same sector or with shared missions may use a common environment for collaboration and shared requirements, such as a group of public agencies or research institutions.
Participants may share costs and tailor governance to common needs. They must also agree on access, responsibility, funding, and security controls. A is shared rather than dedicated to one organization and may offer less flexibility than a private environment.
Hybrid clouds
A combines two or more distinct cloud infrastructures—public, private, or community—and connects them so that data or applications can move between environments. Using multiple environments without meaningful integration does not necessarily make them a .
Hybrid arrangements can keep some workloads in an existing private environment while moving others to , use public-cloud capacity for temporary demand spikes, or place different workloads according to operational or compliance requirements. They can balance control with access to public-cloud services, but connecting environments securely and reliably takes planning.
Teams must manage integration, identity, networking, data movement, and security across environments. Moving data or applications can also be constrained by compatibility, cost, or policy.
Choosing a deployment model
Choose a deployment model by considering each workload’s control, security, compliance, cost, and integration needs. The models are deployment choices, not service types such as IaaS, PaaS, or SaaS. Security depends on how responsibilities and controls are implemented.
Public: Customers use services from a provider’s cloud infrastructure. It offers fast access to services and capacity that can scale with demand, with less control over underlying infrastructure and a need to manage costs and security settings.
Private: One organization uses the environment exclusively. It can provide greater control or specialized configuration, with greater responsibility for operating, staffing, and maintaining the environment.
Community: Organizations with shared concerns use the environment. It supports shared environments and governance for a defined community, while requiring coordination of governance, costs, and responsibilities.
Hybrid: Workloads span connected cloud infrastructures. It allows workloads to be placed in different environments or existing systems to connect to cloud services, with greater integration and cross-environment management complexity.