Computing can be either a public or private cloud. Public cloud platforms (e.g., AWS and Microsoft Azure) centralize resources in data centers distributed around the globe, and users can access them over the public Internet. These resources are made available to customers through metered services, with cloud providers responsible for varying degrees of back-end maintenance. Private clouds are hosted in enterprise data centers or hosted data center facilities. While they are not as powerful as large-scale public clouds. However, they do have some elasticity, and developers and managers in an organization can still access resources using self-service portals. In theory, private clouds offer better control and security, but this requires effort from an organization's IT team. Cloud deployment models include private clouds, public clouds, hybrids of the two, and combinations of multiple cloud platforms. Public and private clouds can also be linked to create a hybrid cloud, or two or more public clouds can be connected to create a multi-cloud architecture. Broadly speaking, cloud computing is also categorized into three layers: Infrastructure as a Service (IaaS), Platform as a Service (PaaS), and Software as a Service (SaaS). Infrastructure as a Service (IaaS) applies to the underlying building blocks, such as compute, network, and storage. It provides a lot of flexibility for application development, but also requires a lot of overhead; Platform-as-a-Service (PaaS) abstracts these lower-tier elements and provides a sandbox environment for application developers. Software-as-a-Service (SaaS), which typically requires no more hands-on cloud computing model, consists of licensed software delivered as a Web application.