On-PremiseOn-premise cloud is when an organization deploys cloud infrastructure in its own data centers or server rooms. In this case, scale is limited by constraints such as space, electricity and the amount of hardware available at a point in time.Infrastructure as a service is cloud infrastructure that is offered to customers or internal clients. Large vendors in this industry have significant resources such as a large number of geographically distributed data centers. Such vendors have incentive to use such resources close to their capacity to maximize return on capital employed. As such, cloud platforms do have limits based on the resources available and how this is allocated to services and customers.
Resource QuotasInfrastructure as a service providers typically manage resources by assigning each customer a resource quota. For example, a small business account might be restricted to 32 virtual machines when they first sign up.
Reserved CapacityGenerally speaking, customers are not guaranteed to be able to scale to their resource quota. If there is a peak in demand, there may be a delay in your ability to get more instances from your cloud provider. A provider may offer reserved capacity such that customers pay a fee in order to guaranty a minimum number of resources will be available to them. Vertical scale is the process of adding resources to an instance. For example, a customer who redeploys a server from a 4-cpu virtual machine to a 16-cpu virtual machine.software service may be designed such that an unlimited number of instances can be deployed at the same time to serve an unlimited number of users. It is common to automatically scale cloud software using an API that can create and destroy instances based on monitoring parameters or custom business rules.
Side-by-side ScaleSide-by-side scale is the deployment of multiple instances of the same software for different purposes such as a production, test and development instance of a server.
Geographical ScaleCloud platforms typically allow software to be deployed to geographically distributed data centers for purposes such as resilience and reduced latency. For example, a content delivery network that is used to server media files from a data center that is closest to each user.
A class of technology that can be scaled without technical limitation.