Cloud computing is a new trend model for enabling workloads that use resources from a normally huge resource pool that is operated by a cloud service provider. These resources include servers, storage, network resources, applications, services, or even functions. These can be rapidly deployed, operated, and automated with little effort and the prices are calculated on a per-minute basis. This cloud model is composed of five essential characteristics, three service models, and four deployment models.

Cloud offerings are mainly categorized into the following service models:

• Infrastructure as a Service (IaaS): This describes a model where the cloud provider enables the consumer to create and configure resources from the computing layer upwards, without any need to care or know about the hardware layer. That includes virtual machines, networks, appliances, and lots of other infrastructure-related resources and services. The most popular IaaS resources in Azure contain virtual machines, virtual networks (internal and external), container services, and storage.
• Platform as a Service (PaaS): This gives the consumer an environment from the operating system upwards. So, the consumer is not responsible for the underlying IaaS infrastructure. Examples are operating systems, databases, or development frameworks. Microsoft Azure contains many PaaS resources such as SQL databases, Azure app services, or cloud services.
• Software as a Service (SaaS): This is the model with the lowest level of control and required management. A SaaS application is reachable from multiple clients and consumers, and the owning consumer doesn't have any control over the backend, except for some application-related management tasks. Examples of SaaS applications are Office 365, Visual Studio Online, the Outlook website, OneDrive, and even the Amazon website itself is a SaaS application with Amazon as its own consumer.

A comparison of service model responsibilities is shown in the following diagram: