2. Cloud Computing¶
Cloud Computing is a new paradigm for the use and management of computing resources and computer services delivered on demand via the Internet. Cloud services are offered by means of standardized catalogues. They guarantee services that can be easily and automatically scaled up depending on load peaks (agility, scalability, elasticity), and can operate simultaneously and securely on the data and systems of different users (multi-tenant).
Typically, Cloud services, according to the model of computational resources offered, are divided into three service models:
1. Infrastructural system services so called Infrastructure-as-a-Service (laaS) for the provision, for example, of virtualized servers and data storage space;
2. Computational platforms services so called Platform-as-a-Service (PaaS) for the provision of pre-configured and managed environments for the development of specific applications, e.g. for software development, data management or containerised applications;
3. Application services so called Software-as-a-Service (SaaS) for the delivery of applications to end users, e.g. e-mail or other remote collaboration systems.
These different service models allow users of Cloud services to avoid many of the basic management activities of a data centre’s infrastructure (such as management of buildings and physical technological components) and, to simplify the management of initial and operational configurations of applications and platforms, thus. allowing considerable economic savings and greater flexibility in managing the organisations’ demand for new computing resources.
Services are typically provided by Cloud Service Providers (CSPs) who guarantee their operation according to contractually determined levels (Service-Level Agreements, SLAs).
The distribution model of Cloud services can be organised in the following categories: Public Cloud, Private Cloud, Hybrid Cloud and Multi-Cloud.
2.1 Public Cloud¶
In the Public Cloud, the infrastructure is owned by a CSP which, with full control, makes its systems available to users, companies and public bodies in different geographical areas (or regions) of the world, sharing processing capacity, applications and storage. Such a deployment allows users of cloud services to benefit from resilient computing capacities that can be scaled according to actual needs. In the area of public cloud CSPs, a small group of non-European companies (mainly of US origin) operate as market leaders. These companies offer Cloud services with almost unlimited computing capacity through highly sophisticated technological solutions, so-called ‘hyperscaler’, with high ease of use, configurability and interoperability.
2.2 Private Cloud¶
The Private Cloud consists of a Cloud environment reserved for an individual customer for its exclusive use.
This may be on-premise, i.e. based on infrastructures that are entirely in the domain of the customer, which holds control and full responsibility for the maintenance and security management of the hosted data and services, or it may be managed at a third party’s data centre, where the customer is provided with dedicated resources.
One of the advantages of a Private Cloud is certainly the greater control that the customer can exert (in terms of choice and contractual arrangements) over the characteristics of the Cloud infrastructure and services, especially with regards to security. However, this solution, particularly in the case of on-premise clouds, presents some disadvantages as the infrastructure may not be able to guarantee adequate scalability to handle unforeseen peaks in demand.
2.3 Hybrid Cloud¶
A combination of the Public and Private Cloud models, the Hybrid Cloud, is a single environment created from several connected environments in which, depending on need, resources from both a private and a public Cloud are made available to users. This model extends the capabilities of a private cloud to use on demand the large-scale resources available on a Public Cloud, t in order to manage, for example, sudden peaks in workload; furthermore, it guarantees savings in terms of the transmission bandwidth needed to exchange data, compared to what would be possible with a connection to a data centre.
2.4 Multi-Cloud¶
Multi Cloud refers to a model in which several clouds of the same type (public or private) offered by different providers are used simultaneously to implement certain services or applications.
Unlike the Hybrid Cloud, which involves the creation of a single infrastructure that transparently uses different types of Cloud (public or private), the Multi Cloud model is based on the use of different public or private Cloud environments that are not interconnected. In a Hybrid Cloud environment, the distribution of the use of computational resources between private and public is typically semi-automated and transparent to the user, whereas a Multi Cloud environment presents itself as a set of distinct computational resources that can potentially be integrated at application level.