The term Cloud, or more precisely Cloud Computing, is used to define a way to access and supply IT services on the basis of the absence of an hardware infrastructure directly controlled by the service provider (more precisely it is defined as Service-Oriented Architecture or SOA).
The necessary resources to the service are not staticly stored, but are managed on-demand, based on the service level and on the customers’ needs, who can set necessary resources thanks to automatisms that grant resources supply through self-service systems.
At the base of Cloud Computing there are virtualization systems that allow to abstract from the use of the software from the hardware on which it works. Thanks to Virtualization it is possible to divide and manage the resources by inserting them in pools that can grant a certain level of service. The customer can then have the control over the power he has at his disposal, by interacting with these pools: by increasing and decreasing pools minimum and maximum levels the customer can intervene on what represents the hardware he has to his disposal, without the need to know which device is actually providing the service.
Another advantage of Cloud systems is the High Availability: being the service completely untied from the hardware, in case of malfunctions or problems to the devices the service can be easily moved to another hardware. If it is done in an automatic way, times of breakdown are reduced to the minimum. In the end, Cloud systems allow to use the minimum of the necessary resources to the amount of work: this results in the decrease in the infrastructure and energy costs.
Some experts in the world of informatics believe, however, that Cloud Computing systems can be a trap: the use of these technologies distances the end user from his data (from photos to work sheets) which are entrusted to third parts. The risk of this behaviour is the los of control over the data themselves and over the access to data, so problems connected to privacy.