The concept of cloud computing began with large-scale Internet service providers, such as Google (Google Cloud Services), Amazon AWS (2006), Microsoft (Microsoft Azure) or Alibaba Cloud") and others that built their own infrastructure. From all of them an architecture emerged: a system of horizontally distributed resources, introduced as massively scaled virtual IT services and managed as continuously configured and pooled resources. This architectural model was immortalized by George Gilder") in his October article 2006 in Wired magazine titled The information factories. Server farms, which Gilder wrote about, were similar in their architecture to mesh computing, but while networks were used for loosely coupled technical processing applications, a system composed of subsystems with a certain autonomy of action, which maintained a continuous interrelation between them, this new cloud model was being applied to Internet services.[2]​.

The fundamental concept of delivering computing resources over a global network has its roots in the 1960s. The idea of ​​an "intergalactic computer network" was introduced in the 1960s by JCR Licklider, whose vision was that everyone could be interconnected and be able to access programs and data from anywhere. According to Margaret Lewis, director of product marketing at AMD, "it's a vision that looks a lot like what we call cloud computing."

Other experts attribute the scientific concept of cloud computing to John McCarthy "John McCarthy (scientist)"), who proposed the idea of ​​computing as a public service, similar to service companies dating back to the 1960s. In 1960 he said: "someday computing may be organized as a public service."[3]​.

Since the 1960s, cloud computing has developed along a number of lines. Web 2.0 is the most recent evolution. However, because the Internet didn't start offering significant bandwidth until the 1990s, cloud computing has been something of a late bloomer. One of the early milestones of cloud computing is the arrival of Salesforce.com in 1999, which pioneered the concept of delivering business applications through a simple web page. The services firm paved the way for both specialists and traditional software companies to publish their applications over the Internet.

The next development was Amazon Web Services in 2002, which provided a set of cloud-based services, including storage, computing, and even human intelligence through Amazon Mechanical Turk. Later in 2006, Amazon launched its Elastic Compute Cloud (EC2) as a commercial service that allows small businesses and individuals to rent computers on which to run their own computing applications.

“Amazon EC2/S3 was the first to offer fully accessible cloud infrastructure services,” according to Jeremy Allaire, CEO of Brightcove, which provides its online video SaaS platform to UK television stations and newspapers.

George Gilder said in 2006: "The desktop PC is dead. Welcome to the Internet cloud, where a huge number of facilities around the planet will store all the data you will ever use in your life."

Another important milestone came in 2009, when Google and others began offering browser-based applications. «The most important contribution to cloud computing has been the emergence of "killer apps" from technology giants such as Microsoft and Google. “When these companies conduct their services in a way that is safe and easy for the consumer, the ‘passing the buck’ effect itself creates a feeling of greater acceptance of online services,” according to Dan Germain, chief technology office at IT service provider Cobweb Solutions.

Another of the key factors that have allowed cloud computing to evolve has been, according to British cloud computing pioneer Jamie Turner, virtualization technologies, the development of universal high-speed bandwidth and universal software interoperability standards. Turner added: “As cloud computing spreads, its reach goes beyond a handful of Google Docs users. We can only begin to imagine its scope and scope. Almost anything can be used in the cloud.

There are several characteristics of cloud computing, and depending on the author or source they may vary. For example, NIST[4]​ specifies that there are essential characteristics of a cloud:

In a broader sense and based on various sources, cloud computing has the following key characteristics:.

Contenido

La computación en la nube establece su arquitectura a partir de una fragmentación entre aplicación informática, plataforma "Plataforma (informática)") y hardware, dando como resultado según NIST[4]​ los siguientes métodos de entrega: software como servicio, plataforma como servicio e infraestructura como servicio.[5]​.

Software as a service (SaaS)

Software as a service is at the highest layer and characterizes a complete application offered as an on-demand service through multitenancy—a single instance of the software running on the provider's infrastructure serves multiple customer organizations. The applications that provide this service model are accessible through a web browser—or any application designed for this purpose—and the user has no control over them, although in some cases they are allowed to make some configurations. This eliminates the need for the client to install the application on their own computers, avoiding the costs of support and maintenance of hardware and software.

Platform as a Service (PaaS)

The middle layer, platform as a service, is the encapsulation "Encapsulation (computing)") of an abstraction of a development environment and the packaging of a series of modules or plugins that provide, typically, horizontal functionality (data persistence, authentication, messaging, etc.). In this way, a platform as a service archetype could consist of an environment containing a basic stack of systems, components or APIs preconfigured and ready to be integrated into a specific development technology (for example, a Linux system, a web server, and a programming environment such as Perl or Ruby). Platform-as-a-service offerings can serve all phases of the software development and testing cycle, or they can be specialized in any particular area, such as content management. PaaS simplifies the user experience by minimizing technical management, allowing users to focus on their applications without worrying about network control, servers, operating system or storage.[7]​.

Commercial examples are Google App Engine, which serves applications from the Google infrastructure; Microsoft Azure, a cloud platform that allows the development and execution of applications coded in various languages ​​and technologies such as .NET, Java, Go and PHP or the G Platform, developed in Perl. PaaS services like these allow for great flexibility, but may be restricted by the capabilities available through the provider. In this service model, the user is offered the development platform and programming tools so they can develop their own applications and control the application, but they do not control the infrastructure.

Infrastructure as a service (IaaS)

Infrastructure as a service—also called in some cases hardware as a service, HaaS)—[8]​ is located at the bottom layer and is a means of delivering basic storage and computing capabilities as standardized services on the network. Servers, storage systems, connections, routers, and other systems are concentrated (for example through virtualization technology) to handle specific types of workloads—from batch processing to server/storage surge during peak loads. The best-known commercial example is Amazon Web Services, whose EC2" and S3 services offer essential compute and storage services (respectively). Another example is Joyent, whose core product is a line of virtualized servers, which provide highly scalable, on-demand infrastructure for running websites, including complex web applications written in Python, Ruby, PHP, and Java.

Types of PaaS

There are several types of platforms as a service, including public, private and hybrid.[9]​ Paas was originally intended for public clouds, before expanding to private and hybrid clouds.[9]​

Public PaaS are derived from software as a service and are located between it and infrastructure as a service.[10]​.

Private PaaS are commonly downloaded and installed from a company's local infrastructure, or from a public cloud. Once the software is installed on one or more machines, private PaaS organizes the application and database components onto a single platform for hosting.[11] PaaS providers include Apprenda, which began on the Microsoft .NET platform; Red Hat's OpenShift and its Cloud Foundry Pivotal.[12]​ Apprenda and Microsoft were considered the only two PaaS that provided superior .NET support.[9]​ Now accompanied by the announced[13]​ partnership between Microsoft and IBM[14]​.

A hybrid PaaS is typically a deployment consisting of a mix of public and private deployments. An example here is IBM Bluemix[15]​ which is delivered as a single integrated cloud platform across public, dedicated and on-premises deployment models.

Started in 2012, mobile PaaS (mPaaS) provides development capabilities for mobile application designers and developers.[16]​ The Yankee Group identified mPaas as one of its themes for 2014, naming several providers including Kinvey, CloudMine, AnyPresence, FeedHenry, FatFractal and Point.io.[17][18]​.

Open PaaS does not include hosting, but instead provides open source software that allows a PaaS provider to run applications in an open source environment. For example, AppScale allows a user to deploy some applications written for Google App Engine to their own servers, providing data warehouse access from a standard SQL or NoSQL database. Some open platforms allow the developer to use any programming language, database, operating system or server to deploy their applications.[19]​[20]​.

In 2014, Forrester Research defined Public Enterprise Platforms for Rapid Developers as an emerging trend, naming several vendors including Mendix, Salesforce.com, OutSystems, and Acquia Acquia").[21]

The advantages of platforms as a service are that it allows for higher levels of programming with extremely reduced complexity; The overall development of the application can be more efficient, since you have an integrated infrastructure; and the maintenance and improvement of the application is simpler[22]​ It can also be useful in situations in which several developers are working on the same project and that involve parties that are not located close to each other.[23]​ They can provide new features without the need to incorporate more staff. Development options are offered on various platforms to handle different applications. It is possible to collaborate on projects where team members are in different locations, since the development environment is accessed through the Internet. Thanks to platforms as a service, the lifecycle of web applications can be managed effectively[24].

Cloud development layers are part of the set of tools and services used to build modern distributed applications. These layers allow you to implement business logic, coordinate microservices, store large-scale data, and process event streams without needing to manage the underlying infrastructure.[26]​.

One of the main layers corresponds to the cloud-native architecture, which is based on independent microservices that are deployed autonomously and communicate through APIs. This model facilitates horizontal scalability and modular application development.[27]​.

Another relevant layer is serverless computing, where services such as AWS Lambda allow functions to be executed in response to events without managing servers. Logic from multiple functions can be coordinated using orchestration engines such as AWS Step Functions, which control workflow and fault tolerance[28].

In the field of data, managed databases and data lakes allow you to store and process structured and unstructured information. Platforms like Snowflake take a decoupled approach between compute and storage, allowing for simultaneous analysis by multiple users or systems.[29]​.

Finally, the messaging and event streaming layer includes systems such as Apache Kafka, which implement a publish–subscribe model to process data in real time through distributed topics. This layer is essential for applications that require high availability and continuous information processing.[30]​.

These development layers complement the SaaS, PaaS and IaaS models, and represent the modern way in which scalable, resilient and distributed applications are built in public, private and hybrid cloud environments.[31]​.

According to NIST[4]​ we can identify the following implementation models:

Within these clouds we can find virtual private clouds,

[33]​ considered as public domain clouds but that improve data security. This data is encrypted through the implementation of a VPN.

Some large companies like Amazon already allow you to provision a section of your AWS with this VPN possibility.

Additionally, we can identify a model that is increasingly present in the market:

The use of multiclouds presents a series of advantages such as: greater performance, in the event of a cloud failing, others can be operated; data transfer is carried out through various platforms, accelerating such an operation; Security is greater since several hosts can be used, costs are reduced and they allow diffusion with the regulations of different countries.[37]​.

Multicloud can be confused with hybrid cloud. Its main difference is that the multicloud combines services from different providers not connected to each other and the hybrid cloud is a mixture of local private cloud and third-party public cloud but connected to each other.[38]​.

Advantages

The main advantages of cloud computing are:

Disadvantages

The disadvantages of cloud computing are generally related to reliability, availability of services, access, security and, above all, storage. Below are details:

Cloud computing is usually confused with grid computing (a form of distributed computing whereby a "virtual supercomputer" is composed of a linked set of loosely coupled computers that work in concert to perform very large tasks).[42]

Since cloud computing does not allow users to physically own the storage devices for their data (with the exception of the ability to copy the data to an external storage device, such as a USB flash drive or hard drive), it leaves the responsibility for data storage and control in the hands of the provider.

Cloud computing has been criticized for limiting the freedom of users and making them dependent on the service provider.[43] Some critics claim that it is only possible to use the applications and services that the provider is willing to offer. Thus, The Times compares cloud computing with the centralized systems of the 1950s and 1960s, in which users connected through "gregarious" terminals with central computers. Users were generally not free to install new applications, and needed administrator approval to perform certain tasks. In short, both freedom and creativity were limited. The Times argues that cloud computing is a return to that era and numerous experts support the theory.[44]​.

Similarly, Richard Stallman, founder of the Free Software Foundation, believes that cloud computing endangers users' freedoms because they leave their privacy and personal data in the hands of third parties. He has stated that cloud computing is "simply a trap designed to force more people into proprietary, locked-in systems that will cost them more and more as time goes on."[45]​.

La seguridad en la computación en la nube puede ser tan buena o mejor que la que existía en los sistemas tradicionales, porque los proveedores son capaces de proporcionar recursos que resuelvan problemas de seguridad que muchos clientes no pueden afrontar. Sin embargo, la seguridad todavía sigue siendo un asunto importante, cuando los datos tienen un matiz confidencial. Esto atrasa la adopción de la computación en la nube hasta cierto punto.

Security as a service

In the cloud environment, security is provided by the providers. Two methods can be distinguished: The first method is that anyone can change their delivery methods included in the cloud services. The second method is that cloud service providers provide security only as a cloud service, with security information from the companies. Security services are classified into categories, according to the Cloud Security Alliance organization:

Browser security

In the cloud environment, remote servers are used for computing. Client nodes are used only for input/output operations, and for authorization and authentication of information in the cloud. A standard web browser is a commonly used platform for all users around the world. This can be categorized into two different types: software as a service, web applications, or Web 2.0. Transport Layer Security (TLS) is typically used for data encryption and host computer authentication.

Authentication or identification

In the cloud environment, the basis for access control is authentication, access control is more important than ever since the cloud and all its data are accessible to everyone through the internet. Trusted Platform Module (TPM) is widely used and a stronger authentication system than username and password. Trusted Computing Group (TCG) is a standard for user authorization and other real-time communication security tools between provider and customer.

Loss of governance

In cloud infrastructures, the customer necessarily cedes control to the cloud provider in various matters, which negatively influence security. At the same time, the service level agreement is typically not intended to provide these types of services on the cloud provider's side, leaving a gap in security defenses.

Dependence

This is a small offering on such tools, procedures or standards for data formats or service interfaces that could ensure data, applications and service portability. This can make it difficult for the customer to migrate from one provider to another, or migrate data and services back to another computing environment, locking them into the chosen ecosystem. This introduces a particular dependency on the cloud provider for the provision of the service, especially with regard to data portability, the most fundamental aspect.

Data protection

In Spain, the Spanish Data Protection Agency and the National Institute of Communication Technologies (INTECO) are the organizations in charge of privacy and the protection of personal data; key aspects of operating in the cloud.[47]​ Cloud computing puts data protection at risk for cloud users and their providers. In many cases, it causes difficulties for the provider (in the role of information controller) to ensure the practical effectiveness of the cloud provider's data management and to ensure that the data is on the right path. This problem is often exacerbated in cases of multiple data transfers, for example between federated systems. On the other hand, some cloud providers provide information on their data collection practices. There are also some certification offerings in data processing, security activities, and data controls that take place; example, SAS70 certification"). Internet data streams are laced with malware and decoy packets to lure the user into unknown involvement in criminal activities.

Algunas limitaciones que están retrasando un poco a la computación en la nube son algunas de las siguientes:.

Data loss/leak

Having data in the cloud has numerous advantages but it can be compromised in different ways. They can be modified or deleted without a backup or backup, they can be detached from their context or accessed by unauthorized people.

Difficulty in assessing the reliability of suppliers

If Cloud providers are part of our society, viewed statistically, someone on their staff or a subcontractor might not be trustworthy. This is why good human resources background investigation procedures, strong information security policies and procedures must be taken.

Strength of authentication mechanisms

There is a lot of data, applications and resources stored in the cloud. Cloud computing is very weak in authentication mechanisms, therefore the attacker can easily obtain the client user account and access the virtual machine.[48]​.

Many universities, institutes, providers and government institutions are investing in cloud computing:[49]​[50]​.

The concept of cloud computing began with large-scale Internet service providers, such as Google (Google Cloud Services), Amazon AWS (2006), Microsoft (Microsoft Azure) or Alibaba Cloud") and others that built their own infrastructure. From all of them an architecture emerged: a system of horizontally distributed resources, introduced as massively scaled virtual IT services and managed as continuously configured and pooled resources. This architectural model was immortalized by George Gilder") in his October article 2006 in Wired magazine titled The information factories. Server farms, which Gilder wrote about, were similar in their architecture to mesh computing, but while networks were used for loosely coupled technical processing applications, a system composed of subsystems with a certain autonomy of action, which maintained a continuous interrelation between them, this new cloud model was being applied to Internet services.[2]​.

The fundamental concept of delivering computing resources over a global network has its roots in the 1960s. The idea of ​​an "intergalactic computer network" was introduced in the 1960s by JCR Licklider, whose vision was that everyone could be interconnected and be able to access programs and data from anywhere. According to Margaret Lewis, director of product marketing at AMD, "it's a vision that looks a lot like what we call cloud computing."

Other experts attribute the scientific concept of cloud computing to John McCarthy "John McCarthy (scientist)"), who proposed the idea of ​​computing as a public service, similar to service companies dating back to the 1960s. In 1960 he said: "someday computing may be organized as a public service."[3]​.

Since the 1960s, cloud computing has developed along a number of lines. Web 2.0 is the most recent evolution. However, because the Internet didn't start offering significant bandwidth until the 1990s, cloud computing has been something of a late bloomer. One of the early milestones of cloud computing is the arrival of Salesforce.com in 1999, which pioneered the concept of delivering business applications through a simple web page. The services firm paved the way for both specialists and traditional software companies to publish their applications over the Internet.

The next development was Amazon Web Services in 2002, which provided a set of cloud-based services, including storage, computing, and even human intelligence through Amazon Mechanical Turk. Later in 2006, Amazon launched its Elastic Compute Cloud (EC2) as a commercial service that allows small businesses and individuals to rent computers on which to run their own computing applications.

“Amazon EC2/S3 was the first to offer fully accessible cloud infrastructure services,” according to Jeremy Allaire, CEO of Brightcove, which provides its online video SaaS platform to UK television stations and newspapers.

George Gilder said in 2006: "The desktop PC is dead. Welcome to the Internet cloud, where a huge number of facilities around the planet will store all the data you will ever use in your life."

Another important milestone came in 2009, when Google and others began offering browser-based applications. «The most important contribution to cloud computing has been the emergence of "killer apps" from technology giants such as Microsoft and Google. “When these companies conduct their services in a way that is safe and easy for the consumer, the ‘passing the buck’ effect itself creates a feeling of greater acceptance of online services,” according to Dan Germain, chief technology office at IT service provider Cobweb Solutions.

Another of the key factors that have allowed cloud computing to evolve has been, according to British cloud computing pioneer Jamie Turner, virtualization technologies, the development of universal high-speed bandwidth and universal software interoperability standards. Turner added: “As cloud computing spreads, its reach goes beyond a handful of Google Docs users. We can only begin to imagine its scope and scope. Almost anything can be used in the cloud.

There are several characteristics of cloud computing, and depending on the author or source they may vary. For example, NIST[4]​ specifies that there are essential characteristics of a cloud:

In a broader sense and based on various sources, cloud computing has the following key characteristics:.

Contenido

La computación en la nube establece su arquitectura a partir de una fragmentación entre aplicación informática, plataforma "Plataforma (informática)") y hardware, dando como resultado según NIST[4]​ los siguientes métodos de entrega: software como servicio, plataforma como servicio e infraestructura como servicio.[5]​.

Software as a service (SaaS)

Software as a service is at the highest layer and characterizes a complete application offered as an on-demand service through multitenancy—a single instance of the software running on the provider's infrastructure serves multiple customer organizations. The applications that provide this service model are accessible through a web browser—or any application designed for this purpose—and the user has no control over them, although in some cases they are allowed to make some configurations. This eliminates the need for the client to install the application on their own computers, avoiding the costs of support and maintenance of hardware and software.

Platform as a Service (PaaS)

The middle layer, platform as a service, is the encapsulation "Encapsulation (computing)") of an abstraction of a development environment and the packaging of a series of modules or plugins that provide, typically, horizontal functionality (data persistence, authentication, messaging, etc.). In this way, a platform as a service archetype could consist of an environment containing a basic stack of systems, components or APIs preconfigured and ready to be integrated into a specific development technology (for example, a Linux system, a web server, and a programming environment such as Perl or Ruby). Platform-as-a-service offerings can serve all phases of the software development and testing cycle, or they can be specialized in any particular area, such as content management. PaaS simplifies the user experience by minimizing technical management, allowing users to focus on their applications without worrying about network control, servers, operating system or storage.[7]​.

Commercial examples are Google App Engine, which serves applications from the Google infrastructure; Microsoft Azure, a cloud platform that allows the development and execution of applications coded in various languages ​​and technologies such as .NET, Java, Go and PHP or the G Platform, developed in Perl. PaaS services like these allow for great flexibility, but may be restricted by the capabilities available through the provider. In this service model, the user is offered the development platform and programming tools so they can develop their own applications and control the application, but they do not control the infrastructure.

Infrastructure as a service (IaaS)

Infrastructure as a service—also called in some cases hardware as a service, HaaS)—[8]​ is located at the bottom layer and is a means of delivering basic storage and computing capabilities as standardized services on the network. Servers, storage systems, connections, routers, and other systems are concentrated (for example through virtualization technology) to handle specific types of workloads—from batch processing to server/storage surge during peak loads. The best-known commercial example is Amazon Web Services, whose EC2" and S3 services offer essential compute and storage services (respectively). Another example is Joyent, whose core product is a line of virtualized servers, which provide highly scalable, on-demand infrastructure for running websites, including complex web applications written in Python, Ruby, PHP, and Java.

Types of PaaS

There are several types of platforms as a service, including public, private and hybrid.[9]​ Paas was originally intended for public clouds, before expanding to private and hybrid clouds.[9]​

Public PaaS are derived from software as a service and are located between it and infrastructure as a service.[10]​.

Private PaaS are commonly downloaded and installed from a company's local infrastructure, or from a public cloud. Once the software is installed on one or more machines, private PaaS organizes the application and database components onto a single platform for hosting.[11] PaaS providers include Apprenda, which began on the Microsoft .NET platform; Red Hat's OpenShift and its Cloud Foundry Pivotal.[12]​ Apprenda and Microsoft were considered the only two PaaS that provided superior .NET support.[9]​ Now accompanied by the announced[13]​ partnership between Microsoft and IBM[14]​.

A hybrid PaaS is typically a deployment consisting of a mix of public and private deployments. An example here is IBM Bluemix[15]​ which is delivered as a single integrated cloud platform across public, dedicated and on-premises deployment models.

Started in 2012, mobile PaaS (mPaaS) provides development capabilities for mobile application designers and developers.[16]​ The Yankee Group identified mPaas as one of its themes for 2014, naming several providers including Kinvey, CloudMine, AnyPresence, FeedHenry, FatFractal and Point.io.[17][18]​.

Open PaaS does not include hosting, but instead provides open source software that allows a PaaS provider to run applications in an open source environment. For example, AppScale allows a user to deploy some applications written for Google App Engine to their own servers, providing data warehouse access from a standard SQL or NoSQL database. Some open platforms allow the developer to use any programming language, database, operating system or server to deploy their applications.[19]​[20]​.

In 2014, Forrester Research defined Public Enterprise Platforms for Rapid Developers as an emerging trend, naming several vendors including Mendix, Salesforce.com, OutSystems, and Acquia Acquia").[21]

The advantages of platforms as a service are that it allows for higher levels of programming with extremely reduced complexity; The overall development of the application can be more efficient, since you have an integrated infrastructure; and the maintenance and improvement of the application is simpler[22]​ It can also be useful in situations in which several developers are working on the same project and that involve parties that are not located close to each other.[23]​ They can provide new features without the need to incorporate more staff. Development options are offered on various platforms to handle different applications. It is possible to collaborate on projects where team members are in different locations, since the development environment is accessed through the Internet. Thanks to platforms as a service, the lifecycle of web applications can be managed effectively[24].

Cloud development layers are part of the set of tools and services used to build modern distributed applications. These layers allow you to implement business logic, coordinate microservices, store large-scale data, and process event streams without needing to manage the underlying infrastructure.[26]​.

One of the main layers corresponds to the cloud-native architecture, which is based on independent microservices that are deployed autonomously and communicate through APIs. This model facilitates horizontal scalability and modular application development.[27]​.

Another relevant layer is serverless computing, where services such as AWS Lambda allow functions to be executed in response to events without managing servers. Logic from multiple functions can be coordinated using orchestration engines such as AWS Step Functions, which control workflow and fault tolerance[28].

In the field of data, managed databases and data lakes allow you to store and process structured and unstructured information. Platforms like Snowflake take a decoupled approach between compute and storage, allowing for simultaneous analysis by multiple users or systems.[29]​.

Finally, the messaging and event streaming layer includes systems such as Apache Kafka, which implement a publish–subscribe model to process data in real time through distributed topics. This layer is essential for applications that require high availability and continuous information processing.[30]​.

These development layers complement the SaaS, PaaS and IaaS models, and represent the modern way in which scalable, resilient and distributed applications are built in public, private and hybrid cloud environments.[31]​.

According to NIST[4]​ we can identify the following implementation models:

Within these clouds we can find virtual private clouds,

[33]​ considered as public domain clouds but that improve data security. This data is encrypted through the implementation of a VPN.

Some large companies like Amazon already allow you to provision a section of your AWS with this VPN possibility.

Additionally, we can identify a model that is increasingly present in the market:

The use of multiclouds presents a series of advantages such as: greater performance, in the event of a cloud failing, others can be operated; data transfer is carried out through various platforms, accelerating such an operation; Security is greater since several hosts can be used, costs are reduced and they allow diffusion with the regulations of different countries.[37]​.

Multicloud can be confused with hybrid cloud. Its main difference is that the multicloud combines services from different providers not connected to each other and the hybrid cloud is a mixture of local private cloud and third-party public cloud but connected to each other.[38]​.

Advantages

The main advantages of cloud computing are:

Disadvantages

The disadvantages of cloud computing are generally related to reliability, availability of services, access, security and, above all, storage. Below are details:

Cloud computing is usually confused with grid computing (a form of distributed computing whereby a "virtual supercomputer" is composed of a linked set of loosely coupled computers that work in concert to perform very large tasks).[42]

Since cloud computing does not allow users to physically own the storage devices for their data (with the exception of the ability to copy the data to an external storage device, such as a USB flash drive or hard drive), it leaves the responsibility for data storage and control in the hands of the provider.

Cloud computing has been criticized for limiting the freedom of users and making them dependent on the service provider.[43] Some critics claim that it is only possible to use the applications and services that the provider is willing to offer. Thus, The Times compares cloud computing with the centralized systems of the 1950s and 1960s, in which users connected through "gregarious" terminals with central computers. Users were generally not free to install new applications, and needed administrator approval to perform certain tasks. In short, both freedom and creativity were limited. The Times argues that cloud computing is a return to that era and numerous experts support the theory.[44]​.

Similarly, Richard Stallman, founder of the Free Software Foundation, believes that cloud computing endangers users' freedoms because they leave their privacy and personal data in the hands of third parties. He has stated that cloud computing is "simply a trap designed to force more people into proprietary, locked-in systems that will cost them more and more as time goes on."[45]​.

La seguridad en la computación en la nube puede ser tan buena o mejor que la que existía en los sistemas tradicionales, porque los proveedores son capaces de proporcionar recursos que resuelvan problemas de seguridad que muchos clientes no pueden afrontar. Sin embargo, la seguridad todavía sigue siendo un asunto importante, cuando los datos tienen un matiz confidencial. Esto atrasa la adopción de la computación en la nube hasta cierto punto.

Security as a service

In the cloud environment, security is provided by the providers. Two methods can be distinguished: The first method is that anyone can change their delivery methods included in the cloud services. The second method is that cloud service providers provide security only as a cloud service, with security information from the companies. Security services are classified into categories, according to the Cloud Security Alliance organization:

Browser security

In the cloud environment, remote servers are used for computing. Client nodes are used only for input/output operations, and for authorization and authentication of information in the cloud. A standard web browser is a commonly used platform for all users around the world. This can be categorized into two different types: software as a service, web applications, or Web 2.0. Transport Layer Security (TLS) is typically used for data encryption and host computer authentication.

Authentication or identification

In the cloud environment, the basis for access control is authentication, access control is more important than ever since the cloud and all its data are accessible to everyone through the internet. Trusted Platform Module (TPM) is widely used and a stronger authentication system than username and password. Trusted Computing Group (TCG) is a standard for user authorization and other real-time communication security tools between provider and customer.

Loss of governance

In cloud infrastructures, the customer necessarily cedes control to the cloud provider in various matters, which negatively influence security. At the same time, the service level agreement is typically not intended to provide these types of services on the cloud provider's side, leaving a gap in security defenses.

Dependence

This is a small offering on such tools, procedures or standards for data formats or service interfaces that could ensure data, applications and service portability. This can make it difficult for the customer to migrate from one provider to another, or migrate data and services back to another computing environment, locking them into the chosen ecosystem. This introduces a particular dependency on the cloud provider for the provision of the service, especially with regard to data portability, the most fundamental aspect.

Data protection

In Spain, the Spanish Data Protection Agency and the National Institute of Communication Technologies (INTECO) are the organizations in charge of privacy and the protection of personal data; key aspects of operating in the cloud.[47]​ Cloud computing puts data protection at risk for cloud users and their providers. In many cases, it causes difficulties for the provider (in the role of information controller) to ensure the practical effectiveness of the cloud provider's data management and to ensure that the data is on the right path. This problem is often exacerbated in cases of multiple data transfers, for example between federated systems. On the other hand, some cloud providers provide information on their data collection practices. There are also some certification offerings in data processing, security activities, and data controls that take place; example, SAS70 certification"). Internet data streams are laced with malware and decoy packets to lure the user into unknown involvement in criminal activities.

Algunas limitaciones que están retrasando un poco a la computación en la nube son algunas de las siguientes:.

Data loss/leak

Having data in the cloud has numerous advantages but it can be compromised in different ways. They can be modified or deleted without a backup or backup, they can be detached from their context or accessed by unauthorized people.

Difficulty in assessing the reliability of suppliers

If Cloud providers are part of our society, viewed statistically, someone on their staff or a subcontractor might not be trustworthy. This is why good human resources background investigation procedures, strong information security policies and procedures must be taken.

Strength of authentication mechanisms

There is a lot of data, applications and resources stored in the cloud. Cloud computing is very weak in authentication mechanisms, therefore the attacker can easily obtain the client user account and access the virtual machine.[48]​.

Many universities, institutes, providers and government institutions are investing in cloud computing:[49]​[50]​.