What is the cloud, how does it work and what is it for?
The cloud, in computing terms, refers to the delivery of technology services over the Internet. These services include storage, processing, data management, and applications, among other technological resources. Instead of maintaining physical hardware and software on-premises, individuals and organisations access these resources remotely through a global network of servers managed by cloud service providers.
The cloud works by virtualising computational resources. Through virtualisation, a physical server can be hosted on multiple virtual servers. Cloud service providers operate large data centres full of these virtualised servers. Users access these resources via the Internet, either through a web interface, mobile applications or APIs. Information is available wherever you go and whenever you need it.
The cloud has multiple applications, including:
- Storage: saving files and data (ej. Google Drive, Dropbox).
- Computation: running applications and processing data (ej. Amazon Web Services, Microsoft Azure).
- Databases: managing relational and non-relational databases (ej. Amazon RDS, MongoDB Atlas).
- Development and testing: providing development and test environments for applications.
- Artificial Intelligence and data analysis: running AI models and big data analytics.
Cloud types
Public cloud
The public cloud is a cloud computing model in which services are provided by external providers over the Internet. These services may include storage, data processing, networking and other IT resources. Prominent examples of public cloud providers include Amazon Web Services (AWS), Google Cloud Platform (GCP) y Microsoft Azure.
Advantages: resources can be scaled quickly as needed, services are accessible from anywhere with an Internet connection and are pay-as-you-go, eliminating the need to invest in proprietary infrastructure.
Disadvantages: data is stored off-site, which can raise security concerns and may lead to dependence on a single provider and its services.
Private cloud
The private cloud is an infrastructure dedicated exclusively to a single organisation. It can be hosted on company premises or in a service provider’s data centre. This model offers full control over data and security.
Advantages: greater control over data security and privacy, and infrastructure and services are customised to meet specific needs.
Disadvantages: higher initial investment in infrastructure and maintenance and less flexibility to scale compared to the public cloud.
Hybrid cloud
Hybrid cloud combines public and private cloud infrastructure, enabling portability of data and applications between the two. This model provides greater flexibility, taking advantage of the benefits of both clouds.
Advantages: the ability to move workloads between public and private clouds as needed, using the public cloud for fluctuating tasks and the private cloud for sensitive data, and facilitating disaster recovery and business continuity.
Disadvantages: managing and integrating different environments can be complex, ensuring security and interoperability between clouds can be a challenge.
Multi-cloud
The multi-cloud approach involves the use of multiple public cloud services from different providers for various tasks. This allows organisations to avoid dependence on a single provider and select the best services for their specific needs.
Advantages: risk of supplier failure is reduced by distributing workloads, selection of suppliers is based on their specific performance and costs, you can take advantage of the best features and services from multiple providers.
Disadvantages: managing multiple providers and services can be complex, and integration and interoperability between different cloud platforms can be a challenge.
Main cloud delivery models
SaaS (Software como Service)
Saas is a software delivery model where applications are made available over the Internet. Users can access the applications via a web browser, without the need to install and manage the applications on their own devices. Examples: Gmail, Salesforce y Microsoft Office 365.
Characteristics:
- Accessibility: applications are available from anywhere with an Internet connection.
- Maintenance: the service provider is responsible for software updates and maintenance.
- Subscription model: generally a periodic (monthly or annual) fee is paid for the use of the service.
- Scalability: facilitates expansion or reduction of usage according to the user’s needs.
PaaS (Platform as a Service)
PaaS provides a platform that allows developers to build, deploy and manage applications without having to manage the underlying infrastructure. This includes development environments, databases and application management tools. Examples: App Engine, Heroku y Microsoft Azure App Services.
Characteristics:
- Development environment: provides pre-configured environments for development and deployment.
- Integrated services: includes databases, authentication services and other tools necessary for application development.
- Simplified management: the provider manages the underlying infrastructure, such as servers and networks.
IaaS (Infrastructure as a Service)
IaaS provides basic computing resources, such as virtual servers, storage and networking, which users can provision and manage through a user interface or API. Examples: AWS EC2, Google Compute Engine y Microsoft Azure VMs.
Characteristics:
- Virtualised resources: provides virtualised computing resources that can be scaled as needed.
- Complete control: users have full control over the operating system, applications and settings.
- Pay-per-use: payment model based on actual use of resources.
FaaS (Function as a Service)
FaaS, also known as serverless computing, allows developers to execute code fragments in response to events without having to manage the underlying infrastructure. Providers automatically manage the provisioning and scaling of the required resources. Examples: AWS Lambda, Google Cloud Functions y Azure Functions.
Characteristics:
- Event-driven execution: code is executed in response to specific events, such as changes in a database or HTTP requests.
- Automatic scaling: scales automatically according to the workload.
- Pay-as-you-run: you pay only for code execution time, not for downtime.
