Cloud computing is a model for providing computing and network resources, as well as data storage and other IT resources on demand through the Internet. The peculiarity of this model is that resources can be obtained with payment upon their use. The second name for cloud computing is cloud computing, or calculations that are performed in the cloud, and this service is available as an online service.
The history of cloud computing
The first attempts to combine computing resources and distribute them, like pieces of a big pie, between different users, were made back in the 60s of the last century. But in the form in which we know what it is, cloud computing appeared thanks to the company SalesForce, which first began to provide a subscription-based CRM system. Further development of this area is associated with Amazon, which today is the largest provider of cloud services on the AWS platform. Its main competitors in the global market are Azure Microsoft and Google Clouds. In Russia, such services are offered by Mail.Ru Cloud Solutions, Selectel, Rostelecom, Yandex – this is brief, in fact there are many times more of them.
How cloud computing works
Cloud computing involves two participants (in some deployment models, this may be one company). The first (provider) builds a data center, places servers and data storage systems there, lays network equipment and deploys software that manages the IT economy and correctly distributes it among users. The second – the user, the client company – enters into an agreement with the provider and requests a certain pool of resources from him: any number of processors, amount of memory for storage, Internet channel width, number of IP addresses, resource management tools (for example, a balancer).
Characteristics of cloud computing
Allocate resources at the request of the user
The work is based on the principle of self-service. This is when the client himself decides how many resources and when he needs. You can request resources yourself through your personal account on the platform of the cloud provider. At the same time, the client does not need to contact the provider’s managers by phone, e-mail or other channels. Instead, he receives the necessary computing power a couple of minutes after manipulations in his personal account.
This principle is convenient for companies that are faced with an irregular load on their services. For example, at night their clients have minimal activity, so they need fewer resources. And, on the contrary, the need for them increases sharply before the New Year holidays, during sales days, after the launch of large-scale advertising campaigns.
Payment for consumed resources
Own data centers are usually built with a margin so that they can cope with growing loads without upgrading equipment. Cloud computing technologies get rid of this. Payment for use is charged monthly after the fact – when the processor power, storage capacity and other parameters of the leased resources are calculated. This is convenient, since the company does not overpay for unused capacity, is spared from the capital costs that inevitably arise during the construction of its data center, and can plan the costs of IT infrastructure in the long term.
Seamless pooling of resources
The provider can have any number of servers, storage systems, switches and incoming Internet channels, but this is not essential for users. Special software combines the available resources so that the platform looks like a single entity. And her clients receive from him their portion of any size. If a provider wants to increase the number of servers in a rack, purchase a batch of fast SSD drives, this will not affect its customers and services: the scaling will go unnoticed, and the total capacity of the data center will increase.
Flexible scaling options
A company that has chosen cloud computing gets all the tools for rapid growth without being tied to the size of its IT infrastructure. Since the infrastructure is in the cloud and is easily scalable, nothing holds back the pace of development.
Another benefit that the business receives is the savings in human resources. The company does not need to maintain a staff of technical specialists, but these are jobs, wages, social security and regular professional development. One or two engineers are enough to deal with the operation of services in the cloud, and the provider takes over the tasks listed above. He also makes sure that the equipment works stably and client services do not stop. Clients themselves can focus on priority tasks while their IT fleet adapts flexibly to their current needs.
Deploying capacity in minutes
If the company opens one or more branches, the scale of the business has changed – this will not be a reason for an inventory of IT equipment. Instead, the system engineer will change the amount of computing resources in the personal account on the cloud platform. The changes take effect in a few minutes. This allows you to improve the quality of customer service and not be tied to a physical “hardware” for short- and long-term planning.
Cloud provider services can be quantified. This is important for companies that strictly control financial flows. The account for the use of resources in the cloud is billed at the end of the month. It can be used to conclude which resources were most in demand and in what periods, as well as to predict future costs.
High availability of computing resources
Providers and their customers enter into an agreement that describes the procedure for providing computing resources in the cloud and the obligations / responsibilities of the parties. The task of the provider is to make sure that as many customers as possible are interested, and for this, services must be provided without delays and stops, and storage systems must be fault-tolerant.
The business understands that even a half-hour downtime will lead to an outflow of customers and negatively affect the reputation, so it chooses providers with data centers that have a high level of reliability (determined by the Tier parameter) and have a clear SLA – an agreement on the provision of services. The latter prescribes what time during the year the client’s services may be unavailable. This is usually the time needed for regular service.
In the cloud model, technical support works 24/7. The engineers, who can be contacted online or by phone, have a high level of competence and will quickly respond to problems that arise in working with cloud resources.
Benefits for providers and their customers
Cloud computing services are a model of interaction in which both parties are interested. Providers can sell computing power in the cloud so that they do not have unclaimed resources (servers, data storage systems.). Their clients receive other benefits: actual payment for the consumed resources, flexible scaling depending on the needs, avoiding capital expenditures for maintaining the IT infrastructure.
Companies are moving to a cloud computing model to get rid of buying and maintaining hardware where they need to quickly deploy new environments for development and testing, get dynamic load distribution, high elasticity, scalability and independence from physical hardware are required, transparent control over the costs of IT infrastructure.
Cloud computing has three basic resource deployment models.
- Public cloud, or Public cloud
Many unrelated users have access to IT resources, and a third-party provider manages the infrastructure. Companies receive computing power in the cloud on the basis of an agreement concluded with a provider. It ensures that users do not intersect with each other in their work, do not claim the same pool of resources and do not have access to each other’s data;
- A private cloud, or Private cloud,
IT resources are used by one company, including its subordinate structures: remote offices, branches, retail outlets (from retailers). In some cases, the list includes contractors that the company contacts to perform one-time work or outsource tasks to them. Both the company itself and the provider with whom it has entered into an appropriate agreement can manage the operation of a private cloud. In the same way, IT resources can be located on the territory of the company or on the premises of the provider’s data center;
- Hybrid cloud or Hybrid cloud
The company’s IT infrastructure is divided between public and private clouds. Part of the computing resources is under its full control in a private cloud, another part – in a public one. Between them, links have been established that allow services from different clouds to exchange data in real time.
Sometimes in this classification there is also a common cloud (community cloud or Community cloud). This is a type of public cloud designed for use by members of the same community or users united by common goals.
Cloud Computing as a Service
What is the concept of cloud computing? Usually this definition is often used in relation to IaaS, PaaS and SaaS. These are all customer service models in the cloud. It is under these names that they are offered by cloud providers, they are also cloud service providers.
Infrastructure as a service, where the user only leases computing resources from the provider. This is the so-called lite version of cloud computing and is the easiest and cheapest model to implement. The provider guarantees users the performance of their services at the server level. In its purest form, IaaS works on the principle of a virtual data center, when a supplier installs special software on one or more machines, and users deploy any number of virtual machines through its interface. Also examples of IaaS are dedicated (Dedicated) and virtual private / dedicated (VPS / VDS) servers.
The advantages of IaaS are in the ease of renting, the cost affordable even for small businesses, the ability to install any operating systems and application programs. Of the minuses, companies are forced to turn to the services of computer specialists to manage a virtual IT infrastructure, and computing and other capabilities directly depend on physical servers and the software that the provider installs on them.
IaaS can be used by any company that offers infrastructure rental services in the cloud. Among them are a large number of Russian cloud providers. IaaS is ideal for hosting websites of any size and content, application testing and development, data storage (backup), high performance computing and big data analysis.
Platform as a service, where a provider offers users access to a software platform. The user rents the processing power of processors, a certain amount of disk drives for their databases, receives a data transmission channel, white IP addresses, additional tools like a balancer, and others. It can use any number of virtual machines and install software based on needs. This is the optimal solution for large companies that are looking for reliability and scalability in the deployment of solutions and their subsequent operation.
The advantages of PaaS include the elasticity of the model for providing computing resources in the cloud, high deployment speed, a large list of additional services, tools in the cloud and connected services, a flexible payment system that linearly depends on the resources consumed. The downside is that the client company needs to have at least one technically savvy specialist on staff who will deal with the IT infrastructure.
The most famous examples of PaaS are Amazon services (AWS platform), Windows Azure, Oracle Cloud Platform. Their services are used by companies that are looking for a stable environment for developing and testing applications, are engaged in business analytics and forecasting.
Software as a service, where the user receives a ready-made set of applications from the service provider. The easiest to understand definition of this concept is when you work in the cloud using only the necessary software and do not think about what computing and other resources are behind it.
The advantage of SaaS is that the technical side of the issue is completely hidden from the user, and the terms for ordering and receiving the service are minimal – you can start using cloud computing in a few minutes after its activation. Cons – a limited list of software available in this format, and limited options for customizing the program. In most cases, the user is highly dependent on the stability of the SaaS provider and the service is more expensive when it comes to deploying large solutions in the cloud (backed by large customers).
The most popular examples of SaaS are Microsoft Office 365 and Bitrix 24 office applications. Software as a service is used for email delivery, CRM (customer relationship management) and ERP (company resource planning) systems, and enterprise document management systems.