Features of cloud computing
Although the industry has not yet reached a consensus on how to define cloud computing, there have been many in-depth discussions on the characteristics of cloud computing. To determine whether a computing service is a cloud computing service, you can analyze the following cloud computing features.
Figure 1.11 Cloud Computing Features
1.3.1 on-demand service
Cloud computing provides information technology as a service. The concept of such a service is from the consumer (User) perspective, rather than from the service provider to consider the problem, so a basic feature is to require on-demand services, that is, users can instantly obtain services as needed. From this perspective, cloud computing is centrally supplied and billed on demand just like tap water, electricity, and gas in our public services. For users, this on-demand service method can lower the entry and exit thresholds for their service usage, thus reducing the possibility of Service binding and investment risks.
1.3.2 resource pool
One benefit of cloud computing is that it can improve the utilization of resources, which generally needs to be achieved through sharing, and sharing needs to first concentrate resources in a public resource pool. Therefore, cloud computing needs to centralize computing resources into a single resource pool to provide services to different users through excessive primary leasing. For users, the location of specific physical resources is transparent to them, and users generally cannot control the location of specific resources. However, for regulatory compliance and data transmission performance considerations, cloud computing service providers sometimes provide high-level location information, such as in which country or region or in which data center. In cloud computing, resources in this resource pool can be of various types. Based on the type of resources in the resource pool, we divide the cloud computing service model into three categories: the so-called SPI model (see section 1.4 "cloud computing SPI service model ").
1.3.3 high scalability
For Internet applications, the number of users and the demand for resources have changed greatly, making it difficult to estimate the system capacity accurately in advance. For traditional applications, there are also great differences in the application modes used by users in different time periods. The high scalability of the cloud computing platform can meet the requirements of application load and demand changes. The resource pool of the cloud computing platform is relatively large compared to the needs of a single user. Considering that a large number of different users share a single resource pool, the resource usage modes between them are generally complementary, therefore, for a user's needs, cloud computing is highly scalable, and its resources are almost unlimited. In addition, when designing the architecture, the cloud computing platform must consider how users can smoothly expand their resource demands, such as computing and storage resources.
1.3.4 elastic Service
Elastic Service (ECS) allows you to dynamically adjust cloud computing resource allocation based on application access conditions, including requirements for adding or reducing resources. Because of this, cloud computing has very good application effects for applications with non-constant requirements, such as large fluctuations in demand and periodic requirements. In the cloud computing environment, resource scaling can be divided into two categories: one is predictable in advance, such as some seasonal needs; the other is to make real-time dynamic adjustments based entirely on Some preset rules. Regardless of the type, the cloud computing platform is required to provide elastic services. For cloud computing users, they feel that they can obtain computing resources at any time as needed.
1.3.5 self-service, automation, and Virtualization
For people who are used to using ATM or online banking, or who prefer to use the airport to print boarding passes, the flexibility and methods provided by self-service are the main reasons why they like self-service. For service providers, self-service not only improves customer satisfaction, but also reduces service costs. If you print your boarding pass over the Internet, you can save the costs of manual service, printing, and boarding pass printing for airlines, at the same time, you are satisfied with the service you get. Similar to these daily services, self-service in cloud computing is also a way to reduce service costs and improve service convenience.
The concept of automation itself is better understood, and the combination of self-service and automation is because of the internal connection between them. For cloud computing service providers, self-service requires a user interface that is as simple as possible to simplify user operations and reduce the difficulty of using services. Only in this way can self-service providers be accepted by users. In addition, because of man-machine interaction, the service response speed is higher. All of this must be achieved through background automation, that is to say, background automation is the guarantee of the front-end self-service. Therefore, in this sense, self-service is the purpose, while automation is the means.
Virtualization is a way to free up resources from hardware constraints, making dynamic resource allocation possible.
The relationship between these concepts is shown in Figure 1.12.
Figure 1.12 relationship between self-service, automation, and Virtualization
Cloud computing self-service is generally embodied through web portals, just like in Amazon's cloud computing service
Users can use a self-service portal to reserve the required computing resources as easily as ordering a book on an e-commerce website. Automation is generally implemented through programs and a large number of automated scripts, so that the operation background platform triggered by the user on the front-end self-service interface can be automated and respond in a timely manner, thus ensuring a good user experience.
1.3.6 convenient network access
It is an ideal to allow users to access the desired information service from any device anywhere. The development of cloud computing has brought us closer to this goal. Another feature of cloud computing is convenient network access. The centralized supply of resources must be accessed by users through the network. The supply of cloud computing services is the same as that of the grid and water pipeline network after the centralized supply of power and tap water: users can access cloud computing services over networks using different devices, including PCs, mobile phones, and tablets. The development of rich terminal devices combined with network technology will solve the "last mile" problem of information services.
1.3.7 service measurable
Drucker, the father of management, once said, "If you cannot measure it, you cannot manage it ". As a service, cloud computing must measure the service. Generally, a service level agreement (SLA) is required between the service provider and the user ). For private clouds, internal fees can be calculated based on service conditions. For public clouds, service measurement is the prerequisite for billing, and then billing can be performed based on actual usage.