Shulou(Shulou.com)11/24 Report--

About two years ago, an editor of the brain pole body left, and the "culprit" was a topic about the origin of the cloud, which directly caused his psychological line of defense to collapse. It is not that he is too fragile. When most people see the original complex content of the cloud, I am afraid they will doubt life.

Click on sandboxie of the Cloud Native Computing Foundation (Cloud Native Computing Foundation,CNCF) and you will see hundreds of different applications and solutions that dazzle developers; cloud manufacturers have more than 100 cloud native products.

About the popular science of cloud origin in the market, the painting style often goes like this: "understanding the four elements of cloud origin", "counting the five major characteristics of cloud origin", "talking about the seven principles of cloud native architecture", "the 12 elements of cloud native application". They all look professional, but Chinese and English are mixed and unfamiliar with numerous concepts, and different cloud manufacturers and different organizations have different opinions. Party An and readers just want to find out what the hot cloud is, not to take the postgraduate entrance examination. In the end, most ordinary people who don't understand have to lie flat and let the technicians get high.

The concept of cloud origin has been popular for as long as it takes to popularize science. There are many reasons why it is difficult to explain:

First, the technical system is numerous and complicated. Cloud native is not a specific technology, but a series of specific technologies and products built around the application logic of cloud computing. Different cloud manufacturers often give their own explanations from different angles. Take a look at this NCNF cloud native knowledge graph, who is not confused?

Another reason is that the cloud itself is still evolving. Its definition, characteristics and core technologies are constantly developing. For example, the Cloud Native Computing Foundation (CNCF) was established in 2015 to modify the cloud native definition and further improve the definition in 2018. Today, it is still 12 elements, but in a twinkling of an eye, it has become 15 elements; it has just merged with Docker, and then it has announced that it will be abandoned. In such a rapidly changing situation, it is a bit difficult to remember all kinds of cloud nouns clearly.

It is really difficult to understand cloud origin from a clear and accepted objective definition. However, there is a popular saying in the IT world: "language is only a tool, thinking is the most important." By the same token, thinking is the most important thing to understand the origin of the cloud.

The White Paper on Cloud Native Development (2020) also points out that Cloud Native is an idea for cloud application design and a best practice path that can give full play to cloud effectiveness.

What kind of thought is cloud origin? We think that cloud nativism is actually the localization strategy of enterprises on the cloud.

You may have seen many overseas companies coming to China, or Chinese brands going out to sea, all need to make a series of integrated adjustments in the light of the economy, culture and living customs of local consumers. Similarly, cloud native, born in the cloud in the first place, is more attractive and is considered to give better play to the advantages of cloud services than simply migrating data and applications to a cloud. This is very important to many enterprises, especially the non-digital native entity enterprises.

Cloud Origin, like a native born and raised in the cloud world, is regarded as a high-quality idol and a potential stock for the future. So, how far has Yunyuan developed since its popularity in 2019?

From virtual fire to real red: there are many concepts of cloud origin in the past three years. Here we comb through the more objective timeline to see how cloud origin has been excavated and sought after by a group of star scouts and become Internet celebrities.

Most people go back to 2010, and CTO Paul Fremantle, a WSO2 company, also mentioned the concept of "Cloud Native" in its blog. However, with regard to container, one of the core technologies native to the cloud, it can be traced back to as early as 2003, when Google proposed an extensible distributed file system, Google File System.

In 2013, Pivotal company CTO Matt Stine tweeted that cloud nativism is a collection of technologies. In the same year, the open source container software Docker was released, and the container technology represented by Docker began to rise.

In 2014, Google released the Kuberbetes open source project, which is by far the most widely used tool for container orchestration.

In 2015, Matt Stine proposed cloud native technical features in his book Migration to Cloud Native Application Architecture, including 12 factors, micro services, agile infrastructure, API-based collaboration, anti-vulnerability, and so on. In the same year, CNCF was established to define cloud natively as containerized encapsulation, automated management, and micro-service-oriented.

In 2017, Matt Stine resummarized the six characteristics of cloud native, while the Pivotal website gave the latest definition of cloud native, which can be summarized into four key points: container, micro-service, DevOps, and continuous delivery.

In 2018, CNCF added two new definitions: service grid (Service Mesh) and declarative API. During the same period, almost all major cloud computing providers such as Aliyun, Huawei Cloud and Tencent Cloud joined CNCF,Kubernetes and containers as the established standards for cloud vendors.

After a long stream, Yunyuan ushered in the first year of concept explosion in 2019. Mainstream cloud manufacturers have begun to announce the full implementation of cloud, and vigorously promote and open source cloud native related technologies. According to the White Paper on Cloud Primary Development, China's cloud primary market has reached 35.02 billion yuan in 2019, with large and medium-sized Internet enterprises leading the development of cloud primary industry.

The primary heat of the cloud will continue to heat up in 2020. The survey shows that the use of CNCF projects in Chinese communities has increased by 64%. Two projects from China have graduated from Harbor and TiKV, with a leap-forward growth in usage.

Global cloud native applications continued to rise in 2021, with Forrester data showing that the use of both container and serverless technologies in the organization grew by more than 75% in a year.

This is a bit like a group of brokerage companies launched the "cloud original" this trainee, did not expect to really hit the aesthetic point of the B-end customers, ushered in the explosive growth. The popularity of "cloud origin" has been going on, and it has changed from virtual fire to real red.

Which aesthetic points of the enterprise are poked by cloud origin? Enterprises will not pay the bill with real money because of an illusory concept or trend. Obviously, the cloud native can break out, or related technologies or products, poked the aesthetic point of cloud computing customers, and become a digital rigid demand.

In fact, the adoption of cloud nativism by enterprises is the same as building local teams, and they all see some advantages of localization.

1. Quick response. Of course, the best way for overseas enterprises to adapt to the Chinese market is to hire Chinese people to respond to the needs of users quickly. Similarly, more and more enterprises need to work through digital technology and online software. At this time, they need faster response, faster expansion, and more agile cloud services to replace traditional IT infrastructure such as traditional computer rooms and IDC. It is not enough to use the cloud to replace the server. Some enterprises also want to move their business and services to the cloud, which requires that application development, deployment and maintenance are all done on the cloud. The advantages of native cloud technologies such as containers are naturally welcomed.

two。 Reduce costs. The cost of hiring locals is also much lower than when overseas executives are sent to China. In the process of cloud hosting, application development, operation and maintenance, enterprises have more and more demand for IT personnel, but traditional physical enterprises are obviously not the first choice for IT technical personnel, and personnel costs are also difficult for enterprises to bear. At this time, cloud native through a series of products, tools, methods, can accelerate the software development process, reduce the investment of related resources, and reduce the pressure of operation and maintenance personnel. This is very attractive for enterprises to promote digitization.

3. Scene-oriented scheme. Of course, the localization strategy is not to pick up a local at random, but to build a team according to the characteristics of the industry, such as retail, automobile, manufacturing and so on. In the past, cloud shopping malls mostly provided infrastructure such as cloud servers, CDN and message middleware, but for some entity enterprises with digital transformation, they were more inclined to buy solutions, packaging and purchasing various capabilities such as cloud servers, hardware, API, AI, edge computing and so on, which is a business model that provides scene-based products and solutions to the needs of enterprises. It also makes the cloud, which includes infrastructure and application services, a new opportunity for cloud vendors.

4. Degrees of freedom. In the process of cloud deployment, enterprises instinctively do not want to be completely bound by a cloud. Whether from the point of view of cost reduction, disaster recovery and security, it is a general trend to deploy business loads to different cloud platforms, while hybrid multi-cloud requires applications to be distributed and migrated on different platforms, public clouds, private clouds and hybrid clouds, while a series of cloud native transformations based on Kubernetes and containers Can better deal with the problem of migration.

In other words, the market needs new cloud computing, and the cloud native technologies and methodologies precisely meet the needs of enterprise customers, which is the inevitable trend of the development of the cloud market.

What kind of business ability does the high-quality idol cloud need? You may ask, what kind of cloud origin is a high-quality idol? In fact, cloud native business capabilities include not only basic singing and dancing RAP (cross-cutting) microservices, containers, DevOps, but also newly added AI, security and other capabilities. Here to help you sort it out briefly.

The first is micro-services.

Just like idols have to look at their faces, almost every cloud native definition includes micro-services, which can be said to be the foundation of cloud native life. Corresponding to the micro-service is the single service, that is, the service deployed to a host or a process. If the host goes down, the service cannot be used. On the other hand, microservices are a group of small services, each of which runs as an independent process and implements only one function. In this way, the development is more agile, and each component of the micro-service can be independent and autonomous, and only iteratively improve a certain module. Second, the system is more fault-tolerant and reliable. When one or a micro-service fails, other micro-services will not be affected.

Microservices were proposed as early as 2014, so why is it valued with the native popularity of the cloud? The answer is the development of container technology.

Although the micro-service is good, it runs on the virtual machine, and the virtual machine communicates with each other, forming a complex relationship, which is very troublesome to manage, and there will be a waste of resources. Container technology packages the resources of an application in a container in the way of a container, and then through a container choreographer like Kubernetes, it can well manage resources and application data, and become the best carrier of micro-services. In 2016, Kubernetes won among the three choreographers to become the most mainstream container choreographer, the first "graduation" project incubated by CNCF, and the de facto standard for cloud native infrastructure.

With the increase of micro-services, services and services need to communicate through internal or external networks. How to manage these services leads to the emergence of Service Grid (Service Mesh).

Service Mesh is a lightweight network agent, which provides a proxy instance Sidecar, which deploys the corresponding services together in the environment and forms a network-like dependency relationship. This architecture enables the service grid to handle any communication problems between services, such as load balancing, service discovery, etc., while providing all the index data to the operators to observe and control the traffic. Higher security is achieved through encrypted communication between services, authentication services and so on.

Therefore, Service Mesh has also become the architectural foundation of the next generation of micro services.

The development of micro services is inseparable from the efforts of developers. Traditional IT infrastructure also requires developers to consider infrastructure and underlying resources such as servers, while the popularity of cloud services has also promoted the emergence of Serverless fully managed computing services. Serverless decouples R & D operations and maintenance from infrastructure, infrastructure and servers, eliminates the development and operation and maintenance of underlying infrastructure, and can scale automatically. It greatly reduces the difficulty of resource capacity planning, developers only need to focus on the core business and write code.

In addition to the development process, cloud native has also brought a qualitative leap for DevOps. DevOps (Development+Operations), bridging the gap between development and operation and maintenance, can greatly improve the quality of software development, shorten the development cycle, and ensure continuous delivery. Compared with traditional IT infrastructure, cloud native container technology and Kubernetes orchestration make cross-platform applications possible; Service Mesh improves the flexibility of operation, maintenance and control; and Serverless makes operation and maintenance transparent to development and automatically scale up and down according to resources. DevOps in turn promotes the development of container technology.

As more and more applications and services migrate to micro-service and containerization, API gateway becomes the entrance of micro-service cluster, connecting micro-service-oriented business architecture. The cloud native API design philosophy is that declarative API-- is required to declare the desired running state to the system. If the expected state is inconsistent, cloud native tools such as Kubernetes will take corresponding actions according to the declaration.

With declarative API, once a component in the system fails, you only need to look at the declared status to determine what needs to be done. The traditional process design describes a series of operations that achieve the desired state. When a failure occurs, it is often necessary to find out which step of the operation has gone wrong one by one. Obviously, declarative API greatly reduces the workload of developers and operators.

The gradual landing of cloud native technology has also affected many traditional cloud computing technologies, such as meeting the needs of native elastic scaling, intelligent operation and maintenance, cross-platform migration and so on. Traditional databases also need to be transformed and delivered in the native form of the cloud, so DBaaS (DataBase as a Service, database as a service) emerged.

When introducing cloud natives, enterprises essentially want to enlighten them on the cloud and gain the ability to run AI models for training and reasoning, which drives cloud vendors to develop cloud-based AI systems to flexibly schedule machine learning algorithms through container technology and orchestration and scheduling capabilities, so as to meet the needs of AI for high-performance computing power, improve the efficiency of AI training and reasoning, and make AI services more stable and efficient in cloud application.

How to ensure security after going to the cloud, cloud native also brings a completely different way to solve the problem. Compared with the fragmentation and complexity of the traditional security mechanism, the cloud native security system blurs the physical boundary between servers and hardware terminals, and enables technological trends such as zero trust to fall to the ground. it can achieve unified identity access, unified global management, unified network security connection, covering the whole process, omni-directional and all links of the cloud on the enterprise. It has given birth to cloud native security products and technical solutions, such as cloud firewall, network micro-isolation, Web application firewall, application layer authentication, cloud security center, etc.

It is not difficult to see that cloud native is indeed a treasure quality idol, which represents a kind of scalable and efficient, which can bring a lot of practical and effective benefits to enterprises' digitization and cloud. However, it may take some time to become top-notch.

If not, it is not far: how many steps are there on the way to the top of the original cloud? Successful card position in the customer's aesthetic orientation, but also has a strong business ability, cloud's original popularity seems to be just around the corner. However, to see objectively, the cloud original to the industry in the process, but also need to overcome some practical problems.

On the one hand, although the original cloud is good, it will take time for the scale to land.

First of all, not all industries and enterprises are optimistic about cloud natives. The public cloud is the native base of the cloud, but for some data-sensitive industries and enterprises, they still prefer to deploy on local servers or private clouds, such as finance and healthcare. The proportion of developers in telecommunications, government / defense areas using public cloud is relatively low.

Secondly, the original transformation of the cloud in the physical industry requires mental training. As mentioned in the book "aiming at the Cloud: best practices for leading the Future of Enterprise IT", cloud migration is hard work. It took a full seven years for Netflix, a digital native company, to complete the full migration. In the process, it chose the cloud native method, reconstructed almost all of its own technical methods, and fundamentally changed the way the enterprise operates. The comprehensive transformation of Netflix into a cloud native enterprise is a time-consuming and laborious project, not to mention a large number of non-digital native physical enterprises. These enterprises have often invested a lot of assets in the construction of physical machines or virtual machines. In order to move to cloud origin, they need to change in terms of cost, organizational structure, business process, personnel training, corporate culture, and so on. The risk and cost of this process are relatively high.

Of course, over the past three years, we have also seen that various head cloud manufacturers are actively laying out the cloud native ecology, constantly lowering the cognitive threshold of cloud native through personnel training, industry co-creation, industry-education integration and other ways. For example, Aliyun's cloud native talent training cooperation program, Huawei Cloud's global cloud native communication platform "Creative Society", Tencent Cloud's cloud generation proficiency standard system and "Cloud Native Open Source White Paper" and so on, along with cloud native open source openness and standardization, to expand enterprises' cloud native practices and cultivate the minds of users.

On the other hand, white-hot leads to homogenization, and cloud native competition enters the differentiated track.

For cloud manufacturers, they need new hot spots to win new users in the increasingly homogeneous market competition, and as "cloud native" has become a new tuyere in the market, it has soon become the focus of fierce competition among cloud manufacturers.

A large number of cloud native technologies are open source, and head cloud manufacturers invest in research and development, and they can quickly create their own cloud original products. Combined with a large number of cloud technologies and products precipitated by them, it is easy to create all kinds of "the most complete cloud original product systems". The situation of spending indiscriminately and attracting people's eyes.

For example, each company will emphasize that they have cloud native capabilities such as micro-services, containerization, Mesh, Serverless, DevOps, etc., which can meet the needs of enterprises for agility, flexibility, reliability, low code or even no code. On the contrary, there is no memory point, and has entered the strange circle of homogenized competition. In the end, corporate customers still see who has a low price and who has a good channel relationship.

However, in the past two years, some cloud vendors have also begun to build differentiated cloud native capabilities. For example, from the perspective of underlying hardware, there have been chips specifically aimed at cloud native, which fully release server performance so as to reduce user costs; some manufacturers combine AI, knowledge computing and other technologies with cloud native, to continuously optimize the performance of cloud application AI, accelerate the landing of industry AI, and specifically meet the needs of enterprise business intelligence. Some cloud manufacturers use their own core technologies, such as AI, film and television rendering, database, storage and so on, to build more characteristic cloud native solutions.

In the face of the digital age, non-digital native enterprises are easy to produce estrangement and anxiety. Cloud native provides a very different kind of digital native thinking, which is rooted in the cloud, grows in the cloud, applies to the cloud, and runs in the cloud. Closely glued with the transformation anxiety of the enterprise, this abstract concept has been pushed into flames.

Predictably, as cloud natives are more embedded in the industry and cloud services gradually become ubiquitous, one day we will not care what cloud natives are, because all enterprises are cloud native enterprises. all cloud technologies are cloud native technologies.

Until then, the fast-growing cloud is still worthy of our concern and expectation.

This article comes from the official account of Wechat: brain polar body (ID:unity007), author: Tibetan fox

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.