Shulou(Shulou.com)06/02 Report--

On April 26, 1956, a crane loaded 58 aluminum truck carriages onto an old oil tanker moored in the harbor, and no one would have thought that a revolution would begin. The birth of containers not only greatly reduces the transportation cost of goods, but also makes the originally chaotic transport environment neat and orderly, the whole logistics system more efficient, and finally changes the world trade.

In the IT world, containers also play the role of "containers", except that the content of "transportation" has been changed from goods to applications. It is lighter than a virtual machine, starts quickly and takes up less resources. It can be run in the same way in almost any scene, and the unified scheduling of development, testing and production environment can be realized, and the same running environment can be created.

If the container provides a set of standardized logistics system, then the container provides a set of standardized application release way.

Around the container, a variety of new architectures have been born to meet the needs of different scenarios. In the market, more and more enterprises apply container technology to practice, and while containers are popular, they also expose a variety of disadvantages of container technology:

First of all, although container technology simplifies the complex process of continuous product delivery, the realization of containerization is complex, and it needs the support of many technologies, such as container management, orchestration, application packaging, container-to-container network, data snapshot and so on.

Second, when enterprises want to build a network solution, they need to be proficient in all kinds of advanced network functions provided by Linux, this technical threshold is too high. Especially for Docker users who focus on business development, such operations tend to be too complex. Moreover, due to the deployment of containers in virtual machines, both cloud platform and Docker platform have their own virtualized network implementation solutions, and their functions overlap and will be nested with each other, resulting in serious network performance loss, even up to 80%.

Therefore, although container technology is gradually being recognized and applied, its network performance and configuration complexity have always been criticized by everyone.

So, how to solve these problems?

In order to solve the above problems, Qingyun QingCloud reconstructs the basic network and storage layer to realize the unified architecture of container and virtualization, and makes a layer of abstraction based on the underlying resources, namely QingCloud Application Framework, which isolates the underlying hardware resources and provides interfaces to upper-layer applications, providing an infrastructure that supports both virtualization and containerization.

Then, SDN (Software defined Network) and SDS (Software defined Storage) are used to provide the ultimate network and storage performance support for the container operating environment. Through the SDN network pass-through feature (SDN Passthrough), the container directly uses the SDN of the cloud platform to avoid the packet capacity loss caused by the overlap of the two-layer networks and save the performance that would have to be lost by multiple virtualization.

At the same time, thanks to the elastic scalability of the Qingyun platform, the container platform deployed on the virtual host will also have the ability to scale flexibly, and users can scale the entire system horizontally and vertically. In the application layer, the Qingyun container platform is delivered through the QingCloud AppCenter framework, deployed with one click, continuously upgraded, and provides application lifecycle management features such as creation, expansion, health monitoring and user management, as well as comprehensive service monitoring and logging features to help users save a lot of OPS costs.

Generally speaking, when an enterprise uses container services in Qingyun QingCloud, it can enjoy the complete enterprise-class container service platform provided by Qingyun, including a series of container applications and services including Kubernetes, etcd, Harbor and public image repositories, covering image repositories, container choreography and management, so as to quickly develop, deploy and upgrade container applications.

Give me a few examples.

Break the heterogeneous environment and realize the application of CI/CD

In the traditional DevOps architecture, the development, test and production environments are often heterogeneous, and there is no unified standard process for service deployment, which directly contradicts the realization of efficient application continuous integration / delivery. The container orchestration and scheduling platform based on Docker and Kubernetes realizes a unified application deployment environment and provides standardized deployment and update functions.

Docker deployment, management and orchestration

Docker is a very mainstream container application. Through the QingCloud container platform, users can easily deploy Docker instances, manage their own and public Docker images, and realize the scheduling and orchestration of Docker through Kubernetes on QingCloud, thus building an application system based on Docker architecture.

Build a micro-service architecture

With the rapid growth of business and the expansion of team size, single applications become more and more difficult to meet the needs of agile development and vertical scaling, and need to evolve to micro-service architecture. With the use of micro-services, some technical debts are bound to be transferred from development to operation and maintenance. The traditional manual scheduling mode is difficult to cope with the operation and maintenance complexity caused by the expansion of the number of services. At this time, container scheduling system becomes an inevitable choice. Moreover, when microservices are used in conjunction with containers, the advantages of microservices architecture will be further magnified. Kubernetes's definition of service specification, rolling upgrade and automatic scalability meet the requirements of micro-service architecture for operation and maintenance system, and reduce the overall cost of adopting micro-service architecture.

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.