Build cloud Code delivery chain _codepipeline Practice Sharing _ Architecture

Absrtact: In the 2017 online technology summit-the first Alibaba research and development performance carnival, the engineers from Aliyun flying Research and Development department Mo Yuan shared the "build cloud code delivery chain, Codepipeline practice sharing." On the basis of cloud computing and cloud platform, continuous integration process and devops, he shares in detail the performance and practice of Alibaba Cloud Codepipeline superior to Jenkins.

In the 2017 online technology summit-the first Alibaba research and development performance Carnival, Mo Yuan, an engineer from Aliyun's Flying Research and Development division, shared the "build cloud code delivery chain, Codepipeline practice sharing." On the basis of cloud computing and cloud platform, continuous integration process and devops, he shares in detail the performance and practice of Alibaba Cloud Codepipeline superior to Jenkins.

The following content is compiled from the live video.

Live Video: https://yq.aliyun.com/edu/lesson/549

PDF Download: https://yq.aliyun.com/attachment/download/?id=1844

Cloud computing and cloud platforms

The Cloud computing field is very similar to other areas, and when an area becomes more and more mature, the direction of development will evolve towards refinement. Ofo, for example, solves the problem of how to get out of the city's public transport to solve the last kilometer. Cloud computing is also becoming more mature, the cloud solves the problem of infrastructure delivery, but not enough for developers, cloud native architecture design and application has become the "last kilometer" problem of cloud computing.

The diagram above is a cloud native architecture design Netflag defined by the company. Mainly divided into three parts: the framework (left), the recommended use of micro-service architecture, Container Runtime (right), the proposed container, resource lifecycle Automation; Infrastructure automation (middle), It expresses the idea of the infrastructure, the code, and the operation and maintenance automation of the piping.

To transform the diagram just now, we can sum it up to four main points: DevOps, continuous Delivery, microservices, containers. But for developers, micro-services and containers are not required, but are related to their own models, programming frameworks, while devops and continuous integration are common models. As a cloud computing vendor, we're dealing with how developers can solve their own problems under a common model.

The above figure is familiar, and tells about the Cloud Platform's SaaS, PaaS, IaaS layer. At the bottom is the IAAS layer, which provides the resources, the PAAs layer provides the capability, and the SaaS layer provides the service. The bottom of the pyramid is the most solid, the more on the smaller, and DevOps is the spire. Therefore, only the infrastructure of the lowest level of the environment becomes very solid, heavy, to ensure that the top of the devops have a more efficient ability. Continuous integration process

Common DevOps processes are divided into code development, the submission of code to the Code warehouse, triggered by the web hook, the continuous integration system deployed to the test environment, the pre-launch environment, the formal environment.

The above diagram is a continuous integration flowchart based on container. The difference from the previous diagram is that there is a mirrored warehouse at the top of the continuous integration system, and the continuous integration system pushes the mirror to the mirrored warehouse, which is deployed by pulling the mirror from the deployment environment.

Original link