Design And Implementation Of Agricultural Products E-commerce Cloud Platform Based On Container Engine

With the vigorous development of digital agriculture,cloud computing technology has become the backbone of serving new agricultural information.Nowadays,it has been widely used in agricultural production,supply chain and sales process,such as intelligent agriculture,agricultural informatization,etc.,and the functions provided are constantly improving.With the deepening of agricultural reform and the gradual opening of the ’ last kilometer ’,the ecommerce platform of agricultural products has become an important means of upgrading the agricultural industry.However,with the rapid increase of access terminals and business volume,the demand for business-intensive computing cannot be fully met on the agricultural e-commerce cloud platform of the traditional virtual machine architecture.Aiming at this problem,this thesis reconstructs the agricultural e-commerce cloud platform of the internship unit with the help of cloud computing technology represented by Docker container.Taking the later system expansion and service operation as the starting point,through demand analysis and design,a portable and flexible agricultural product e-commerce cloud platform based on container engine is realized.The main work is as follows:1、The architecture design of agricultural e-commerce cloud platform based on container engine.The high availability design of master node,slave node and application realizes the high availability of the whole cluster and ensures the continuity of business.At the same time,the cloud platform database is designed and optimized,and the design of cluster management function and application management function is completed,which is conducive to the expansion and contraction of cloud platform cluster nodes and the control of application scaling,and better meets the needs of high concurrent access,load balancing and other scenarios.2、Construction of application automation deployment model based on continuous integration.It provides support for applications from four aspects : application construction,deployment,monitoring,and performance,reduces the upgrade and maintenance costs of traditional application development models,accelerates the efficiency of application development and deployment,and improves the Dev Ops culture of the platform.Through the open source monitoring technology such as c Advisor and Prometheus,the cloud platform resource monitoring function is constructed,which can monitor the running state and resource usage of the cloud platform cluster in real time.Through the visual panel,the anomaly can be found quickly and intuitively,and the alarm function is provided to notify the user actively,which provides decision support for cluster resource management,operation and maintenance troubleshooting.3、Design and application of dynamic scheduling component based on pluggable.In view of the fact that the original scheduling strategy does not fully consider the dynamic update of resources,it is re-optimized.The cloud platform resource monitoring function monitors the use of node resources to perform dynamic scheduling.After scheduling performance test comparison,the performance and reliability of dynamic scheduling are tested,and the resource utilization rate is improved to a certain extent.At the same time,considering the special needs of the platform to bind the business to specific nodes,the directional scheduling function is realized in combination with label management.The two scheduling methods of dynamic scheduling and directional scheduling can be used in combination with the original scheduling method,which greatly satisfies the universality of cloud platform container scheduling.