Design And Implementation Of Agricultural Cloud Platform Resource Monitoring System Based On Docker Cluster

With the popularization of the Internet and the development of cloud computing technology,more and more applications are deployed based on cloud platform technology.The traditional cloud platform 1.0 is usually built with virtual machine technology,which can reduce software operation and maintenance costs to a certain extent and improve security.However,this cloud platform does not perform well in terms of virtual machine startup time and deployment of new applications,while cloud platform 2.0 using Docker container technology has faster startup time and consumes less resources,especially in terms of microservices.At present,with the massive application of a new generation of information technology represented by the Internet of Things in agricultural production practice,various agricultural information systems are increasing day by day,and it is usually necessary to integrate and manage these systems with the help of cloud platforms.Especially with the research and construction of smart agriculture,the number of terminals and the amount of data generated by the agricultural cloud platform has grown rapidly,and the contradiction between business-intensive computing and low-computing equipment on the business side has become more common.The computing power of the cloud is used to complete the computing work.In the smart agriculture scenario,in addition to statically loading application services on the cloud platform,terminal devices also need to dynamically load services on the cloud platform to make real-time decisions on short-cycle data.The virtual machinebased cloud platform cannot handle this well.application scenarios.The cloud platform 2.0 technology based on Docker containers is more suitable for this scenario due to its own advantages.The current Docker container technology still has the problem that the resource management mechanism is not perfect,and the current management system is not flexible enough in dynamic resource allocation,and cannot maximize the overall utilization of the system on the premise of ensuring the quality of application services.The Central Plains Rural Information Port is a comprehensive service platform of Henan Province’s national rural informatization demonstration province based on virtual machine technology and Docker container cluster technology.This paper takes the Docker container cluster of the agricultural cloud platform of the Central Plains Rural Information Port as the research object,and focuses on the management of Docker container resources.,carry out the container resource monitoring module design,build the container resource prediction model,and establish a resource monitoring and management system based on the Docker cluster cloud platform on this basis to realize the real-time monitoring of container resources,resource elastic management and automatic deployment of application services on the agricultural cloud platform.The main research contents include:(1)Design a cloud platform resource monitoring and management architecture based on Docker cluster.Using c Advisor,Prometheus and other technologies to build a container cluster cloud platform resource monitoring architecture,collect system resources used by each container in real time,and store these data persistently in the Influx DB database to provide decision-making data for elastic resource management.(2)Build a container resource prediction algorithm based on deep learning.According to the time series monitoring data of the container cluster,such as CPU and memory,a two-layer LSTM,Attention-LSTM,GRU-LSTM and other models are constructed to predict resources,and according to the prediction results,the elastic scaling of the container cluster and the timely alarm of abnormal data are realized.In the application scenario of the agricultural cloud platform,higher requirements are placed on the operating efficiency of the resource prediction model,and the prediction time of the GRU-LSTM prediction model is 80% more efficient than that of the Attention-LSTM model,and the GRU-LSTM model predicts the memory usage rate.The MSE of GRU-LSTM is0.032,while the MSE of the two-layer LSTM model in predicting memory usage is 298.33.In contrast,GRU-LSTM predicts better.Therefore,this paper selects GRU-LSTM as the prediction model used by the container resource data module in the system.(3)Design the application automation deployment model.The image management and application automatic deployment model based on Jenkins technology is designed,and the test results show that it can realize the effective management of platform images and the automatic deployment of platform applications.(4)Based on the above research results,an agricultural cloud platform resource monitoring and management system based on Docker cluster was constructed,and the simulation experiment of agricultural cloud platform resource management in the Central Plains Rural Information Port was carried out.The experimental results show that the system can realize the monitoring and visualization of container resources.Display,resource prediction and elastic scaling scheduling,automatic deployment of applications and other management functions.The agricultural cloud platform resource monitoring and management system based on Docker container cluster constructed in this paper realizes the effective management of platform resources,can improve the overall resource utilization of the platform,and can provide reference ideas and technical tool support for agricultural cloud platform resource management.