Design And Implementation Of Resource Allocation And Container Cluster Management System Based On Docker

In recent years,"cloud computing" has become more and more hot at one time,and more and more computing has begun to join the cloud.Due to the dynamic scalability of its configuration and the virtualization of resources,many benefits can be brought.Many Internet companies Cloud applications will be applied to the cloud,and cloud computing will become a battleground for Internet companies.Looking at the computer field from a development perspective,the computing and networking environments that users need in the future must be extremely complex,so we must be able to manage it more effectively.This way developers don't have to worry about complex operating environments,and can put 100%of their energy into the core business logic.In addition,we hope to provide users with a continuous integrated development environment based on this.On the basis of providing users with server and computing environment,we also hope that users can easily and quickly apply for resources and environment,and for containers.Object management.A few years ago,there were still many cloud computing platforms running on virtual machines.Although the deployment method of physical machines has been greatly improved,there are still some problems in resource utilization and performance.The big data computing framework is a typical application scenario on the cloud platform.Based on this cloud,when dealing with massive data,the big data framework cannot cope with this processing environment because of its resource rate utilization rate and performance constraints..So it's not the most efficient way to deliver an application.Docker is a container engine with lightweight and better computing performance,more suitable for micro-services,and more efficient isolation and control of resources.Therefore,Docker has a good application prospect in the field of HPC and cloud computing.Since Docker has the above advantages,it is completely possible to use Docker as the underlying to completely replace the virtual machine,so that efficient virtualization can be realized,and progress can be made in the start and stop of the application.On the basis of this,if resources can be redistributed,And with efficient container orchestration,a Docker-based distributed container cluster management system that can perform resource scheduling can be implemented.This paper mainly implements a distributed container orchestration and deployment system based on Docker.It has only two core components,which is simple to deploy and can be expanded arbitrarily.It provides scheduling for different resource dimensions,and is also responsible for orchestrating containers.It implements three efficient container allocation algorithms to enable efficient and transparent container deployment and orchestration behavior,and develops a web API to provide users with interfaces to access.So that even if the user does not understand the internal implementation details,it can be quickly and easily implemented.On this basis,the workflow is unified through the upper support components to reduce the complexity of operation and maintenance,providing a continuous integrated development envirornment integrating development,deployment,and operation and maintenance,for organizations and individual developers seeking efficient operation and maintenance solutions.The words are more friendly.Essentially,it provides the same container orchestration and scheduling capabilities as k8s/swarm,but with two components,which minimizes the threshold for learning,usage,and deployment,and is more pure and simple than mesos,and has less complicated structure.Just need a machine and a docker.