Font Size: a A A

Design And Implementation Of Docker-based OpenStack Cloud Platform Automation Deployment Scheme

Posted on:2020-01-16Degree:MasterType:Thesis
Country:ChinaCandidate:J QinFull Text:PDF
GTID:2428330572967240Subject:Information and Signal Processing
Abstract/Summary:PDF Full Text Request
In the process of automatic installation and deployment of OpenStack cloud platform,the installation and deployment process is complicated due to network complexity,server diversity,cloud platform complexity and other factors.Even if the cloud platform is installed successfully,a large number of problems such as late operation and maintenance work,high coupling between components and difficulty in upgrading will affect the cost and efficiency of cloud platform in production environment.With the advent of 5G era,major operators have begun to invest heavily in the construction of cloud DC,and edge computing has become a hot star.However,edge cloud needs a large number of simplified,convenient and customized cloud platforms to support,and traditional deployment methods cannot meet its needs.The existing containerized deployment scheme in the community is only in the development stage and cannot successfully install and deploy a set of available multi-node and multi-role cloud platform system.After fully studying the automatic installation and deployment of various cloud platforms,this thesis selects the containerization scheme to automate the deployment of OpenStack cloud platform.This thesis designs and implements a complete container installation and deployment system of cloud platform,and realizes this system by combining ansible,docker,Helm,Kubernetes,kolla and other technologies.The system implementation is mainly divided into three modules: First,the design and implementation of hardware resource management module.Since container deployment may involve a large number of servers,complex network environment and manual configuration and installation mode increase the unreliability of cloud platform.By designing this module,the purpose of optimizing configuration and automatic batch setting network can be achieved.The second is the design and implementation of Kubernetes cluster management module.This module contains Kubernetes automatic installation and local warehouse management functions.The offline and highly available performance of Kubernetes in production environment can be achieved by implementing custom components.Third,the design and implementation of OpenStack management module.This module combines a custom kolla image with a cloud platform installation to make the installed cloud platform fully customizable.In addition,the decoupling of OpenStack components is achieved by encapsulating each component of OpenStack in its corresponding Charts,so that it can complete the installation anduninstallation of OpenStack through simple operations.This system can successfully install and deploy self-customized OpenStack components in batch automation in the way of containers.A set of cloud platform system customized by enterprises can be successfully installed and deployed.Through testing,it is convenient to manage and operate the cloud platform,achieving the purpose of container deployment of cloud platform.
Keywords/Search Tags:Automatic deployment, Cloud platform, Container deployment, Build images
PDF Full Text Request
Related items