基于云平台的任务管理系统的设计与实现

发布时间：2018-05-21 15:36

本文选题：OpenStack云平台 + 任务管理系统　；参考：《沈阳工业大学》2017年硕士论文

【摘要】：随着人工智能、大数据、云计算的日益发展,OpenStack作为当今基础设施即服务的开源云平台之一,以其节约企业成本为特点逐渐得以被广泛应用。本课题基于OpenStack云平台,采用微服务架构实现基于云计算平台的分布式系统,并结合实际项目需求设计一款任务管理系统来提高公司的工作效率,强化员工之间的协同办公能力。本文首先根据OpenStack各个功能模块,在实验室环境搭建OpenStack私有云平台,为分布式任务管理系统提供充足物理资源。其次结合实际任务管理系统项目,以微服务架构设计思想从功能、业务、监控等各个方面对任务管理系统进行需求分析,在保证功能不变的情况下,根据任务管理系统中的不同业务将系统分解为多个可管理的分支或不同的服务。最后结合Spring Cloud、Swagger UI等技术手段,建立一套前后端分离的微服务系统架构,并在系统中引入负载均衡能力、熔断机制以及服务网关等概念提高了分布式任务管理系统的并发性、容错性、安全性。针对微服务之间的异步进程通信问题,利用Kafka分布式消息队列提高并发时的消息吞吐量以及解决因同步等待引起的大量系统时间消耗问题;针对分布式任务管理系统结合Kafka而产生的问题,采用Hash算法解决消息的绝对有序性,以及采用时间轮算法解决时间维度的消息回溯问题。最终使用AngularJS框架构建整个系统的Web前端页面,利用Zipkin对分布式调用链问题进行监控分析,并实现了任务管理系统中日程任务管理、人力资源管理管理、后台管理、消息提醒等功能。通过实际应用表明,基于云平台的分布式任务管理系统对业务进行了松耦合,使系统更加容易扩展、可靠性高,并实现敏捷开发,具有较好的工程实践意义。
[Abstract]:With the development of artificial intelligence, big data and cloud computing, OpenStack, as one of the open source cloud platforms of infrastructure as service, has been widely used because of its cost saving. This topic is based on OpenStack cloud platform, using micro-service architecture to realize the distributed system based on cloud computing platform, and design a task management system according to the actual project requirements to improve the efficiency of the company and strengthen the ability of co-working among employees. Firstly, according to the function modules of OpenStack, the private cloud platform of OpenStack is built in the laboratory environment, which provides sufficient physical resources for the distributed task management system. Secondly, combining with the actual task management system project, using the micro-service architecture design idea to analyze the requirement of the task management system from the aspects of function, business, monitoring and so on, under the condition that the function remains unchanged, The system is decomposed into multiple manageable branches or different services according to the different services in the task management system. Finally, combined with Spring CloudSawagger UI and other technical means, a set of micro-service system architecture with separate front and rear end is established, and the concepts of load balancing ability, fuse mechanism and service gateway are introduced into the system to improve the concurrency of distributed task management system. Fault tolerance, security. Aiming at the problem of asynchronous process communication between microservices, Kafka distributed message queue is used to improve the throughput of concurrent messages and to solve the problem of system time consumption caused by synchronous wait. Aiming at the problem of distributed task management system combined with Kafka, Hash algorithm is used to solve the absolute order of message, and time round algorithm is used to solve the problem of message backtracking in time dimension. Finally, the AngularJS framework is used to build the Web front-end page of the whole system, and the distributed call chain problem is monitored and analyzed by using Zipkin, and the scheduling task management, human resource management and background management are realized in the task management system. Message alerting, etc. The practical application shows that the distributed task management system based on cloud platform is loosely coupled to the business, which makes the system easier to expand, more reliable, and realizes agile development, which has good engineering practical significance.
【学位授予单位】：沈阳工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP311.52;TP393.09

【相似文献】