Design And Implementation Of A High Availability Micro-service System

The topic of this paper comes from the cooperation project between the laboratory and a research institute “High-Availability Micro-service System Construction Technology Research”.The project builds a domain-specific micro-services infrastructure platform which is based on the open source micro-services framework Spring Cloud.The system includes functions,such as services registration,configuration management,service blowing,load balancing,gateway distribution and more.In the initial stage of platform construction,it provides basic management,security and fault tolerance for application services.However,with the increase of the number of instances,the platform exposes many problems,such as the inability to effectively collect the running status of the platform;operation and maintenance personnel management and control services are difficult;system fault tolerance schemes are single.As a result,system operation and maintenance costs have increased significantly,and system availability has been greatly challenged.For above problems,this paper designs and implements “High Availability Micro-service System” by analyzing the application background,system location and operating environment of the micro-service infrastructure platform.The system provides some functions such as service deployment,cluster management,and monitoring distributed collection,monitoring evaluation,instance migration scheduling,and fault-tolerant policy configuration.The control operations of micro-service entities such as clusters,nodes,services,hubs,and instances in the platform are implemented.The specific research content includes the following three points.1.Design and implement micro-service management moduleThe physical model in the platform is obtained by analyzing and modeling each object existing in the platform.According to the relationship between the model and the model and the platform management process,the micro-service management module is designed and implemented.The module implements management of clusters,nodes,services,hubs,and instances,and provides one-click cluster deployment and exception instance migration scheduling for the micro-service platform.In addition,the micro-service management and control module provides users with functions such as system display and user management,which improves the operation convenience of the high-availability micro-service system and reduces the operation and maintenance time and labor costs.2.Design and implement micro-service monitoring moduleThrough analyzing the application environment of the platform and the operation of the service instance,the micro-service monitoring module is designed and implemented to monitor nodes and instances in the cluster.The module mainly includes functions such as collecting and evaluating the status of nodes and instances in the cluster,and persistent storage and visual display.By integrating the evaluation module and the persistent storage module in the node state collection component,combined with the micro-service management module,the distributed collection scheme of the platform is implemented.3.Design and implement micro-service fault tolerant moduleAiming at the single fault-tolerant scheme in the micro-service infrastructure platform,the fault-tolerant strategy in the traditional single-mode mode is used to analyze and select,and the micro-service fault-tolerant module is designed and implemented.This module provides fault-tolerant policy integration and fault-tolerant policy configuration.By integrating the fault-tolerant policy module into the micro-service infrastructure platform by annotation,the fault-tolerant configuration module configures and manages the integrated fault-tolerant policy,and finally implement fault-tolerant management and control of the micro-service infrastructure platform.In order to test the system function,five server nodes are used as the test environment,and the functions of nodes and instances running state collection,the fault-tolerant policy configuration,the cluster deployment,and the instance migration are tested in the100 Mbps LAN.In addition,the system response time and the number of service running instances are tested by the performance test tool Jmeter.Experiments show that in the face of 800 concurrently per second,the platform can still respond normally within 500 ms,and the service instance runs stably.In summary,the high-availability micro-service system constructed in this paper can effectively improve the efficiency and compression time of operation and maintenance personnel.At the same time,it improve the reliability of the micro-service infrastructure platform.