英文论文2

1 Introduction
In recent years, with the accelerated advancement of enterprise digital transformation, traditional monolithic architecture has increasingly exposed its inherent limitations when dealing with increasingly complex business requirements. Monolithic architecture concentrates all business logic within a single application. As business scale expands, problems such as high code coupling, escalating maintenance costs, and lengthy deployment cycles become increasingly prominent. Concurrently, the maturation of cloud computing technologies and the widespread adoption of DevOps practices have established a solid technical foundation for the implementation of microservices architecture [1].
Microservices architecture, as an emerging software architectural style, is fundamentally characterized by the decomposition of a single application into a suite of small, independently deployable services. Each service is built around specific business capabilities and communicates through lightweight mechanisms. This architectural style not only enhances system flexibility and maintainability, but also effectively supports continuous delivery and continuous deployment practices central to DevOps [2]. In the System Architect qualification examination, microservices architecture and its applications first appeared as an essay topic in 2016 and was again selected as an examination topic in 2021, fully demonstrating the technology's significant standing and sustained influence within the field of software engineering.
This paper takes the vehicle dispatch management system for a large logistics enterprise — a project in which the author participated — as a practical case study to deeply explore the application methods and key technologies of microservices architecture in enterprise-grade application systems. The system handles an average daily volume exceeding 100,000 vehicle dispatch requests, imposing stringent requirements on system availability, scalability, and real-time responsiveness. The author served as the system architect for this project, responsible for overall architecture design and technology selection, with a focus on resolving core technical challenges including service decomposition strategy, inter-service communication mechanisms, and distributed data management.
2 Related Work
In recent years, significant progress has been made in both theoretical research and engineering practice regarding microservices architecture design patterns. Gintoro et al., in a paper published in IEEE Access, conducted a quantitative evaluation of 110 microservices across 15 open-source applications using the ISO 25002:2024 standard as a reference framework. Their research found that the API Gateway pattern demonstrated the most outstanding performance in improving system modifiability, achieving a 49.7% reduction in Change Impact Factor (CIF) and a 17.6% improvement in the Modifiability Index, and proposed the "Worst First" principle for refactoring prioritization [3]. Rocha et al. performed a systematic mapping study of 2,653 microservices architecture publications from 2013 to 2024, ultimately selecting 76 high-quality papers for in-depth analysis. Their work established a systematic taxonomy of microservices architecture patterns, providing architects with a scientific basis for pattern selection decisions [4]. Gurram, in research published in 2026, proposed a cloud-native reference framework that combines microservice modularity with event-driven agility, employing the Strangler Fig pattern to achieve phased migration from legacy systems to cloud-native architecture, thereby offering a viable technical pathway for the smooth transition of traditional enterprise applications to microservices architecture [5]. Seiger and Malburg presented a detailed case study of refactoring a smart factory cyber-physical system from monolithic to microservices architecture, with particular attention to reconciling non-functional requirements such as fault tolerance and security within resource-constrained embedded environments [6].
The aforementioned research provides a solid theoretical foundation for the practical implementation of microservices architecture in enterprise applications. However, questions regarding how to determine appropriate service granularity based on specific business contexts and how to ensure data consistency in distributed environments still warrant further exploration. Building upon the above research and incorporating practical project experience, this paper proposes a microservices architecture design methodology tailored specifically for enterprise application systems.
3 System Architecture Design
3.1 Business Background and Requirements Analysis
This system serves a nationwide logistics enterprise with the following core business requirements: (1) real-time monitoring and management of the operational status of over 5,000 transport vehicles, including real-time acquisition and display of critical parameters such as vehicle location, speed, and fuel consumption; (2) support for high-concurrency vehicle dispatch requests, processing over 500 dispatch instructions per second during peak periods while maintaining system stability; (3) real-time acquisition, persistent storage, and retrospective analysis of vehicle trajectory data, supporting historical trajectory playback for up to 90 days; (4) data integration with the enterprise's existing ERP and CRM systems to enable seamless business data flow; (5) multi-terminal access supporting Web browsers, mobile applications (Android/iOS), and third-party enterprise system integration.
The non-functional requirements include: core service availability of 99.9% or above (annual downtime not exceeding 8.76 hours); end-to-end request latency not exceeding 200ms (P99 percentile); horizontal elastic scaling capability to accommodate business peaks; eventual consistency guarantees for critical business data; and comprehensive monitoring, alerting, and audit logging capabilities.
3.2 Overall Architecture Design
Based on the aforementioned requirements analysis, this paper adopts a microservices architecture solution centered on Spring Cloud Alibaba. The overall architecture employs a layered design, consisting — from bottom to top — of the infrastructure layer, data layer, service layer, gateway layer, and presentation layer.
The infrastructure layer employs Docker containerization technology and Kubernetes cluster orchestration. Application environments are standardized and made reproducible through container images, while Kubernetes' automated orchestration capabilities enable automatic deployment, elastic scaling, and fault self-healing of services. The data layer adopts a Database per Service deployment strategy, where each microservice possesses an independent database instance, thereby avoiding coupling at the data level. Storage engines are selected based on the read-write characteristics of different services: MySQL for core business data, MongoDB document storage for trajectory data, and Redis for caching.
The service layer comprises five core microservice modules: the User Service, responsible for user authentication, authorization management, and organizational structure; the Vehicle Management Service, responsible for full lifecycle management of vehicle information and real-time vehicle status maintenance; the Dispatch Service, responsible for creation, assignment, execution tracking, and completion confirmation of dispatch tasks, serving as the system's core business module; the Track Service, responsible for receiving and persistently storing real-time location data reported by vehicle GPS terminals; and the Analytics Service, responsible for offline analysis and report generation based on trajectory and dispatch data. Inter-service communication employs RESTful APIs for synchronous calls and RocketMQ message queues for asynchronous communication.
The gateway layer is built upon Spring Cloud Gateway, a reactive API gateway that uniformly handles routing, authentication, authorization, rate limiting, and request logging for all external requests. The presentation layer supports multi-channel access through Web browsers (Vue.js front-end framework), mobile applications (native Android/iOS apps), and third-party enterprise systems (via open API interfaces).