金蝶天燕 Apusic
Home›Customers›Success Stories›China Railway Beijing Group: A Xinchuang Trio with Zero Business Interruption and Zero Performance Degradation
Transportation

China Railway Beijing Group: A Xinchuang Trio with Zero Business Interruption and Zero Performance Degradation

The Beijing Group completed the Xinchuang migration of multiple core business systems with Kingdee Apusic's AAS, AMDC and ALB trio — no business interruption, no performance degradation.

Overview

China Railway Beijing Group Co., Ltd., founded in 1953, is an extra-large state-owned railway transport enterprise under China State Railway Group. Its jurisdiction covers Beijing, Tianjin and Hebei ("two municipalities and one province") as well as parts of Shandong, Henan and Shanxi, with 251 main lines and more than 10,000 kilometers of operating mileage, undertaking important railway transport tasks in the core Beijing-Tianjin-Hebei region and the Bohai Rim economic circle.

As a national critical infrastructure sector, the information system security of the railway industry directly affects national transport security and public service stability. In 2022, the SASAC of the State Council issued Document No. 79, explicitly requiring all central and state-owned enterprises to complete Xinchuang domestic renovation of their information systems by 2027. China Railway Beijing Group actively responded to national policy, advancing the Xinchuang upgrade and renovation of multiple core business application systems, with the goal of migrating the underlying foundational software from a non-domestic technology stack to an independently controllable domestic environment. Kingdee Apusic participates in the project with three full-stack middleware products — the AAS application server, the AMDC distributed cache and the ALB load balancer — providing foundational software support for the Xinchuang migration and stable operation of multiple Beijing Group business application systems.

Customer Requirements

Full-stack Xinchuang adaptation and domestic replacement

Xinchuang domestic renovation covers core foundational software including operating systems, databases and middleware. As the critical link between the operating system and upper-layer business applications, middleware must be deeply adapted to domestic chips, domestic operating systems and domestic databases, forming a complete Xinchuang technology closed loop and ensuring that business applications remain functionally consistent and perform stably after migration.

High availability and business continuity

Railway transport operates uninterrupted 7×24 hours, and core business systems such as passenger dispatching, freight management and equipment monitoring tolerate no interruption. The middleware must provide high-availability clustering, load balancing and automatic failover, continuously ensuring stable system operation during and after the Xinchuang migration, avoiding business shutdown caused by foundational software replacement.

High-performance data processing and cache acceleration

Railway business systems involve large volumes of real-time data interaction, including high-concurrency access scenarios such as train operation status, ticket information and dispatching instructions, placing heavy pressure on databases under traditional architectures. The system needed a distributed caching mechanism to improve response speed through hot data caching and session sharing, reduce backend database load and ensure user experience in high-concurrency scenarios.

Products and Solutions

Overall approach

To address these requirements, Kingdee Apusic provides a full-stack middleware solution centered on the AAS application server, the AMDC distributed cache and the ALB load balancer, with "smooth migration, stable replacement and performance assurance" as the core principles: AAS provides business applications with a domestic enterprise-grade runtime environment; AMDC provides high-performance cache acceleration for hot data; and ALB provides intelligent traffic scheduling and fault isolation for application clusters — reproducing and optimizing the original systems' operating capability on an independently controllable technology stack, ensuring no business interruption and no performance degradation during the Xinchuang migration.

AAS application server — the stable foundation of Xinchuang operation

AAS is an enterprise-grade application server supporting Jakarta EE 8/9 standards, featuring technological leadership, security and stability, high availability and subscription-based delivery, with deep adaptation to the domestic software and hardware ecosystem: deeply adapted to domestic chips such as Kunpeng and Phytium and domestic operating systems such as Kylin and UOS, and compatible with mainstream domestic databases such as DM and Kingbase. It provides comprehensive migration tools and technical support to help business systems transition smoothly from non-domestic middleware to AAS; supports cluster deployment and automatic failover to meet the railway's 7×24 uninterrupted service requirements; and has built-in national cryptography algorithms providing encrypted transmission and storage security for business data.

ALB load balancer — intelligent traffic scheduling and fault isolation

ALB is a high-performance, high-availability and highly scalable load balancer, deeply integrated with AAS and supporting active health checks and fault masking: it distributes user requests evenly across multiple nodes of the AAS application cluster to avoid single-node overload; monitors backend node status in real time, automatically detecting faulty nodes and removing them from the service pool; supports dynamic adjustment of load policies and backend node configuration at runtime without service interruption during scaling or maintenance; and itself supports cluster deployment, eliminating single points of failure at the load balancing layer.

AMDC distributed cache — a high-performance data acceleration engine

AMDC is a high-performance distributed cache compatible with the Redis protocol, supporting master-slave, sentinel and cluster high-availability deployment modes, with national cryptography encryption and automatic failover: it caches frequently accessed hot data such as train information, ticket data and user sessions in memory, significantly reducing database access pressure; cluster mode deployment provides node auto-discovery, failure detection and automatic failover, ensuring cache service availability when nodes fail; it provides unified distributed caching services for multiple business applications, supporting cross-system data sharing and high-speed access, and supports national cryptography encrypted transmission.

The trio working together to build the Xinchuang middleware foundation

AAS, ALB and AMDC are not deployed independently but form a complete enterprise-grade application support system: ALB handles inbound traffic scheduling, AAS carries business applications, and AMDC accelerates hot data. Deeply adapted and seamlessly integrated, the three together form the middleware cornerstone of the Beijing Group's business system domestic renovation, achieving end-to-end domestic replacement of application runtime, traffic scheduling and data acceleration on an independently controllable technology stack.

Customer Value

The project supported the Beijing Group in migrating the underlying foundational software of multiple core business application systems — including passenger dispatching, freight management and equipment monitoring — to a domestic environment, implementing SASAC's mandatory requirement to complete Xinchuang domestic renovation by 2027 and forming a Xinchuang technology closed loop connecting the operating system, database and middleware.

During migration, AAS's migration tools and cluster high-availability capabilities, together with ALB's traffic scheduling and fault isolation, ensured the smooth switchover of business applications and reduced the impact of foundational software replacement on railway transport operations. After migration, the trio works together on the independently controllable technology stack to maintain complete capabilities in application hosting, traffic distribution and data acceleration — with no performance degradation.