Scaling a Fast‑Growing Supply Chain Platform: Architecture and Ops Insights
This article details how a rapidly expanding B2B fresh‑food company restructured its R&D organization, adopted a matrix management model, and built a comprehensive distributed infrastructure—including task scheduling, service discovery, messaging, logging, file storage, CDN, configuration, sharding, search, caching, and monitoring—to support nationwide warehouse operations and future growth.
Background: A company spun off its B2B fresh‑food business into an independent entity, rebuilding the R&D team and creating a new supply‑chain platform to support large‑scale expansion across roughly 70 warehouses and billions in revenue.
R&D Scale
The initial R&D team was planned at about 100 people within six months, growing to around 180 later. The organization uses a matrix structure with groups for finance, factories, warehousing, procurement, major B sales, minor B sales, data analysis, and architecture.
Product Director – oversees all product personnel, processes, usage, and experience.
Project Manager – coordinates project schedules and resource allocation.
Chief Architect – designs technical architecture, builds the tech platform, and ensures stability.
Data Manager – manages data and business reports, builds a data platform, and supports decision‑making.
Application Architecture
Business Architecture
Technical Architecture
Architecture Evolution
R&D Infrastructure
Distributed Infrastructure Components
Task Scheduling: xxl-job for background and asynchronous tasks.
Service Registry: eureka-server for registration, load balancing, failover, and health checks.
Message Queue: rocketmq for asynchronous decoupling and cross‑service communication.
Log Center: ELK stack for unified log management and analysis.
File Service: Qiniu for file upload/download, lifecycle management, and dynamic image resizing.
CDN: Tencent CDN for accelerating static assets.
Config Center: Apollo for centralized configuration and gray releases.
Sharding Middleware: sharding-jdbc to abstract database sharding and master‑slave separation.
Search Service: Elasticsearch for full‑text and similarity queries.
Cache: Redis for distributed caching to speed up data access.
Distributed Monitoring System
Call‑Chain Monitoring: Pinpoint (future migration to SkyWalking) for tracing service interactions and performance metrics.
Log Monitoring: ELK for log aggregation and analysis.
Server Monitoring: Zabbix for Linux host performance.
Database Monitoring: PMM for MySQL performance, slow‑query analysis.
Application Performance Monitoring: CAT for request tracing, latency analysis, and custom reports.
BSF Distributed Framework
Goal: Enable developers to focus on business logic by abstracting middleware, providing high‑performance, easy‑to‑use foundational services and tools, achieving up to ten‑fold efficiency gains.
Open‑source repository: https://gitee.com/chejiangyi/free-bsf-all
The base framework (bsf) and business framework (business) are separated, allowing common libraries to be isolated. On top of them, a standard project scaffold csx-b2b-demo accelerates service development.
Future Directions
Refactor business services into a split‑architecture.
Implement big‑data solutions and read/write separation.
Integrate AI applications into business processes.
Support JDK 17 and 21 in BSF.
Develop form and workflow engines.
Signed-in readers can open the original source through BestHub's protected redirect.
This article has been distilled and summarized from source material, then republished for learning and reference. If you believe it infringes your rights, please contactand we will review it promptly.
21CTO
21CTO (21CTO.com) offers developers community, training, and services, making it your go‑to learning and service platform.
How this landed with the community
Was this worth your time?
0 Comments
Thoughtful readers leave field notes, pushback, and hard-won operational detail here.