Lichu Saobei (Wuhan Lichu Business Service Co., Ltd.) serves 700,000 merchants across more than 400 cities in China, processing 12 million transactions daily with an annual volume of CNY 200 billion. Having successfully delivered aggregate payment solutions across hundreds of industries, Lichu Saobei faced new scaling challenges as its business grew. A standalone ApsaraDB RDS instance could no longer keep up — 2 TB of data accumulated in just three months, and a standalone MySQL database had become a bottleneck for analytics. Alibaba Cloud addressed these challenges with a multi-product cloud database architecture that now supports up to 100 TB of storage, 100,000 transactions per second (TPS), and millions of queries per second (QPS).
Challenges
Lichu Saobei's rapid growth surfaced three compounding problems:
-
Storage exhaustion. The standalone ApsaraDB RDS instance was nearly full after generating 2 TB of data over three months, with no straightforward path to scale further.
-
High concurrency under load. As the merchant base and daily transaction volumes grew, read and write performance degraded significantly.
-
Analytics at scale. The volume of data being queried grew exponentially, and a standalone MySQL database could not meet the requirements for large-scale analysis and reporting.
Solution
Alibaba Cloud designed a distributed architecture combining six services, each with a distinct role:
| Service | Role |
|---|---|
| PolarDB-X 1.0 | Handles high concurrency and horizontal partitioning. Fully compatible with MySQL syntax, enabling migration from a standalone instance to a distributed architecture. |
| PolarDB | Archives large data volumes. Its compute-storage separation architecture and full MySQL compatibility let the original business run without modification. |
| ApsaraDB for Redis | Caches hot data in memory, significantly improving read efficiency. |
| ApsaraDB RDS (read-only instances) | Provides automatic load balancing for read traffic. |
| AnalyticDB for MySQL | Runs real-time queries on large datasets, enabling instant operational insights and precision marketing. |
| Data Transmission Service (DTS) | Synchronizes data in real time across services. |
Benefits
Linear scalability. Combining ApsaraDB for Redis caching with PolarDB-X 1.0 horizontal and vertical partitioning, the system now supports 100 TB of storage, 100,000 TPS, and millions of QPS — enough capacity to handle up to 10 times Lichu Saobei's current business volume. Rapid business growth no longer risks exhausting database storage or degrading transaction performance.
Faster data-driven decisions. DTS keeps data synchronized in real time across the stack. With AnalyticDB for MySQL and PolarDB handling analytics workloads, report generation time dropped from minutes to seconds — most reports complete within 10 seconds, and even complex reports finish within a minute. Business behavior data is immediately available for decision-making.
24/7 high availability. The architecture provides continuous access to specification upgrade and downgrade, smooth scaling, SQL audit, and read-only instances. The security and stability meet the requirements of the finance industry.