The customer is a premier logistics service provider with a significant footprint across Southeast Asia and China. Founded in 2015, their network spans 13 countries, catering to millions of households and businesses annually. An efficient billing management system at the heart of their operation is pivotal for handling transactions and maintaining smooth logistics workflows.
The Challenge: Scalability Limitation Coupled with Complex Data Stack
Initially, the billing management system of the logistics service provider was deployed on PolarDB for MySQL, a cloud-based database service. However, as the company’s business flourished, its data infrastructure ran into bottlenecks under increased demand. The core issues were:
- Disrupted user experiences due to architectural limitations: PolarDB can only scale vertically, which requires substantial investment in higher-specification hardware and maintenance downtime when upgrading. This architectural bottleneck hindered horizontal scaling for increased data load and user requests, especially in high-concurrency scenarios.
- Complex data synchronization: To facilitate data computation and transition, the customer had to leverage various data synchronization tools– Kettle to sync PolarDB data to ElasticSearch, DTS to sync data to AnalyticDB for computing and reporting functionalities, and ElasticSearch for aggregated queries. This functional solution introduced further architectural complexity and extended data transition cycles, adding to the regular system optimization and maintenance.
The Solution: A Streamlined Data Pipeline Based on TiDB
After rigorous testing and application compatibility verification, the logistics service provider opted for TiDB, a distributed SQL database, to address their challenges and enhance the data architectural capabilities.
The critical implementation highlights include the following:
- A horizontally scalable data infrastructure: Unlike PolarDB, TiDB supports horizontal scalability, effectively resolving the system’s performance bottleneck.
- Consolidated data stack with streamlined data processing: TiDB’s Hybrid Transactional/Analytical Processing (HTAP) capability allowed the replacement of the complex data stacks (PolarDB, AnalyticDB, DTS, CANA) with a single-stack solution. They used TiCDC, the change data capture tool of TiDB, to sync data to ElasticSearch for detailed queries in the front end.
- Efficient hotspot data management: TiDB’s ability to scatter hotspot data and auto-balance loads enhanced the handling of data peaks, optimizing resource utilization and minimizing performance issues.
Benefits of Transitioning to TiDB
TiDB provided a robust, scalable solution that addressed the core challenges faced by the logistics service provider. Specifically, the TiDB implementation in their data infrastructure has brought the following benefits:
- Improved performance with lowered operational burden
The horizontal scalability of TiDB allowed for the dynamic addition of resources, eliminating the need for database sharding and reducing development and operational burdens. In addition, TiDB’s robust transaction consistency ensures developers can focus on business logic without worrying about intricate database operations. - Streamlined data processing workflows with cost-efficiency
The consolidation into TiDB simplified the overall system architecture, reducing the need for costly hardware upgrades, data warehouse maintenance, and associated downtime. These factors all brought substantial cost savings and improved overall operational efficiency. - Improved decision-making with real-time analytics
TiDB’s HTAP capabilities allow for streamlined data processing workflows and more efficient daily and monthly summarizations. - Mitigation of data sync risks
The consolidated solution eliminated the need for multiple data sync links, reducing data processing complexity and lowered risks of data inconsistency during synchronization.
The adoption of TiDB revolutionized our client’s billing management platform, overcoming major operational hurdles with its scalable and real-time data processing capabilities. This strategic move streamlined efficiency and reinforced system reliability, proving essential in the fast-evolving logistics sector.
