Kuaicheng Logistics X TiDB: Leading the New Format of Logistics Technology with Digital Engines

With the continuous development of the logistics industry, various types of data involved in production, transportation, warehousing, and transportation flow in logistics companies have grown rapidly, showing the characteristics of large data volume, multiple data types, and difficulty in realizing value. As a leading enterprise in China's bulk logistics, Shanxi Kuaicheng Logistics Technology Co., Ltd. ("Kaicheng Logistics" for short) uses TiDB one-stack data service platform to achieve full-process refined operations, accelerate the realization of the value of massive data, and further drive industrial innovation .

Shanxi Kuaicheng Logistics Technology Co., Ltd. (hereinafter referred to as "Kaicheng Logistics") is a platform-based technology company with a new generation of information technology as the core of the "Internet + bulk logistics industry chain ecology". Its network freight ranks among the top three in the country. Kuaicheng Logistics insists on data-driven, and takes the core business of “Internet freight” as the starting point to build a “digital commodity logistics industry chain”. At present, the platform serves more than 580,000 vehicles, more than 600,000 registered drivers, and nearly 30,000 cargo owners. The number of national annual freight orders exceeds 7 million, and the annual freight value exceeds 10 billion yuan.

Real-time trading and massive analysis bottlenecks highlight

Kuaicheng Logistics has various business data types, including structured data of vehicle waybills, time series data of driving trajectories, driver behavior data, image data of order invoices, etc. During the peak period of business, the platform generates a large number of orders and grab orders. It has high-concurrency bearer capability, and a large number of settlement services consistency . Logistics needs to conduct real-time statistics and analysis of on the trajectory data of bulk cargo in order to make real-time decisions on route optimization and tariff adjustment.

Kuaicheng Logistics originally used MySQL cluster, which was restricted in terms of multi-dimensional query and related query, and performance problems became the bottleneck restricting business development. **Especially when the MySQL master database performs a large-scale update operation, the synchronization delay problem of the master-slave cluster is more prominent. Considering the continuous growth of business data volume, Kuaicheng Logistics considers choosing a database system that can not only support transactions, but also can flexibly expand to meet the needs of massive data queries.

Build a new generation of real-time data platform

undergone comparative testing and application compatibility verification, TiDB database has excellent performance in scalability, query performance under massive data scale, transaction integrity, etc. , Kuaicheng Logistics decided to use TiDB distributed database to build a new generation of data service platform.

Kuaicheng Logistics deploys TiDB clusters on the public cloud, migrates core business to TiDB, supports multiple business applications such as order acceptance, order grabbing, orders, driving trajectory, contract and invoice, as well as integrated operation management and office management applications. In addition, TiDB is seamlessly connected with the data lake to provide data sources for various types of big data analysis. Based on the high availability of the TiDB cluster itself, Kuaicheng Logistics has established a multi-level disaster recovery system through MySQL and public clouds to ensure business continuity in all aspects.

Figure 1: The logical architecture of Kuaicheng Logistics' new-generation data platform

Kuaicheng Logistics combines advanced technologies such as big data, artificial intelligence, and the Internet of Things to connect drivers, vehicles and goods with numbers to create an innovative format of "data logistics operator". With the in-depth application of TiDB's new-generation data platform, the data-driven full-process refined operation of begun to show benefits. Overload control, further optimize the return plan, reduce the no-load rate of trucks, and provide a more reasonable dynamic tariff adjustment strategy in abnormal weather. TiDB is mainly manifested in the following aspects:

• Multi-scene support

A TiDB data platform supports multiple business scenarios, and the complete HTAP capability supports transactional transactions (OLTP) and real-time analysis (OLAP) of massive data at the same time. On the basis of meeting data consistency, TiDB supports high concurrent reading and writing, and provides minute-level statistical analysis, which is helpful for more flexible business decisions and changes. TiDB has a highly open data ecosystem with complete offline and real-time data synchronization tools. It can build a real-time or offline data warehouse system with big data ecosystems such as Flink, Spark, and BI.

• development efficiency of

The future-oriented cloud-native distributed architecture provides business unaware auto-scaling capabilities, which can independently expand computing or storage, without the need to implement distributed transactions through applications. TiDB is non-intrusive to application development and data model design, supports sensitive development and online business changes, and approximately doubles the development efficiency of the original database system.

• Operation and maintenance costs reduced by 50%

TiDB supports all development programming languages and ORM frameworks such as Java, Python, Golang that can be connected to MySQL, and provides online migration tools. TiDB's built-in graphical TiDB Dashboard and Prometheus monitoring system provide complete closed-loop monitoring capabilities and fault analysis capabilities, reducing operation and maintenance costs by about 50%.

Figure 2: TiDB provides stable support during the peak period of order grabbing

In the process of digital transformation, various businesses have more urgent requirements for "massive, real-time, online" data. Real-time recommendation, precise marketing, and real-time decision-making have become key capabilities in digital scenarios, and they can keenly identify, perceive and guide user needs. , Improving user experience will bring continuous competitive advantages for enterprises. distributed HTAP database is an inevitable product of this trend. uses one platform to simultaneously solve the problems of massive data transactions and real-time analysis, making the realization of data value more efficient and simpler.

💡 Welcome to click [Read the original text] to see more TiDB user cases.