Independently Developed MPP Database Construction Project

Domestically-Built MPP Database Project

Project Background

The project aims to build a data platform aligned with the bank’s future business development by introducing fully domestically-built MPP database software, graphical reporting tools, data governance subsystems, and other applications, drawing on financial industry localization practices and the current situation of a commercial bank in Hainan.

Solution

The scope of this project includes big data foundation platform software, MPP database software, data application development and implementation, data reporting tools or BI software, and more. It adopts a lakehouse architecture that combines a Hadoop cluster with a GBase 8a MPP database cluster, building a low-cost, high-performance, large-capacity, and easily scalable data foundation platform. This enables centralized collection, storage, processing, analysis, and application of structured, semi-structured, and unstructured data. Based on this technical architecture, the data warehouse is reconstructed and historical data storage and usage are standardized, forming a foundational data system that provides various data applications to support business management.

 

Technology Architecture Diagram of the Data Platform for a Commercial Bank in Hainan

The project consists of four layers: data source layer, data middle platform (data exchange layer, data lake, data warehouse, data access layer, and data service platform), business middle platform, and data application layer. The overall solution uses the GBase 8a MPP analytical database to build the big data platform. The GBase 8a MPP fast loading tool enables rapid data ingestion, synchronizing data from core banking and credit business systems to the data center in near real time, effectively supporting various historical data analysis applications. Leveraging GBase 8a MPP’s columnar storage, intelligent indexing, high compression, and other technologies, disk I/O is significantly reduced, resulting in substantial improvements in query and statistical analysis compared to the legacy system. It supports ad-hoc queries and multidimensional analysis, and powers complex BI analytics and reporting.

Overall Architecture Diagram of the Data Platform for a Commercial Bank in Hainan

 

Results

Achieve unified management of data resources, comprehensively enhance data service capabilities, fully unlock data value, meet the needs of customer marketing, risk control, operational management, and external regulatory compliance, drive data governance, significantly improve the bank's data resource management level and comprehensive data asset application capabilities, and continuously foster business innovation, marketing innovation, service innovation, and management innovation.