CFFEX Real-Time Trading Surveillance System Project

As the China Financial Futures Exchange (CFFEX) continues to grow its business—especially after listing options products and introducing a market maker system—trading volumes have surged, and order data is set to increase explosively in the long term. However, the existing Oracle database does not support node expansion, providing limited performance improvement and failing to meet the demands of ongoing business growth. This will degrade the performance of interactive real-time query and analysis applications within the exchange, affecting the efficiency of real-time surveillance operations.

The project deploys GBase 8a MPP Cluster as its data management platform, supporting various analytical applications for real-time analysis and detection of abnormal transactions. The GBase 8a cluster combines high-performance real-time data ingestion with real-time analytics capabilities. Transaction data is synchronized in real time to the GBase 8a cluster through Kafka in the data synchronization layer.

Multiple Cluster Deployments: Deployed on standard x86 servers, substantially reducing hardware and software procurement and maintenance costs;

Extreme Performance: Delivers 10 TB/hour data loading capacity, fully satisfying the online monitoring system's requirement for ingesting nearly 100 MB of data and tens of thousands of records per second;

Significantly Enhanced Analytics: Real-time analysis efficiency has improved exponentially over the legacy system, from minutes down to seconds or milliseconds;

High Scalability: The cluster leverages a shared-nothing architecture, supporting online, on-demand horizontal scaling without service disruption, and meeting business systems’ storage and computing demands.