Tianjin Rural Commercial Bank Data Warehouse Replacement and Upgrade Project Using Independently Developed Technology

1 Project Overview

• Project Background

Tianjin Rural Commercial Bank’s original data warehouse system served as a critical foundational platform. It was deployed on a Netezza appliance. Given that data is a key asset for banks and data analytics is an effective way to extract value from it, strengthening independent control over data assets and related equipment is a priority in the drive to localize financial IT infrastructure. Therefore, the bank decided to replace the Netezza appliance and upgrade the existing data warehouse system.

The original data warehouse primarily stored data from various business systems, including the core banking system, personal loan system, online banking system, and retail system. Data volumes from each system were growing continuously, and the data was complex—comprising over 2,000 small, complex tables and multiple large fact tables. Choosing a replacement data warehouse system that could match the original Netezza system was the main challenge. Additionally, after the Netezza replacement, the system needed to support incremental data growth over the next five years without significant architectural changes. Consequently, the bank decided to procure a domestic database that meets IT application innovation requirements to replace the existing Netezza system and build a next-generation data warehouse.

• Business Requirements

To establish a high-performance, flexibly scalable domestic data warehouse system that resolves the Netezza replacement challenge and accommodates future data growth, the upgrade project had to meet the following requirements:

• Chinese-developed Platform: The architecture supports a localized hardware and software stack, including operating systems and servers.

• Seamless Migration: Replace the legacy Netezza appliance, migrate historical workloads, support batch data import, and provide custom function capabilities to ensure a seamless transition.

• Business Scalability: Provide reliable storage, management, and utilization for incremental data over the next five years, with the platform able to elastically scale as data grows.

• Performance: Improve ad-hoc query performance for business data.

• Implementation Requirements

• Since data is a vital bank asset and analytics creates value, achieving independent control over data assets and related equipment was a primary objective of this domestic replacement project.

• Support data-at-rest encryption, database password encryption, encrypted data compression, and relevant encryption functions.

• Support deployment on domestic servers and operating systems.

2 Solution

The bank’s new data warehouse system was built on a GBase 8a MPP Cluster of over ten nodes, handling all structured data storage and computation, including data processing, transformation, and analysis. A data extraction tool migrated data from the original Netezza database to GBase 8a MPP Cluster, achieving seamless data migration. This not only preserved the original Netezza functionality but also enhanced storage capacity and computational performance. Working closely with the application vendor, the bank successfully completed the business migration, deployment, and adaptation tasks.

System Architecture Diagram of Tianjin Rural Commercial Bank Data Warehouse

3 Results and Value

By adopting GBase 8a MPP Cluster for its data warehouse, Tianjin Rural Commercial Bank achieved its expected goals and realized both technical and business value:

Full Localization: Built on an entirely Chinese-developed platform, leveraging GBase 8a MPP’s advanced architecture and high-performance data processing, the solution completely replaced the traditional data warehouse appliance.

Massive Storage: Compared to the Netezza appliance, GBase 8a MPP offers high-compression-ratio storage for massive data volumes, significantly reducing storage footprint.

Superior Performance: GBase 8a MPP loads TB-scale data at up to 20 TB/hour, delivering outstanding performance in high-speed data loading, export, and processing. The cluster’s batch processing performance surpassed that of the traditional Netezza appliance.

High Availability: GBase 8a MPP’s multi-level high availability technology ensures the availability of the bank’s data warehouse, reporting, auditing, and other systems, enabling 24/7 stable operation with no single point of failure.

Security and Control: The core data warehouse is built on a domestic distributed database with fully independent intellectual property rights. Its self-developed, secure, and controllable nature significantly enhances the data security and business security of the bank’s accumulated data assets.

Quality Service Assurance: GBase’s comprehensive, integrated product services provide end-to-end protection for the user.