DTCC Interview | GBase Logical Data Warehouse on Cloud Empowers Enterprises to Achieve New Breakthroughs in Digital Transformation

During the 12th China Database Technology Conference (DTCC 2021), Zhang Shaoyong, General Manager of the Data Intelligence Product Division at GBase, was invited to give a thematic interview on "GBase Logical Data Warehouse on Cloud Empowers Industries to Meet New Challenges in Digital Transformation."

Host: Why did you choose this topic, and what can the audience take away from it?

Zhang Shaoyong: Logical data warehouses actually solve real-world problems such as massive data volumes, complex data types, and deep value mining. The core technologies are data virtualization and data federation, which unify the management of various business systems. With the development of cloud, the original functional architecture has been adapted to changes in hardware infrastructure, bringing the logical data warehouse to the cloud.

Host: Regarding cloud data warehouses, what can we learn from Snowflake?

Zhang Shaoyong: Snowflake's advantages: ease of use and flexible billing. On the "virtual data warehouse concept at the compute layer": GBase 8a MPP virtual clusters, where each sub-cluster is physically separate but logically unified, with fault isolation, elastic scalability, and unified cluster management. GBase foresaw this direction early, building the GBase 8a MPP on-premises virtual data warehouse, and migrated the entire local computing of the original virtual clusters to the cloud within half a year. We can learn from Snowflake's architectural computing approach and its business model.

Host: What is the current demand for cloud data warehouses domestically and internationally?

Zhang Shaoyong: In China, demand mainly comes from two key industries: finance and telecom. In the financial sector, which includes large state-owned banks and joint-stock banks, non-sensitive data is first placed in private clouds to ensure security, then gradually moved to the cloud. GBase is currently adapting and verifying with a large state-owned bank in real business scenarios to drive implementation. Small and medium-sized city commercial banks adopt industry public clouds with shared storage and rent computing units.

Host: Why are new concepts like distributed databases so hot?

Zhang Shaoyong: Business drives technology. Concepts emerge as needs evolve; a standalone machine's analysis capability was insufficient for growing data volumes, so proprietary solutions evolved toward open approaches, giving rise to new concepts. When relational databases are transformed into distributed architectures, they gain distributed characteristics such as consistency, high availability, and scalability while preserving the existing relational ecosystem, scale-out support, and transaction capabilities. This solves users' original application challenges without reducing compatibility, driving their development. Early transactional databases used sharding and table splitting, requiring middleware that necessitated application modifications, thus evolving to distributed transactional databases.

Host: Some say this is the best era for domestic databases. Do you agree?

Zhang Shaoyong: I agree. Policies in China are currently promoting the development of domestic databases. Users' attitudes have shifted from rejection to willingness to try, and from willingness to try to usability. The push for domestic databases has been ongoing for 10 years, and we are about to reach the stage where they become truly user-friendly.