Interview with Tencent Cloud’s Zhang Zhigang on Lakehouse Architecture and Cloud‑Native Integration

On the eve of the 8th GIAC Global Internet Architecture Conference (July 22‑23, 2022, Shenzhen), High‑Availability Architecture interviewed Zhang Zhigang, a senior architect from Tencent Cloud, about the increasingly popular combination of lakehouse architecture and cloud‑native technologies.

Interviewee background: Zhang currently leads cloud‑based big‑data architecture optimization and scenario design at Tencent Cloud. Prior to joining Tencent, he worked on Hadoop platform design and optimization, contributing to data integration platforms, offline and real‑time analytics systems.

Lakehouse architecture overview: A lakehouse merges the data lake’s storage foundation with data‑warehouse‑style management, offering two construction approaches: adding warehouse modeling capabilities on top of a lake, or integrating lake storage into the warehouse’s source layer. Core technologies include unified metadata, engine acceleration, and storage optimization, with metadata unification being the most critical.

Industry pain points addressed: Traditional data management (databases, warehouses, lakes) faces limitations such as high cost, complex governance, and performance bottlenecks. Lakehouses inherit warehouse governance to solve lake‑related issues like data quality degradation and “data swamp” problems, while leveraging mature projects such as Hudi, Delta, and Iceberg to provide atomic transactions, consistency, and metadata performance.

Impact on the big‑data ecosystem: Lakehouses streamline data development, reduce reliance on specialized architects, and enable faster, more flexible business‑driven data solutions. They also foster smoother collaboration among roles—data scientists, analysts, and engineers—by providing a unified platform for data integration and management.

Cloud‑native synergy: Combining cloud‑native principles (containerization, micro‑services, dynamic management) with lakehouse designs improves robustness, portability across clouds, multi‑tenant isolation, and overall stability. Cloud providers’ object‑storage foundations and standardized protocols further enhance performance and operational efficiency.

Advice for big‑data practitioners: Embrace cloud‑native concepts, deepen understanding of container orchestration, DevOps, and continuous delivery, and broaden skill sets beyond a single platform to increase adaptability and value in the evolving data landscape.

Closing remarks: Zhang expressed enthusiasm for sharing practical lakehouse solutions at GIAC, wishing participants continued success in big‑data endeavors and a prosperous conference.

For more details about the GIAC conference agenda, including sessions on data‑intelligent platforms and lakehouse topics, click the original article link.