Future of Databases & Big Data: Insights from the First Techo TVP Summit

The first Techo TVP Developer Summit, organized by Tencent Cloud TVP on April 24, 2021 in Shenzhen, brought together more than 500 developers to discuss the evolving landscape of databases and big data. The event featured two main tracks—Database and Big Data & Data Analytics—and included technical talks, case studies, and a panel on career development.

Morning: Database Track

Opening remarks were delivered by Lin Xiaobin, General Manager of Tencent Cloud Database, followed by presentations from industry experts including Gai Guoqiang (CEO of Yunhe Enmo and the first Oracle ACE in China), Li Haixiang (Tencent Cloud Database Expert Engineer), and Hu Panpan (WeBank Database Platform Manager). Key topics covered cloud‑native databases, HTAP capabilities, and the performance of Tencent's TDSQL during the national census.

Gai Guoqiang’s talk, “Database Technology Evolution and Future,” traced the historical development of database systems, highlighted the convergence of commercial and open‑source solutions, and emphasized cloud as the primary arena for future database innovation.

Li Haixiang presented “Evolution of Distributed Databases,” reviewing the history of distributed database architectures, consistency models, and the HTAP‑enabled strong consistency techniques used in Tencent’s TDSQL 3TS open‑source system.

Hu Panpan shared a real‑world case study of TDSQL in WeBank’s core systems, describing a multi‑region, multi‑center deployment that now handles over 600 million financial transactions per day and peaks of more than 100 k TPS, while outlining future directions such as hardware localization, cloud‑native containerization, and AI‑ops.

Zhang Qinglin, Technical Director of Tencent Cloud Database, detailed the architecture of TDSQL‑C, a storage‑compute‑separated product offering massive storage, intelligent scaling, linear performance, and compatibility with MySQL and PostgreSQL. He highlighted breakthroughs in serverless scenarios, I/O‑bound performance optimization, and seamless backup, and outlined two future focus areas: ultra‑simple cloud‑native database operations and low‑code database development.

Afternoon: Big Data & Data Analytics Track

Opening remarks were given by Nie Jing, General Manager of Tencent Cloud Big Data Products, who discussed the rapid evolution of data analysis, the rise of real‑time data warehouses, and the importance of cloud as an enabler for data platforms.

Guo Wei, CTO of Yoyi and Tencent Cloud TVP, delivered “ClickHouse Latest Technologies and Applications,” explaining why ClickHouse achieves superior performance through vectorized computation, columnar storage, and a strong community, and illustrated its use cases with Tencent Music, Sina, Ximalaya, and Bilibili.

Chen Long, Lead of Tencent Cloud Elastic MapReduce, defined cloud‑native computing, identified four pillars (industrialized delivery, cost‑quantified storage/computing, adaptive load, data‑centric design), and described how Tencent’s data lake solution restructures Hadoop clusters into master, router, core, and task layers to improve performance and cost efficiency.

Du Li, Senior Big Data Engineer, addressed challenges of Flink SQL, presenting extensions such as Windowing Table‑valued Functions, new window types, and optimizations that reduced downstream data volume by up to 30× and improved throughput by nearly 20% for million‑row workloads.

Chang Lei, CEO of Odd Number Technology, explored next‑generation cloud‑native data warehouses, comparing Snowflake and OushuDB architectures, and introduced the “lake‑warehouse integration” concept that unifies data processing, governance, and AI/ML within a single platform.

Yu Huali, Senior Big Data Engineer, highlighted the advantages of cloud‑native data lakes—lower cost, higher performance, serverless operation, and unified metadata—and noted Tencent’s two ready‑to‑use data lake products that simplify building and analyzing data lakes.

Panel Discussion: From Industry to Career

The closing round‑table, “From Industry to Career, Data Now and Future,” featured Li Haixiang, Chen Long, Guo Wei, and Chang Lei. Speakers offered career advice, emphasizing deep specialization in kernel development, data architecture, or data science, and encouraging developers to become “dragon‑style” data engineers who blend engineering, product, and operational skills.

Moderator Wen Ming summarized that active participation in events like Techo TVP, continuous technical research, and open‑source contributions are essential for personal growth and advancing the data ecosystem.

Conclusion

The summit received positive feedback, was live‑streamed across multiple channels, and reinforced Tencent Cloud’s commitment to technology democratization, open‑source collaboration, and the mission of “using technology to impact the world.”