Key Features and Optimizations of Tencent Cloud‑Native Database TDSQL‑C
The article details Tencent Cloud's TDSQL‑C database, explaining traditional database limitations, its cloud‑native architecture, four standout capabilities such as petabyte storage and serverless billing, and the extensive kernel optimizations that enable rapid scaling, backup, indexing, and future cross‑region reads.
On November 4, 2021, Tencent Cloud’s data‑technology director Zhang Qinglin presented a session on the cloud‑native database TDSQL‑C, covering its architecture, storage, elastic scaling, snapshot backup, and serverless capabilities.
Traditional database architectures suffer from local storage limits, single‑master‑multiple‑replica bottlenecks, and difficulty scaling horizontally, prompting the design of a new cloud‑native solution that automatically expands storage, eliminates master‑slave latency, and supports seamless horizontal scaling.
Four major characteristics of TDSQL‑C are highlighted: (1) support for petabyte‑level storage via network‑attached storage, (2) second‑level elastic scaling through decoupled compute and storage, (3) second‑level snapshot backup and point‑in‑time recovery, and (4) a serverless, pay‑as‑you‑go model that charges only for actual compute and storage usage.
Key product optimizations include:
Ultra‑fast backup and restore using parallel snapshot extraction and per‑cell redo‑log application, achieving backup times one‑tenth of competitors.
Three‑stage index creation: parallel scanning, parallel merge‑sort, and batch B+‑tree construction to reduce I/O and sorting overhead.
Accelerated instance start‑stop by parallelizing buffer‑pool and transaction‑system initialization, cutting startup/shutdown time by an order of magnitude.
Extreme elasticity with on‑demand log‑driven storage expansion and rapid compute node scaling.
Instant column modification (DDL) that allows second‑level column changes without blocking reads/writes.
Secondary cache independent of the buffer pool, leveraging non‑volatile memory to mitigate network‑storage I/O bottlenecks.
These innovations collectively improve reliability, availability, and performance for demanding workloads such as finance and gaming.
The future roadmap for TDSQL‑C focuses on two areas: achieving ultra‑high write performance through a faster log store, and providing cross‑region read services with financial‑grade reliability and disaster‑recovery capabilities.
Tencent Database Technology
Tencent's Database R&D team supports internal services such as WeChat Pay, WeChat Red Packets, Tencent Advertising, and Tencent Music, and provides external support on Tencent Cloud for TencentDB products like CynosDB, CDB, and TDSQL. This public account aims to promote and share professional database knowledge, growing together with database enthusiasts.
How this landed with the community
Was this worth your time?
0 Comments
Thoughtful readers leave field notes, pushback, and hard-won operational detail here.