Award-Winning DevOps Product “Developer Lab” and Tencent Cloud Distributed Database (DCDB) – Technical Overview

On November 17, the Global Operations Conference (GOPS) and the 2nd China Operations Industry Annual Summit, organized by the Efficient Operations Community, was held. The event, the first dedicated operations industry conference in China, targeted operations engineers from both internet and traditional industries to disseminate advanced technical ideas and share best practices.

During the conference, Tencent Cloud’s “Developer Lab” and the “DCDB” product received the 2017 “Most Innovative DevOps Product” award and the “Most Influential Database Product” award, respectively, presented by the China Academy of Information and Communications Technology.

Developer Lab – Most Innovative DevOps Product

The Developer Lab integrates protocols and technologies such as SSH, WebSocket, RDP, and Canvas to create an immersive experimental environment with strong openness. It also incorporates a scheduling algorithm for resource optimization.

Technical Highlights:

1. Online IDE close to user habits By integrating low‑level protocols (SSH, Telnet, RDP) into the browser, the platform provides a remote terminal, code editor, and remote desktop (with clipboard sharing). Users can access Linux or Windows directory trees, perform visual file operations, and upload/download files directly from a web page, eliminating the need for external tools like XShell.

2. Tight coupling of tutorial and user terminal The web layer merges the online IDE with tutorial content, allowing users to follow step‑by‑step instructions on the same page. A step checker validates progress using configurable commands and regex rules, enhancing guidance.

3. Lightweight markup language for tutorial generation Developer Lab defines a script language called LDS for flexible tutorial creation. Tutorials are authored in Markdown, compiled to LDS, and can be edited, warned, generated, and debugged in real time. The extended Markdown parser enables interaction with other components such as the directory tree and editor.

4. Flexible resource scheduling A backend scheduling algorithm automatically allocates, recycles, and pools laboratory resources, reducing server costs and allowing more users to experience the lab simultaneously.

DCDB – Most Influential Database Product

DCDB adopts a cluster architecture. Its core modules include the Decision Scheduling Cluster (Tschedule), Database Node Group (SET), Access Gateway Cluster (TProxy), and Configuration Scheduling Cluster (TzooKeeper).

Logical Instance Represents a complete database instance from a business perspective.

Physical Sharding A master‑slave group of database nodes (SET) that horizontally partitions data.

Database Node Group (SET) Combines a MySQL‑compatible engine, monitoring, and data collection (Tagent). Typically deployed as a master‑slave architecture across racks, each DataNode runs heartbeat monitoring and Tagent for cluster robustness.

Scheduling Job Cluster (TScheduler) Automates tasks such as backups, monitoring collection, report generation, and business workflow execution. It integrates Schedule, Zookeeper, and OSS to activate resource plans within time windows, similar to Oracle’s DBMS_SCHEDULER.

Configuration Scheduling Cluster (TzooKeeper) Provides configuration maintenance, leader election, routing synchronization, and supports creation, deletion, and replacement of sharded node groups. It coordinates DDL operations across at least three groups deployed across data centers.

Operations Support System (OSS) A customized platform for business operation and management, integrating network, system, and monitoring services.

Access Gateway Cluster (TProxy) Handles SQL parsing, routing, and load balancing at the network layer (distinct from Tencent Cloud’s TGW). It pulls node status from TzooKeeper, provides sharding routing, monitors SQL execution, performs security authentication, and offers a virtual IP service via TGW. The design simplifies inter‑node communication and hardware requirements, enabling stable deployment even on simple x86 servers.