Highlights of the Cloud+ Community Developer Conference – Suzhou: IoT, Mini‑Program Performance, Game Networking, and Distributed Service Monitoring

On November 24, the Cloud+ Community Developer Conference – Suzhou station "Beyond the City Walls: IoT, Mini‑Programs, Micro‑services" concluded successfully at Tongcheng Building in Suzhou. The event featured technical experts from Tencent Cloud, Tongcheng Elong, Tetrate and others, discussing hot topics such as the Internet of Things, mini‑programs, and micro‑services.

The conference was hosted by Wang Xiaobo, TVP of Tencent Cloud and CTO of the Ticket Business Group at Tongcheng Elong.

Tencent Cloud IoT Platform Practice

Speaker: Xiao Ying, R&D Director of Tencent Cloud IoT Product Center.

Xiao Ying highlighted that among B2B, IoT, and blockchain, IoT is the biggest transformation point. He described the IoT concept as the interconnection of all things, where any product or device can communicate with each other. He introduced Tencent’s IoT product portfolio, including TencentOS tiny, edge computing services, IoT security service TID, and IoT SIM cards.

IoT architecture is divided into four layers: cloud, management, edge, and device. Emphasis was placed on end‑to‑end security, authentication methods, and data privacy, as well as the importance of disaster recovery and using common components for better team collaboration.

Exploration and Practice of Mini‑Program Performance Monitoring at Tongcheng Elong

Speaker: Niu Tifa, Senior Architect at Tongcheng Elong.

Niu Tifa started with the mini‑program runtime architecture, explaining the logical layer (JScore) and the rendering layer (WebView). He detailed the mini‑program lifecycle and stressed that measuring code execution time is the core of performance monitoring.

The presentation covered three core areas: startup acceleration, rendering mechanism, and the steData mechanism. Four key performance indicators were introduced: basic data, stage‑wise page timing, API request latency, and script exceptions. The monitoring system collects these metrics via an SDK embedded in the mini‑program, stores aggregated data in Druid, uses Elasticsearch for detailed queries, visualizes metrics on dashboards, and triggers real‑time alerts.

How to Quickly Build a Multiplayer Game

Speaker: Lin Jiewen, Tencent Cloud Expert Engineer.

Lin Jiewen broke down the functional requirements of a multiplayer game: gameplay, data synchronization, room management, matchmaking, chat, etc., noting the high development cost and complexity.

The Tencent Multiplayer Engine enables developers to implement data sync, matchmaking, and room management by invoking a few SDK APIs, while cloud development simplifies storage, lobby creation, and even some features via configuration only. The engine’s architecture includes a high‑defense layer, modular routing to service clusters, micro‑service deployment with isolated ingress layers, a lobby service built on cloud functions, storage, databases, and AI, and a custom service logic cluster managed with Kubernetes. Developers can also connect their own services by binding a VPC in the console.

How to Monitor Large‑Scale Distributed Services

Speaker: Wu Sheng, Founding Engineer at Tetrate.io.

Wu Sheng compared metrics, tracing, and logging, noting that tracing provides rich correlation but incurs high cost. He introduced Apache SkyWalking as an APM solution focusing on monitoring indicators, topology maps, alerts, and multi‑dimensional monitoring.

The talk covered SkyWalking’s community, features, automatic probes, and core concepts, followed by a deep dive into performance monitoring indices such as response time heatmaps, percentiles, success rates, traffic, and average response time. Percentiles reveal latency distribution, while heatmaps highlight request concentration.

— Follow Cloud+ Community for more exciting events —