Mobile QQ Operations: Business Overview, Network Fault Cases, Backend Architecture Optimization, Global Intelligent Scheduling, and Mobile Network Performance Enhancements

Guo Zhiwen, QQ mobile operations leader, shares the evolution of Mobile QQ from 2008 to 2017, covering rapid user growth, network fault cases in Chongqing and Hong Kong, and the challenges of 2G/3G/4G environments.

Case studies illustrate how packet loss and latency issues were diagnosed and resolved through cooperation with carriers, gateway troubleshooting, and real‑time monitoring.

The backend architecture progressed from a single data center to dual‑center (Shenzhen‑Tianjin) and finally a three‑center (Beijing‑Shanghai‑Shenzhen) deployment, eliminating cross‑network traffic, reducing packet loss to ~1%, and saving billions in settlement fees.

Global intelligent scheduling aggregates billions of connection metrics, automatically reroutes traffic within minutes, and fine‑tunes routing down to IP‑port level, achieving up to 5‑minute response to network degradation.

Mobile‑side optimizations include signaling channel pre‑activation, IP‑direct routing to bypass DNS hijacking, and server‑side logical aggregation that cuts message‑send latency by ~30% and login time from 8.6 s to 1.9 s.

Overseas acceleration points in North America, Europe, Southeast Asia and Oceania further improve latency for the small (<2%) overseas QQ user base.