0 views collected around this technical thread.
This article details Qunar's high‑concurrency flight pricing architecture, covering the origin of pricing data, two‑level caching design, a cache‑management system, CQRS‑based policy storage, data compression techniques, and performance improvements achieved through these optimizations.
This article examines the rapid growth of Qunar's domestic flight pricing system, identifies response time bottlenecks caused by increased request volume and complex calculations, and details a series of architectural and algorithmic optimizations—including caching, search structure changes, election logic refinement, and fault‑tolerant mechanisms—that reduced average latency by up to 40%.
This article describes how Qunar’s senior architect designed a high‑availability, high‑performance, real‑time flight ticket pricing search system, discussing user requirements, challenges, and the backend architecture choices such as on‑demand computation, caching, MQ‑Redis integration, and layered system design.