Qunar Tech Salon is a learning and exchange platform for Qunar engineers and industry peers. We share cutting-edge technology trends and topics, providing a free platform for mid-to-senior technical professionals to exchange and learn.
This article presents a comprehensive overview of Qunar Travel's microservice architecture, covering background motivations, best‑practice service discovery and communication patterns, development efficiency improvements, governance techniques, and ServiceMesh experiments, all illustrated with real‑world data and diagrams.
This article defines software performance as executor + (process + IO + computation), explains when performance should be considered, and presents a comprehensive framework of executor, I/O‑type, and generic optimizations—including concurrency, sharding, caching, compression, async, batching, and algorithmic improvements—supported by cross‑domain case studies.
This article describes how Qunar Travel has evolved its testing environment governance from a fixed 10‑machine setup to a template‑driven, soft‑routing architecture (Environment 3.0), improving delivery speed, reliability, business connectivity, and reducing operational costs through automated sync, smart recommendations, and continuous business checks.
Easy Dump is a Bash‑based utility that streamlines PostgreSQL backups by wrapping pg_dump and pg_dumpall, offering multiple dump modes, parallel execution, table exclusion, automatic configuration generation, and detailed success or failure notifications for reliable database migration and maintenance.
This article explains the background, benefits, and measurement criteria of Kubernetes Horizontal‑Pod‑Autoscaler (HPA), describes the creation of metric tables and SQL queries for collecting scaling events and CPU usage, and presents a Python‑based workflow that aggregates the data, stores daily reports, validates results, and sends automated email summaries.
This article details Qunar Travel's mature chaos engineering platform built on chaosblade, covering value analysis, system architecture, shutdown and dependency drills, automated closed‑loop testing, attack‑defense exercises, and the measurable reliability improvements achieved across thousands of services.
This article describes how a multi‑layered cache system—including a DAO‑level Redis cache, shared cache components, and global cache planning—was designed, implemented, and deployed to reduce database QPS by over 80%, cut response latency, and improve overall system stability in a high‑traffic e‑commerce platform.
This article introduces and compares three MySQL online schema‑change solutions—pt‑online‑schema‑change, gh‑ost, and native MySQL ONLINE DDL—detailing their workflows, limitations, risks, monitoring features, and practical considerations for safely altering large tables in production environments.
This article shares the author’s experience and practical guidance on writing technical articles, covering why it matters, how to overcome initial hesitation, methods for continuous improvement, and detailed tips on content completeness, logical flow, structure, visual design, and daily habits.
This article describes the architecture, challenges, and performance optimizations of Qunar.com's Watcher monitoring platform, covering massive metric collection, master‑worker redesign, Graphite/Whisper storage enhancements, and future migration to Go‑based cloud‑native solutions.
