Ctrip Hotel DevOps Testing Practices and the Moss Platform
This article examines how Ctrip Hotel implements DevOps testing—from business and technical drivers, through continuous delivery challenges, to the design of the Moss platform, its toolchain, data center, and future automation goals—highlighting practical solutions for rapid, high‑quality software delivery.
Author Wang Xingfu, senior test manager at Ctrip Hotel R&D, introduces the growing adoption of DevOps in large internet companies and raises questions about the role of QA within the DevOps paradigm.
Business and technology transformations—driven by cloud, mobile, AI, and blockchain—have forced a shift from long release cycles to rapid, continuous delivery, making flexibility and speed essential for competitive advantage.
Ctrip has long pursued continuous delivery, containerizing deployments and building an automated CI/CD system (Ctrip CD). While the process is highly automated, issues remain such as long feedback loops, excessive manual steps, and a lack of integrated monitoring across development, testing, and production.
The DevOps testing workflow at Ctrip, built on GitLab CI/CD, is divided into three stages: (1) static code analysis, security scanning, and unit testing at commit time; (2) interface, fault‑injection, comparison, and performance testing after deployment to a test environment; (3) traffic‑replay and verification testing in a production‑like environment, followed by production release and monitoring.
To support these practices, Ctrip created the Moss platform, a self‑built visual DevOps testing platform that consolidates configuration, execution, and reporting for developers, testers, and product owners, reducing learning curves and promoting cross‑team visibility.
The platform’s toolchain integrates open‑source frameworks (SonarQube, JUnit, Selenium, Appium, etc.) with internally developed extensions (Cobra, Buffalo, UTP, CAS, MOCK, ATL) to provide static scanning, unit testing, API testing, UI automation, data mocking, and performance testing.
Monitoring is extended beyond traditional Ops metrics to include business‑level indicators, with a dedicated DevOps data center that aggregates metrics from all tools into MongoDB and Neo4j, offering visual dashboards and searchable data for continuous improvement.
In the conclusion, the author reflects on how DevOps reshapes testing foundations, emphasizing high automation, rapid feedback, standardized processes, and metric‑driven optimization, while outlining future work such as enhanced data‑driven test strategies, quality visualization, integrated monitoring, and a Chrome extension for the Moss platform.
Ctrip Technology
Official Ctrip Technology account, sharing and discussing growth.
How this landed with the community
Was this worth your time?
0 Comments
Thoughtful readers leave field notes, pushback, and hard-won operational detail here.