Abstract: Some problems in urban rail transit (hereinafter referred to as ′urban rail′) information system conventional technical architecture are solved by virtualization technology with its advantages such as high availability, flexibility and scalability. However, issues such as architecture bloating, information silos, low resource utilization rate and high operation and maintenance costs are not addressed. Additionally, the conventional virtualization technology can only provide basic resource for urban rail operation which doesn′t fully utilize the powerful service capabilities of cloud computing. With the continuous development of cloud computing technology, cloud native technologies such as containers and microservices are widely used to build agile architectures. By analyzing the characteristics of urban rail information system, the application history of cloud computing technology in urban rail information system, and the technical advantages of containers and microservices in cloud native architectures, it is pointed out that containers cloud technology has higher service availability and resource elasticity than conventional virtualization technology while supporting the storage and analysis of big data, realizing elastic decoupling and efficient resource utilization for conventional virtualization technology application in combination with microservices technology. A strong backend support is provided for building a ′safe, efficient, convenient and economical′ smart urban rail transit.