Abstract: The China railway ticketing and reservation system (referred to as the ticket system) has high requirements for the database in terms of function, compatibility, security, and performance. Focused on the problem of difficulty in fully reproducing the production environment in business simulation scenarios during database selection, this paper conducted research on database selection for ticket systems based on traffic mirroring, analyzed the development stages and selection and adaptation requirements of the ticket system database. Based on this, the paper designed a traffic mirroring network architecture and a ticket system database adaptation and selection testing process based on traffic mirroring, used traffic mirroring, traffic analysis, feature extraction and other technologies to make simulation testing as close as possible to the real production environment, and rigorously analyzed the operation test results based on evaluation indicators, comprehensively compared the actual performance of the database in different aspects. It can provide reasonable and valuable reference suggestions for the database selection work of the ticket system in different business scenarios.