The successful use of cloud computing in different areas, such as analytics and workflow management, makes the railway operators to move their systems to the cloud. The main motivations for this transition are the increase of reliability, efficiency and safety of the transport service. The cloud-based infrastructure is more flexible for deployment and reduces management efforts through services and automation. The focus in this paper is on the cloud management as an essential functionality for the resilience of the cloud platforms for critical railway systems. Standing on analysis of different fault management use cases, the basic cloud resource fault management functions are identified and an approach to design them as RESTful services is proposed. The physical and logical entities involved in alarm management are represented as uniquely identified resources which can be manipulated by HTTP methods. The approach feasibility is provided by modelling the alarm status from the fault management application and service points of view.