Software Fault Tree Construction And Application Of Public Safety System

The increasing complexity of today’s information systems, as well as changing needs of a growing number of integration between various systems, making the system reliability as a constraint to the development of a complex system bottleneck. The way of how to timely monitoring system components status, display each component status with good user experience UI, diagnose and auto recovery components are important and efficient method to improve the reliability of complex systems.Software fault tree construction and application is one of the good ways to monitoring, diagnosing and recovering the complex information systems. Algorithm to be set between various nodes by their relationship, so that when a failure occurs at current node, it will iterate report its status to parent node till the top node. All system components status will be displayed by fault tree node health status. All fault tree node health status to be detected by calling diagnoses method. An email will be sent to administrators when an alert or error generated, and auto recovery method will be fired when current node health status is turned to error.This paper gives a brief introduction to software fault tree construction and application based on cluster servers, various windows event log collecting to diagnose each component status, and auto recovering when error occurs, eventually improves system reliability. It also discussed system both hardware and software liabilities improve methods, and evaluate the quantitative reliability.Finally, in order to validate the feasibility of the design solutions, it applied to Premier One project owned by Motorola Solutions Company. The result shows that it ultimately achieves 99.9999% reliability by diagnosing each component status, and auto recovering failure node. Since actual business was changing very time, this paper also takes scalability into consideration.