Research And Implementation Of Arbitration Technology Of A Transaction Oriented Fault-Tolerant Computer

The reliability of the computer system is very important in the industry of transaction processing oriented such as bank. The immeasurable losses may be caused by the system failures. Therefore such industries generally adopted fault-tolerant computer. The arbitration mechanism of traditional fault-tolerant computers is implemented through heartbeat detection. This single detection mechanism cannot effectively position the faulty machine and cannot satisfy the requirement of high reliability. Therefore, the research on how to enhance the system reliability of arbitration mechanisms has become an important research topic.In this paper an arbitration system and arbitration algorithm based on arbitration processor are proposed. Arbitration processor uses TMR (Three Module Redundant) system and the chip level fault-tolerant design technology to improve reliability. Arbitration algorithm uses hierarchical mode, which can effectively reduce the arbitration overhead. When a fault occurs, the arbitration system can locate the faulty unit quickly. So it can provide uninterrupted service. Thus, it solves the problems of lower coverage rate and improves the reliability of systems.Traditional fault tolerant computer uses hierarchical detection method to detect fault, it is an important method to greatly reduce detection costs and effectively improve scalability of system. It however suffers from the single point failure, and the Leader node (the management node of system) thus becomes reliability bottleneck of the system. Therefore, this paper puts forward an arbitration method with double Leader nodes, which can effectively solve the problem the single point failure. Using the detection and heartbeat detection, the fault coverage rate and the successful failure detection rate are improved.Lastly, this paper analyzes the reliability of the arbitration system by establishing a model, and further analyzes the reliability of the fault-tolerant computer system. Through the analysis the fault-tolerant computer system has a very high reliability.