Research on Failure Detector in Remote Disaster-tolerant System
He Jun, Tao Wang, Tang Lei, Wen Chuanhua · 2010
Failure detector is one of the key components for remote disaster-tolerant systems. In this paper, a flexible hierarchical architecture is designed and the detection component is divided from decision-making. In doing so, failure detector can be applied to remote disaster-tolerant system and can meet the requirements of different applications. An adaptable failure detection algorithm is put forward considering message delay and message loss and clock synchronization. This algorithm can adapt itself to the change of network condition and applications' requirements. The algorithm has been implemented in a remote disaster-tolerant system and the test results showed that the error rate is greatly decreased using the algorithm.