Fault detector of fault-tolerant distributed systems based on self-adaptive heartbeat algorithm

The detection of failures is one of the fundamental issues for fault tolerant distributed systems. In order to improve the performance of the fault detection, a novel failure detector called self-adaptive heartbeat detector (SA-HD) was proposed. SA-HD employed a self-adaptive heartbeat detecting algorithm based on pull mode, in which the performance of the failure detection and the network resource caused by heartbeat detecting were both considered. The algorithm could change the frequency of transmitting the heartbeat messages and enhance the adaptability for the network environment, especially in a heavy network payload. The model of SA-HD was built and simulations were carried out to compare the performance between SA-HD and the approach based on push mode. The simulation analysis and experiment results show that the performance of SA-HD is better than the latter's.