Abstract：Failure detection is a key technology to realize high availability. It usually uses overtime mechanism to determine whether a process fails or not. But old failure detectors without adaptive mechanism can not meet the requirements of Quality of Service (QoS) of applications all the time because of the instability of the network. Adaptive failure detection requires that failure detectors can dynamically adjust the detection quality according to the variations of applications and the load of the network. A new failure detection model based on the predicted message delay was proposed on the basis of researching and analyzing the existing adaptive failure detection algorithms. An adaptive failure detection algorithm named PA-AFD based on the prediction from historical message delay was realized. The experimental results and analysis show that the algorithm can relieve the affect of message delay on the failure detection while ensuring the accuracy and integrity of detection.
2010, Vol. 30 
