When the Monitor Agent fails over, it must repoll all the monitored agents to ascertain their current status. This might take a few moments, depending on the number of agents being monitored. However, no action is necessary on your part as the Monitor Agent starts on the next node in the cluster.