<div dir="ltr">Hi, Have anybody successfully used the new feature "Predictive Dependency Checks". From my understanding, when nagios detects a Host/Service down/critical status, it does an actual check of parents Host/Service instead of using a "cached" status in previous version and decided whether or not a notification should be sent out. This is the most feature and reason I am testing right now to upgrade from 2.x -> 3.0, however I found it is not working for me. I've read the documentation carefully and got things configured as below: --- In nagios.cfg ------ cfg_file=/etc/nagios/objects/host_dependency.cfg cached_host_check_horizon=0 cached_service_check_horizon=0 enable_predictive_host_dependency_checks=1 enable_predictive_service_dependency_checks=1 -------------------------- --- In host_dependency.cfg ------------- define hostdependency{ host_name parent_host_A dependent_host_name child_host_B notification_failure_criteria d,u } ------------------------------------------ I expected when child_host_B is down, it checks parent_host_A's host state for real, if it's DOWN, nagios won't send notification for child_host_B down. But now, It seems to me, it still looks for cached status for parent_host_A (instead of doing an actual check). For instance, even the parent is down, as far as the cached state is UP, it will send notification of child_host_B. Any ideas how to get this work? Thanks. Mizuki </div>