Besides running a check every check_interval, Naemon may run on demand host checks of a host. This may happen in the following situations:
- When a service associated with the host changes state.
- As needed as part of the host reachability logic.
- As needed for predictive host dependency checks.
Most often, this is seen when a service changes state. This is needed to ensure that the host goes into hard critical before its associated services, when there is a full server outage. The reason for this is that if the host is in a hard critical, before any of its services, we only send one host notification. Without the on-demand checks, a situation may occur where multiple services get to a hard critical state before the host, thus generating a notification storm, one notification for each service.