is there a way to check n times a service before changing its state to WARNING/CRITICAL?
By example, I have a service that checks the bandwidth.
I would like to be notified only if its value has been above the critical limit for n checks.
Would that be flapping detection?
When the check command returns warning or critical return code, the state of service is changed to soft state.
Nagios doesn’t notify soft states. After period defined by parameter retry_interval the service is checked again.
This repeats max_checks_attempts times. Then the state of service is changed to hard and you got notification.