I have a situation where we have a long-period check_interval (1 hour) for an expensive (compute-wise) check. Once it goes Hard, checks are lightweight in nature until it is resolved.
Unfortunately, with the 1h check interval, the situation might be automatically or manually resolved in (for example) 5 minutes, but it will be another 55m until Icinga notices. (Unless someone hits Check now).
Ideally I would like to change the check_interval from 60m to 5m until the state goes to OK again. Then, go back to 60m.
Is there a way to do this?