Stopped Metrics Make Alert Get Recovered

Hi everyone,
I created an alert to monitor the disk and alert me if it is below 3 GB (free space). The problem is the alert gets recovered when the computer I monitor is down. Then when the computer is turned on, I will have to deal with the same alerts which are kinda frustrating. The computers are scattered across different time zones and will be shut down at night. Any recommendation on how to make it not recover when the data stopped coming to elastic? Btw I use metric threshold for alerting.