Workaround to avoid false alerts in server monitoring
A server monitoring set up that has jobs running periodically throughout the day at unspecified times, and for an unspecified amount of time. As these jobs are running, they use up the server's CPU usage to 100%. This triggers false alerts, leading to wrong assumptions.
Hover on the hamburger icon beside the server monitor display name
Click on Edit and go to the Edit Server Monitor page
Under Configuration Profiles , click on the pencil icon beside Threshold and Availability
In the Edit Threshold Profile window, under Threshold Configuration , select CPU Utilization threshold
In the drop down under Strategy , select Average Time
Mention the average time (in minutes) as to when you should receive a Trouble or Critical alert
Save the changes
In case we know the exact time period as to when the CPU will be utilized to the maximum, we can set up a maintenance period or a "Downtime Notification Delay" in the Notification Profile.