We are monitoring a deployment in which some hosts are dedicated to batch jobs. With our current setting we are receiving problem notifications when a batch occurs because the CPU utilization reaches 100% for some minutes.
Will the automatic alerting recognize this pattern or do I need a custom setting?
Using alerting based on custom settings we can change the threshold on the CPU utilization for the alert but we can not modify the number of samples.
Ideally we would like to specify a rule like "Alert if CPU utilization is over 95% for 15 minutes (samples)"
Is this possible?