Noise is the normal variability that can occur while a resource is being monitored. For example, a CPU might run at 75% utilization with infrequent spikes of 100% utilization. The infrequent spikes are noise. You probably do not need to be alerted to those spikes because they do not reflect the true overall performance of the CPU. To avoid unnecessary alerts like these, you can set the Throw alert after and Clear alert after properties appropriately.

For example, you determine how many times a condition must occur before an alert is generated. Setting a low number, such as 1, means you are more susceptible to noise. Setting this value to a higher number, such as 3, reduces the likelihood of noise alerts, but also means an event must occur three times in a row before an alert is generated. The higher the setting of the Throw alert after property, the longer it takes to generate an alert. For example, with a Throw alert after property set to 3, three sample periods occur before the alert is generated.