In certain environments, it is acceptable for some metrics to exceed the configured alert thresholds for very short periods of time, for example one refresh. Typically these exceptions last less than a minute. If you are monitoring many SQL Server instances or if the amount of data collected is very high, this could produce a significant number of alerts. The Advanced Alert Configuration window allows you to limit the number of these alerts that are generated by letting you input the number of minutes a threshold violation occurs before an alert is raised, reducing the amount of alert "noise" you receive.
The amount of time you enter here depends on how critical the metric is in your environment and the amount of time SQLdm waits between refreshes. If you enter a time less than that of the refresh, the second consecutive refresh where the alert is outside the acceptable threshold range prompts the alert.
The following table displays the length of time SQLdm waits before an alert is raised when the scheduled refresh occurs every six minutes and various values are entered:
Time entered on the Advanced Alert Configuration window | Alert is generated |
|---|---|
1-5 minutes | Second Refresh occurs six minutes after the first encounter of the problem |
6-11 minutes | Third Refresh occurs 18 minutes after the first encounter of the problem |
12-17 minutes | Fourth Refresh occurs 24 minutes after the first encounter of the problem |