Put simply, the limit value defines how many consecutive alerts are required to raise an actual alert. For example, if the limit is 3, the first one or two alerts won't cause any action; i.e. they won't be shown as alerts in the monitoring page and no emails will be sent. Only if there is a third consecutive value that is also above the threshold (or below in case of reversed performance counters) will an alert be raised. Thus, the first alerts before the limit are not really alerts but simply values that exceed the threshold. If the recorded values stay above (or below) the threshold after the limit is reached, then the alert will remain active.
Consider the following example: the performance counter is CPU usage %, the threshold is 80 and the limit is 3. The recorded performance counter values are shown in the top row and the resulting action in the bottom row:
| 76 % | 82 % | 85 % | 86 % | 81 % | 79 % |
| no alert, below threshold | no alert, threshold exceeded | no alert, threshold exceeded | alert raised | alert raised | no alert, below threshold |
That is, with a limit of three, the four consecutive threshold crossings cause two alerts in total.