I am monitoring some applications that send errors via snmp traps. Traps are only sent when there is an errors. Since most of the trigger functions are only evaluated when the item is updated this caused the trigger to stay on until another error was sent, could be days. The other issue is keeping an alert visible long enough t be seen and reported by my team.
Here is an example and my solution. Hope it helps.
I only care if I receive multiple errors in a period of time. I created an iformation level trigger for the count
{item}.count(120)<4
This trigger will stay active until the count is r in 2 minutes.
The next trigger on the same item is a high level trigger.
{item}.nodata(600)=0
This trigger is dependent on the first trigger and can obly be active if the count is greater than 3. The nodata function will continue to update while the count funtion will not. This insure I get the alert and that it clears after 10 minutes.
Here is an example and my solution. Hope it helps.
I only care if I receive multiple errors in a period of time. I created an iformation level trigger for the count
{item}.count(120)<4
This trigger will stay active until the count is r in 2 minutes.
The next trigger on the same item is a high level trigger.
{item}.nodata(600)=0
This trigger is dependent on the first trigger and can obly be active if the count is greater than 3. The nodata function will continue to update while the count funtion will not. This insure I get the alert and that it clears after 10 minutes.