To detect if a host is down and not sending data it seems the best way is to use agent.ping with a nodata() trigger. Couldn't it be possible to implement nodata() to allow to check for the number of checks no data was received for?
For example, if agent.ping gets data every minute and my trigger is for nodata(60) then this would be one check. If I updated agent.ping to 120 seconds then wouldn't nodata(60) trigger prematurely?
If nodata(#1) was set then it would coincide with the update interval.
Would this be possible or am I overlooking a reason why this isn't needed?
Ultimately, I would just like to set nodata(#2) to trigger if no data was received for the last 2 checks to prevent flapping, regardless of the update interval.
For example, if agent.ping gets data every minute and my trigger is for nodata(60) then this would be one check. If I updated agent.ping to 120 seconds then wouldn't nodata(60) trigger prematurely?
If nodata(#1) was set then it would coincide with the update interval.
Would this be possible or am I overlooking a reason why this isn't needed?
Ultimately, I would just like to set nodata(#2) to trigger if no data was received for the last 2 checks to prevent flapping, regardless of the update interval.