There's a feature in a competing monitoring solution that would be nice to see in SAM. As shown below, under normal circumstances you poll an application every 5 minutes. But if a check fails, it would be great to start polling more frequently to better verify that there is a real problem. This would help reduce false positives while still keeping polling frequencies and alert trigger times set to reasonable levels. Right now, we'd have to poll everything ever minute and delay our alert for 5 minutes to get similar results, but we'd prefer not to poll everything that frequently, unless an issue is actually detected.
