we're doing it kinda wrong, we should mention stale results when a check switches to OK/FAIL to do this, we need to track the age of a failed check
280 B
280 B
for targets: run checks
compare against cache: 1. count stale checks 2. count changed checks
if changed checks: send notification with changed + stale checks else if stale checks & last stale notification > staletime: send notification with stale checks