It would make sense to check a server more frequently if it is down. So for example if we have services that check a server every 10 or every 30 minutes and they happen to be down, they will be marked as being down for 10 or 30 minutes even though they’ve only been down for a few seconds or 1 minute max.