#2 alert continuity--warning/threshold

[Adam Guyot]

The threshold is not reset to zero if the service is actually found to be up
and running. Say the threshold is 4. It fails twice in a row and then connects
the next time through. It could then go for days w/ out a failure but the number of failures would still be set at 2. Then it would take only 2 more failures. So maybe I need to put in a consecutive failure option