Action (with escalations) starts working on trigger-firing as expected. When the trigger is going back to normal, the action is not aborted and continues with it's escalations (1) although there is definitely a 'trigger-back-to-OK'-event.

So the check/host/service is up, but the action thinks it is down.

When the same trigger on the same machine raises again the action starts as expected, Now lets say the action has 4 steps and the problem is resolved after escal-step 1, the 'trigger-back-to-OK'-event sends immediately recovery-messages (2) to ALL recepients of all steps executed in (1).

This causes many false-positives. First at (1), and then at (2) again.

I'll attach screenshots for better explanation.