Faulty behavior re: maintenance?

[ebonweaver]

Either there is a fault in the software, or I'm missing some tidbit. Pretty simple scenario:
1. Item - simple TCP
2. Trigger - simple check of last value of item
3. Action - email me when trigger status is ok or problem and not in maintenance time.

As expected, there is logic somewhere that automatically says don't email if the value of the trigger is the same as the last one, only email if the status has changed. Great, but now we hit maintenance. The logic I'd expect at that point is once outside maintenance, if the trigger value is still the same as before maintenance, don't email me. However, that's not what happens. Every time we come out of maintenance where the system was actually down, I get emails for every thing I have set up this way telling me it's ok. I don't want that, I only want to know if it's in problem state at that point indicating it failed to come up after maintenance. As is, I'm getting notice it's ok, but I won't get notice if it's problem because it already was problem but I didn't get notice because of maintenance.

So there is something faulty about this logic, or there is an option I'm not seeing to have the action act as expected. Can anyone offer workarounds as I'm sure this is a common issue?

[LenR]

I'm not sure I understand your action, but I think a action only initiates when an item value changes, so f it went down during maintenance, you won't get notice (not in maintenance) and when maintenance period ends, you won't get a notice because the item won't change (it's down).

[ebonweaver]

That's not quite accurate.
Item is up, trigger=ok
Maintenance period starts
server reboots due to MS updates, so item is now down, trigger=problem
No action is taken because maintenance is in effect
server comes back up after reboot, so item is up again, trigger=ok
Maintenance period ends, item still up, trigger still ok
action is taken to alert me status is now OK

This is not my expectation. Because status was ok BEFORE maintenance, and is still ok AFTER maintenance (and ok before maintenance ended), no action should be taken as status has not changed. However, it sees it changed DURING maintenance, so it takes action once maintenance is over, despite the fact changes during that period should be ignored, and status has not changed in the last several checks. Effectively, it's holding the action until maintenance is over, instead of ignoring the action during that period.