"Alert after Every" not behaving as I expect...

Topics: User Forum
Dec 29, 2007 at 12:58 AM
I have the retention scheme job run twice per day (every 720 cycles, 1 cycle per minute). I set the alert to "Alert after Every 2 Events)" so that, presumably, I'd get a "ping" from PolyMon telling me that it is alive ever 24 hours (1440 cycles or 1440 minutes). However, it sends out notifications every 36 hours instead. Which, to me, seems to be every THIRD cycle not second, which I would expect to read "every 3 events." Am I reading this wrong? If so, can this be cleared up? If not, do we have a ">" where a ">=" should be in the code? :-) :-)

Can anybody confirm?
Jan 18, 2008 at 10:35 PM
I'm guessing that your "1 cycle per minute" takes Polymon more than 1 minute to complete. I'm not sure, but it seems to me that it might be that it actually counts the cycles so is doing it every 2 events, but the events are more than 720 minutes apart.

Here's a query to check on how long it is taking for your 1 minute monitors. Hope it helps.

min(mcs.EventDT) as 'OldestEntry',
max(mcs.EventDT) as 'NewestEntry',
datediff(second, min(mcs.EventDT), max(mcs.EventDT)) as 'DiffSec',
datediff(second, min(mcs.EventDT), getdate()) as 'DiffSecRT',
datediff(second, max(mcs.EventDT), getdate()) as 'TimeSinceEntry',
count(*) as 'Enabled1MinuteMonitors',
count(*)/ convert(float,datediff(second, min(mcs.EventDT), max(mcs.EventDT))) as 'MaxPerSec',
count(*)/ convert(float,datediff(second, min(mcs.EventDT), max(mcs.EventDT))) *60 as 'MaxPerMin'
dbo.MonitorCurrentStatus mcs --with (nolock)
join dbo.Monitor m --with (nolock)
on m.MonitorID = mcs.MonitorID
m.TriggerMod = 1
and m.IsEnabled = 1