count does not work as I expect it to....

[frater]

For years I've been using the item "packetloss" which I wrote myself.
It works great and has helped me a lot finding the culprit of a bad internet experience.

The triggers are set in a way that would give me info on a constant bad connection, but I found out that some connections just have a lot of packetloss for a short period of time.
That's why I started counting the times I have 100% packetloss.

It works partially...
I'm getting the connections that have had 100% packetloss in the last hour, but that message does not go away after it has been OK for more than an hour.

Anyone knows how that's possible??


ITEM

Name: Packetloss on $1
Type: External check
key: packetloss[{HOST.CONN}]
Type: Numeric (unsigned)
Units: %
Update interval: 1m
History: 180d
Trend: 365d

Trigger

Site {HOST.CONN} has had 100% packetloss in the last hour
{Template_Standalone:packetloss[{HOST.CONN}].count(1h,100)}>2
Dependency: Template_Standalone: Site {HOST.CONN} has 100% packetloss (Internet down)

[frater]

I set the "OK Event Generation" to "None" never realizing that this would create such a situation.

This trigger was the result of a clone from a trigger having a "Recovery Expression". Because that didn't apply to this trigger I set it to "None".
This was wrong...

Now I realize I had to set it to "Expression"