I am running into an issue with NPM group status notifications.
We run a WAN with multiple sites and technicians where nodes are grouped into locations/campuses etc. I have tried setting up group notifications with 'group status is "not equal" to up' and a reset notification when the when it no longer applies.
The above worked great but it had a flaw, in that if for example a group went from a warning state to a down state it would not send another notification. Thusly it did the same when it would go from a down state to a warning state etc.
I wanted an alert that would send an email if a group changed its status at any stage and stayed in the status for over 10 minutes (so it didnt pick up VM reboots etc).
So I setup three different alerts per group..
One for Status is Up, One for Status is Down and One for Status is Warning.
This works great too, however if there are any slight fluctuations in groups it will send a "Group is Up". For example, a node goes into warning because it is rebooting, sending the group into a warning state. Because the group left the "Up" state and returned to the "Up" state within a couple of minutes, no warning email was sent because it didnt last for 10 minutes, however we always receive a "Group is Up" message because it returns to this state for over 10 minutes.
*Sigh*... so I'm back to square one with a half-baked solution. So what I want is a per-group status notification that shows the transitions between different statuses but only if they have existed in that state for over 10 minutes.
Group Up -> Group Warning -> Group Down -> Group Warning -> Group Up etc.
Has anyone found a solution to this? A fair amount of searching and I've come up empty.
Many thanks for your help.