I hope someone can help with KNM dependencies.
we have 2 cases:
Case 1: most of the servers
I think setting KNM dependency to ping on the same asset will work here. (There is a catch though most of the service monitors are set up to trigger alarm in 5 min (test interval 60, alarm generation 5)- the same as ping alarm settings)
So I understand most of the alarms will trigger anyway, unless I pump up alert generation settings for the rest of the monitors to , say 10 min)
Case 2: some critical devices
These are set up to raise additional alarm "30 min down" when ping fails for 30 min and one of the actions is to send emails to on call engineer.
Now things got more complicated. If I set up dependency for ping monitor then my "30 min down" alert would not trigger and would not sent email.
Should I assign then dependency to "30 min down" monitor instead of ping ?
Would dependency block all the noise right after 1st test failure for the monitor configured for dependency? Or would that start blocking alerts only when alert is triggered after a number of failed tests (alert generation)
Also we have one of the alerts based on event viewer events for server reboot which we don't really want to suppress.
any advice here?