Hello,
Looking for some advice on a popular topping of non reporting hosts. Perhaps someone already came across something like this, or has a better way of doing it.
We have device pairs that report differently, and I am looking for a way to alert if a device stops reporting based on expected reporting cadence for a particular device. For example, have a CSV with device name/IP and a column for the expected reporting threshold that can be used to generate an alert if it is exceeded.
Example:
FW1-primary, 2m
FW1-secondary, 4h
So the search can look at the second column, and if it's been more than 4 hours since FW1-secondary sent an event, an alert can be generated. TIA!
I would recommend doing it like this:
Lookup (demolookup):
host,seconds
somehostname,60
Your alert search:
| metadata type=hosts | table host recentTime | lookup demolookup host | eval threshold=now()-seconds | where recentTime<threshold
I would recommend doing it like this:
Lookup (demolookup):
host,seconds
somehostname,60
Your alert search:
| metadata type=hosts | table host recentTime | lookup demolookup host | eval threshold=now()-seconds | where recentTime<threshold