There are plenty of ways to specify the exact time range or maximum range between two events in a search. But I need to specify a minimum.
My search is: index=antivirus INFECTION dedup 1 infection host | top host limit="10"
It correctly finds my top 10 infected hosts by distinct viral infection and host. However, I want to make sure there is over X amount of time between each event, because I want to catch Bob the Bittorrenter who gets a new infection every week, and not Sue the Surfer who downloads one bad file and gets 8 infections in less than minute.
I've tried building a timechart, "transaction infection,host maxspan=X", and specifying a bucket on _time, but everything seems to give me the opposite of what I need.
I'm not sure why | transaction infection,host maxspan=X | top host
doesn't work for you. Perhaps | transaction infection,host maxpause=X | top host
works better (though similarly, but both of those will group together infections into a single event if they are close enough, leaving you with blocks of infections separated by the time range.
Possibly you are interpreting the results backwards, but since transaction groups together things that are closer than your max ranges (whether the overall range or the time between infections), you don't look inside the transaction, you look at each separate transaction and count those.
I'm not sure why | transaction infection,host maxspan=X | top host
doesn't work for you. Perhaps | transaction infection,host maxpause=X | top host
works better (though similarly, but both of those will group together infections into a single event if they are close enough, leaving you with blocks of infections separated by the time range.
Possibly you are interpreting the results backwards, but since transaction groups together things that are closer than your max ranges (whether the overall range or the time between infections), you don't look inside the transaction, you look at each separate transaction and count those.