Is there a way to have a monitor only notify you if a process state changes (from running to not running or from not running to running)? Let me explain. Right now I set up a monitor to alert me when a process stops (transitions from running to not running) but at that point it will continue to alert me that it isn't running. I only want to be alerted once when it goes from running to not running. Is that possible?
Are you monitoring it as a Service or as a Process
Process Monitoring has an ignore alarm for 'x' minutes option .. so I guess you could try setting the Ignore value to something huge so you'd only get an alert on the first transition.
Not use if the alarm interval resets when the process restart though
May pay to test
Or you could have the monitor set trigger a script when the process stops.
Have the script update a custom field .. and then also have the script check if there is already a value inthe field.
If there is no value make the script create an Alarm/Ticket and populate the field
If there is already a value then just exit the script with no Alarm/Ticket
On Process state change have the script delete the value from the field
I have a "Smart Monitor" procedure that does this kind of alert - part of our RMM Suite.
It consists of a VBScript and PSTools - the script, config file, and PSList command are deployed by Kaseya and then the script is invoked. The script loops constantly, sleeping for several minutes and then using pslist to check for a specific process (several, actually, defined in a config file). For each process that I'm monitoring, I have a set of actions that the script can perform, such as restarting the process up to N times, then logging an alert to the event log to indicate that it is not running.
Usually, we use it to alert when console-based apps have failed, but we have some customers with very long running processes (5+ days) where this is used to alert only (0 restarts) when the app has completed.
Basically, move the complex logic out of Kaseya - use Kaseya to deploy and execute the script, then monitor for the results, either by event log or email.
With a small mod, you can check for commands that should not normally be running, ignoring when they aren't in the PSList output and alerting when they are. The script can maintain state, only alerting when the transition is detected.