Approach for monitoring not always on devices

[jriker1]

Looking for some input/insight on the best way to monitor something that is turned on and off as needed. Or maybe I just shouldn't monitor it. In this particular case, can be two items:

1. A network color laser printer. This printer is turned on and off as needed to conserve power and because it's not the primary printer used. Just used for color prints. I would like to know the ink status of it, but would not like to have any alerts should it be off and a check fails.

2. I have one server that we turn on and off for development purposes. it is not used much. Same kind of deal as 1 above, would not like to have any active alerts come across or anything or show horrible critical red messages in the dashboard if it's not on and checks fail.

Thoughts?

Thanks.

JR

[rhassing]

For the server you could use passive checks and let the server send the updates when it is switched on.

For the printer you could check and see if it send snmptraps and create a passive checks on your nagios server

[jriker1]

Thanks for the reply. I think I need a lesson on the difference between passive and active monitoring. Is there a description anywhere on the difference and implications?

Thanks.

JR

[rhassing]

There is some information available on the Nagios website:

https://assets.nagios.com/downloads/nag ... hecks.html

https://assets.nagios.com/downloads/nag ... hecks.html

[jriker1]

For the printer you could check and see if it send snmptraps and create a passive checks on your nagios server

So I understand this piece correctly. Is the concept that nagios would check if the system is "pinging" and if so run it's checks, otherwise it wouldn't run them? Could you provide an example of a passive check in this situation from Nagios?

Thanks.

JR

[hsmith]

Passive checks do nothing more than run the host check on the system, and then send the result to Nagios. Not ping-like mechanism involved. What would you like an example of?

[rhassing]

Sean Falzon had a talk about passive checks last year at the World conference.
https://youtu.be/k1g4TAK6ipo

What he explains in his talk is that he has a client installed on laptops or desktop PC's which will send their information when a system is online.

Passive checks are always initiated from a client, so its not, if it is pingable. The clients always sends the info when it can...

[rkennedy]

Thanks @rhassing!

@jriker1 let us know if you have any further questions.

[jriker1]

I am listening to the video now but as the part I originally highlighted I was responding to in my last post, I am talking about the printer. So the printer isn't going to have NSCA or NSClient++ installed to send checks to Nagios. That's where I was curious about the ping concept. So think passive is the right answer, just need an intermediary I would assume to make this happen and that intermediary needs to know when the printer comes online.

"Oh it's up, check on it's ink status". "Hey Nagios, here's the latest from the printer as it came on for a bit".

JR

[rkennedy]

I believe what @rhassing refered to was with SNMP passive checks.

Does the printer have SNMP and SNMP trap capabilities?