To reduce poller times and to finish within 1 minute I am currently fine tuning poller modules for every device which is really cumbersome and not a way this can scale. Like why waste time to even try polling aruba-controller, cisco-cbqos, cisco-asa-firewall, … on an APC UPS?
As it’s really hard to discover the needed modules in an automated way without burning valuable timeslots I would love to see a way I can pre-define poller modules per Device Group so I don’t have to switch them on/off in a manual way device by device
Under /opt/librenms/includes/definitions/asa.yaml I can see poller_modules and discovery_modules being switched on/off but not for e.g. apc.yaml. Not sure if we can build a GUI for this or should just make it mandatory for have the poller_modules beeing defined inside the yaml OS file oing forward?
I am running into this problem now. I have a wide variety of devices being monitored on my LibreNMS instance, and there are some modules I’d like to be polled (i.e.: STP on switches) and when I don’t need to have it polled (i.e.: STP on routers). Having a GUI feature to enable/disable polling per device group (whether it’s by OS type, whatever) can be very handy.
I know that this is an old thread, but has there been any type of follow-up on this in similar developments? I tried to do a search, but to no avail…
This would indeed be very nice as for example, even within a “junos” family, we only have a few devices that would require module X or module Y, so it could indeed be a very nice bump in terms of performance.
Hi,
sorry I didn’t find support to get it done and I am really not a software developer so can’t get it done on my own.
Yes, to bad but I am more then happy to help and test if we can find someone doing the coding part.
I would like to see something like this as well.
Have a device/OS based profile that you can assign to (similar) devices where you can specify and update what needs monitoring and what doesn’t in bulk but obviously doesn’t touch per device overridden values.
