Netonix Forums

I've recently installed LibreNMS to monitor network devices. I've noticed I'll frequently see Netonix devices being reported as down, even though I know they are not. (If they were down I'd also receive node down alerts on the devices plugged into them, which I do not.) Has anyone else seen this, and if so did you find a reason (or a way to make it not happen)?

Thanks
Sam

I have this too, it's down for 5 minutes. It missed one or several polling rounds. Rest of the network stays reachable, so it's not the polling machine.
The switch is "local" to the polling machine via fiber. The netonix replaces a netgear switch that was hooked up the same for two years without missing any polling rounds, so it's not infrastructure. It must be SNMP service in netonix.
I'm also having lots of spikes as well, probably also because of missed polling rounds. Spikes go up to 50Gbit or so.

Gaps are not a big issue to me, the spikes are, since they make the graphs unreadable.

Right now we are trying to finish up v1.3.5

After we get that out we will setup several NMS servers and see if we can replicate.

ColoCenter wrote:...

Gaps are not a big issue to me, the spikes are, since they make the graphs unreadable.

To remove erroneous peaks from RRD files, you can use this shell script :
http://cricket.sourceforge.net/contrib/files/killspike2