Hi
I use PRTG (
https://www.paessler.com/prtg) to monitor some devices behind routers.NR is a great client to use as I dont have to worry about setting up rules on firewalls/open ports etc.The clients are mainly Raspberry Pi's, with a few OpenWRT routers as well
All in all, it works well, but sometimes the ping sensor which monitors the PI (using the NR private IP as the destination), which is normally <20ms, will shoot up to >500ms, and stay there.PRTG alerts me to this, and the only way to fix it is to either reboot the device, or restart the nrclient service on the device.
I have tried various things but only the above will fix the issue-I have set up other pings on the PRTG server, to the public IP of the router (of which the Pi is behind), and through this, that ping remains <20ms-so the issue is not bandwidth or routing-it always seem to be the NR process on the Pi itself.Using the
top command, the Pi seems ok from a memory and CPU process point of view-,yes the NR service jumps up and down from about 1-7% CPU, but this shouldn't make a difference in the latency measurement.You know when things are not right, as when you SSH into the Pi, the console is very sluggish-restart NR,everything is ok then
Any idea on what could be causing this, or a way I can further investigate? It can be fine for days/even weeks, and then suddenly will shoot up-looking at the syslogs,I cant find any reason why this would happen.
The attached picture shows what I mean, hopefully!
cheers
cabs