Rrd reporting unsubstantiated packet loss

[clarknova]

2.1-RELEASE (i386)
built on Wed Sep 11 18:16:44 EDT 2013
FreeBSD 8.3-RELEASE-p11

As seen in the attached snap, my rrd graph is showing a solid 2% packet loss on the WAN gateway. However pinging the WAN gateway manually from pfsense never shows any packet loss, so why is rrd recording it?

[NOYB]

What are the monitor settings?
(System - Routing - Edit)

[clarknova]

The settings look default to me.

[NOYB]

Do a packet capture to see whether or not all the monitor ping replies are really making it back.

[clarknova]

Yup. 115 requests, 115 replies. Graph still showing a solid 2.0% packet loss. I'm guessing there's some bad math happening in the box somewhere.

[NOYB]

apinger service is what reports data to the quality rrd.

apinger seems to be temperamental sometimes.

Some things that might correct it.
  Stop / restart apinger
  Reboot
  Reset RRDs (will lose existing data for all RRD graphs)

[clarknova]

Thanks for the suggestions. I now have another pfsense reporting a solid 30% packet loss. What is the recommended way to start apinger?

[scurrier]

Resurrecting a bit of an old thread here.  Just wanted to throw my hat in the ring and say that I am seeing some behavior that I don't understand from the quality plots as well.  I am using 8.8.8.8 (google DNS) as my quality IP so that it leaves my ISP's network and I can measure quality to the real internet, not just to the ISP's gateway.

I am not seeing the constant packet loss like what's reported in this thread.

My plots instead will have strange discontinuities triggered by (for example) shutting down and restarting pfSense.  In these cases the quality ping might jump from 40ms to 80ms and stay there, and not go back until I disturb the setup somehow, usually with a reboot.  During the sustained 80ms, I can use pfSense to ping the exact same 8.8.8.8 that the plot should be recording and get a completely normal ping of around 40ms.  Yet, the quality readouts continue to report 80ms.  So, it seems like something is wrong here.  Or perhaps I am misunderstanding.

[Balrog30]

I'd just like to throw in that I'm seeing the same behaviour as the original post mentions, and I found this with google, so others may find this if the have the same problem.

To restart apinger go to Status > Services and apinger should be the first in the list.  On the right hand side there are a few buttons and the first one is used to restart the service.  Previously we would reboot the machine because we thought it was reporting actual packet loss and couldn't find any other way to correct it, just rebooting this service is a much simpler solution.

[Derelict]

It's apinger.  I think I restart it by turning gateway monitoring off then back on for the interface.