rrd graphs showing zero data for given interval

[jericho_g]

We haven't upgraded MRTG, to my knowledge. Requested output:

[root@clesitonag1 ~]# env LANG=C LC_ALL=C /usr/bin/mrtg
Usage: mrtg <config-file>

mrtg-2.16.2 - Multi Router Traffic Grapher

Copyright 1995-2006 by Tobias Oetiker
Licensed under the Gnu GPL.

If you want to know more about this tool, you might want
to read the docs. You can find everything on the
mrtg website:

http://oss.oetiker.ch/mrtg/

[root@clesitonag1 ~]# ll /var/lock/mrtg
total 0

[slansing]

The previous command was missing a bit of information, this should be the correct one, please run it and show us the output:

Code: Select all

LANG=C LC_ALL=C /usr/bin/mrtg /etc/mrtg/mrtg.cfg –check

Though there should be nothing wrong with your configuration file, since the graphs are being properly displayed, and the devices are clearly being checked. It's more of how the information is being returned, and subsequently averaged out for the graphs. How long ago had you added the example service you showed in post 1?

[jericho_g]

The graph shows data going back to August, 2013. Command output attached.

[slansing]

Once again, this is probably not related to how you are seeing the graphs displayed, but:

It looks like you have a number of interfaces which are not reporting correctly, they may either be unreachable, or no longer in use. To clean up your mrtg.cfg a bit you can follow the below document to remove old services (definitions in the config file):

http://assets.nagios.com/downloads/nagi ... Router.pdf

How many network devices are you monitoring via MRTG? Hosts, and services?

[jericho_g]

I'll review that PDF. We have 129 network devices defined in the mrtg.cfg, with graphs for their interfaces. So the number will be far larger than that.

[slansing]

Sounds good, how often are you checking those devices with Nagios? The reason I ask is that MRTG will sometimes run into issues where it completely drops checks that were supposed to occur due to timeouts on old devices that have not been removed. I'd definitely remove the devices you are currently NOT checking, as that may resolve the dropped data issue. The other route would be to split your mrtg.cfg into two so they run concurrently.

[jericho_g]

The PDF shows how to add a service. Is it sufficient to just delete old entries from the mrtg.cfg file? Or is there a service that needs to be restarted for changes to take effect?

[slansing]

You should be able to follow the "Remove Previous MRTG Configuration" section then continue on from there, verifying your mrtg configuration, be sure to remove the host(s)/services via the CCM as well. Let us know if you have issues.

[jericho_g]

Ok, I cleaned up my MRTG config and removed any orphan entries. Now the MRTG check is error-free. The graphs are definitely cleaner now since I applied the new config on Monday. However, today I noticed 2 dropouts at the same time across multiple hosts, around 11:30am:

mrtg_03262014.jpg

Any ideas as to why, or where to check?

Thanks
-Jericho

[abrist]

Well, start with the same logs you have probably checked 50 times at this point

Code: Select all

tail -25 /usr/local/nagios/var/perfdata.log
tail -25 /usr/local/nagios/var/npcd.log