[collectd] collectd-nagios problems with ds queries

Ulrich Habel rhaen at cpan.org
Wed Jul 27 18:29:35 CEST 2011


I am trying to implement a monitoring system with nagios and a data
collection system with collectd 5.0.0. Right now all the data from the
hosts gets collected on one server using the network socket which
works great. Now I am trying to use collectd-nagios to query the rrd
files for the df plugin. In order to know how full a disk really is I
tried to use this query:

monitor:~# collectd-nagios -s /var/run/collectd-unixsock -H
monitor.xxxxx.de -g percentage -n df/df-root -d free -d used

This however leads to the following error:

ERROR: Retrieving values from the daemon failed: Server error: No such value.

If I look into the rrd directories I get the following information:

monitor:~# ls -l /var/lib/collectd/rrd/monitor.xxxxx.de/df-root
insgesamt 456
-rw-r--r-- 1 root root 154888 27. Jul 15:06 df_complex-free.rrd
-rw-r--r-- 1 root root 154888 27. Jul 15:05 df_complex-reserved.rrd
-rw-r--r-- 1 root root 154888 27. Jul 15:04 df_complex-used.rrd

Ok, it's clear that this method of querying data is likely to fail. Am
I missing something? Is the method to query the rrds correct?

I noticed the difference in data recording in some plugins e.g. load
collects all its data in one rrd db. Three values are stored inside
the db. The df plugin creates 3 rrd files but stores only 1 value in
the db. Is this a design flaw? I am asking this as I am about to write
a small frontend to generate the graphs. Life would be easier if there
would be a consistent schema to store all the data.


More information about the collectd mailing list