Nagios Support Forum

Running Nagios XI 2014 R2.5 and deploying a cpu plugin that I modified to use sar data. Running the check locally seems to work fine, but when nrds runs the passive check, it shows up in Nagios without the data.
Entry from nrfs.cfg --> command[check_cpu_linux]=/usr/local/nagios/libexec/check_cpu_linux.sh -w 10 -c 5 -t id
Manually running above check:
[nagios@devserver libexec]$ /usr/local/nagios/libexec/check_cpu_linux.sh -w 10 -c 5 -t id
OK : CPU IDLE: 99% | Id=99%;10;5

XI Service Detail Status Information displays:
OK : CPU IDLE: %

I can't figure out what I'm doing wrong here.

Looks like this is related to sar not running (or not running in the same fashion) via cron job.

Did you check to see if the config was updated on the client? Are the version numbers the same? Can you run the following commands on the remote box and show us the output?
Also, in Nagios XI, go to:

Home->Service Detail-><your service>->Advanced tab and show us a screenshot
Admin->NRDS Config Manager-><your config>->Edit and show us a screenshot

The nagios config is good and the plugin works when it is run directly, or if I run the following line from the nagios crontab /usr/local/nrdp/clients/nrds/nrds.pl -H 'devserver' 2>&1. It is only when the cron job is run non-interactively that the CPU IDLE value is not returned. Im guessing it is a difference in the output of sar when run non-interactively, Will have to do some more troubleshooting of the plugin.

Does the LOG_FILE feature of in the NRDS Config under Additional Settings work? I have specified /logs/techserv/nagios/nrds.log and have verified that the nagios user can touch this file on the client, but nothing seems to get written here.

What are the permissions on the log file? Can you run the following command and show us the output?

As info, I had to change my plugin to parse a different column (minus 1) to get the %IDLE Time when tty is not present. I'm still not sure what is going on but am moving on for now.

Here is my nrds.log and an example nrds.cfg file.

[nagios@deva511a ~]$ ls -lrt /logs/techserv/nagios
total 0
-rw-rw-r-- 1 nagios nagios 0 Mar 31 15:25 nrds.log

CONFIG_VERSION="0.2"
CONFIG_OS="Linux"
CONFIG_NAME="Linux_standard_tomcat"
URL="https://nagiosxi/nrdp/"
TOKEN="mytokenvalue"
PLUGIN_DIR="/usr/local/nagios/libexec"
SEND_NRDP="/usr/local/nrdp/clients/send_nrdp.sh"
TMPDIR="/usr/local/nrdp/clients/tmp"
COMMAND_PREFIX=""
LOG_FILE="/logs/techserv/nagios/nrds.log"
UPDATE_CONFIG="1"
UPDATE_PLUGINS="1"

command[check_cpu_linux]=/usr/local/nagios/libexec/check_cpu_linux.sh -w 10 -c 5 -t id
command[check_swap_linux]=/usr/local/nagios/libexec/check_swap_linux -w 60 -c 80
command[check_all_linux]=/usr/local/nagios/libexec/check_all_linux -f /usr/local/nagios/etc/check_all_linux.cfg
command[check_fsspace_linux]=/usr/local/nagios/libexec/check_fsspace_linux -f /usr/local/nagios/etc/check_fsspace_linux.cfg
command[check_heap_used]=/usr/local/nagios/libexec/check_jmxeval.sh /usr/local/nagios/etc/jmxeval-heap_used.xml
command[check_catalina_log]=/usr/local/nagios/libexec/check_catalina_log.sh

I had to talk to the NRDS developer about this one. It seems like the "LOG_FILE" setting is not used in the Linux version. It's left there, because it is used in the Windows NRDS client (default log file: C:\Program Files\Nagios\NRDS_Win\logs\NRDS_Debug.log).

You can probably modify the cron entry to redirect the output to a file, i.e. Every time the cron job is run the log will be overwritten. I would not recommend using ">>" instead of ">". If you append the output, this log could grow out of control, unless you trim it.

As info, I had to change my plugin to parse a different column (minus 1) to get the %IDLE Time when tty is not present. I'm still not sure what is going on but am moving on for now.

Is the issue resolved after the mods you made?

Yes, the initial problem I was having seems to have been plugin related. Thanks for the nrds logging info.

Are we ok to mark this as resolved and lock the topic?

ok with me