2012년 10월 9일 화요일

Nagios: NRPE - Unable to read output

I added a script that checks CPU usage for remote Linux machines. I downloaded the file and give execution permission to it.

# Download the script for CPU usage 
http://verypowerful.info/home/nagiosplugin-checkcpuperformance

# Giving a permission 
$ chmod a+x /usr/lib64/nagios/plugins/check_cpu_perf.sh

# Command definition in nrpe.cfg on the remote server.
$ vi /etc/nagios/nrpe.cfg 
command[check_cpu]=/usr/lib64/nagios/plugins/check_cpu_perf.sh $ARG1$ $ARG2$


While I was testing this script on nagios monitoring server, I got the following error: 
$ /usr/lib64/nagios/plugins/check_nrpe -H 192.168.20.174 -c check_cpu -a 30 20
NRPE: Unable to read output

At the start, I searched why this happened and how this solved. 
Most of suggestions on those results were: 
1) Add the command that you want execute into nrpe.cfg 
command[check_cpu]=sudo /usr/lib64/nagios/plugins/check_cpu_perf.sh
2) Add nagios user and the script into /etc/sudoers
nagios ALL=(ALL) NOPASSWD:/usr/lib64/nagios/plugins/check_cpu_perf.sh
3) Comment out the section in /etc/sudoers
#Defaults requiretty

However, my case was different to these. It caused because of SELINUX.
SELINUX was enabled on my remote server.


$ vi /etc/selinux/config 
....
# SELINUX= can take one of these three values:
#     enforcing - SELinux security policy is enforced.
#     permissive - SELinux prints warnings instead of enforcing.
#     disabled - No SELinux policy is loaded.
# SELINUX=enforcing 
SELINUX=disabled 

# Needs to reboot to apply this.
$ reboot

I executed check_cpu again on Nagios Monitoring host
$ /usr/lib64/nagios/plugins/check_nrpe -H 192.168.20.174 -c check_cpu -a 30 20
OK: CPU Idle = 99.79% | CpuUser=0.12; CpuNice=0.00; CpuSystem=0.08; CpuIowait=0.01; CpuSteal=0.00; CpuIdle=99.79:20:10

It worked. If you can't find any solution about the error "Unable to read output", It's good to check SELINUX.  


In addition, when you have the following error message, It needs to enable "dont_blame_nrpe " option.
CHECK_NRPE: Received 0 bytes from daemon.  Check the remote server logs for error messages.

vi /etc/nagios/nrpe.cfg
...
dont_blame_nrpe=1

References: 
1. http://stackoverflow.com/questions/10542648/python-unable-to-call-sudo-using-popen

댓글 2개: