Unable to access /var/log files

[djs]

Hello all (unfortunatly I cannot post in the Nagios core forum - so have had to post here)

I am trying to cat a log file held in /var/logs/ e.g /var/log/messages.
I appreciate this is likely to be a large file but bare with me! I can create any other file by doing a touch file1 put some random stuff init.
When I run the command ./check_nrpe -H 00.00.00.001-c check_log it will return the contents of the file i created. (Brilliant)
If i move the file i created in to /var/log/ it works (again brilliant)
If i try and access /var/log/messages it complains saying no output from nrpe. I have played around with permissions (etc/sudoers) created a group and tried adding the user nrpe to that group.

If anybody has any ideas as to what can be done to access files in /var that would be great!

Thanks

[abrist]

/var/log/messages is usually owned by root. You will need a sudoers rule to allow the nagios user to read the file. Can you post your additions to the sudoers file?

[djs]

Hi,

The only edit i added to the /etc/sudoers file was:
nrpe ALL=NOPASSWD: ALL

[djs]

Just an update.

I can run the script to cat the file when i am logged into nagios (su nrpe -s /bin/bash) however on the webpage this displays nothing to me. Well it displays a '\n'

The script I am running consists of:

Code: Select all

#!/bin/bash
echo $(pwd)
echo $(whoami)
#echo $(hostname)
echo $(ls -l /var/log/messages)
echo $(cat /var/log/messages)
echo $(head /var/log/messages)

This is being called in /etc/nagios/object/commands.cfg

Code: Select all

define command{
  command_name    view_log_messages
  command_line       sh /etc/nagios/plugins/viewmessages
}

Any help would be great.

Thanks

[lmiltchev]

I am a bit confused about your use of check_nrpe... You are using:

Code: Select all

./check_nrpe -H 00.00.00.001-c check_log

This implies you are running a check against a remote box (this is what check_nrpe is used for). On the other hand, you have a command, defined locally (on the nagios server):

Code: Select all

define command{
  command_name    view_log_messages
  command_line       sh /etc/nagios/plugins/viewmessages
}

The command "check_log" should be defined on the client and on the nagios box, you should have a service defined with the following directive:

Code: Select all

check_command			check_nrpe!check_log!

Can you elaborate? Where are you actually running the check - locally or on the remote box?

[djs]

Sorry i forgot to edit my original post. Ive been trying all sorts to get it working.

The script is located locally on the same box and i can log into the user nrpe/nagios shell and run it (again locally) however on the nagios website/dashboard this does not display the information i was previosulay able to get from logging into the nrpe/nagios user.

Basically it works fine from the shell however as soon as it goes on the gui it fails

[lmiltchev]

I sent you a PM. Let me know if this worked for you so that I can lock the post. Thanks!

Update: Issue resolved after disabling SELinux.