Nagios check_ssh returns usage information instead of status

Question

I installed Nagios on a Ubuntu Desktop (Nagios server) and I want to monitor a Ubuntu server instance (monitored client). I can connect via SSH between both machines and SSH is not blocked. The nagios standard services such as PING, and check_users work, but check_ssh remains in an UNKNOWN state from the very beginning. The state information delivers "Usage: " which is an indicator that the parameters are wrong.

I can execute the check manually, at the nagios server (Ubuntu Desktop)

/usr/local/nagios/libexec/check_ssh -H 192.168.0.2

SSH OK - OpenSSH_6.6.1p1 Ubuntu-2ubuntu2.3 (protocol 2.0) | time=0,012856s;;;0,000000;10,000000

and also at the host (Ubuntu Server)

/usr/lib/nagios/plugins/check_ssh 192.168.0.2

SSH OK - OpenSSH_6.6.1p1 Ubuntu-2ubuntu2.3 (protocol 2.0) | time=0.011613s;;;0.000000;10.000000

This is the host configuration:

define host {
        use                          linux-server
        host_name                    backup
        alias                        Backup Server
        address                      192.168.0.2 
        register                     1
}

The configuration of the host is standard:

define service {
      host_name                       backup
      service_description             Check SSH
      check_command                   check_ssh
      max_check_attempts              2
      check_interval                  2
      retry_interval                  2
      check_period                    24x7
      check_freshness                 1
      contact_groups                  admins
      notification_interval           2
      notification_period             24x7
      notifications_enabled           1
      register                        1
}

I also tried adding the IP of the host manually:

define service {
      host_name                       backup
      service_description             Check SSH
      check_command                   check_ssh!192.168.0.2
      max_check_attempts              2
      check_interval                  2
      retry_interval                  2
      check_period                    24x7
      check_freshness                 1
      contact_groups                  admins
      notification_interval           2
      notification_period             24x7
      notifications_enabled           1
      register                        1
}

What am I missing here?

What I also tried based on the hints

Based on the comment of Dan I could figure out that the service actually tries to submit the parameters in this configuration:

ssh_check $ARG1$ '$HOSTADDRESS$' 

Verifying the available ssh_check configuration on my backup system shows these possibilities, however the standard ssh_check command did not work nevertheless.

cat /etc/nagios-plugins/config/ssh.cfg 
# 'check_ssh' command definition
define command{
    command_name    check_ssh
    command_line    /usr/lib/nagios/plugins/check_ssh '$HOSTADDRESS$'
    }

# 'check_ssh_port' command definition
define command{
    command_name    check_ssh_port
    command_line    /usr/lib/nagios/plugins/check_ssh -p '$ARG1$' '$HOSTADDRESS$'
    }

####
# use these checks, if you want to test IPv4 connectivity on IPv6 enabled systems
####

# 'check_ssh_4' command definition
define command{
        command_name    check_ssh_4
        command_line    /usr/lib/nagios/plugins/check_ssh -4 '$HOSTADDRESS$'
        }

# 'check_ssh_port_4' command definition
define command{
    command_name    check_ssh_port_4
    command_line    /usr/lib/nagios/plugins/check_ssh -4 -p '$ARG1$' '$HOSTADDRESS$'
    }

I further tried to submit the IP address of the backup server as both parameters, but did not suceeed.

define service {
      host_name                       backup
      service_description             Check SSH
      check_command                   check_ssh!192.168.0.2!192.168.0.2
      max_check_attempts              2
      check_interval                  2
      retry_interval                  2
      check_period                    24x7
      check_freshness                 1
      contact_groups                  admins
      notification_interval           2
      notification_period             24x7
      notifications_enabled           1
      register                        1
}

Solution

I could not figure out how to address the service commands defined at the client nagios plugin configuration (/etc/nagios-plugins/config/ssh.cfg ). When I set for instance ssh_check_4, my nagios server complained about an undefined service. I ended up configuring a new service command:

define service {
      host_name                       backup
      service_description             Check SSH
      check_command                   check_ssh_fix
      max_check_attempts              2
      check_interval                  2
      retry_interval                  2
      check_period                    24x7
      check_freshness                 1
      contact_groups                  admins
      notification_interval           2
      notification_period             24x7
      notifications_enabled           1
      register                        1
}

define command{
  command_name  check_ssh_fix
  command_line  /usr/lib/nagios/plugins/check_ssh '$HOSTADDRESS$' 
}

Thanks!

Answer 1

There should be a defined command for "check_ssh"

Like so, e.g. on a Debian/Ubuntu system /etc/nagios-plugins/config/ssh.cfg

define command{
  command_name  check_ssh
  command_line  /usr/lib/nagios/plugins/check_ssh '$HOSTADDRESS$'
}

So by default the hostaddress is passed to the command, without you doing anything other than use check_ssh.

I guess your command looks like this:

define command{
  command_name  check_ssh
  command_line  /usr/lib/nagios/plugins/check_ssh '$HOSTADDRESS$' $ARG1$
}

Therefor the following command would be run:

/usr/lib/nagios/plugins/check_ssh '1.2.3.4' 1.2.3.4

Answer 2

Check that the version of nagios-client is compatible with your monitoring system. I had similar issue when locally was running ok. But it was not updating my monitoring (centron). I downgrade nagios-client to the previous version and it started to work ok.