Nagios Event Handler/Command not executing

monitoringnagios

I'm trying to configure my Nagios setup to automatically log a ticket in our ticketing system when a host goes down (PING service goes into a HARD CRITICAL state). I've got a script that will create tickets that runs successfully with 'sudo -u nagios'. I've got a command and event handler set up that executes when the host state fails, or at least appears to. However, the script associate with the command doesn't seem to be executing, or the logging I've added isn't working. Any ideas where to look next? The log file isn't created when the event handler is invoked.

command:

define command {
        command_name    make-ticket
        command_line    /etc/nagios/commands/make-ticket-wrapper "$SERVICESTATE$" "$SERVICESTATETYPE$" "$HOSTNAME" "$HOSTADDRESS$ "$HOSTSTATE" "$HOSTGROUPALIAS" "$SERVICEDESC"
}

example host and service

define host {
    use             generic-switch
    host_name       test
    alias           test
    address         192.168.100.13
    }

define service {
    use                     generic-service
    host_name               test
    service_description     PING
    check_command           check_ping!200.0,20%!600.0,60%
    normal_check_interval   5
    retry_check_interval    1
    event_handler           make-ticket
    }

wrapper script:

#!/bin/bash
#
# Cut a ticket
#

set -x

LOGFILE="/tmp/tickets.log"

touch $LOGFILE
echo Running make-ticket-wrapper `date` >> $LOGFILE
echo Params: $* >> $LOGFILE

TICKET="/etc/nagios/commands/make-ticket"

SERVICESTATE="$1"
SERVICESTATETYPE="$2"
HOSTNAME="$3"
HOSTADDRESS="$4"
HOSTSTATE="$5"
HOSTGROUPALIAS="$6"
SERVICEDESC="$7"

echo "SERVICESTATE=$SERVICESTATE" >> $LOGFILE
echo "SERVICESTATETYPE=$SERVICESTATETYPE" >> $LOGFILE
echo "HOSTNAME=$HOSTNAME" >> $LOGFILE
echo "HOSTADDRESS=$HOSTADDRESS" >> $LOGFILE
echo "HOSTSTATE=$HOSTSTATE" >> $LOGFILE
echo "HOSTGROUPALIAS=$HOSTGROUPALIAS" >> $LOGFILE
echo "SERVICEDESC=$SERVICEDESC" >> $LOGFILE

$TICKET "$SERVICESTATE" "$SERVICESTATETYPE" "$HOSTNAME" "$HOSTADDRESS" "$HOSTSTATE" "$HOSTGROUPALIAS" "$SERVICEDESC" 2>&1 | tee -a $LOGFILE

log:

[1422560163] HOST ALERT: test;DOWN;SOFT;1;(Host Check Timed Out)
[1422560253] HOST ALERT: test;DOWN;SOFT;2;(Host Check Timed Out)
[1422560353] HOST ALERT: test;DOWN;SOFT;3;(Host Check Timed Out)
[1422560433] SERVICE ALERT: test;PING;CRITICAL;HARD;1;PING CRITICAL - Packet loss = 100%
[1422560433] SERVICE EVENT HANDLER: test;PING;CRITICAL;HARD;1;make-ticket

Best Answer

command_line /etc/nagios/commands/make-ticket-wrapper "$SERVICESTATE$" "$SERVICESTATETYPE$" "$HOSTNAME" "$HOSTADDRESS$ "$HOSTSTATE" "$HOSTGROUPALIAS" "$SERVICEDESC"

You're missing a double quote after $HOSTADDRESS$, and missing a $ at the end of HOSTNAME, HOSTSTATE, HOSTGROUPALIAS, and SERVICEDESC.

When you miss a double quote, it screws up the rest of the line, causing the shell script to miss some arguments.

When you leave out the trailing dollar sign, Nagios doesn't substitute the macro, so it gets passed to the shell where it's resolved as an empty variable.

Related Topic