#! /bin/bash

###
# Nagios plugin skeleton
# http://nagiosplug.sourceforge.net/developer-guidelines.html
#
# Check dns resolution of the hostname spacified agains the current ip.
#
# @AUTHOR: Patrik Dufresne (http://patrikdufresne.com)
# Copyright 2012 Patrik Dufresne
# Last modified 2012-11-02
# Please send all comments, suggestions, bugs and patches to (info AT patrikdufresne DOT com)
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, version 2 of the License.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program.  If not, see <http://www.gnu.org/licenses/>.
###

# You should provide a meaningful VERSION
VERSION=0.1
# Who can be contacted about this?
AUTHOR="Patrik Dufresne"
# Name what is being checked to be printed out next to OK/WARNING/CRITICAL/UNKNOWN
SERVICE="DYNDNS"

# Replacement for the exit function, will cleanup any tempfiles or such
# before exiting.
function cleanup {
	exit $1
}

declare -rx PROGNAME=${0##*/}
declare -rx PROGPATH=${0%/*}/

if [ -r "${PROGPATH}utils.sh" ] ; then
	source "${PROGPATH}utils.sh"
else
	echo "Can't find utils.sh. This plugin needs to be run from the same directory as utils.sh which is most likely something like /usr/lib/nagios/plugins or /usr/lib64/nagios/plugins"
	printf "Currently being run from %s\n" "$PROGPATH"
	# Since we couldn't define STATE_UNKNOWN since reading utils.sh
    # failed, we use 3 here but everywhere else after this use cleanup $STATE
	cleanup 3
fi

# Set STATE to UNKNOWN as soon as we can (right after reading in util.sh
# where the STATES are defined)
STATE=$STATE_UNKNOWN

# make sure that any external commands are installed, in the PATH and
# executable. The following example is stupid because of course date is
# installed but it's the only command this trivial check really uses
CURL=/usr/bin/curl
if [ ! -x "$CURL" ] ; then
	echo "The utility $CURL is not installed. Exiting."
	cleanup $STATE_UNKNOWN
fi

HOST=/usr/bin/host
if [ ! -x "$HOST" ] ; then
        echo "The utility host is not installed. Exiting."
        cleanup $STATE_UNKNOWN
fi

# provide a quick one liner of how to use the program
function usage {
	printf " %s %s for Nagios - Usage %s -H <dns name> \
[-t timeout] [-v [-v [-v]]]\n" "$PROGNAME" "$VERSION" "$PROGNAME"
	cleanup $STATE_UNKNOWN
}

# provide detailed explanations of the command line syntax
function longhelp {
	# put your long help here
	printf "%s plugin version %s for Nagios by %s
  -h, --help          Display this message.
  -H, --hostname      Define the dns name to verify.
  -t, --timeout=sec   Set script timeout in seconds.
  -v, --verbose       Up the verbosity level by one.
  --verbosity=val     Set the verbosity level to val.
  -V, --version       Print version information.
" "$PROGNAME" "$VERSION" "$AUTHOR"
	cleanup $STATE_UNKNOWN
}

if [ $# -eq 0 ] ; then
	usage
fi

# use getopt, trust me on this one. It's the easiest way
getopt -T
if [ $? -ne 4 ] ; then
	printf "%s: getopt is in compatibility mode.\n" "$SCRIPT"
	cleanup $STATE_UNKNOWN
fi

# Tell it which switches and longswitches you'll take and place a trailing
# colon (:) on the ones take arguments. Nagios guidelines require you to
# use all the ones specified below with the exception of --verbosity which I've
# added to circumvent the awkward -v -v -v syntax. Getopt takes care of
# positional parameters and errors for missing expected arguments so we can
# shift later without checking
RESULT=`getopt --name "$SCRIPT" --options "-h,-V,-v,-t:,-H:" --longoptions "help,version,verbose,verbosity:,timeout:,hostname:" -- "$@"`

# make the result of getopt your new argument list ($@)
eval set -- "$RESULT"

declare WARNING
declare CRITICAL
# all scripts should have a mechanism to terminate themselves if they are
# running for too long. Scripts you might think of as innocuous could end
# up waiting forever on I/O, especially if a disk is failing
declare -i TIMELIMIT=15
# Nagios defines behavior for VERBOSITY 0 (default) through 3
declare -i VERBOSITY=0

while [ $# -gt 0 ] ; do
	case "$1" in
		-h | --help)
			longhelp;;
		-V | --version)
			print_revision "$PROGNAME" "$VERSION"
			cleanup $STATE;;
		-v | --verbose)
			VERBOSITY=$(($VERBOSITY + 1));;
		--verbosity)
			shift
			VERBOSITY=$1;;
		-t | --timeout)
			shift
			TIMELIMIT=$1;;
		-H | --hostname)
			shift
			HOSTNAME=$1;;
		--)
			shift
			break;;
		*)
			echo "Option $1 not supported. Ignored." >&2;;
	esac
	shift
done

#Verbosity level	Type of output
#0			Single line, minimal output. Summary
#1			Single line, additional information (eg list processes that fail)
#2			Multi line, configuration debug output (eg ps command used)
#3			Lots of detail for plugin problem diagnosis
if [ $VERBOSITY -gt 2 ] ; then
	shopt -o -s xtrace
fi

# what needs to happen in the event of a timeout
function timeout {
	echo "UNKNOWN - script timed out after $TIMELIMIT seconds."
	cleanup $STATE_UNKNOWN
}

# since we've processed the options which potentially set the timeout limit,
# we can setup a timeout trap now
trap timeout USR1
	# what we're doing here sending a USR1 signal back to this process which
    # we just set a trap to catch and run the timeout function the syntax of
    # this is important and very odd - if you know of a better way to do this, 
    # please email me what we're doing is starting another process in the
    # background that sleeps for TIMELIMIT seconds and then uses pgrep when
    # it 'wakes up' to see if a process with our number, name and user exists,
    # only then will the USR1 signal be sent we have to use pgrep so that we
    # don't sent a USR1 signal to just any program. The only risk we run with
    # this is sending USR1 to another instance of this script that just happens
    # to get assigned the same process ID it should be reasonable to assume
    # that your Nagios check interval is greater than the specified timeout
	# still, if you havea better idea...
    ( sleep $TIMELIMIT; if [ `pgrep -U $USER -f "$SCRIPT" | grep -c ^$$$` -gt 0 ] ; then kill -USR1 $$ ; fi; ) </dev/null &>/dev/null &

    # Check that sensor is provided
    if [ -z "$HOSTNAME" ] ; then
        usage
    fi

    # Get the hostname value
    DYNDNS_IP=$($CURL --max-time $TIMELIMIT  checkip.dyndns.org 2>&1)
    IP_STATE=$?
    if [ $? -eq 0 ]
    then
        DYNDNS_IP=`expr "$DYNDNS_IP" : '^.*<body>Current IP Address: \([0-9][0-9]*\.[0-9][0-9]*\.[0-9][0-9]*\.[0-9][0-9]*\)</body>.*$'`
        IP_STATE=$?
    fi

    HOSTNAME_IP=`$HOST "$HOSTNAME"`
    HOST_SATE=$?
    if [ $? -eq 0 ]
    then
        HOSTNAME_IP=`expr "$HOSTNAME_IP" : '.* has address \([0-9][0-9]*\.[0-9][0-9]*\.[0-9][0-9]*\.[0-9][0-9]*\)$'`
        HOST_SATE=$?
    fi
    

    # Once we're done doing work that could take any real time, we can end the
    # trap because from here on out it's just comparisons and string
    # concatenation
trap - USR1

# Check the state

if [ $IP_STATE -ne 0 ] ; then
	STATE=$STATE_UNKNOWN
        OUT="Cannot parse result from checkip.dyndns.org"
else
    if [ $HOST_SATE -ne 0 ] ; then
        STATE=$STATE_CRITICAL
        OUT="Could not resolve hostname $HOSTNAME"
    else 
        if [ "$HOSTNAME_IP" == "$DYNDNS_IP" ] ; then
        STATE=$STATE_OK
        OUT="hostname resolved to $HOSTNAME_IP"
        else
        STATE=$STATE_CRITICAL
        OUT="hostname resolution doesn't matches : $HOSTNAME_IP <> $DYNDNS_IP"
        fi
    fi
fi

case $STATE in
	$STATE_OK)
		printf "%s OK - %s\n" "$SERVICE" "$OUT";;
	$STATE_WARNING)
		printf "%s WARNING - %s\n" "$SERVICE" "$OUT";;
	$STATE_CRITICAL)
		printf "%s CRITICAL - %s\n" "$SERVICE" "$OUT";;
	$STATE_UNKNOWN)
		printf "%s UNKNOWN - %s\n" "$SERVICE" "$OUT";;
esac

cleanup $STATE