Need Help with Development of Nagios Plugin

monkinsane · 04-15-2009, 07:25 AM

Hi There,

I am in the process of developing a script that returns the output of the nagios "check_nt" plugin in a format that is recognized by Splunk.

I have pretty much finished it when I ran into a "snag". When I run the plugin from the command-line, it outputs exactly like I want it to.

However, from inside Nagios - The plugin returns {no output from plugin}. If I remove the actual check results from the plugin's output and replace it with a plain text string - nagios displays this string in the status information area.

I am baffled, as this is not my first plugin I wrote - and according to me - it should work.

Here is my Code:

Code:

#!/bin/sh

PROGNAME='Check NT Perf - Results Splunk Compatible'

SCP="./check_nt"

HOST=$1
PORT=$2
VAR=$3
PARAM=$4
WARN=$5
CRIT=$6



UPTIME=`$SCP -H $HOST -p $PORT -v $VAR|sed 's/System Uptime - //'|sed 's/day(s)/./'|sed 's/hour(s)/./'|sed 's/minute(s)//'`

CPU=`$SCP -H $HOST -p $PORT -v $VAR -l $PARAM|sed 's/CPU Load//'|sed 's/%//'`

MEM=`$SCP -H $HOST -p $PORT -v $VAR $WARN $CRIT|sed 's/Memory usage: total:[0-9][0-9][0-9][0-9].[0-9][0-9] Mb//'|sed 's/ - used: //'|sed 's/[0-9][0-9][0-9][0-9].[0-9][0-9] Mb//'|sed 's/[0-9][0-9][0-9].[0-9][0-9] Mb//'|sed 's/(//'|sed 's/%)/./'|sed 's/- free://'|sed 's/[0-9][0-9][0-9][0-9].[0-9][0-9] Mb//'|sed 's/([0-9][0-9]%)//'|sed 's/. [0-9]//'|sed 's/.//'|sed 's/0.00;0.00;0.00;//'|sed 's/;[0-9][0-9][0-9][0-9].[0-9][0-9]//'`

DRIVE=`$SCP -H $HOST -p $PORT -v $VAR -l $PARAM $WARN $CRIT`

							  
if [ "$VAR" = "UPTIME" ] ; then

	OUTPUT="${OUTPUT} 'System_Uptime'=$UPTIME |Format: days.hours.minutes"
	STAT=0

	elif [ "$VAR" = "CPULOAD" ] ; then

		OUTPUT="${OUTPUT} 'CPU_Load'=$CPU |"
		#STAT=0
	
	elif [ "$VAR" = "MEMUSE" ] ; then

		OUTPUT="${OUTPUT}'Memory_Used_Percent'= `echo $MEM`"
		#STAT=0

	elif [ "$VAR" = "USEDDISKSPACE" ] ; then

		OPT="`echo $DRIVE > drv.tmp`"
		OUTPUT="${OUTPUT} `cat drv.tmp |sed \"s/([0-9][0-9]%)//\"|sed \"s/total:/'Total Space'=/\"|sed \"s/used:/'Used Space'=/\"|sed \"s/free/'Free Space'=/\"|sed 's/-//'|sed \"s/([0-9][0-9]%)//\"|sed \"s/([0-9]%)//\"|sed \"s/([0-9][0-9][0-9]%)//\"`"
		STAT=0
	
	else

		OUTPUT="${OUTPUT} ERROR - Invalid Variable Specified!"
		STAT=2

fi

echo $OUTPUT 
#echo $DRIVE
exit $STAT

The Command-Line output is as follows:

Code:

c:\ 'Total Space'= 136.72 Gb - 'Used Space'= 83.98 Gb - 'Free Space'= 52.73 Gb | 'c:\ Used Space'=83.98Gb;109.37;0.00;0.00;136.72

Any help would be greatly appriciated.

Regards,

Monkinsane

datopdog · 04-15-2009, 07:35 AM

The problem is you are using a relative path instead of the full path.

monkinsane · 04-15-2009, 07:40 AM

Quote:

Originally Posted by datopdog

The problem is you are using a relative path instead of the full path.

Hi TopDog,

Thanx for the quick reply.

I changed to full paths, still no joy. Any suggestions?

It does work if I run it from command line.

datopdog · 04-15-2009, 07:43 AM

Can you post the modified plugin and the check_command you are using to call it from nagios.

monkinsane · 04-15-2009, 09:56 AM

Quote:

Originally Posted by datopdog

Can you post the modified plugin and the check_command you are using to call it from nagios.

Thanx for the assist.

Modified Code:

Code:

#!/bin/sh

PROGNAME='Check NT Perf - Results Splunk Compatible'

SCP="/usr/local/nagios/libexec/check_nt"
SED="/bin/sed"

HOST=$1
PORT=$2
VAR=$3
PARAM=$4
WARN=$5
CRIT=$6

UPTIME=`$SCP -H $HOST -p $PORT -v $VAR|$SED 's/System Uptime - //'|$SED 's/day(s)/./'|$SED 's/hour(s)/./'|$SED 's/minute(s)//'`

CPU=`$SCP -H $HOST -p $PORT -v $VAR -l $PARAM|$SED 's/CPU Load//'|$SED 's/%//'`

MEM=`$SCP -H $HOST -p $PORT -v $VAR $WARN $CRIT|$SED 's/Memory usage: total:[0-9][0-9][0-9][0-9].[0-9][0-9] Mb//'|$SED 's/ - used: //'|$SED 's/[0-9][0-9][0-9][0-9].[0-9][0-9] Mb//'|$SED 's/[0-9][0-9][0-9].[0-9][0-9] Mb//'|$SED 's/(//'|$SED 's/%)/./'|$SED 's/- free://'|$SED 's/[0-9][0-9][0-9][0-9].[0-9][0-9] Mb//'|$SED 's/([0-9][0-9]%)//'|$SED 's/. [0-9]//'|$SED 's/.//'|$SED 's/0.00;0.00;0.00;//'|$SED 's/;[0-9][0-9][0-9][0-9].[0-9][0-9]//'`

DRIVE=`$SCP -H $HOST -p $PORT -v $VAR -l $PARAM $WARN $CRIT`

							  
if [ "$VAR" = "UPTIME" ] ; then

	OUTPUT="${OUTPUT} 'System_Uptime'=$UPTIME |Format: days.hours.minutes"
	STAT=0

	elif [ "$VAR" = "CPULOAD" ] ; then

		OUTPUT="${OUTPUT} 'CPU_Load'=$CPU |"
		#STAT=0
	
	elif [ "$VAR" = "MEMUSE" ] ; then

		OUTPUT="${OUTPUT}'Memory_Used_Percent'= `echo $MEM`"
		#STAT=0

	elif [ "$VAR" = "USEDDISKSPACE" ] ; then

		OPT="`echo $DRIVE > drv.tmp`"
		OUTPUT="${OUTPUT} `cat drv.tmp |$SED \"s/([0-9][0-9]%)//\"|$SED \"s/total:/'Total Space'=/\"|$SED \"s/used:/'Used Space'=/\"|$SED \"s/free/'Free Space'=/\"|$SED 's/-//'|$SED \"s/([0-9][0-9]%)//\"|$SED \"s/([0-9]%)//\"|$SED \"s/([0-9][0-9][0-9]%)//\"`"
		STAT=0
	
	else

		OUTPUT="${OUTPUT} ERROR - Invalid Variable Specified!"
		STAT=2

fi

echo $OUTPUT 
#echo $DRIVE
#echo $MEMHEAD
exit $STAT

Nagios Command:

Code:

define command {
  command_name                    check_nt_perf
  command_line                    $USER1$/check_nt_perf.sh $HOSTADDRESS$ $ARG1$
}

Service Command:

Code:

check_command         check_nt_perf!12489 USEDDISKSPACE c -w 80 -c 90

Executed from Konsole:

Code:

check_nt_perf.sh 10.0.0.1 12489 USEDDISKSPACE c -w 80 -c 90
                 [ip] [port] [Variable] [par] [warning] [critical]

datopdog · 04-16-2009, 02:28 AM

Okay, it is because your plugin is only coded to return a result when checking for UPTIME, nothing else.

That is what the if statement does.

monkinsane · 04-16-2009, 08:07 AM

Quote:

Originally Posted by datopdog

Okay, it is because your plugin is only coded to return a result when checking for UPTIME, nothing else.

That is what the if statement does.

Well,

It does return the output from the uptime check. Kudo's for that.

I Changed the elif statements to if statements and ended each - still no joy.

Even if I move the Disk Space Check, for example, to the top - it still gives me {no output returned from plugin}

New Code:

Code:

#!/bin/sh

PROGNAME='Check NT Perf - Results Splunk Compatible'

SCP="/usr/local/nagios/libexec/check_nt"
SED="/bin/sed"

HOST=$1
PORT=$2
VAR=$3
PARAM=$4
WARN=$5
CRIT=$6

UPTIME=`$SCP -H $HOST -p $PORT -v $VAR|$SED 's/System Uptime - //'|$SED 's/day(s)/./'|$SED 's/hour(s)/./'|$SED 's/minute(s)//'`

CPU=`$SCP -H $HOST -p $PORT -v $VAR -l $PARAM|$SED 's/CPU Load//'|$SED 's/%//'`

MEM=`$SCP -H $HOST -p $PORT -v $VAR $WARN $CRIT|$SED 's/Memory usage: total:[0-9][0-9][0-9][0-9].[0-9][0-9] Mb//'|$SED 's/ - used: //'|$SED 's/[0-9][0-9][0-9][0-9].[0-9][0-9] Mb//'|$SED 's/[0-9][0-9][0-9].[0-9][0-9] Mb//'|$SED 's/(//'|$SED 's/%)/./'|$SED 's/- free://'|$SED 's/[0-9][0-9][0-9][0-9].[0-9][0-9] Mb//'|$SED 's/([0-9][0-9]%)//'|$SED 's/. [0-9]//'|$SED 's/.//'|$SED 's/0.00;0.00;0.00;//'|$SED 's/;[0-9][0-9][0-9][0-9].[0-9][0-9]//'`

DRIVE=`$SCP -H $HOST -p $PORT -v $VAR -l $PARAM $WARN $CRIT`

							  
if [ "$VAR" = "UPTIME" ] ; then

	OUTPUT="${OUTPUT} 'System_Uptime'=$UPTIME |Format: days.hours.minutes"
	STAT=0

fi

	if [ "$VAR" = "CPULOAD" ] ; then

		OUTPUT="${OUTPUT} 'CPU_Load'=$CPU |"
		#STAT=0
	
fi

	if [ "$VAR" = "MEMUSE" ] ; then

		OUTPUT="${OUTPUT}'Memory_Used_Percent'= `echo $MEM`"
		#STAT=0

fi	

	if [ "$VAR" = "USEDDISKSPACE" ] ; then

		OPT="`echo $DRIVE > drv.tmp`"
		OUTPUT="${OUTPUT} `cat drv.tmp |$SED \"s/([0-9][0-9]%)//\"|$SED \"s/total:/'Total Space'=/\"|$SED \"s/used:/'Used Space'=/\"|$SED \"s/free/'Free Space'=/\"|$SED 's/-//'|$SED \"s/([0-9][0-9]%)//\"|$SED \"s/([0-9]%)//\"|$SED \"s/([0-9][0-9][0-9]%)//\"`"
		STAT=0
	
fi	

#		OUTPUT="${OUTPUT} ERROR - Invalid Variable Specified!"
#		STAT=2

#fi

echo $OUTPUT 
#echo $DRIVE
#echo $MEMHEAD
exit $STAT

Thanks for the advice so far.

datopdog · 04-16-2009, 08:09 AM

In cases like this you are supposed to use a switch.

monkinsane · 04-16-2009, 09:42 AM

Quote:

Originally Posted by datopdog

In cases like this you are supposed to use a switch.

What I don't get is that some of my other plugins are coded like this - and they work fine.

I think it might be something to do with the fact that the UPTIME check does not require you to specify warning or critical states. It merely displays the uptime of the server. The others have parameters you need to specify.

Anyways, gonna try and figure it out, since I'm fairly certain it is not because of the if statements, if it was none of my other plugins would work.

Thanks for your assistance.

datopdog · 04-17-2009, 01:40 AM

Quote:

Originally Posted by monkinsane

What I don't get is that some of my other plugins are coded like this - and they work fine.

I think it might be something to do with the fact that the UPTIME check does not require you to specify warning or critical states. It merely displays the uptime of the server. The others have parameters you need to specify.

Anyways, gonna try and figure it out, since I'm fairly certain it is not because of the if statements, if it was none of my other plugins would work.

Thanks for your assistance.

You are wrong, i cannot tell you why your other plugins work but from the plugin code here, the only thing it will display is the uptime as USEDISKSPACE will never get matched, come on that is clear to see.

monkinsane · 04-23-2009, 04:55 AM

Quote:

Originally Posted by datopdog

You are wrong, i cannot tell you why your other plugins work but from the plugin code here, the only thing it will display is the uptime as USEDISKSPACE will never get matched, come on that is clear to see.

USEDDISKSPACE does get matched. Command-Line The plugin works perfectly - it returns the values from the different checks flawlessly in a console - the problem is that info does not get passed to the nagios process for some odd reason.

Why don't you copy the code and run it command line - you'll see that it does match the different checks.

So I am NOT wrong, the problem is that info does not get passed to the nagios process for some odd reason.