Nagios - Status UNKNOWN for bash script

Question

I am trying to monitor my EC2 loadblancer through Nagios using a bash script. Below is the script which I am trying to implement with Nagios.

#!/bin/sh

ST_OK=0
ST_WR=1
ST_CR=2
ST_UK=3


LB_NAME="xxx"
AWS_REGION="us-west-2"
PROFILE="default"


CMD=$(/usr/bin/aws elb describe-instance-health --region ${AWS_REGION} --load-balancer-name ${LB_NAME} --profile ${PROFILE})

if [ $? -eq 0 ]; then

    IN_SERVICE_COUNT=$(echo ${CMD} | jq -c '.InstanceStates[].State' | grep InService |wc -l)
    TOTAL_COUNT=$(echo ${CMD} | jq -c '.InstanceStates[].State' | wc -l)

    if [ ${IN_SERVICE_COUNT} -eq 0 ]; then
                    NAGIOS_STATE=CRITICAL
                    EXIT_CODE=$ST_CR
    elif [ ${TOTAL_COUNT} -eq ${IN_SERVICE_COUNT} ]; then
                    NAGIOS_STATE=OK
                    EXIT_CODE=$ST_OK
    elif [ ${IN_SERVICE_COUNT} -lt ${TOTAL_COUNT} ]; then
                    NAGIOS_STATE=WARNING
                    EXIT_CODE=$ST_WR
    fi
    echo "${NAGIOS_STATE}: ELB:${LB_NAME} is running fine. Total #instances:${TOTAL_COUNT} Healthy instances:${IN_SERVICE_COUNT}"
    else
    echo "Failed to retrieve ELB Instances health from AWS"
    EXIT_CODE=$ST_UK
fi
exit ${EXIT_CODE}

The above script is working fine for me while running manually. Also I have ran it with nagios user and I am able to get the result like below:

OK: ELB:xxx is running fine Total:18 Healthy:18

So, I don't think any permission issue. I have configured AWS credentials for nagios user. But in the nagios interface I am always getting status "UNKNOWN".

Below is the code for command.cfg

define command {
    command_name    check_elb_status
    command_line    /usr/local/nagios/libexec/check_elb_status.sh
}

Below is the code for host file:

define service{
    use generic-service
    host_name Prod-ELB
    service_description Prod ELB Status
    check_command check_elb_status
}

The same script I have used with NRPE from a different host and I am able to get the result:

Code for nrpe.cfg

command[check_elb_sts]=/usr/local/nagios/libexec/check_elb_status.sh

Code for host file

define service{
   use generic-service
   host_name xxx
   service_description Prod ELB Status
   check_command check_nrpe!check_elb_sts
}

Don't know why the script is not able to give result while using on Nagios host. Please help to resolve the issue.

Answer 1

Somehow the bash script was not able to locate the config file for AWS cli. Now I have configured the AWS_CONFIG_FILE location in the bash script and it's working fine.