Monitoring service status on linux server

[deek]

Hello All,

I need to monitor a service memcached-monitor.service and memcached.service which is under the path /opt/airwatch/memcached . So when i do systemctl status xxx the output shows active and running .
Is there a way to monitor this in nagios . Alert should be triggered when the service is not running .

Capture_memchached.PNG

Capture_memchched_1.PNG

[dchurch]

You could use NCPA to monitor the service is running. NCPA has a built-in check for a whether a named service is running.

1. Install NCPA on the target machine. NCPA is a FOSS program we maintain: https://github.com/NagiosEnterprises/ncpa
2. Pick a password, open ports, etc to make the check work.
3. Set up a check like so inside nagios XI

Code: Select all

./check_ncpa.py -H <host ip or FQDN> -t Str0ngT0k3n -M services -q service=memcached,status=running

More documentation here: https://support.nagios.com/kb/article/s ... ce_started

[deek]

Thats great .
But is there any other way .

[dchurch]

Sure, you could use NSCA to send a passive check to the Nagios server, indicating whether the memcached service is running.

The difference between an active check and a passive check is an active check is a check that the Nagios monitoring engine initiates, and a passive one is one that's initiated on the remote server and sent to the listening monitoring engine.

Here's also a step-by-step guide to setting up NSCA: http://nagios.sourceforge.net/download/ ... _Setup.pdf

Note that if you're only sending a passive check, you can skip the parts about using xinetd and probably don't have to configure your firewall to allow any incoming ports.

[deek]

Thanks for that .

Can we use the below script ? Is this script proper

#!/usr/bin/env bash

# Author: Jon Schipp
# 2015-03-09 [Pascal Hegy] - Add sudo for linux
# 2015-03-09 [Pascal Hegy] - Change USER variable to USERNAME to avoid the use and confusion with the USER env variable
# 2017-08-30 [Roberto Leibman] - Reordered checks to make sure dead and inactive get checked first
# 2018-04-25 [Robin Gierse] - Update check via systemctl for Linux with grep to produce better output for systemctl
# 2019-03-15 [nem / liberodark] - Add support for check all failed services in linux

########
# Examples:

# 1.) List services for osx
# $ ./check_service.sh -l -o osx
#
# 2.) Check status of SSH service on a linux machine
# $ ./check_service.sh -o linux -s sshd

# 3.) Manually select service management tool and service
# $ ./check_service.sh -o linux -t "service rsyslog status"
# Exemple for check all failed services
# $ ./check_service.sh -o linux -t "systemctl list-units --state=failed"

# Nagios Exit Codes
OK=0
WARNING=1
CRITICAL=2
UNKNOWN=3

# Weather or not we can trust the exit code from the service management tool.
# Defaults to 0, put to 1 for systemd. Otherwise we must rely on parsing the
# output from the service management tool.
TRUST_EXIT_CODE=0

usage()
{
cat <<EOF
Check status of system services for Linux, FreeBSD, OSX, and AIX.
Options:
-s <service> Specify service name
-l List services
-o <os> OS type, "linux/osx/freebsd/aix"
-u <user> User if you need to ``sudo -u'' for launchctl (def: nagios, linux and osx only)
-t <tool> Manually specify service management tool (def: autodetect) with status and service
e.g. ``-t "service nagios status"''
EOF
}

argcheck() {
# if less than n argument
if [ $ARGC -lt $1 ]; then
echo "Missing arguments! Use \`\`-h'' for help."
exit 1
fi
}

os_check() {
if [ "$OS" == null ]; then
unamestr=$(uname)
if [[ $unamestr == 'Linux' ]]; then
OS='linux'
elif [[ $unamestr == 'FreeBSD' ]]; then
OS='freebsd'
elif [[ $unamestr == 'Darwin' ]]; then
OS='osx'
else
echo "OS not recognized, Use \`-o\` and specify the OS as an argument"
exit 3
fi
fi
}



determine_service_tool() {
if [[ $OS == linux ]]; then
if command -v systemctl >/dev/null 2>&1; then
SERVICETOOL="systemctl status $SERVICE | grep -i Active"
LISTTOOL="systemctl"
if [ $USERNAME ]; then
SERVICETOOL="sudo -u $USERNAME systemctl status $SERVICE"
LISTTOOL="sudo -u $USERNAME systemctl"
fi
TRUST_EXIT_CODE=1
elif command -v service >/dev/null 2>&1; then
SERVICETOOL="service $SERVICE status"
LISTTOOL="service --status-all"
if [ $USERNAME ]; then
SERVICETOOL="sudo -u $USERNAME service $SERVICE status"
LISTTOOL="sudo -u $USERNAME service --status-all"
fi
elif command -v initctl >/dev/null 2>&1; then
SERVICETOOL="status $SERVICE"
LISTTOOL="initctl list"
if [ $USERNAME ]; then
SERVICETOOL="sudo -u $USERNAME status $SERVICE"
LISTTOOL="sudo -u $USERNAME initctl list"
fi
elif command -v chkconfig >/dev/null 2>&1; then
SERVICETOOL=chkconfig
LISTTOOL="chkconfig --list"
if [ $USERNAME ]; then
SERVICETOOL="sudo -u $USERNAME chkconfig"
LISTTOOL="sudo -u $USERNAME chkconfig --list"
fi
elif [ -f /etc/init.d/$SERVICE ] || [ -d /etc/init.d ]; then
SERVICETOOL="/etc/init.d/$SERVICE status | tail -1"
LISTTOOL="ls -1 /etc/init.d/"
if [ $USERNAME ]; then
SERVICETOOL="sudo -u $USERNAME /etc/init.d/$SERVICE status | tail -1"
LISTTOOL="sudo -u $USERNAME ls -1 /etc/init.d/"
fi
else
echo "Unable to determine the system's service tool!"
exit 1
fi
fi

if [[ $OS == freebsd ]]; then
if command -v service >/dev/null 2>&1; then
SERVICETOOL="service $SERVICE status"
LISTTOOL="service -l"
elif [ -f /etc/rc.d/$SERVICE ] || [ -d /etc/rc.d ]; then
SERVICETOOL="/etc/rc.d/$SERVICE status"
LISTTOOL="ls -1 /etc/rc.d/"
else
echo "Unable to determine the system's service tool!"
exit 1
fi
fi

if [[ $OS == osx ]]; then
if [ -f /usr/sbin/serveradmin >/dev/null 2>&1 ] && serveradmin list | grep "$SERVICE" 2>&1 >/dev/null; then
SERVICETOOL="serveradmin status $SERVICE"
LISTTOOL="serveradmin list"
elif [ -f /Applications/Server.app/Contents/ServerRoot/usr/sbin/serveradmin >/dev/null 2>&1 ] && \
/Applications/Server.app/Contents/ServerRoot/usr/sbin/serveradmin list | \
grep "$SERVICE" 2>&1 >/dev/null; then
SERVICETOOL="/Applications/Server.app/Contents/ServerRoot/usr/sbin/serveradmin status $SERVICE"
LISTTOOL="/Applications/Server.app/Contents/ServerRoot/usr/sbin/serveradmin list"
elif command -v launchctl >/dev/null 2>&1; then
SERVICETOOL="launchctl list | grep -v ^- | grep $SERVICE || echo $SERVICE not running! "
LISTTOOL="launchctl list"
if [ $USERNAME ]; then
SERVICETOOL="sudo -u $USERNAME launchctl list | grep -v ^- | grep $SERVICE || echo $SERVICE not running! "
LISTTOOL="sudo -u $USERNAME launchctl list"
fi
elif command -v service >/dev/null 2>&1; then
SERVICETOOL="service --test-if-configured-on $SERVICE"
LISTTOOL="service list"
else
echo "Unable to determine the system's service tool!"
exit 1
fi
fi

if [[ $OS == aix ]]; then
if command -v lssrc >/dev/null 2>&1; then
SERVICETOOL="lssrc -s $SERVICE | grep -v Subsystem"
LISTTOOL="lssrc -a"
else
echo "Unable to determine the system's service tool!"
exit 1
fi
fi
}

ARGC=$#
LIST=0
MANUAL=0
OS=null
SERVICETOOL=null
LISTTOOL=null
SERVICE=".*"
#USERNAME=nagios

argcheck 1

while getopts "hlst:u:" OPTION
do
case $OPTION in
h)
usage
exit 0
;;
l)
LIST=1
;;
s)
SERVICE="$OPTARG"
;;
o)
if [[ "$OPTARG" == linux ]]; then
OS="$OPTARG"
elif [[ "$OPTARG" == osx ]]; then
OS="$OPTARG"
elif [[ "$OPTARG" == freebsd ]]; then
OS="$OPTARG"
elif [[ "$OPTARG" == aix ]]; then
OS="$OPTARG"
else
echo "Unknown type!"
exit 1
fi
;;
t)
MANUAL=1
MANUALSERVICETOOL="$OPTARG"
;;
u)
USERNAME="$OPTARG"
;;
\?)
exit 1
;;
esac
done

os_check

if [ $MANUAL -eq 1 ]; then
SERVICETOOL=$MANUALSERVICETOOL
else
determine_service_tool
fi

# -l conflicts with -t
if [ $MANUAL -eq 1 ] && [ $LIST -eq 1 ]; then
echo "Options conflict: \`\`-t'' and \`\`-l''"
exit 2
fi

if [ $LIST -eq 1 ]; then
if [[ $LISTTOOL != null ]]; then
$LISTTOOL
exit 0
else
echo "OS not specified! Use \`\`-o''"
exit 2
fi
fi

# Check the status of a service
STATUS_MSG=$(eval "$SERVICETOOL" 2>&1)
EXIT_CODE=$?

## Exit code from the service tool - if it's non-zero, we should
## probably return CRITICAL. (though, in some cases UNKNOWN would
## probably be more appropriate)
[ $EXIT_CODE -ne 0 ] && echo "$STATUS_MSG" && exit $CRITICAL

## For systemd and most systems, $EXIT_CODE can be trusted - if it's 0, the service is running.
## Ref https://github.com/jonschipp/nagios-plugins/issues/15
[ $TRUST_EXIT_CODE -eq 1 ] && [ $EXIT_CODE -eq 0 ] && echo "$STATUS_MSG" && exit $OK

case $STATUS_MSG in

*stop*)
echo "$STATUS_MSG"
exit $CRITICAL
;;
*STOPPED*)
echo "$STATUS_MSG"
exit $CRITICAL
;;
*not*running*)
echo "$STATUS_MSG"
exit $CRITICAL
;;
*NOT*running*)
echo "$STATUS_MSG"
exit $CRITICAL
;;
*NOT*RUNNING*)
echo "$STATUS_MSG"
exit $CRITICAL
;;
#*inactive*)
# echo "$STATUS_MSG"
# exit $CRITICAL
# ;;
*dead*)
echo "$STATUS_MSG"
exit $CRITICAL
;;
*running*)
echo "$STATUS_MSG"
exit $OK
;;
*RUNNING*)
echo "$STATUS_MSG"
exit $OK
;;
*SUCCESS*)
echo "$STATUS_MSG"
exit $OK
;;
*[eE]rr*)
echo "Error in command: $STATUS_MSG"
exit $CRITICAL
;;
*[fF]ailed*)
echo "$STATUS_MSG"
exit $CRITICAL
;;
*[eE]nable*)
echo "$STATUS_MSG"
exit $OK
;;
*[dD]isable*)
echo "$STATUS_MSG"
exit $CRITICAL
;;
*[cC]annot*)
echo "$STATUS_MSG"
exit $CRITICAL
;;
*[aA]ctive*)
echo "$STATUS_MSG"
exit $OK
;;
*Subsystem*not*on*file)
echo "$STATUS_MSG"
exit $CRITICAL
;;
[1-9][1-9]*)
echo "$SERVICE running: $STATUS_MSG"
exit $OK
;;
"")
echo "$SERVICE is not running: no output from service command"
exit $CRITICAL
;;
*)
echo "Unknown status: $STATUS_MSG"
echo "Is there a typo in the command or service configuration?: $STATUS_MSG"
exit $UNKNOWN
;;
*0\ loaded*)
echo "$STATUS_MSG"
exit $OK
;;
esac

[dchurch]

Sure, it looks fine. Let me know if you run into any issues.

[deek]

You can close the ticket

Thank you so much

[scottwilkerson]

You can close the ticket

Thank you so much

Locking thread