This article applies as of PRTG 22

How to monitor, automatically restart, and report the current status of a Linux service

To monitor a service, restart it automatically, and push the current status to PRTG with HTTP, use the script below.

• The script basically checks if the desired service is available or not.
• If the service is up, the script sends a report to the PRTG core server with the value "1".
• Value "1" means that the service is up and value "0" means that the service is down or could not properly restart.
• The script also sends sensor messages like $service is running, $service is down, $service restarted properly, and $service can't restart properly.

How to use the script

Save the following script on the target system, where it can be executed by CRON.

#!/bin/bash
#    ____  ____  ____________
#   / __ \/ __ \/_  __/ ____/
#  / /_/ / /_/ / / / / / __  
# / ____/ _, _/ / / / /_/ /  
#/_/   /_/ |_| /_/  \____/                         
#    NETWORK MONITOR
#-------------------
#(c) 2016 Dariusz Gorka, Paessler AG
#
#This script checks if a certain service is running.
#The script also tries to restart the service if it is not started.
#

#Enter the correct process name. (ps -e)
service=$1
#Enter the server address of your PRTG, including HTTPS/HTTP and the sensor port.
prtghost=$2
#Enter the Identification Token of the HTTP Push Data Advanced sensor.
identtoken=$3

#Check if process is running
if (( $(pgrep -x $service | wc -l) > 0 ))
  then
    #Send response to PRTG that the service is running.
    wget -O/dev/null "$prtghost/$identtoken?content=<prtg><result><channel>$service status</channel><value>1</value></result><text>Service: $service is running!</text></prtg>"
  else
    #Send response to PRTG that the service is not started.
    wget -O/dev/null "$prtghost/$identtoken?content=<prtg><result><channel>$service status</channel><value>0</value></result><text>Service: $service is down, but will restart!</text></prtg>"
    #Try to restart the service
    /etc/init.d/$service start

    #Check if restart was successfully
    if (( $(ps -ef | grep -v grep | grep $service | wc -l) > 0 ))
      then
        #Send response to PRTG that the restart was successfully
        wget -O/dev/null "$prtghost/$identtoken?content=<prtg><result><channel>$service status</channel><value>1</value></result><text>Service: $service restarted properly!</text></prtg>"
      else
        #Send response to PRTG that the restart was not successful
        wget -O/dev/null "$prtghost/$identtoken?content=<prtg><result><channel>$service status</channel><value>0</value></result><text>Service: $service can't restart properly! Please take action!</text></prtg>"
      fi
fi

Create the CRONTAB below on your Linux server to use the script. To open the CRONTAB file, enter the command crontab -e as root and paste the code into the last line. */5 * * * * /PATH/TO/THESCRIPT/script.sh <SERVICENAME> <PRTGHOST> <IDENT-TOKEN>

Note: Keep in mind that you need to adjust the path and the parameters of the crontab, like in the examples below:

• <SERVICENAME>: snmpd
• <PRTGHOST>: https://example.com:5050 (use the Port-Host combination of the configured sensor here)
• <IDENT-TOKEN>: prtg-is-great

The response will be sent to the HTTP Push Data Advanced sensor. Create this sensor with the following settings:

• Sensor Name: <NAME OF THE SERVICE>
• Port: 5050 (Use a different port or Identification Token for every sensor.)
• Identification Token: <YOUR-IDENT-TOKEN>
• You can leave the default settings for all other options.

Set a limit on the sensor's $service status channel. To set this limit, select Enable alerting based on limits in the channel's settings and enter "1" in the Lower Error Limit (#) field.