Simple Watchdog script

 I will give an example of a simple script for restarting services if they are not running and sending a notification by email:

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

#!/bin/bash SERVICES="ssh apache2 zabbix-server zabbix-agent mysql smbd nmbd asterisk" DATE=$(date '+%d-%m-%Y %H:%M:%S')   for SERVICE in ${SERVICES}  do    service $SERVICE status 2>&1>/dev/null     if [ $? -ne 0 ];       then         service $SERVICE stop         sleep 3         service $SERVICE start         echo -e "Starting $SERVICE"         (echo "Subject:Restarting $SERVICE"; echo "$DATE $SERVICE is not running on $HOSTNAME! Restarting!";) | sendmail test@ixnfo.com       else         echo -e "$SERVICE OK"     fi done

In the script, specify the email on which you want to receive notifications, and in the “SERVICES” list of services through the space that you want to check.
I first stopped the service in order to surely stop all possible processes, paused for 3 seconds and then started, because with the restart command, some highly loaded services could not start correctly and hang with the status “active (exited)”.

To make the script automatically run for example every 10 minutes, add a line to /etc/crontab:

1	*/10 * * * * root /dir/watchdog.sh > /dev/null 2>&1