Create a new file called process.txt in the /root/scripts folder (you might need to also make the /root/scripts folder)
Get a list of the processes you want to monitor from the ps -ef output
and add them to this file.
The case statement in the middle is what is used to restart them, so you need to change these to suite the specific processes you are monitoring and wanting to restart.
Add an entry to the crontab
This will attempt to restart the process if it's down. It won't tell you why it's down, and if it won't restart, you will get notified every 1 minute until you can figure why it's down and broken.
Get a list of the processes you want to monitor from the ps -ef output
Bash:
/usr/sbin/mysqld
/usr/bin/memcached
/usr/libexec/postfix/master
/bin/java
Bash:
#!/bin/bash
echo "Checking all processes are running....";
for i in `/bin/cat /root/scripts/process.txt`
do
if
[ $(ps aux | grep -e "$i" | grep -v grep | wc -l | tr -s "\n") -eq 0 ];
then
echo -e "$i is down...attempting restart" | mail -s "$i is down, restarting" "MYEMAILADDRESS"
case "$i" in
*java* ) /usr/bin/systemctl restart elasticsearch.service;;
*memcached* ) /usr/bin/systemctl restart memcached.service;;
*mysqld* ) /usr/bin/systemctl restart mariadb.service;;
*postfix* ) /usr/bin/systemctl restart postfix.service;;
esac
else
printf "$i running\n";
fi
doneThe case statement in the middle is what is used to restart them, so you need to change these to suite the specific processes you are monitoring and wanting to restart.
Add an entry to the crontab
Code:
*/1 * * * * root /root/scripts/check.sh > /dev/null 2> /dev/nullThis will attempt to restart the process if it's down. It won't tell you why it's down, and if it won't restart, you will get notified every 1 minute until you can figure why it's down and broken.
