In production sites, we normally need to ensure that everything must go fine. However, in many cases, some server services are terminated which causes your website broken/died. For example, sometimes hhvm goes down in my production server which cause my commercial system stop working. This is hurt and it affects directly to my pocket :-(. So, I must think a way to monitor a service periodly and restart it if it is broken. In this short tutorial, I will guide how to write a bash script to check if a service is running.
First, create the following bash script, name it script_check_hhvm.sh for example:
case "$(pidof hhvm | wc -w)" in
0) echo "hhvm not running, restarting oscam1: $(date)" >> $logfile
service hhvm start
*) echo "multiple instances of hhvm running. Stopping & restarting hhvm: $(date)" >> $logfile
service hhvm restart
In the above script, if hhvm is stopped, there is no hhvm process and we will start it by service hhvm start command. When there is more than 1 hhvm process, it means that there is some problem in our system and we will restart it.
Then, setup it as a cron run every 2 minutes via crontab:
*/2 * * * * sh /root/script_check_hhvm.sh[/bash]
It’s all! You can easily replace the command actions for your cases with the above script and enjoy it 🙂