How to restart a process out of crontab on a Linux/Unix
I
am a new Linux user and not familiar with crontab syntax. How do I use a cron job to restart server process if not running on a Linux or Unix-like system?
There are many ways to make sure that the essential server process such as HTTPD/Nginx/PHP-FPM/MySQL remains up and running all the time.
This tutorial shows how to create and use a simple bash shell script to restart failed services using crontab.
Step 1 – Create a Bash Shell Script
Here is a simple shell script to check for httpd pid using the pgrep command. It will restart if httpd is not running on a CentOS/RHEL using the systemctl command. Create a file named /root/bin/restart-httpd
# vi /root/bin/restart-httpd
Append the following code:
#!/bin/bash
# Apache Process Monitor
# Restart Apache Web Server When It Goes Down
# RHEL / CentOS 7.x httpd restart command
RESTART= /bin/systemctl restart httpd
#Path to pgrep command
PGREP= /usr/bin/pgrep
# Httpd daemon name: Under RHEL/CentOS/Fedora it is httpd
HTTPD= httpd
# Find httpd pid
$PGREP ${HTTPD}
if [ $? -ne 0 ] # if apache not running
then
# restart apache
$RESTART
fi
Save and close the file. Set permissions:
# chmod +x /root/bin/restart-httpd
# /root/bin/restart-httpd
Step 2 – Setup the cron job to restart process
Cron allows Linux and Unix users to run commands or scripts at a given date and time. You can schedule scripts to be executed periodically. Type the following command:
# crontab -e
Append the following code to restart httpd
restart httpd if not running. check for httpd every 5 mins
*/5 * * * * /root/bin/restart-httpd >/dev/null 2>&1
Save and close the file.
A note about Linux and systemd based system
Systemd can handle keeping the service alive for you. You need to use the cp command:
# cp /lib/systemd/system/httpd.service /etc/systemd/system/httpd.service
Edit the file using vi command:
# vi /etc/systemd/system/httpd.service
Add Restart=always and RestartSec=5 in the [Service] section as follows:
[Unit]
Description=The Apache HTTP Server
After=network.target remote-fs.target nss-lookup.target
Documentation=man:httpd(8)
Documentation=man:apachectl(8)
[Service]
########################################
Make sure httpd always restart if failed ##
Configures the time to sleep before ##
restarting a service with 5 seconds too ##
########################################
Restart=always
RestartSec=5
########################################
Type=notify
EnvironmentFile=/etc/sysconfig/httpd
ExecStart=/usr/sbin/httpd $OPTIONS -DFOREGROUND
ExecReload=/usr/sbin/httpd $OPTIONS -k graceful
ExecStop=/bin/kill -WINCH ${MAINPID}
# We want systemd to give httpd some time to finish gracefully, but still want
# it to kill httpd after TimeoutStopSec if something went wrong during the
# graceful stop. Normally, Systemd sends SIGTERM signal right after the
# ExecStop, which would kill httpd. We are sending useless SIGCONT here to give
# httpd time to finish.
KillSignal=SIGCONT
PrivateTmp=true
[Install]
WantedBy=multi-user.target
Reload configuration:
# systemctl daemon-reload
# systemctl restart httpd
Test it by killing httpd using the kill command:
# kill -9 $(pgrep httpd)
After 5 seconds httpd will restart again. Verify with the ps command:
# ps aux | grep httpd
Sample outputs:
root 8052 0.0 0.0 221960 7580 ? Ss 13:46 0:00 /usr/sbin/httpd -DFOREGROUND
apache 8053 0.0 0.0 221960 6096 ? S 13:46 0:00 /usr/sbin/httpd -DFOREGROUND
apache 8054 0.0 0.0 221960 6096 ? S 13:46 0:00 /usr/sbin/httpd -DFOREGROUND
apache 8055 0.0 0.0 221960 6096 ? S 13:46 0:00 /usr/sbin/httpd -DFOREGROUND
apache 8056 0.0 0.0 221960 6096 ? S 13:46 0:00 /usr/sbin/httpd -DFOREGROUND
apache 8057 0.0 0.0 221960 6096 ? S 13:46 0:00 /usr/sbin/httpd -DFOREGROUND
root 8061 0.0 0.0 112644 1720 ? S+ 13:46 0:00 grep –color=auto httpd
Conclusion
This is easy to setup and uses method to make sure an app/service remain online if killed due to some issues. A better and more reliable solution is monit monitoring software for restarting services such as mysql, apache and sendmail under a UNIX / Linux operating systems:
Monit is a free and open source software that acts as process supervision. It comes with the ability to restart services which have failed. You can use Systemd, daemontools or any other such tool for the same purpose.
