Here is a simple shell script tested on CentOS / RHEL / Fedora / Debian / Ubuntu Linux. Should work under any other UNIX liker operating system. It will check for httpd pid using pgrep command
pgrep command
pgrep looks through the currently running processes and lists the process IDs which matches the selection criteria to screen. If no process found it will simply return exit status 0 (zero).
Download the script and set cronjob as follows:
*/5 * * * * /path/to/script.sh >/dev/null 2>&1
Sample script
#!/bin/bash # Apache Process Monitor # Restart Apache Web Server When It Goes Down # ------------------------------------------------------------------------- # Copyright (c) 2003 nixCraft project <http://cyberciti.biz/fb/> # This script is licensed under GNU GPL version 2.0 or above # ------------------------------------------------------------------------- # This script is part of nixCraft shell script collection (NSSC) # Visit http://bash.cyberciti.biz/ for more information. # ------------------------------------------------------------------------- # RHEL / CentOS / Fedora Linux restart command RESTART="/sbin/service httpd restart" # uncomment if you are using Debian / Ubuntu Linux #RESTART="/etc/init.d/apache2 restart" #path to pgrep command PGREP="/usr/bin/pgrep" # Httpd daemon name, # Under RHEL/CentOS/Fedora it is httpd # Under Debian 4.x it is apache2 HTTPD="httpd" # find httpd pid $PGREP ${HTTPD} if [ $? -ne 0 ] # if apache not running then # restart apache $RESTART fi
A better and more reliable solution is monit monitoring software for restarting services such as mysql, apache and sendmail under UNIX / Linux operating systems.
🐧 Get the latest tutorials on SysAdmin, Linux/Unix, Open Source, and DevOps topics via:
- RSS feed or Weekly email newsletter
- 27 comments... add one ↓
Related Shell Scripts:
| Category | List of Unix and Linux commands |
|---|---|
| Disk space analyzers | ncdu • pydf |
| File Management | cat |
| Firewall | Alpine Awall • CentOS 8 • OpenSUSE • RHEL 8 • Ubuntu 16.04 • Ubuntu 18.04 • Ubuntu 20.04 |
| Network Utilities | NetHogs • dig • host • ip • nmap |
| OpenVPN | CentOS 7 • CentOS 8 • Debian 10 • Debian 8/9 • Ubuntu 18.04 • Ubuntu 20.04 |
| Package Manager | apk • apt |
| Processes Management | bg • chroot • cron • disown • fg • jobs • killall • kill • pidof • pstree • pwdx • time |
| Searching | grep • whereis • which |
| User Information | groups • id • lastcomm • last • lid/libuser-lid • logname • members • users • whoami • who • w |
| WireGuard VPN | Alpine • CentOS 8 • Debian 10 • Firewall • Ubuntu 20.04 |
Why won’t you just use ‘monit’?
I am not a big fan of greping httpd in process list to see if httpd server is running or not. There can be several httpd services installed. Two options are either the status module, or this bash script to the extent that you known the full path to your httpd bin folder:
function isapup(){
pid=$($1/bin/apachectl -S | grep PidFile | awk ‘{print $2}’) # retrieve the expected path of the process id file
pid=${pid:1:`expr ${#pid} – 2`} # get rid of the quotes
up=$(ls $pid 2> /dev/null) # see if process id file exists
if [ ! “$up” ]
then
return 1
fi
}
isapup /opt/httpd/2.4.0
if [ $? -eq 1 ] # if isapup returned 1
then
echo Starting Apache
/opt/httpd/2.4.0/bin/apachectl start
fi
Had to use PGREP=”/usr/bin/pgrep -l -x” to get it to work on CentOs (5.4)
If adapting this for other processes add an -x flag to force pgrep to match the process name exactly
$PGREP ${HTTPD}
The name of my script contains the name of the the process to pgrep was giving a false positive when the process wasn’t running.
You made mistake in crontab entry. It will not work.
You should specify user which will execute the script:
This crontab is correct (CentOS):
*/5 * * * * root /path/to/script.sh >/dev/null 2>&1
I’m looking to modify this to tie into the screensaver and end the session running/ logout current user for a kiosk installation. In other words, how could this script be modified to check for screensaver initiating and then logout/restart the session instead of restarting apache?
Thanks in advance,
Taylor G.
I put this script to github so everybody may clone and change it.
http://github.com/andreashe/keepapacherunning
Tell me if this is not ok and I will remove it.
Maybe someone want to build in some autodetections for differences in distries.
It is a GPLed script, so I don’t see any problem as long as anyone follows GPL.
I saved the script to httpdmonitor.sh and then when it gets called from the script pgrep was not doing a whole word search and returned trued even thought httpd was turned off. This is on an older redhat system. change pgrep to be ‘pgrep -l -x’. Now it appears to work well and ignores httpdmonitor.sh
FYI for anyone else who runs into this snag.
Regards
I got a problem, sometimes my httpd stop but the PID still alive, and i need to restart manually the httpd, how can i fix this?? When it happens, i run the httpdmonitor and appears only one PID, so i need to restart apache, then i run the httpdmonitor again, and i can see many PID from apache.
Please help me, i´m needing to restart the apache almost 6 times a day.
Thanks
Try ed it for checking deamon of one custom script and works like a charm.Those who says how they get command not found i guess they don’t have command PGREP installed on their OS or they set wrong PATH.
i’m getting
“[0: command not found”
for the
if [ $? -ne 0 ] # if apache not running
any ideas?
I make run this script and wen i stop my apache, this $PGREP ${HTTPD}
still geting cero and never restart the apache service, there are another way to see if apache is running?
./root/monitoringapache.sh: line 5: /sbin/service httpd restart: No such file or directory
pgrep: No matching criteria specified
Usage: pgrep [-flvx] [-d DELIM] [-n|-o] [-P PPIDLIST] [-g PGRPLIST] [-s SIDLIST]
[-u EUIDLIST] [-U UIDLIST] [-G GIDLIST] [-t TERMLIST] [PATTERN]
./root/monitoringapache.sh: line 14: $: command not found
./root/monitoringapache.sh: line 16: [127: command not found
line 14 is
$PGREP $ { HTTPD }
line 16 is
if [ $? -ne 0]
I have this troubles im working on redhat please help me
@shuji
Follow this:
vi /usr/bin/httpdmonitor
copy the script in to that, after that chmod 777 /usr/bin/httpdmonitor
Now set corntab as:
crontab -e
*/5 * * * * /usr/bin/httpdmonitor >/dev/null 2>&1
This is good script.
Thanks Vivek-
Shrii
Thanks-
hi all,
i wan to ask, i already run the command ” $ crontab -e ” at the terminal.
so is it i copy the whole sample scripts into the crontab file and then save it?
that all??
for your information im newbie in unix.
Please help. :(
regards
Shuji
Nice script, works great.
I just found another good monitoring script posted on this blog: Link
The script tests the webserver’s functionality by downloading a file.
Pasci many thx , good job, easy
and also thx to NixCraft for this shell script
—
# find httpd pid
$PGREP ${HTTPD}
if [ $? -ne 0 ] # if apache not running
then
# restart apache
$RESTART
fi
—
can be done in one line..
$PGREP ${HTTPD} || $RESTART
Regards pasci
Ben,
A small bug is fixed and script should work now.
what bug was fixed? is the orginal source updated?
PGREP httpd returns nothing when httpd is not running, and I think that is causing the check to restart it to fail… Is there any way to test not that it returns 0 but empty or null?
Hubert,
Thanks for the heads up. The script has been updated to ‘/sbin/service’ from ‘service’.
In CentOs 5 You should use /etc/init.d/httpd restart.
Crontab would’t execute service httpd restart
regards
Hubert
A minor change is to be done in the above script,
The “echo $?” statement will give the exit status of the last command. In this case, if httpd is not running, the “pgrep httpd” exit status will be “non zero”.
So, it should be
if [ $? -ne 0 ] #instead of -eq
then
# restart apache
$RESTART
fi