Understanding a service well enough to troubleshoot it is an important skill as a System Administrator. In this lab, we’ll go over troubleshooting an Apache installation.
Learning Objectives
Successfully complete this lab by achieving the following learning objectives:
- Ensure Apache starts correctly.
The first thing to check is why Apache isn’t starting in the first place.
Run:
systemctl start httpdIt returns:
Failed to start httpd.service: Unit not found.It looks like the junior admin didn’t think to install Apache. Let’s install it:
yum install httpd -y systemctl start httpdIt returns:
Job for httpd.service failed because the control process exited with error code. See "systemctl status httpd.service" and "journalctl -xe" for details.Looking at
systemctl status httpd -l, we see errors like:Apr 20 16:20:10 server1 httpd[1451]: (98)Address already in use: AH00072: make_sock: could not bind to address [::]:80So something is already running on port 80.
Runningss -lpshows us thatnginxis running and listening on port 80 and port 443. We need to stopnginxand starthttpd:systemctl stop nginx systemctl start httpdFinally, Apache is running.
- Ensure files are being served correctly.
If everything is configured properly, simply running
curl localhostshould get our webpage.It doesn’t, unfortunately. Now we need to configure Apache to use the different DocumentRoot we’re using here. We need to edit
/etc/httpd/conf/httpd.confand replace all instances of/var/www/htmland/var/www/with/opt/website/.
We also need to make sure SSL is enabled.
First, we need to install themod_sslpackage:yum install mod_ssl -yFortunately, there’s a sample configuration file in the
/rootdirectory. Copy that to/etc/httpd/conf.d/, and we should be good to go after restarting Apache:systemctl restart httpd curl https://localhost -k(The
-kis used to ignore alerts due to a self-signed certificate)The result doesn’t look like our index.html (you can
cat /opt/website/index.htmlto verify).Looking at
/var/log/httpd/error_log, it looks like a permission issue. Using justls -l /opt/website, everything looks fine – but maybe it’s SELinux. Check with:# ls -lZ /opt/website/ -rw-r--r--. root root system_u:object_r:admin_home_t:s0 index.html drwxr-xr-x. root root system_u:object_r:admin_home_t:s0 storageThat’s it.
admin_home_tis for admin home files (likeroot‘s home directory) Compare that to the default/var/wwwdirectory:# ls -lZ /var/www drwxr-xr-x. root root system_u:object_r:httpd_sys_script_exec_t:s0 cgi-bin drwxr-xr-x. root root system_u:object_r:httpd_sys_content_t:s0 htmlSo all we need to do is:
chcon -R system_u:object_r:httpd_sys_content_t:s0 /opt/website/*And our
curlshould work fine. Good job!
