Custom HTTPD configuration not working on Apache/Nginx combination

[ericovk]

We are running a Apache/Nginx combination pretty much out of the box. When I try to install Magento 2, I have to rename the public_html folder to pub. Therefore I added this rule to the "Custom4" field in HTTPD customizations at the admin level.

DocumentRoot /home/username/domains/domain.com/pub

DA says the syntax is okay, but on restart of Nginx, it won't start due to a misconfiguration.

● nginx.service - The nginx HTTP and reverse proxy server
   Loaded: loaded (/etc/systemd/system/nginx.service; enabled; vendor preset: disabled)
   Active: failed (Result: exit-code) since Thu 2017-04-20 14:06:29 CEST; 4s ago
  Process: 2004 ExecStartPre=/usr/sbin/nginx -t -c /etc/nginx/nginx.conf (code=exited, status=1/FAILURE)


Apr 20 14:06:29 servername systemd[1]: Starting The nginx HTTP and reverse proxy server...
Apr 20 14:06:29 servername nginx[2004]: nginx: [emerg] unexpected "}" in /usr/local/directadmin/data/users/username/nginx.conf:39
Apr 20 14:06:29 servername nginx[2004]: nginx: configuration file /etc/nginx/nginx.conf test failed
Apr 20 14:06:29 servername systemd[1]: nginx.service: control process exited, code=exited status=1
Apr 20 14:06:29 servername systemd[1]: Failed to start The nginx HTTP and reverse proxy server.
Apr 20 14:06:29 servername systemd[1]: Unit nginx.service entered failed state.
Apr 20 14:06:29 servername systemd[1]: nginx.service failed.

After commenting the DocumentRoot rules manually, Nginx starts, but of course the changes I made are gone.

How can I set up this DocumentRoot folder properly at /pub in stead of /public_html?

 

[ericovk]

Bump, I haven't figured it out myself.

 

[zEitEr]

Hello,

Add into CUSTOM or CUSTOM1 for nginx and apache:

|?DOCROOT=`HOME`/domains/`DOMAIN`/public_html/sub|

Related: https://help.directadmin.com/item.php?id=2

 

[ericovk]

Thanks, but it isn't working on this server. It still loads the public_html folder in stead.

I tried it with a hardcoded path, but it still lacks working.

|?DOCROOT=/home/mag2/domains/mag2.domain.com/pub|

edit: I tried this on my other server with Apache only, and over there it worked as expected. But only on the Apache / Nginx combo the server just isn't willing to load another folder.

 

[zEitEr]

Are you sure that you added the line into both NGINX and Apache?
Did you make sure that NGINX and Apache restarted?

killall -9 nginx
killall -9 httpd

and then start nginx/apache.

How are your configs in nginx and apache for the domain looking like?

 

[ericovk]

The only thing I added was your piece of code at custom1

|?DOCROOT=`HOME`/domains/`DOMAIN`/public_html/sub|

I can't find any trails for an error for this strange behaviour. Searched my error logs, but I only found this:

tail /var/log/httpd/error_log
[Tue Apr 25 07:53:30.011882 2017] [mpm_event:notice] [pid 22837:tid 140257394391168] AH00492: caught SIGWINCH, shutting down gracefully
[Tue Apr 25 07:55:14.248289 2017] [ssl:warn] [pid 1200:tid 140714591922304] AH01909: localhost:443:0 server certificate does NOT include an ID which matches the server name
[Tue Apr 25 07:55:14.249590 2017] [ssl:warn] [pid 1200:tid 140714591922304] AH01909: www.example.com:8081:0 server certificate does NOT include an ID which matches the server name
[Tue Apr 25 07:55:14.249752 2017] [suexec:notice] [pid 1200:tid 140714591922304] AH01232: suEXEC mechanism enabled (wrapper: /usr/sbin/suexec)
[Tue Apr 25 07:55:14.284503 2017] [auth_digest:notice] [pid 1200:tid 140714591922304] AH01757: generating secret for digest authentication ...
[Tue Apr 25 07:55:15.016453 2017] [ssl:warn] [pid 1200:tid 140714591922304] AH01909: localhost:443:0 server certificate does NOT include an ID which matches the server name
[Tue Apr 25 07:55:15.017131 2017] [ssl:warn] [pid 1200:tid 140714591922304] AH01909: www.example.com:8081:0 server certificate does NOT include an ID which matches the server name
[Tue Apr 25 07:55:15.017293 2017] [lbmethod_heartbeat:notice] [pid 1200:tid 140714591922304] AH02282: No slotmem from mod_heartmonitor
[Tue Apr 25 07:55:15.021970 2017] [mpm_event:notice] [pid 1200:tid 140714591922304] AH00489: Apache/2.4.25 (Unix) OpenSSL/1.0.1e-fips configured -- resuming normal operations
[Tue Apr 25 07:55:15.022021 2017] [core:notice] [pid 1200:tid 140714591922304] AH00094: Command line: '/usr/sbin/httpd -D FOREGROUND'

Seems okay to me. Besides that I don't know why the server is giving me error notices on localhost:443 because the server has a valid Letsencrypt SSL certificate.


The content of the HTTPD config looks like this:


Contents of the httpd.conf file for magento2.domain.com

# Auto generated nginx config file by DirectAdmin version 1.51.3
# Modifying this file is not recommended as any changes you make will be
# overwritten when the user makes any changes to their website


# For global config changes that affect all Users, see this guide:
# http://help.directadmin.com/item.php?id=558
# For local config changes that only affect one User, see this guide:
# http://help.directadmin.com/item.php?id=3




server
{
	listen 46.145.192.200:80;
	server_name magento2.domain.com www.magento2.domain.com ;
	access_log /var/log/nginx/domains/magento2.domain.com.log;
	access_log /var/log/nginx/domains/magento2.domain.com.bytes bytes;
	error_log /var/log/nginx/domains/magento2.domain.com.error.log;
	root /home/mag2/domains/magento2.domain.com/public_html/pub;
	index index.php index.html index.htm;
	include /usr/local/directadmin/data/users/mag2/nginx_php.conf;
	location /
	{
		# access_log off;
		proxy_buffering off;
		proxy_pass http://46.145.192.200:8080;
		proxy_set_header X-Client-IP      $remote_addr;
		proxy_set_header X-Accel-Internal /nginx_static_files;
		proxy_set_header Host             $host;
		proxy_set_header X-Forwarded-For  $proxy_add_x_forwarded_for;
	}
	location /nginx_static_files/
	{
		# access_log  /var/log/nginx/access_log_proxy;
		alias       /home/mag2/domains/magento2.domain.com/public_html/pub/;
		internal;
	}
	include /etc/nginx/webapps.conf;
}


server
{
	listen 46.145.192.200:443 ssl;
	server_name magento2.domain.com www.magento2.domain.com ;
	access_log /var/log/nginx/domains/magento2.domain.com.log;
	access_log /var/log/nginx/domains/magento2.domain.com.bytes bytes;
	error_log /var/log/nginx/domains/magento2.domain.com.error.log;
	root /home/mag2/domains/magento2.domain.com/public_html/pub;
	index index.php index.html index.htm;
	ssl on;
	ssl_certificate /etc/httpd/conf/ssl.crt/server.crt.combined;
	ssl_certificate_key /etc/httpd/conf/ssl.key/server.key;
	include /usr/local/directadmin/data/users/mag2/nginx_php.conf;
	location /
	{
		# access_log off;
		proxy_buffering off;
		proxy_pass https://46.145.192.200:8081;
		proxy_set_header X-Client-IP      $remote_addr;
		proxy_set_header X-Accel-Internal /nginx_static_files;
		proxy_set_header Host             $host;
		proxy_set_header X-Forwarded-For  $proxy_add_x_forwarded_for;
	}
	location /nginx_static_files/
	{
		# access_log  /var/log/nginx/access_log_proxy;
		alias       /home/mag2/domains/magento2.domain.com/public_html/pub/;
		internal;
	}
	include /etc/nginx/webapps.ssl.conf;
}

Configuration Check

nginx: the configuration file /etc/nginx/nginx.conf syntax is ok
nginx: configuration file /etc/nginx/nginx.conf test is successful

Everything looks ok, but the public_html is still the folder that's being loaded in stead of the pub folder.

 

[zEitEr]

You posted content only from NGINX. What about Apache? You run NGINX+Apache and should update config for the domain in both NGINX and Apache.

Are you sure that you added the line into both NGINX and Apache?

 

[ericovk]

Aha! That was the trick. I expected DA to handle this automatically. After adding the same rule under both customization (and restarted Nginx and HTTPD) everything works as expected. Great! Thanks again Alex!