Using gitlab's nginx to serve another app

Question

Hello I have installed Gitlab using this
https://gitlab.com/gitlab-org/omnibus-gitlab/blob/master/README.md#installation

Now I want to use nginx to serve another content other than gitlab application how can I do this

• Where are the config files that I need to modify
• How can I point a directory like /var/www so that nginx knows that is the root for another app.

Update(forgot to mention I'm running this under Red Hat 6.5, Debian/Ubuntu solution welcome)

Answer 1

Here I am using

- gitlab.example.com to serve gitlab.example.com over https.
- example.com over http to serve another content other than gitlab application.

Gitlab installed from deb package is using chef to provision ngnix, so you have to modify chef recipies and add new vhost template into chef cookbooks directory

You can find all chef cookbooks here: /opt/gitlab/embedded/cookbooks/gitlab/

open /opt/gitlab/embedded/cookbooks/gitlab/recipes/nginx.rb

change:

nginx_vars = node['gitlab']['nginx'].to_hash.merge({
  :gitlab_http_config => File.join(nginx_etc_dir, "gitlab-http.conf"),
})

to:

nginx_vars = node['gitlab']['nginx'].to_hash.merge({
  :gitlab_http_config => File.join(nginx_etc_dir, "gitlab-http.conf"),
  :examplecom_http_config => File.join(nginx_etc_dir, "examplecom-http.conf"),
})

add this to the same file:

template nginx_vars[:examplecom_http_config] do
  source "nginx-examplecom-http.conf.erb"
  owner "root"
  group "root"
  mode "0644"
  variables(nginx_vars.merge(
    {
      :fqdn => "example.com",
      :port => 80,
    }
  ))
  notifies :restart, 'service[nginx]' if OmnibusHelper.should_notify?("nginx")
end

then in template directory(/opt/gitlab/embedded/cookbooks/gitlab/templates/default), create nginx vhost template file( nginx-examplecom-http.conf.erb) and add this there:

server {
  listen <%= @listen_address %>:<%= @port %>;
  server_name <%= @fqdn %>;
  root /var/www/example.com;

  access_log  <%= @log_directory %>/examplecom_access.log;
  error_log   <%= @log_directory %>/examplecom_error.log;

  location /var/www/example.com {
    # serve static files from defined root folder;.
    # @gitlab is a named location for the upstream fallback, see below
    try_files $uri $uri/index.html $uri.html;
  }

  error_page 502 /502.html;
}

you have to set nginx['redirect_http_to_https'] = false in(/etc/gitlab/gitlab.rb):

external_url "https://gitlab.example.com"
gitlab_rails['gitlab_email_from'] = "info@example.com"
gitlab_rails['gitlab_support_email'] = "support@example.com"



nginx['redirect_http_to_https'] = false
nginx['ssl_certificate'] = "/etc/gitlab/ssl/ssl-unified.crt"
nginx['ssl_certificate_key'] = "/etc/gitlab/ssl/ssl.key"


gitlab_rails['gitlab_default_projects_limit'] = 10

add include <%= @examplecom_http_config %>; into /opt/gitlab/embedded/cookbooks/gitlab/templates/default/nginx.conf.erb :

http {
  sendfile <%= @sendfile %>;
  tcp_nopush <%= @tcp_nopush %>;
  tcp_nodelay <%= @tcp_nodelay %>;

  keepalive_timeout <%= @keepalive_timeout %>;

  gzip <%= @gzip %>;
  gzip_http_version <%= @gzip_http_version %>;
  gzip_comp_level <%= @gzip_comp_level %>;
  gzip_proxied <%= @gzip_proxied %>;
  gzip_types <%= @gzip_types.join(' ') %>;

  include /opt/gitlab/embedded/conf/mime.types;

  include <%= @gitlab_http_config %>;
  include <%= @examplecom_http_config %>;
}

after all those changes run:

gitlab-ctl reconfigure
gitlab-ctl restart

Answer 2

vndr's above solution would work but on the https://gitlab.com/gitlab-org/omnibus-gitlab/blob/master/doc/settings/nginx.md, it said:

Inserting custom settings into the NGINX config

If you need to add custom settings into the NGINX config, for example to include existing server blocks, you can use the following setting.

Example: include a directory to scan for additional config files nginx['custom_nginx_config'] = "include /etc/nginx/conf.d/*.conf;"

So let's check your /opt/gitlab/embedded/cookbooks/gitlab/templates/default/nginx.conf.erb to see if it contains: <%= @custom_nginx_config %> (it looks like the current gitlab-7.5.3_omnibus.5.2.1.ci-1.el6.x86_64.rpm doesn't include it)

If not, then just add it above the line include <%= @gitlab_http_config %>; like:

<%= @custom_nginx_config %>
include <%= @gitlab_http_config %>;

Then open the /etc/gitlab/gitlab.rb to add: nginx['custom_nginx_config'] = "include /etc/nginx/conf.d/*.conf;"

We can make it simply by just add: include /etc/nginx/conf.d/*.conf; instead <%= @custom_nginx_config %>

Then create normal nginx .conf files in /etc/nginx/conf.d/ and gitlab-ctl reconfigure

Answer 3

As I did not wanted to change the config for gitlab Nginx server nor installing/configuring another Nginx and to make sure gitlab would survive an major update, I came to below solution for the Gitlab Omnibus package.

also as per

Gitlab:Ningx =>Inserting custom settings into the NGINX config

edit the /etc/gitlab/gitlab.rb of your gitlab:

nano /etc/gitlab/gitlab.rb

and sroll to nginx['custom_nginx_config'] and modify as below make sure to uncomment

# Example: include a directory to scan for additional config files
nginx['custom_nginx_config'] = "include /etc/nginx/conf.d/*.conf;"

create the new config dir:

mkdir -p /etc/nginx/conf.d/
nano /etc/nginx/conf.d/new_app.conf

and add content to your new config: /etc/nginx/conf.d/new_app.conf

server {
  listen *:80;
  server_name new_app.mycompany.com;
  server_tokens off;
  access_log  /var/log/new_app_access.log;
  error_log   /var/log/new_app_error.log;

  root /var/www/html/new_app/;
  index index.html index.htm;

 }

and reconfigure gitlab to get the new settings inserted

gitlab-ctl reconfigure

to restart nginx after changing your config's or adding more config's in /etc/nginx/conf.d:

gitlab-ctl restart nginx

to check nginx error log:

tail -f /var/log/gitlab/nginx/error.log

and see https://stackoverflow.com/a/39695791/6821811 for redirecting to another application server.

Answer 4

Even through you really can do it, a better practise would be to use upper level separate nginx server to serve both gitlab's nginx and your other custom content. Gitlab's nginx may change it's configuration at any time and it can break your custom content. Also, separate nginx is completely yours for configuring.

Just install those two instances to different ports and proxy gitlab's nginx with upper one. Of cause, it will be an overhead, but completely insignificant one.

Answer 5

I have tried both approaches, and the one that worked for me was to put a clean NGINX on top of gitlab's built in one. its more easy/convenient in the long run.

Depending on your needs here are some crucial things that have to be in place first:

• DNS settings of your network/router/etc. (else this will not work, since the configurations here are based on server names,)
• My setup is trivial one server, multiple sites hosted in the same server IP, and I filter by naming the apps thru NGINX name filter.

Here are the main steps to follow, keep in mind that depending on your needs this could imply more tweaking around, also this is a Ubuntu Server 14.04 .

1. First deactivate the main Nginx (the one bundled with omnibus) edit /etc/gitlab/gitlab.rb

   nginx['enable'] = false ci_nginx['enable'] = false

2. Now your free to install a clean instance of NGINX.

3. Regarding the previous step: sometimes the installer doesn't create the sites-enabled/ and sites-available/ folders, create them, and make sure to include them in the /etc/nginx/nginx.conf file

   include /etc/nginx/sites-enabled/*.conf;

4. In a general nginx workflow you include your site configurations under sites-available/ and then when your ready/happy you make a link to sites-enabled/ folder and restart nginx so the changes are effective
5. Add your Gitlab configuration to Nginx site-available/ folder

here is my conf:

`upstream gitlab-workhorse {
  server unix:/var/opt/gitlab/gitlab-workhorse/socket;
}

## Normal HTTP host
server {
  ## Either remove "default_server" from the listen line below,
  ## or delete the /etc/nginx/sites-enabled/default file. This will cause gitlab
  ## to be served if you visit any address that your server responds to, eg.
  ## the ip address of the server (http://x.x.x.x/)n 0.0.0.0:80 default_server;
  #listen 0.0.0.0:80 default_server;
  listen 0.0.0.0:80 ;
#  listen [::]:80 default_server;
  server_name gitlab.mycompany.com; ## Replace this with something like gitlab.example.com
  server_tokens off; ## Don't show the nginx version number, a security best practice
  root /opt/gitlab/embedded/service/gitlab-rails/public;

  ## See app/controllers/application_controller.rb for headers set

  ## Individual nginx logs for this GitLab vhost
  access_log  /var/log/nginx/gitlab.access.log;
  error_log   /var/log/nginx/gitlab.error.log;

  location / { 
    client_max_body_size 0;
    gzip off;

    ## https://github.com/gitlabhq/gitlabhq/issues/694
    ## Some requests take more than 30 seconds.
    proxy_read_timeout      300;
    proxy_connect_timeout   300;
    proxy_redirect          off;

    proxy_http_version 1.1;

    proxy_set_header    Host                $http_host;
    proxy_set_header    X-Real-IP           $remote_addr;
    proxy_set_header    X-Forwarded-For     $proxy_add_x_forwarded_for;
    proxy_set_header    X-Forwarded-Proto   $scheme;

        proxy_pass http://gitlab-workhorse;
      }
    } 

you can find more detail of the configuration in here more options

6. restart/reload nginx

   sudo service nginx restart

7. restart gitlab omnibus and check your Gitlab configuration

   sudo gitlab-ctl reconfigure

   sudo gitlab-ctl tail (just to check if something is wrong in your gitlab configuration)

8. Add extra (as many as you like) server configurations that you need in /etc/nginx/sites-available/ and eventually when happy/ready add the link to /etc/nginx/sites-enabled/
   here is another example of another app

   upstream app_server {
       server 127.0.0.1:9080 fail_timeout=0;
   }
   server {
       listen 80; 
       server_name jenkins.mycompany.com;    
       access_log            /var/log/nginx/jenkins.access.log;
       location / { 
           proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
           proxy_set_header Host $http_host;
           proxy_redirect off;
           if (!-f $request_filename) {
               proxy_pass http://app_server;
               break;
           }
   
       }   
   }
   

9. remember to always restart/reload nginx so that you see your changes.

   sudo service nginx restart

10. check the logs in case something is wrong /var/log/nginx/anysites*.log

11. Please note that here we are using upstream with different ports and the names(They exist/are real/are registered in the domain of your company) are all pointing to the same IP address, meaning NIGNX will come and find the same IP address but it will not break because of the different ports in the upstreams this is really important

That's how my configuration is working right now I have not had any issues with Gitlab or any other apps.
So hopefully this will help anyone out there.

Answer 6

Those "other content" are declared in NGiNX with "Server Blocks".

The GitLab one is in /etc/nginx/sites-available/gitlab (according to the documentation, and symlined in [/etc/nginx/sites-enabled][3]).
You can add other server blocks in it, similar to this one (you may have to choose a different port number), as illustrated in this process (updated here for Ubuntu 14.04)

server {
        listen   80; ## listen for ipv4; this line is default and implied
        #listen   [::]:80 default ipv6only=on; ## listen for ipv6

        root /var/www/example.com/public_html;
        index index.html index.htm;

        # Make site accessible from http://localhost/
        server_name example.com;
}

The root directive should reference the root folder of your webapp (/var/www or more likely a subfolder of /var/www).

That server block is quite separate from any GitLab config.