How To Use Apache as a Reverse Proxy with mod_proxy on CentOS 7

How To Use Apache as a Reverse Proxy with mod_proxy on CentOS 7


A reverse proxy is a type of proxy server that takes HTTP(S) requests and transparently distributes them to one or more backend servers. Reverse proxies are useful because many modern web applications process incoming HTTP requests using backend application servers which aren’t meant to be accessed by users directly and often only support rudimentary HTTP features.

You can use a reverse proxy to prevent these underlying application servers from being directly accessed. They can also be used to distribute the load from incoming requests to several different application servers, increasing performance at scale and providing fail-safeness. They can fill in the gaps with features the application servers don’t offer, such as caching, compression, or SSL encryption too.

In this tutorial, you’ll set up Apache as a basic reverse proxy using the mod_proxy extension to redirect incoming connections to one or several backend servers running on the same network. This tutorial uses a simple backend written with the with Flask web framework, but you can use any backend server you prefer.


To follow this tutorial, you will need:

Step 1 — Introducing Necessary Apache Modules

The modules that are needed to use Apache as a reverse proxy include mod_proxy itself and several of its add-on modules, which extend its functionality to support different network protocols. Specifically, we will be using:

  • mod_proxy, the main proxy module Apache module for redirecting connections; it allows Apache to act as a gateway to the underlying application servers.
  • mod_proxy_http, which adds support for proxying HTTP connections.
  • mod_proxy_balancer and mod_lbmethod_byrequests, which add load balancing features for multiple backend servers.

All four modules are enabled by default on a fresh CentOS 7 installation. You can verify that they are enabled by running:

  1. httpd -M

The command output will list all enabled Apache modules. The four lines you’re looking for are the aforementioned module names:

. . . proxy_module (shared) . . . lbmethod_byrequests_module (shared) . . . proxy_balancer_module (shared) . . . proxy_http_module (shared) . . .

In case the modules are not enabled, you can enable them by opening the /etc/httpd/conf.modules.d/00-proxy.conf with nano:

  1. sudo nano /etc/httpd/conf.modules.d/00-proxy.conf

and uncommenting lines with necessary modules by removing # sign from the line beginnings so the file looks as follows:

. . . 
LoadModule proxy_module modules/mod_proxy.so
. . . 
LoadModule lbmethod_byrequests_module modules/mod_lbmethod_byrequests.so
. . . 
LoadModule proxy_balancer_module modules/mod_proxy_balancer.so
. . . 
LoadModule proxy_http_module modules/mod_proxy_http.so
. . . 

To put the changes into effect, save the file and restart Apache.

  1. sudo systemctl restart httpd

Apache is now ready to act as a reverse proxy for HTTP requests. In the next step, we will create two very basic backend servers. These will help us verify if the configuration works properly, but if you already have your own backend application(s), you can skip to Step 3.

Step 2 — Creating Backend Test Servers

Running some simple backend servers is an easy way to test if your Apache configuration is working properly. Here, we’ll make two test servers which respond to HTTP requests with printing a line of text. One server will say Hello world! and the other will say Howdy world!.

Note: In non-test setups, backend servers usually all return the same kind of content. However, for this test in particular, having the two servers return different messages makes it easy to check that the load balancing mechanism uses both.

Flask is a Python microframework for building web applications. We’re using Flask to create the test servers because a basic app requires just a few lines of code. You don’t need to know Python to set these up, but if you’d like to learn, you can look at these Python tutorials.

Let’s install the IUS package repository files first. IUS (Inline with Upstream Stable) is a community project that brings up-to-date versions of select software to CentOS, including Python 3.

  1. sudo yum -y install https://centos7.iuscommunity.org/ius-release.rpm

Then install Python 3 and Pip, the recommended Python package manager.

  1. sudo yum -y install python35u python35u-pip

Use Pip to install Flask.

  1. sudo pip3.5 install flask

Now that all the required components are installed, start by creating a new file that will contain the code for the first backend server in the home directory of the current user.

  1. nano ~/backend1.py

Copy the following code into the file, then save and close it.

from flask import Flask
app = Flask(__name__)

def home():
    return 'Hello world!'

The first two lines initialize the Flask framework. There is one function, home(), which returns a line of text (Hello world!). The @app.route('/') line above the home() function definition tells Flask to use home()'s return value as a response to HTTP requests directed at the / root URL of the application.

The second backend server is exactly the same as the first, aside from returning to a different line of text, so start by duplicating the first file.

  1. cp ~/backend1.py ~/backend2.py

Open the newly copied file.

  1. nano ~/backend2.py

Change the message to be returned from Hello world! to Howdy world!, then save and close the file.

from flask import Flask
app = Flask(__name__)

def home():
    return 'Howdy world!'

Use the following command to start the first background server on port 8080. This also redirects Flask’s output to /dev/null because it would cloud the console output further on.

  1. FLASK_APP=~/backend1.py flask run --port=8080 >/dev/null 2>&1 &

Here, we are preceding the flask command by setting FLASK_APP environment variable in the same line. Environment variables are a convenient way to pass information into processes that are spawned from the shell. You can learn more about environment variables in How To Read and Set Environmental and Shell Variables on a Linux VPS.

In this case, using an environment variable makes sure the setting applies only to the command being run and will not stay available afterwards, as we will be passing another filename the same way to tell flask command to start the second server

Similarly, use this command to start the second server on port 8081. Note the different value for the FLASK_APP environment variable.

  1. FLASK_APP=~/backend2.py flask run --port=8081 >/dev/null 2>&1 &

You can test that the two servers are running using curl. Test the first server:

  1. curl

This will output Hello world! in the terminal. Test the second server:

  1. curl

This will output Howdy world! instead.

Note: To close both test servers after you no longer need them, like when you finish this tutorial, you can simply execute killall flask.

In the next step, we’ll modify Apache’s configuration file to enable its use as a reverse proxy.

Step 3 — Modifying the Default Configuration to Enable Reverse Proxy

In this section, we will set up the default Apache virtual host to serve as a reverse proxy for single backend server or an array of load balanced backend servers.

Note: In this tutorial, we’re applying the configuration at the virtual host level. On a default installation of Apache, there are no virtual hosts configured. We will be creating a single, default virtual host that will catch all traffic. However, you can use all those configuration fragments in other virtual hosts as well. To learn more about virtual hosts in Apache, you can read this How To Set Up Apache Virtual Hosts on CentOS 7 tutorial.

If your Apache server acts as both HTTP and HTTPS server, your reverse proxy configuration must be placed in both the HTTP and HTTPS virtual hosts. To learn more about SSL with Apache, you can read this How To Create a SSL Certificate on Apache for CentOS 7 tutorial.

Create the new default virtual host by creating a new empty Apache configuration file in /etc/httpd/conf.d directory using nano or your favorite text editor.

  1. sudo nano /etc/httpd/conf.d/default-site.conf

The first example below explains how to configure the default virtual host to reverse proxy for a single backend server, and the second sets up a load balanced reverse proxy for multiple backend servers.

Example 1 — Reverse Proxying a Single Backend Server

Paste the following contents into the default-site.conf file, so your configuration file looks like this:

<VirtualHost *:80>
    ProxyPreserveHost On

    ProxyPass /
    ProxyPassReverse /

If you followed along with the example servers in Step 2, use as written in the block above. If you have your own application servers, use their addresses instead.

There are three directives here:

  • ProxyPreserveHost makes Apache pass the original Host header to the backend server. This is useful, as it makes the backend server aware of the address used to access the application.
  • ProxyPass is the main proxy configuration directive. In this case, it specifies that everything under the root URL (/) should be mapped to the backend server at the given address. For example, if Apache gets a request for /example, it will connect to http://your_backend_server/example and return the response to the original client.
  • ProxyPassReverse should have the same configuration as ProxyPass. It tells Apache to modify the response headers from backend server. This makes sure that if the backend server returns a location redirect header, the client’s browser will be redirected to the proxy address and not the backend server address, which would not work as intended.

To put these changes into effect, restart Apache.

  1. sudo systemctl restart httpd

Now, if you access http://your_server_ip in a web browser, you will see your backend server response instead of standard Apache welcome page. If you followed Step 2, this means you’ll see Hellow world!.

Example 2 — Load Balancing Across Multiple Backend Servers

If you have multiple backend servers, a good way to distribute the traffic across them when proxying is to use load balancing features of mod_proxy.

Replace all the contents within the VirtualHost block with the following, so your configuration file looks like this:

<VirtualHost *:80>
<Proxy balancer://mycluster>

    ProxyPreserveHost On

    ProxyPass / balancer://mycluster/
    ProxyPassReverse / balancer://mycluster/

The configuration is similar to the previous one, but instead of specifying a single backend server directly, we’ve used an additional Proxy block to define multiple servers. The block is named balancer://mycluster (the name can be freely altered) and consists of one or more BalancerMembers, which specify the underlying backend server addresses. The ProxyPass and ProxyPassReverse directives use the load balancer pool named mycluster instead of a specific server.

If you followed along with the example servers in Step 2, use and for the BalancerMember directives, as written in the block above. If you have your own application servers, use their addresses instead.

To put these changes into effect, restart Apache.

  1. sudo systemctl restart httpd

If you access http://your_server_ip in a web browser, you will see your backend servers’ responses instead of the standard Apache page. If you followed Step 2, refreshing the page multiple times should show Hello world! and Howdy world!, meaning the reverse proxy worked and is load balancing between both servers.


You now know how to set up Apache as a reverse proxy to one or many underlying application servers. mod_proxy can be used effectively to configure reverse proxy to application servers written in a vast array of languages and technologies, such as Python and Django or Ruby and Ruby on Rails. It can be also used to balance traffic between multiple backend servers for sites with lots of traffic or to provide high availability through multiple servers, or to provide secure SSL support to backend servers not supporting SSL natively.

While mod_proxy with mod_proxy_http is the perhaps most commonly used combination of modules, there are several others that support different network protocols. We didn’t use them here, but some other popular modules include:

  • mod_proxy_ftp for FTP.
  • mod_proxy_connect for SSL tunneling.
  • mod_proxy_ajp for AJP (Apache JServ Protocol), like Tomcat-based backends.
  • mod_proxy_wstunnel for web sockets.

To learn more about mod_proxy, you can read the official Apache mod_proxy documentation.

Thanks for learning with the DigitalOcean Community. Check out our offerings for compute, storage, networking, and managed databases.

Learn more about our products

About the authors
Default avatar

Software Engineer, CTO @Makimo

Creating bespoke software ◦ CTO & co-founder at Makimo. I’m a software enginner & a geek. I like making impossible things possible. And I need tea.

Default avatar

staff technical writer

hi! i write do.co/docs now, but i used to be the senior tech editor publishing tutorials here in the community.

Still looking for an answer?

Ask a questionSearch for more help

Was this helpful?

This textbox defaults to using Markdown to format your answer.

You can type !ref in this text area to quickly search our full set of tutorials, documentation & marketplace offerings and insert the link!


So i followed up this instruction except for the part that sets up load balancing.

I used a wordpress as a backend. Both backend and proxy machines are using Apache on Centos7.

This setup works on the main page only. But once you navigate to subfolder, it will return the actual address of the backend server.

http://proxyIP — gives the front page/welcome page of WP.

but when, say, clicking the Hello World post gives me:

http://backendIP/wordpress/<the rest=“” of=“” url=“”>

my understanding that it still should be http://proxyIP/<rest of=“” url=“”>

I am doing this on the same network. So obviously, if i was to do this setup on WAN, it would not show page from my backends IP.

Please help.

Thank you,


Note that on CentOS7, with SELinux enabled (as it should be!), Apache needs to be allowed explicitly to connect to a backend server on the network (i.e., not on localhost):

# /usr/sbin/setsebool -P httpd_can_network_connect 1

Very well explained, thanks alot!!!

This is a great article. For some reason, I was getting logs that said:

[Sun Apr 16 18:14:12.083870 2017] [proxy:error] [pid 17345] (13)Permission denied: AH00957: HTTP: attempt to connect to (localhost) failed
[Sun Apr 16 18:14:12.083986 2017] [proxy:error] [pid 17345] AH00959: ap_proxy_connect_backend disabling worker for (localhost) for 60s
[Sun Apr 16 18:14:12.084082 2017] [proxy_http:error] [pid 17345] [client] AH01114: HTTP: failed to make connection to backend: localhost

Solution at the bottom of this page

Try DigitalOcean for free

Click below to sign up and get $200 of credit to try our products over 60 days!

Sign up

Join the Tech Talk
Success! Thank you! Please check your email for further details.

Please complete your information!

Featured on Community

Get our biweekly newsletter

Sign up for Infrastructure as a Newsletter.

Hollie's Hub for Good

Working on improving health and education, reducing inequality, and spurring economic growth? We'd like to help.

Become a contributor

Get paid to write technical tutorials and select a tech-focused charity to receive a matching donation.

Welcome to the developer cloud

DigitalOcean makes it simple to launch in the cloud and scale up as you grow — whether you're running one virtual machine or ten thousand.

Learn more
DigitalOcean Cloud Control Panel