High Availability using HAProxy - Practical Guide with Sample Project

Last Updated on April 19, 2024 by cscontents

Table of Contents

Introduction

HAProxy or High Availability Proxy is an open-source software that is widely used as a High Availability functionality provider. HA functionality is such a crucial functionality without which modern-day business applications/systems can’t think of going live.

To show the use of HAProxy as a High Availability provider we will do a demo project or a POC (proof of concept) where we are going to deploy two instances of a static application on the Apache HTTPD server and then configure HAProxy in front of them to work as a single point of access which will make the system highly available. It means if any of the application servers goes down HAProxy will route the incoming traffic request to the other application server. That means our system will be highly available.

In this case, we have three options.

Options	Application Instance 1	Application Instance 2	Comment
Option 1	Active	Passive	In this mode of operation, app instance 1 is active and app instance 2 is passive. It means by default HAProxy will send the traffic to only app instance 1 and if app instance 1 goes down, then HAProxy will send traffic to app instance 2.
Option 2	Passive	Active	In this mode of operation, app instance 1 is in passive mode, and app instance 2 is in active mode. It means by default HAProxy will send the traffic to only app instance 2 and if app instance 2 goes down, then HAProxy will send traffic to app instance 1.
Option 3	Active	Active	In this mode of operation, both the app instances are active, and HAProxy will send the traffic request to both servers based on some method like round-robin, etc.

Below is a pictorial representation of the above options.

Prerequisite

To complete this demo project, you need to have the below prerequisites.

3 Linux machines are required. We will be using 3 Ubuntu machines for this demo project.
- On 2 machines we will deploy 2 instances of a sample static application and
- On the 3^rd machine, we will install HAProxy (with HA configuration).
On all the machines, we need a non-root user with proper sudo privileges to perform the activity.
Last but the least, this guide presumes you have experience with Linux commands.

Sections of this Guide

For simplicity, we have divided this guide into a few sections.

Section 1: Installation of Apache HTTPD web server on 2 Linux machines.
Section 2: Deployment of a sample web application (static) in the two web servers.
Section 3: Installation of HAProxy on the 3^rd server which will work as ‘High-Availability Provider’.
Section 4: Configure HAProxy to provide high-availability functionality.
Section 5: Test the whole system and validate the working HAProxy.

Section 1: Installation of Apache HTTPD on Linux machines 1 & 2

Follow the below steps to install the Apache HTTPD web server on both machines (in our case it is an Ubuntu machine).

Step 1: Update the OS

Run the below command to update the OS.

sudo apt update -y

In the case of RHEL / CentOS machines, run the below command to update the OS.

sudo yum update -y

Step 2: Install the Apache HTTPD web server

Run the below command to install the HTTPD web server.

sudo apt install apache2 -y

In the case of RHEL / CentOS machines, the installation command is,

sudo yum install httpd -y

Step 3: Start the HTTPD web server

Run the below command to start the HTTPD web server.

sudo systemctl start apcahe2

In the case of RHEL / CentOS machines, run the below command to start the Apache HTTPD web server.

sudo systemctl start httpd

Step 4: Check the status of the HTTPD web server

Run the below command to check the status of the HTTPD web server.

systemctl status apache2

In the case of RHEL / CentOS machines, run the below command to check the status of the Apache HTTPD web server.

systemctl status httpd

Section 2: Deployment of sample web application (static)

In this section, we will deploy a sample static web application in the Apache HTTPD server (on both machines). It means we will deploy the same static web application on both servers (since we plan to implement load-balancing functionality using the HTTPD web server, for this we need to deploy at least 2 instances of the sample web application).

Follow the below steps on both machines.

Step 1: Create a New HTML Page Under /var/www

First, remove the existing default HTML pages. Run the below commands one by one.

cd /var/www/html

Next,

sudo rm -rf *.html

then create a new index.html page,

sudo vi index.html

To make the work of the load balancer very clear and visible, we will create two different content web pages.

Copy the below content and paste it into the newly opened index.html file in server-1/ machine-1.

<h1>Hello World! I am from Server 1</h1>

Copy the below content and paste it into the newly opened index.html file in server-2/ machine-2.

<h1>Hello World! I am from Server 2</h1>

Finally, save the files.

Step 2: Reload the Apache HTTPD web server

Once we have added a new index.html page, we need to reload the web server to see the changes. Execute the below command to reload the Apache HTTPD web server.

sudo systemctl reload apache2

In the case of RHEL / CentOS machines, run the below command to reload the Apache HTTPD web server.

sudo systemctl reload httpd

Step 3: Validate the Deployed web app

In the previous step, we created a new static HTML web page. Now we need to validate it. By default, the Apache HTTPD web server runs on port 80. Make sure you have allowed inbound network connection to your machine over port 80.

To validate whether the newly deployed static web page is working properly or not, access this URL from any browser – http://your_machine_ip:80

(Please replace ‘your_machine_ip’ by your machine IP address)

Section 3: Installation of HAProxy

In this section, we will install HAProxy on another machine. This HAProxy will provide the High Availability functionality. Follow the below steps to install HAProxy.

Step 1: Update the OS

For Ubuntu,

sudo apt update -y

For RHEL / CentOS,

sudo yum update -y

Step 2: Install HAProxy using the Package Manager

For ubuntu,

sudo apt install haproxy -y

For RHEL / CentOS,

sudo yum install haproxy -y

Step 3: Start HAProxy and check the status

Run the below command to start HAProxy,

systemctl start haproxy

To check the status of HAProxy, run the below command.

systemctl status haproxy

Section 4: Configure HAProxy to Provide high-availability Functionality

In this section, we will configure haproxy to provide HA functionality.

Before moving forward, we need to know a few details regarding load balancing. Load balancing is done either on layer 4 or layer 7 of the OSI model.

Use Layer 4 load balancing for Basic Traffic Distribution:
- Need to distribute traffic across multiple servers for High Availability (HA).
- Decisions are based on simple factors like port numbers or IP addresses.
- Focuses on speed and efficiency. (e.g., HAProxy for basic load balancing)
Use Layer 7 load balancing for Smarter Routing:
- Need to distribute traffic based on content or user data (URLs, cookies).
- Want to optimize traffic for specific application functionalities?
- Requires more complex configuration but offers finer control. (e.g., Routing users to different servers based on their location)

Looking at our use case where we just want to perform basic load balancing to distribute the traffic between two backend servers, we will implement ‘layer 4 load balancing’.

Now, coming to the haproxy configuration (to work as a “layer 4 load balancer”), there are 3 blocks in the haproxy configuration.

“global” block – it is regarding the global configuration of haproxy.
“defaults” block – it is regarding common default configurations like log, mode, etc.
“frontend” block – it says how haproxy handles incoming requests, on which port it will listen, etc.
“backend” block – it says regarding the backend server where haproxy needs to forward the traffic.

Below is the /etc/haproxy/haproxy.cfg file in our case,

#---------------------------------------------------------------------
# Global settings
#---------------------------------------------------------------------
global
    log         127.0.0.1 local2

    chroot      /var/lib/haproxy
    pidfile     /var/run/haproxy.pid
    maxconn     4000
    user        haproxy
    group       haproxy
    daemon

    # turn on stats unix socket
    stats socket /var/lib/haproxy/stats

    # utilize system-wide crypto-policies
    ssl-default-bind-ciphers PROFILE=SYSTEM
    ssl-default-server-ciphers PROFILE=SYSTEM

#---------------------------------------------------------------------
# common defaults that all the 'listen' and 'backend' sections will
#---------------------------------------------------------------------

defaults
    mode                    http
    log                     global
    option                  httplog
    option                  dontlognull
    option http-server-close
    option forwardfor       except 127.0.0.0/8
    option                  redispatch
    retries                 3
    timeout http-request    10s
    timeout queue           1m
    timeout connect         10s
    timeout client          1m
    timeout server          1m
    timeout http-keep-alive 10s
    timeout check           10s
    maxconn                 3000

#---------------------------------------------------------------------
# main frontend which proxys to the backends
#---------------------------------------------------------------------
frontend ft_app_server
    bind *:80
    default_backend bk_app_servers
        reqadd X-Forwarded-Proto:\ http

backend bk_app_servers
        server server1 25.159.237.135:80 check
        server server2 205.178.199.98:80 check

In the above configuration file,

You need to change the server1 IP and server2 IP.
Also, verify the ports on which the web pages are running and on which port your haproxy server is listening.

Section 5: Validate the working HAProxy

In this section, we will validate the working of haproxy as a HA provider.

You can hit the below URL to validate it (Please replace ‘your_haproxy_machine_ip’ by your HAProxy machine IP address).

http://your_haproxy_machine_ip:80

If you keep on refreshing the browser, it will display the below outputs alternatively.

This is all about using Haproxy as an HA functionality provider.

Thank you.

If you are interested in learning DevOps, please have a look at the below articles, which will help you greatly.