HA-Proxy

 

HAProxy Setup Explained (From Beginner to Enterprise)

HAProxy (High Availability Proxy) is a TCP/HTTP load balancer and reverse proxy. In Kubernetes, it is commonly used in on-premises clusters to provide a single endpoint for the Kubernetes API server and, in some architectures, to load balance application traffic.


Why Do We Need HAProxy?

Imagine you have a highly available Kubernetes control plane.

Master-1

192.168.1.11


Master-2

192.168.1.12


Master-3

192.168.1.13


If your developers configure kubectl to connect directly to:

192.168.1.11 - Master-1


and Master-1 fails:

kubectl

   │

   ▼

Master-1 ❌


The Kubernetes API becomes unreachable from the client's perspective, even though Master-2 and Master-3 are healthy.


Solution: Put HAProxy in Front

                                       kubectl

                                           │

                                          ▼

                         HAProxy (192.168.1.10)

                                         │

          ┌──────────┼──────────┐

          ▼                          ▼                            ▼

     Master-1                Master-2            Master-3

      6443                      6443                          6443


Now everyone connects to one IP address:

192.168.1.10

HAProxy forwards requests to a healthy API server.

Step 1 – Provision the HAProxy Server

You need a Linux VM.

ComponentIP
HAProxy192.168.1.10
Master-1192.168.1.11
Master-2192.168.1.12
Master-3192.168.1.13

Recommended OS:

  • Ubuntu
  • RHEL
  • Rocky Linux
  • AlmaLinux

Step 2 – Install HAProxy

Ubuntu

sudo apt update

sudo apt install haproxy -y

RHEL

sudo yum install haproxy -y

Verify:

haproxy -v


Step 3 – Configure HAProxy

Main configuration file:

/etc/haproxy/haproxy.cfg


Example:


global

    daemon

    maxconn 4000


defaults

    mode tcp

    timeout connect 10s

    timeout client 1m

    timeout server 1m


frontend kubernetes-api

    bind *:6443

    default_backend masters


backend masters

    balance roundrobin

    option tcp-check


    server master1 192.168.1.11:6443 check

    server master2 192.168.1.12:6443 check

    server master3 192.168.1.13:6443 check


What Does This Configuration Mean?

Frontend

frontend kubernetes-api

This tells HAProxy:

"Listen for incoming connections."


bind *:6443

Listen on port 6443, the Kubernetes API port. 

Backend


backend masters

Defines the destination servers.

server master1 192.168.1.11:6443

Traffic can be sent to Master-1.

------------


server master2 192.168.1.12:6443

Traffic can be sent to Master-2.

------------

Health Checks

check

HAProxy periodically tests each API server.

Example:

Master-1 ✓

Master-2 ✓

Master-3 ✗

Master-3 is automatically removed from rotation until it becomes healthy again.


Step 4 – Start the Service

sudo systemctl enable haproxy

sudo systemctl start haproxy


Verify:

sudo systemctl status haproxy

Step 5 – Validate the Configuration


> sudo haproxy -c -f /etc/haproxy/haproxy.cfg

Expected:
Configuration file is valid


Step 6 – Configure Kubernetes

During kubeadm init:

Instead of:

192.168.1.11

Use:

192.168.1.10:6443

Example:

kubeadm init \

--control-plane-endpoint=192.168.1.10:6443

Now every node joins using the HAProxy endpoint.


Request Flow

             kubectl get pods

                      │


                     ▼


                 HAProxy


                     │


 ┌─────-─|──────┐

 ▼                ▼               ▼

Master1   Master2      Master3


Real time HAproxy file 

#---------------------------------------------------------------------

# Global Settings

#---------------------------------------------------------------------

global

    log /dev/log local0

    log /dev/log local1 notice


    daemon


    user haproxy

    group haproxy


    maxconn 50000


    stats socket /run/haproxy/admin.sock mode 660 level admin

    stats timeout 30s


    tune.ssl.default-dh-param 2048


#---------------------------------------------------------------------

# Default Settings

#---------------------------------------------------------------------

defaults

    log global


    mode tcp


    option tcplog

    option dontlognull


    timeout connect 10s

    timeout client 60s

    timeout server 60s


    retries 3


#---------------------------------------------------------------------

# Kubernetes API Server Frontend

#---------------------------------------------------------------------

frontend kubernetes-api


    bind *:6443


    default_backend kubernetes-control-plane


#---------------------------------------------------------------------

# Kubernetes API Backend

#---------------------------------------------------------------------

backend kubernetes-control-plane


    mode tcp


    balance roundrobin


    option tcp-check


    default-server inter 10s fall 3 rise 2


    server master01 10.10.1.11:6443 check

    server master02 10.10.1.12:6443 check

    server master03 10.10.1.13:6443 check


#---------------------------------------------------------------------

# Statistics Page

#---------------------------------------------------------------------

listen stats


    bind *:8404


    mode http


    stats enable


    stats uri /stats


    stats refresh 10s


    stats show-legends


    stats auth admin:Admin@123








Comments

Popular posts from this blog

Third Eye View