Setting up a highly available load balancing cluster requires two primary components: HAProxy for traffic distribution and Keepalived for failover management. This setup uses two virtual machines running CentOS 7.

Environment Preparation

Install the EPEL repository and required packages on both nodes:

yum install -y epel-release
yum install -y haproxy keepalived nginx psmisc net-tools vim

HAProxy Configuration

Create the HAProxy configuration file at /etc/haproxy/haproxy.cfg:

global
    maxconn 200000
    chroot /var/lib/haproxy
    user haproxy
    group haproxy
    daemon
    nbproc 2
    pidfile /var/run/haproxy.pid
    log 127.0.0.1 local0 info

defaults
    option http-keep-alive
    maxconn 200000
    mode http
    timeout connect 3000ms
    timeout client 30000ms
    timeout server 30000ms

listen stats_page
    mode http
    bind *:9090
    stats enable
    stats uri /haproxy-stats
    stats auth admin:securepass

frontend web_frontend
    bind 192.168.10.100:80
    mode http
    option httplog
    log global
    default_backend web_servers

backend web_servers
    option forwardfor header X-CLIENT-IP
    option httpchk HEAD / HTTP/1.0
    balance roundrobin
    server app01 192.168.10.101:8080 check inter 2000 rise 30 fall 15
    server app02 192.168.10.102:8080 check inter 2000 rise 30 fall 15

The balance roundrobin directive ensures requests are distributed evenly across backend servers. Alternative algorithms include source for IP-based persistence or leastconn for connection-based distribution.

Backend Web Server Configuration

On the first backend node, create the web content directory and Nginx configuration:

mkdir -p /var/www/html
echo "Server 01" > /var/www/html/index.html

Nginx server block configuration at /etc/nginx/conf.d/default.conf:

server {
    listen 8080;
    server_name localhost;
    location / {
        root /var/www/html;
        index index.html index.htm;
    }
    access_log /var/log/nginx/access.log main;
    error_log /var/log/nginx/error.log warn;
}

On the second backend node, repeat the process with different content:

mkdir -p /var/www/html
echo "Server 02" > /var/www/html/index.html

Start the services:

systemctl restart nginx
systemctl restart haproxy

Keepalived Configuration for Master Node

Configure Keepalived on the primary load balancer:

global_defs {
    notification_email {
        admin@example.org
    }
    notification_email_from keepalived@example.org
    smtp_server 127.0.0.1
    smtp_connect_timeout 30
    router_id lb_master
}

vrrp_instance VI_1 {
    state MASTER
    interface eth0
    virtual_router_id 51
    priority 150
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass MyPass123
    }
    virtual_ipaddress {
        192.168.10.200
    }
}

Enable binding to non-local IP addresses:

echo 1 > /proc/sys/net/ipv4/ip_nonlocal_bind

Update HAProxy to bind to the virtual IP:

frontend web_frontend
    bind 192.168.10.200:80

Start Keepalived:

systemctl start keepalived
systemctl reload haproxy

Keepalived Configuration for Backup Node

On the secondary load balancer, configure Keepalived with a lower priority:

global_defs {
    notification_email {
        admin@example.org
    }
    notification_email_from keepalived@example.org
    smtp_server 127.0.0.1
    smtp_connect_timeout 30
    router_id lb_backup
}

vrrp_instance VI_1 {
    state BACKUP
    interface eth0
    virtual_router_id 51
    priority 100
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass MyPass123
    }
    virtual_ipaddress {
        192.168.10.200
    }
}

Apply the same kernel parameter and HAProxy binding configuration:

echo 1 > /proc/sys/net/ipv4/ip_nonlocal_bind
systemctl restart haproxy
systemctl restart keepalived

Failover Verification

Access the virtual IP (192.168.10.200) from a browser. When stopping Keepalived on the master node, the virtual IP automatically migrates to the backup node. For automated health checks, implement monitoring scripts that trigger failover based on HAProxy or backend service availability.

Disaster Recovery Considerations

Enterprise disaster recovery planning should follow a structured approach:

• Classify systems into core and non-core business applications
• Categorize data by importance and recovery priority
• Build recovery procedures from the infrastructure layer upward
• Conduct regular disaster recovery drills
• Evaluate requirements for cold standby, warm standby, or active-active configurations