How to Configure Nginx as a TCP Load Balancer for Kubernetes Applications

Prerequisites

Assume a Kubernetes cluster is already deployed. Create a CentOS 8 Minimal virtual machine for Nginx (IP 192.168.1.50) and note the master (192.168.1.40) and two worker nodes (192.168.1.41, 192.168.1.42).

Step 1: Install the EPEL repository

dnf install epel-release -y

Step 2: Install Nginx

dnf install nginx -y

rpm -qi nginx

Step 3: Open firewall ports for HTTP and HTTPS

firewall-cmd --permanent --add-service=http
firewall-cmd --permanent --add-service=https
firewall-cmd --reload

Step 4: Set SELinux to permissive and reboot

sed -i s/^SELINUX=.*$/SELINUX=permissive/ /etc/selinux/config
reboot

Step 5: Retrieve NodePort values from the ingress‑nginx namespace

kubectl get all -n ingress-nginx

The output shows that each worker node maps NodePort 32760 to port 80 and NodePort 32375 to port 443. These ports will be used in the Nginx configuration.

Step 6: Configure Nginx for TCP load balancing

vim /etc/nginx/nginx.conf

In the configuration file, comment out the original server block (lines 38‑57) and add the following sections:

upstream backend {
    server 192.168.1.41:32760;
    server 192.168.1.42:32760;
}

server {
    listen 80;
    location / {
        proxy_read_timeout 1800;
        proxy_connect_timeout 1800;
        proxy_send_timeout 1800;
        send_timeout 1800;
        proxy_set_header Accept-Encoding "";
        proxy_set_header X-Forwarded-By $server_addr:$server_port;
        proxy_set_header X-Forwarded-For $remote_addr;
        proxy_set_header X-Forwarded-Proto $scheme;
        proxy_set_header Host $host;
        proxy_set_header X-Real-IP $remote_addr;
        proxy_pass http://backend;
    }
    location /nginx_status {
        stub_status;
    }
}

Step 7: Start and enable the Nginx service

systemctl start nginx
systemctl enable nginx

Step 8: Test the TCP load balancer

Deploy a simple Nginx deployment in the cluster and expose it as a NodePort:

kubectl create deployment nginx-deployment --image=nginx
kubectl expose deployment nginx-deployment --name=nginx-deployment --type=NodePort --port=80

Update the local /etc/hosts file so that nginx-lb.example.com resolves to the Nginx VM (192.168.1.50):

echo "192.168.1.50  nginx-lb.example.com" >> /etc/hosts

Access http://nginx-lb.example.com in a browser. The request is forwarded to the two worker nodes via their NodePort 32760, confirming that Nginx successfully balances TCP traffic across the Kubernetes nodes.

Conclusion

The experiment demonstrates that Nginx can act as a TCP load balancer for Kubernetes workloads, distributing traffic on port 80 to the underlying worker nodes’ NodePort services. This approach extends Nginx’s capabilities beyond HTTP/HTTPS and provides a simple, open‑source solution for TCP load balancing in cloud‑native environments.