Using Load Balancers in QBO - A Step-by-Step Guide

QBO, with its containerized architecture and metal performance, simplifies networking and service management in Kubernetes. One of its standout features is the management of kernel-level load balancing using IPVS (IP Virtual Server), providing a robust and efficient way to handle network traffic.

Why Use Load Balancers in QBO?

Load balancers in QBO ensure:

  • Kernel-Level Efficiency with IPVS: By leveraging IPVS, QBO manages load balancing directly at the kernel level, offering faster and more efficient packet handling compared to user-space alternatives.
  • Transparency: Load balancers are deployed using standard Kubernetes constructs, making the integration seamless and requiring no additional user effort.
  • Efficient Traffic Distribution: Balance incoming requests across multiple nodes or pods with minimal latency.
  • Scalability: Adapt to increased traffic or node configurations effortlessly.

Benefits of IPVS Load Balancers

QBO utilizes IPVS, a highly optimized kernel module for load balancing. The advantages of this approach include:

  • Low Overhead: Since IPVS operates in the kernel space, it minimizes CPU and memory usage compared to user-space load balancers.
  • High Throughput: Capable of handling a large number of concurrent connections with minimal latency.
  • Advanced Scheduling Algorithms: Support for round-robin, least connection, and other algorithms ensures flexible and efficient traffic distribution.
  • Seamless Integration: As the load balancers are managed via Kubernetes constructs, users don’t need to configure IPVS directly, simplifying deployment.

Steps to Use Load Balancers in QBO

1. Verify the QBO CLI Installation

Ensure the QBO CLI is installed:

qbo version | jq .version[]?

2. Add a Kubernetes Cluster

Add a cluster named lb for managing load balancers:

qbo add cluster lb -i hub.docker.com/kindest/node:v1.32.0 | jq

3. Retrieve Kubernetes Nodes

List all nodes in the cluster to verify configuration:

qbo get nodes lb | jq .nodes[]?

4. Configure Kubernetes

Set up the Kubernetes configuration for the cluster:

Cloud Environment:

qbo get cluster lb -k | jq -r '.output[]?.kubeconfig | select( . != null)' > $HOME/.qbo/lb.conf
export KUBECONFIG=$HOME/.qbo/lb.conf

On-Premises:

export KUBECONFIG=/tmp/qbo/lb.conf

5. Deploy and Test Services with Load Balancers

Apply Nginx Test Configurations

The script tests different configurations using Nginx services:

kubectl apply -f nginx/nginx-1-80.yaml
kubectl apply -f nginx/nginx-1-80-8080.yaml
kubectl apply -f nginx/nginx-1-8080.yaml
kubectl apply -f nginx/nginx-2-80.yaml

Check Load Balancer IPs

Monitor the load balancer IP addresses for each service:

kubectl get svc nginx-service-1 -o json | jq -r '.spec.externalIPs[0]'

Validate Service Availability

Use curl to check if the service is accessible via the load balancer:

curl -k -m 3 http://<LB_IP>:<PORT>

6. Cleanup and Reconfigure Services

Delete Services

Remove specific configurations as needed:

kubectl delete -f nginx/nginx-1-8080.yaml
kubectl delete -f nginx/nginx-2-80.yaml

Reapply and Patch Services

Reapply configurations and change service types dynamically:

kubectl apply -f nginx/nginx-1-80.yaml
kubectl patch svc nginx-service-1 -p '{"spec": {"type": "ClusterIP"}}'
kubectl patch svc nginx-service-1 -p '{"spec": {"type": "LoadBalancer"}}'
kubectl patch svc nginx-service-1 -p '{"spec": {"type": "NodePort"}}'
kubectl patch svc nginx-service-1 -p '{"spec": {"type": "LoadBalancer"}}'

Final Notes

QBO’s integration with IPVS for kernel-level load balancing ensures high performance, scalability, and low overhead. By abstracting these capabilities into Kubernetes constructs, QBO allows users to deploy and manage load balancers effortlessly. Whether you’re scaling a service or optimizing traffic distribution, QBO makes the process seamless and efficient.