It's Huy

Introduction

Longhorn is a cloud-native distributed block storage solution for Kubernetes. It enables persistent storage for Kubernetes applications with built-in replication, backup, and recovery features. Unlike traditional storage solutions, Longhorn provides lightweight, highly-available storage with automatic failover and data redundancy. It is especially useful for stateless workloads requiring reliable persistent storage in Kubernetes.

In this guide, we’ll configure Longhorn on an existing Kubernetes cluster with one master node and two worker nodes running Debian 12. We’ll also expose the Longhorn UI for easier management. Since we assume MetalLB is already deployed, we will leverage it for LoadBalancer service exposure.

Prerequisites

Before proceeding, ensure the following:

• A running Kubernetes cluster with one master node and two worker nodes
• kubectl installed on your local machine
• Helm installed on the master node
• SSH access to all nodes
• open-iscsi installed on all nodes (required for Longhorn to function properly)
• MetalLB already deployed for LoadBalancer service exposure

Install open-iscsi on all nodes

Run the following command on each node:

sudo apt update && sudo apt install -y open-iscsi
sudo systemctl enable --now iscsid

Verify that the iscsid service is running:

systemctl status iscsid

Procedure

Step 1: Deploy Longhorn

Add the Helm Repository

helm repo add longhorn https://charts.longhorn.io
helm repo update

Install Longhorn

kubectl create namespace longhorn-system
helm install longhorn longhorn/longhorn --namespace longhorn-system

Step 2: Verify Longhorn Installation

Check if all Longhorn components are running:

kubectl get pods -n longhorn-system

Ensure all pods are in a Running state before proceeding.

Step 3: Expose Longhorn UI

Since MetalLB is deployed, we will expose the Longhorn UI via a LoadBalancer service:

kubectl patch svc longhorn-frontend -n longhorn-system -p '{"spec": {"type": "LoadBalancer"}}'

Get the assigned external IP:

kubectl get svc longhorn-frontend -n longhorn-system

Access the UI using http://<load-balancer-ip>.

Step 4: Configure Longhorn

1. Navigate to the Longhorn UI.
2. Set up the default storage class.
3. Configure backup settings if needed.
4. Attach volumes to workloads as necessary.

Troubleshooting

Issue: Longhorn Pods Not Running on a New Worker Node

If longhorn-csi-plugin and longhorn-manager pods on the new worker node are not running, there is a high chance that the cluster DNS service is not functioning properly. To resolve this issue, restart the coredns service:

kubectl -n kube-system rollout restart deployment coredns

Wait a few moments and check if the Longhorn pods on the new worker node start running:

kubectl get pods -n longhorn-system -o wide

If the issue persists, check the logs of the affected pods for further debugging:

kubectl logs -n longhorn-system <pod-name>

Conclusion

We have successfully installed Longhorn on an existing Kubernetes cluster running Debian 12. Longhorn provides high availability, data replication, and snapshots, making it an excellent choice for cloud-native workloads. By leveraging MetalLB, we have exposed the Longhorn UI using a LoadBalancer service, making it easily accessible. You can now explore its features, including backup and restore, volume management, and monitoring, through its UI.