Importing Data into Longhorn with Busybox

Published on Sep 14, 2024

Recently, I deployed my Kubernetes cluster in my homelab and decided to use Longhorn as the storage provider. One question that came up was: How do I import my existing data into Longhorn volumes?

In this guide, I’ll walk you through the process of exporting data from an existing container, creating a Longhorn volume, and importing that data into the new volume. Let’s get started!

Step 1: Export Data from an Existing Container or Volume

First, we need to export the data from the existing container.

Open a Shell in the container you want to export data from:

sudo docker exec -it gitea /bin/sh

Navigate to the directory where your data is stored and compress it into a tarball for easy copying:

cd /data/
tar -czf gitea.tar.gz .

Exit the shell and use docker cp to copy the tarball to your host

docker cp gitea:/data/gitea.tar.gz .

Now that we have exported the data, let’s move on to creating the Longhorn volume where we will import this data.

Step 2: Create a Volume on Longhorn

There are different ways to create a volume in Longhorn. For simplicity, we’ll use the Longhorn UI to create a Persistent Volume Claim (PVC).

Create a Persistent Volume Claim (PVC)

After creating the PVC via the Longhorn UI, you can verify its existence in the desired Kubernetes namespace:

kubectl get pvc -n gitea
NAME         STATUS   VOLUME       CAPACITY   ACCESS MODES   STORAGECLASS      AGE
gitea-data   Bound    gitea-data   10Gi       RWO            longhorn-static   22s

Once the PVC is bound, you’re ready to import the data into the Longhorn volume.

Step 3: Create a Pod to Import the Data

To copy the data to the Longhorn volume, we’ll need a pod that can mount the volume. For this example, we’ll use a simple busybox pod, but you can use any container you prefer.

Below is a YAML file to create a busybox pod that will sleep for an hour, giving us time to copy and extract the data.

---
apiVersion: v1
kind: Pod
metadata:
  name: busybox
  labels:
    app: busybox
spec:
  containers:
    - image: busybox
      name: busybox
      command:
        - sleep
        - "3600"
      volumeMounts:
        - name: data
          mountPath: "/data"
  volumes:
    - name: data
      persistentVolumeClaim:
        claimName: gitea-data

Apply the YAML file using kubectl:

kubectl apply -f busybox.yaml -n gitea

Verify the Pod is Running

Check that the pod is up and running with the following command:

$ kubectl get pods -n gitea
NAME                     READY   STATUS    RESTARTS   AGE
busybox                  1/1     Running   0          70s
gitea-5df8b6db65-6fkc8   1/1     Running   0          25m

Now that the pod is running, we can proceed with copying the data.

Step 4: Copy the Tar File into the Pod

Use the kubectl cp command to copy the tar file into the busybox pod, specifically to the /data directory:

kubectl cp -n gitea ./gitea.tar.gz busybox:/data/

Open a shell in the busybox pod and extract the Tar File

kubectl exec -n gitea --stdin --tty busybox -- /bin/sh
cd /data
tar -xzf gitea.tar.gz

Step 5: Clean Up

Once the data is successfully imported, delete the Tar file and delete the busybox pod to free up resources. If you used the provided YAML file, simply run:

$ kubectl delete -f busybox.yaml -n gitea

That’s It!

You’ve now imported your existing data into a Longhorn volume. You can deploy your applications and attach the Longhorn volume as needed.

With this approach, you can easily migrate your data from old containers or volumes to new Longhorn-backed PVCs in your Kubernetes cluster.

Feel free to customize this process depending on your setup. Longhorn makes managing persistent storage in Kubernetes seamless, even in homelabs!