Installing Apache Superset on IBM Kubernetes Cluster

Posted on | In , ,
Reading time 4

Superset is a great alternative to Tableau. Installing Superset on an IBM Kubernetes cluster cannot be done just with an out-of-the-box Helm chart. This guide is meant to help you get started.

Before you begin

The following sections assume that all necessary CLI tools are installed, and there is an IBM Kubernetes cluster already running. The following CLI tools are also required:

  1. kubectl
  2. ibmtools
  3. helm

The IBM Clusters tutorial provides instructions for creating a cluster and installing kubectl and ibmtools on your machine.

The official Helm documentation will guide you through installing helm on your machine (scroll down to find your operating system).

You will also need to configure helm with a Chart repository, so that you can download the charts we will be using in this guide.

# Add the official Helm repository and update
helm repo add stable https://kubernetes-charts.storage.googleapis.com/
helm repo update

Install

We are ready to install Superset. We will create a custom StorageClass to make Superset work with with IBM Cloud. We will also create a Persistent Volume Claim (PVC) to be used by the Superset pod.

In the following sections we will first create the PVC and then move on to the Helm chart.

Create StorageClass and PersistentVolumeClaim

We will copy the default class ibmc-file-gold to make it use a specified Group ID (GID), so that the pod can write to the volume after mounting.

In the same file we will define the PersistentVolumeClaim (PVC) which the Superset pod will mount.

Save the following as superset-sc-pvc.yaml:

# Copy of ibmc-file-gold, with a fixed GID
apiVersion: storage.k8s.io/v1beta1
kind: StorageClass
metadata:
  name: ibmc-file-gold-superset
  labels:
    kubernetes.io/cluster-service: "true"
provisioner: ibm.io/ibmc-file
reclaimPolicy: Delete
volumeBindingMode: Immediate
parameters:
  billingType: hourly
  classVersion: "2"
  iopsPerGB: "10"
  sizeRange: '[20-4000]Gi'
  type: Endurance
# Note: `superset` group ID is 1001
  gidAllocate: "true"
  gidFixed: "1001"

---

apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  annotations:
    volume.beta.kubernetes.io/storage-provisioner: ibm.io/ibmc-file
  finalizers:
    - kubernetes.io/pvc-protection
  labels:
    app: superset
    region: eu-gb
    zone: lon02
  name: superset-pvc
  namespace: default
spec:
  accessModes:
    - ReadWriteOnce
  resources:
    requests:
      storage: 8Gi
  storageClassName: ibmc-file-gold
  volumeMode: Filesystem

And create the resource:

kubectl apply -f superset-sc-pvc.yaml

Verify that the resources were created correctly by using describe:

kubectl describe storageclass ibmc-file-gold-superset
kubectl describe pvc superset-pvc

Note: Some errors might not show up until the pod has claimed the PVC. Keep this in mind in case you need to troubleshoot later.

Wait a few minutes until the cluster has provisioned the Persistent Volume.

We now have a Persistent Volume ready to be used by Superset. In the next section we will download and modify the official Helm chart to get our deployment running.

Modifying the Helm Chart

Run the following command to download the Superset Chart locally:

helm pull stable/superset --untar

You should now have a local directory named superset containing the Superset chart definition.

# Change to the Chart directory
cd superset
helm install \
  --set persistence.enabled=true \
  --set persistence.existingClaim=superset-pvc \
  superset \
  ./

Your output should look like this:

NAME: superset
LAST DEPLOYED: Wed Oct  7 15:19:07 2020
NAMESPACE: default
STATUS: deployed
REVISION: 1
TEST SUITE: None
NOTES:
Superset can be accessed via port 9000 on the following DNS name from within your cluster:
superset.default.svc.cluster.local

Initially you can login with username/password: admin/admin.

Initialization will take a few minutes. You can monitor progress by watching the pod’s logs:

kubectl get pods
# Get the pod name. In this example it is `superset-798c9c7fbd-k9lsr`
kubectl logs -f superset-798c9c7fbd-k9lsr

You have now finished setup.

Set admin credentials

Start a shell inside the Superset container

kubectl exec --stdin --tty superset-798c9c7fbd-k9lsr /bin/bash

In the container shell, type the following to reset the admin credentials:

export FLASK_APP=superset
flask fab create-admin superset

You will need to enter the new credentials interactively. When finished, the output should look like this:

Recognized Database Authentications.
Admin User admin created.

That means your admin credentials have been reset. You’re ready to connect to the web UI.

Connect to Web UI

Finally, start a port-forwarding session via kubectl to connect to the login page.

kubectl port-forward svc/superset-demo 8080:9000

Visit localhost:8080 and use the credentials from the previous step.

Congratulations, you have set up Superset on IBM Cloud!