Getting Started

One of the best ways of getting started with a new platform is to try it out. Any big data platform has a lot of moving parts and getting some hands on keyboard time with it helps reinforce learning.

About this guide

Firstly, let’s cover whether this Getting Started guide is right for you. This is intended as a learning tool to discover more about Stackable, its deployment and architecture.

  • If you just want to get up and running quickly there is a quickstart script that will install services on a single node available in the stackable-utils repository on GitHub.

  • If you want to build a production cluster then this is not for you. This tutorial is to familiarise you with the Stackable architecture and is not a guide for building robust clusters.

  • This is intended for use in a private network or lab; it doesn’t enable many security features such as authentication or encryption and should not be directly connected to the Internet. Be careful if you’re deploying in the cloud as your instances may default to using public IPs.

Overview

Stackable is based on Kubernetes and uses this as the control plane to manage clusters. In this guide we will build a simple cluster with 3 services; Apache ZooKeeper, Apache Kafka and Apache NiFi.

Installing Kubernetes

Stackable’s control plane is built around Kubernetes, and we’ll give some brief examples of how to install Kubernetes on your machine.

Installing kubectl

Stackable operators and their services are managed by applying manifest files to the Kubernetes cluster. For this purpose, you need to have the kubectl tool installed. Follow the instructions here for your platform.

Installing Kubernetes using Kind

Kind offers a very quick and easy way to bootstrap your Kubernetes infrastructure in Docker. The big advantage of this is that you can simply remove the Docker containers when you’re finished and clean up easily, making it great for testing and development.

If you don’t already have Docker then visit Docker Website to find out how to install Docker. Kind is a single executable that performs the tasks of installing and configuring Kubernetes for you within Docker containers. The Kind Website has instructions for installing Kind on your system.

Once you have both of these installed then you can build a Kubernetes cluster in Docker. We’re going to create a 2 node cluster to test out Stackable, one node hosting the Kubernetes control plane and the other hosting the Stackable services.

kind create cluster --name quickstart --config - << EOF
---
kind: Cluster
apiVersion: kind.x-k8s.io/v1alpha4
nodes:
- role: control-plane
- role: worker
  kubeadmConfigPatches:
    - |
      kind: JoinConfiguration
      nodeRegistration:
        kubeletExtraArgs:
          node-labels: node=quickstart-1
EOF

Installing Kubernetes using K3s

K3s provides a quick way of installing Kubernetes. On your control node run the following command to install K3s:

curl -sfL https://get.k3s.io | sh -s - --node-label 'node=quickstart-1' --write-kubeconfig-mode 644

So long as you have an Internet connection K3s will download and automatically configure a simple Kubernetes environment.

Create a symlink to the Kubernetes configuration from your home directory to allow tools like Helm to find the correct configuration.

mkdir ~/.kube
ln -s /etc/rancher/k3s/k3s.yaml ~/.kube/config

Testing your Kubernetes installation

To check if everything worked as expected you can use kubectl cluster-info to retrieve the cluster information. The output should look similar to:

Kubernetes control plane is running at https://127.0.0.1:6443
CoreDNS is running at https://127.0.0.1:6443/api/v1/namespaces/kube-system/services/kube-dns:dns/proxy
Metrics-server is running at https://127.0.0.1:6443/api/v1/namespaces/kube-system/services/https:metrics-server:/proxy

Installing Helm

Stackable uses Helm as the package manager for its Kubernetes operators. This greatly simplifies the deployment and management of Kubernetes operators and CRDs. The quickest way to install Helm is to run the following command:

curl -sfL https://raw.githubusercontent.com/helm/helm/main/scripts/get-helm-3 | /bin/bash -

Installing Stackable

Install Stackable Helm Repositories

With Helm installed you can add the Stackable operator repo, where the helm charts to install Stackable operators can be found. There are development, test and a stable repositories available, we’ll be using the stable repo in this guide.

helm repo add stackable-stable https://repo.stackable.tech/repository/helm-stable/

Installing Stackable Operators

The Stackable operators are components that translate the service definitions deployed via Kubernetes into deploy services on the worker nodes. These can be installed on any node that has access to the Kubernetes control plane. In this example we will install them on the controller node.

Stackable operators are installed using Helm charts. Run the following commands to install the operators for ZooKeeper, Kafka and NiFi using the repo configured earlier. The --devel flag will choose the latest available version; alternatively the --version flag can be used to deploy a specific version.

Install operators from the Stackable repository

helm install zookeeper-operator stackable-stable/zookeeper-operator --version=0.9.0
helm install kafka-operator stackable-stable/kafka-operator --version=0.5.0
helm install secret-operator stackable-stable/secret-operator --version=0.2.0
helm install nifi-operator stackable-stable/nifi-operator --version=0.5.0

You can check which operators are installed using helm list:

NAME              	NAMESPACE	REVISION	UPDATED                                	STATUS  	CHART                   	APP VERSION
kafka-operator    	default  	1       	2022-02-15 08:17:26.84659409 +0000 UTC 	deployed	kafka-operator-0.5.0    	0.5.0
nifi-operator     	default  	1       	2022-02-15 08:17:37.93720808 +0000 UTC 	deployed	nifi-operator-0.5.0     	0.5.0
secret-operator   	default  	1       	2022-02-15 08:17:32.700301793 +0000 UTC	deployed	secret-operator-0.2.0   	0.2.0
zookeeper-operator	default  	1       	2022-02-15 08:17:17.893844595 +0000 UTC	deployed	zookeeper-operator-0.9.0	0.9.0

Deploying Stackable Services

At this point you’ve successfully deployed Kubernetes and the Stackable operators we need and are ready to deploy services to the cluster. To do this we provide service descriptions to Kubernetes for each of the services we wish to deploy.

Apache ZooKeeper

We will deploy an Apache ZooKeeper instance to our cluster.

kubectl apply -f - <<EOF
---
apiVersion: zookeeper.stackable.tech/v1alpha1
kind: ZookeeperCluster
metadata:
  name: simple-zk
spec:
  servers:
    roleGroups:
      primary:
        replicas: 1
        config:
          myidOffset: 10
  version: 3.5.8
EOF

Apache Kafka

We will deploy an Apache Kafka broker that depends on the ZooKeeper service we just deployed. The zookeeperReference property below points to the namespace and name we gave to the ZooKeeper service deployed previously.

kubectl apply -f - <<EOF
---
apiVersion: kafka.stackable.tech/v1alpha1
kind: KafkaCluster
metadata:
  name: simple-kafka
spec:
  version: 2.8.1
  zookeeperConfigMapName: simple-kafka-znode
  brokers:
    roleGroups:
      brokers:
        replicas: 1
        selector:
          matchLabels:
            node: quickstart-1
---
apiVersion: zookeeper.stackable.tech/v1alpha1
kind: ZookeeperZnode
metadata:
  name: simple-kafka-znode
spec:
  clusterRef:
    name: simple-zk
    namespace: default
EOF

Apache NiFi

We will next deploy an Apache NiFi server.

kubectl apply -f - <<EOF
---
apiVersion: zookeeper.stackable.tech/v1alpha1
kind: ZookeeperZnode
metadata:
  name: simple-nifi-znode
spec:
  clusterRef:
    name: simple-zk
---
apiVersion: v1
kind: Secret
metadata:
  name: nifi-admin-credentials-simple
stringData:
  username: admin
  password: AdminPassword
---
apiVersion: nifi.stackable.tech/v1alpha1
kind: NifiCluster
metadata:
  name: simple-nifi
spec:
  version: "1.15.0"
  zookeeperConfigMapName: simple-nifi-znode
  authenticationConfig:
    method:
      SingleUser:
        adminCredentialsSecret:
          name: nifi-admin-credentials-simple
        autoGenerate: true
  sensitivePropertiesConfig:
    keySecret: nifi-sensitive-property-key
    autoGenerate: true
  nodes:
    roleGroups:
      default:
        selector:
          matchLabels:
            node: quickstart-1
        config:
          log:
            rootLogLevel: INFO
        replicas: 1
EOF

You can check the status of the services using kubectl get pods. This will retrieve the status of all pods running in the default namespace.

NAME                                             READY   STATUS    RESTARTS   AGE
nifi-operator-deployment-64c98c779c-nw6h8        1/1     Running   0          24m
kafka-operator-deployment-54df9f86c7-psqgd       1/1     Running   0          24m
zookeeper-operator-deployment-767458d4f5-2czb9   1/1     Running   0          24m
secret-operator-daemonset-pddkv                  2/2     Running   0          24m
simple-zk-server-primary-0                       1/1     Running   0          23m
simple-kafka-broker-brokers-0                    2/2     Running   0          21m
simple-nifi-node-default-0                       1/1     Running   0          22m

Since this is the first time that each of these services has been deployed to these nodes, it will take some time to download the software from the Stackable repository and deploy the services. Once all of the pods are in the running state your cluster is ready to use.

Testing your cluster

If all has gone well then you will have successfully deployed a Stackable cluster and used it to start three services that should now be ready for you.

Apache ZooKeeper

We can test ZooKeeper by running the ZooKeeper CLI shell. The easiest way to do this is to run the CLI shell on the pod that is running ZooKeeper.

kubectl exec -i -t simple-zk-server-primary-0 -- bin/zkCli.sh

The shell should connect automatically to the ZooKeeper server running on the pod. You can run the ls / command to see the list of znodes in the root path, which should include those created by Apache Kafka and Apache NiFi.

[zk: localhost:2181(CONNECTED) 0] ls /
[nifi, znode-17b28a7e-0d45-450b-8209-871225c6efa1, zookeeper]

Apache Kafka

To test Kafka we’ll create a topic, and verify that it was created. First create the topic with the following command:

kubectl exec -i -t simple-kafka-broker-brokers-0 -c kafka -- \
  bin/kafka-topics.sh --bootstrap-server localhost:9092 --create --topic demo
Created topic demo.

Now let’s check if it was actually created:

kubectl exec -i -t simple-kafka-broker-brokers-0 -c kafka -- \
  bin/kafka-topics.sh --bootstrap-server localhost:9092 --list
demo

Apache NiFi

Apache NiFi provides a web interface and the easiest way to test it is to view this in a web browser. To access the web interface we first need to get the ip address and port Nifi is listening on. To get the IP address we need to connect to (in this case 172.18.0.2), run:

kubectl get nodes --selector=node=quickstart-1 -o wide
NAME                STATUS   ROLES    AGE   VERSION   INTERNAL-IP   EXTERNAL-IP   OS-IMAGE       KERNEL-VERSION      CONTAINER-RUNTIME
quickstart-worker   Ready    <none>   45m   v1.21.1   172.18.0.2    <none>        Ubuntu 21.04   5.15.0-25-generic   containerd://1.5.2

With the following command we get the port (in this case 30247):

kubectl get svc simple-nifi
NAME          TYPE       CLUSTER-IP    EXTERNAL-IP   PORT(S)          AGE
simple-nifi   NodePort   10.43.75.25   <none>        8443:30247/TCP   49m

Browse to the address of your Kubernetes node on port 30247 e.g. https://172.18.0.2:30247/nifi and you should see the NiFi login screen.

The Apache NiFi web interface login screen

The Apache NiFi operator will automatically generate the admin user credentials with a random password and store it as a Kubernetes secret in order to provide some security out of the box. You can retrieve this password for the admin user with the following kubectl command.

kubectl get secrets nifi-admin-credentials-simple \
-o jsonpath="{.data.password}" | base64 -d && echo

Once you have these credentials you can log in and you should see a blank NiFi canvas.

The Apache NiFi web interface canvas