Get started with AKO locally

This tutorial describes how to create an Aerospike Database Enterprise Edition deployment using the Aerospike Kubernetes Operator (AKO) on a local Kubernetes cluster with Minikube or MicroK8s.

Minikube creates a temporary, single-node cluster that you start and stop as needed. MicroK8s installs as a persistent, lightweight Kubernetes distribution that runs continuously on your machine. The setup and deployment steps are nearly identical — where they differ, choose the tab that applies to your platform.

Tip

If you’re on macOS and want to try AKO quickly, use Minikube. MicroK8s on macOS requires a Multipass VM, which adds extra setup. MicroK8s is simpler on Linux, where it runs natively.

Prerequisites

• Minikube or MicroK8s installed
• Helm
• Git
• Kubectl
• An Aerospike Enterprise Edition feature-key file
  • Download a free Enterprise Edition feature-key file good for 60 days at this link: Get Started with Aerospike

This tutorial assumes basic conceptual knowledge of Kubernetes.

Pre-install

Start your local cluster and verify that your terminal can communicate with it.

Minikube

Start a new Minikube cluster. If you already have one running, you can skip the minikube start command.

# Start a local Kubernetes cluster
minikube start

# Check connectivity and versions
kubectl cluster-info                 # API-server reachable?
kubectl get nodes -o wide            # Node visible?
helm version --short                 # Helm installed?
git --version                        # Git installed?

MicroK8s

MicroK8s runs as a persistent service after installation. Enable the required add-ons and export its kubeconfig so that standard kubectl and helm commands work.

# Enable required add-ons
microk8s enable dns hostpath-storage

# Export kubeconfig for standard kubectl access
mkdir -p ~/.kube
microk8s config > ~/.kube/config

# Check connectivity and versions
kubectl cluster-info                 # API-server reachable?
kubectl get nodes -o wide            # Node visible?
helm version --short                 # Helm installed?
git --version                        # Git installed?

Caution

The microk8s config command overwrites your existing ~/.kube/config. If you manage other clusters, merge the configuration manually or use the KUBECONFIG environment variable instead.

Install AKO

In this section, you use Helm to install AKO on your Kubernetes cluster and configure the Kubernetes namespace to watch for your Aerospike Database deployment.

1. Add the JetStack Helm repository so you can install cert-manager, a utility that AKO relies on to manage certificates.

   helm repo add jetstack https://charts.jetstack.io --force-update

2. Install the cert-manager Helm chart.

   helm install cert-manager jetstack/cert-manager --namespace cert-manager --create-namespace --version v1.17.0 --set crds.enabled=true

3. Add the AKO Helm repository.

   helm repo add aerospike https://aerospike.github.io/aerospike-kubernetes-enterprise

4. Install AKO to your cluster. The watchNamespaces parameter tells AKO which Kubernetes namespace to monitor for AerospikeCluster custom resources.

   helm install aerospike-kubernetes-operator aerospike/aerospike-kubernetes-operator --version=4.3.0 --set watchNamespaces="aerospike"

AKO is now running on your cluster and is ready to create a new Aerospike Database deployment.

Deploy an Aerospike database

In this section, you use kubectl to deploy Aerospike Database. This tutorial uses a pre-built sample configuration for Aerospike Database deployment.

1. Create a dedicated Kubernetes namespace for your Aerospike Database deployment. This must match the watchNamespaces value you set when installing AKO.

   kubectl create namespace aerospike

2. Download the Aerospike Kubernetes Operator repository from GitHub to your local machine. This repository contains sample configuration files for an Aerospike Database deployment. You can modify these files on your local machine before using kubectl to apply the changes to the cluster.

   git clone https://github.com/aerospike/aerospike-kubernetes-operator.git

   The only directory you need to interact with during this tutorial is config/samples/ as shown in the following diagram:

   • Directoryaerospike-kubernetes-operator

     • Directoryapi/

       • …
     • Directorycmd/

       • …
     • Directoryconfig/

       • Directorysamples/

         • …
       • … (sample configuration files)
     • … (other directories)

3. Navigate to the repository and check out the release tag that matches your AKO version.

   cd aerospike-kubernetes-operator
   git checkout v4.3.0

   Your working directory should remain aerospike-kubernetes-operator/ for the rest of this tutorial. All config/samples/... paths are relative to this directory.

4. Copy your feature-key file, typically named features.conf, into the existing config/samples/secrets/ directory. Aerospike Enterprise Edition requires this file to start.

5. Create the Kubernetes Secrets that the Aerospike cluster needs at runtime. The first secret loads everything in the secrets directory, including your feature-key file. The second sets a placeholder initial password for the Aerospike admin user.

   kubectl -n aerospike create secret generic aerospike-secret --from-file=config/samples/secrets
   kubectl -n aerospike create secret generic auth-secret --from-literal=password='admin123'

6. Create a ServiceAccount for AKO in the aerospike Kubernetes namespace. AKO uses this identity when it manages pods, services, and other resources for your database.

   kubectl -n aerospike create serviceaccount aerospike-operator-controller-manager

7. Bind the aerospike-cluster ClusterRole to the ServiceAccount you just created. This grants AKO the permissions it needs to manage Aerospike cluster resources.

   kubectl create clusterrolebinding aerospike-cluster --clusterrole=aerospike-cluster --serviceaccount=aerospike:aerospike-operator-controller-manager

8. Configure storage for the cluster. The sample CR expects a storage class named ssd for its persistent volume. Neither Minikube nor MicroK8s includes this class by default, so you must create it or apply a sample from the repository.

   Minikube

   Create an ssd storage class that uses Minikube’s built-in hostpath provisioner:

   kubectl apply -f - <<EOF
   apiVersion: storage.k8s.io/v1
   kind: StorageClass
   metadata:
     name: ssd
   provisioner: k8s.io/minikube-hostpath
   reclaimPolicy: Delete
   volumeBindingMode: Immediate
   EOF

   MicroK8s

   Apply the MicroK8s filesystem storage class from the sample files in the GitHub repository.

   kubectl apply -f config/samples/storage/microk8s_filesystem_storage_class.yaml

9. Create the Aerospike cluster by applying a sample Custom Resource (CR). AKO watches for this resource and automatically provisions the Aerospike Database pods. This sample stores data in memory, which is the simplest configuration for local development and testing.

   kubectl apply -f config/samples/dim_nostorage_cluster_cr.yaml

   Note

   This CR uses data-in-memory storage, so data does not persist across pod restarts. For persistent storage on MicroK8s, use a storage-backed CR file from the config/samples directory.

   After you apply the CR, the cluster starts initializing or updating.

10. Watch the cluster status. The -w flag streams updates so you can see the phase change in real time.

    kubectl get aerospikeclusters aerocluster -n aerospike -w

    Wait until the PHASE column shows Completed, then press Ctrl+C to stop watching. This can take a few minutes depending on the resources available.

    NAME          SIZE   IMAGE                                           MULTIPODPERHOST   HOSTNETWORK   AGE   PHASE
    aerocluster   2      aerospike/aerospike-server-enterprise:8.1.1.0   true                            2s    InProgress
    aerocluster   2      aerospike/aerospike-server-enterprise:8.1.1.0   true                            21s   InProgress
    aerocluster   2      aerospike/aerospike-server-enterprise:8.1.1.0   true                            21s   InProgress
    aerocluster   2      aerospike/aerospike-server-enterprise:8.1.1.0   true                            27s   InProgress
    aerocluster   2      aerospike/aerospike-server-enterprise:8.1.1.0   true                            28s   Completed

11. Find the access endpoints for your cluster. These are the host:port pairs that client applications use to connect to your Aerospike database.

    kubectl -n aerospike describe aerospikeclusters aerocluster | grep -E 'Access Endpoints|Alternate Access Endpoints' -A1

    You can use any of the Aerospike client libraries to write tests for reading and writing to this Aerospike Database.

You now have a locally running Aerospike Database deployment using AKO!

Cleanup

When you are finished with your local deployment, clean up the cluster resources.

Minikube

Delete the entire Minikube cluster. This removes all Kubernetes resources including the Aerospike deployment.

minikube delete

The next time you need a cluster, start fresh with minikube start.

MicroK8s

Since MicroK8s is a persistent installation, delete the Aerospike cluster and AKO resources rather than the entire Kubernetes environment.

kubectl delete -f config/samples/dim_nostorage_cluster_cr.yaml
helm uninstall aerospike-kubernetes-operator
helm uninstall cert-manager -n cert-manager
kubectl delete clusterrolebinding aerospike-cluster
kubectl delete ns aerospike

Next steps

The biggest difference between the cluster created in this tutorial and a production cluster is in the storage class and CR file configurations. See the AKO Configuration section to learn how to configure your deployment for your own application needs.

Become familiar with the Aerospike Backup Service (ABS) and the monitoring stack. They are separate services that run alongside the database in your cluster. ABS listens for REST requests to perform backups and restores of the database, while the monitoring stack lets you visualize cluster statistics on Grafana dashboards.