Deploy the services-kafka helm chart on Azure Managed Openshift Service

Question

Pulumi · Accepted Answer

To deploy the `services-kafka` Helm chart on Azure Managed OpenShift Service using Pulumi, you will need to set up an Azure Managed OpenShift cluster first, and then use the Helm chart resource in Pulumi to deploy to that cluster.

### Setting up an Azure Managed OpenShift Cluster

Before deploying the Helm chart, you need an operational OpenShift cluster. With Pulumi, you can provision an Azure Red Hat OpenShift (ARO) cluster using the `azure-native.redhatopenshift.OpenShiftCluster` resource.

Here's an overview of the process:
1. Define an OpenShift cluster resource with the necessary configurations like location, resource group, cluster profile (version, pull secret, etc.), network profile, and master/worker node profiles.
2. After the cluster is provisioned, you need to connect to the cluster, usually by obtaining the `kubeconfig` file that contains the credentials to interact with your cluster.

### Deploying the `services-kafka` Helm Chart

Once you have your OpenShift cluster up and running, you can deploy a Helm chart to it using the `kubernetes.helm.v3.Chart` resource in Pulumi.

Steps for deploying the Helm chart:
1. Initialize the Pulumi Kubernetes provider with the OpenShift cluster's kubeconfig.
2. Define a Helm chart resource, specifying the chart name (`services-kafka`), repository, and version.

In the following TypeScript example, we illustrate how to create an Azure Managed OpenShift Service and then deploy the `services-kafka` Helm chart to it. Please note that since the creation of an ARO cluster and its setup can take a considerable amount of time, the operation might exceed Pulumi's default timeout, so it's usually a good idea to handle these operations separately. However, for simplicity, this example assumes they are done within a single Pulumi program.

```typescript
import * as azure from "@pulumi/azure-native";
import * as pulumi from "@pulumi/pulumi";
import * as k8s from "@pulumi/kubernetes";

const clusterName = "aro-cluster";
const resourceGroupName = "aro-resource-group";
const location = "East US"; // Choose the Azure region right for you

// Create an Azure resource group
const resourceGroup = new azure.core.ResourceGroup(clusterName + "-rg", {
  location: location,
  name: resourceGroupName,
});

// Create an Azure Red Hat OpenShift cluster
const openshiftCluster = new azure.redhatopenshift.OpenShiftCluster(clusterName, {
  location,
  resourceName: clusterName,
  resourceGroupName: resourceGroupName,
  clusterProfile: {
    domain: "example",
    version: "4.7", // Specify the version of OpenShift you wish to deploy
  },
  masterProfile: {
    vmSize: "Standard_D2s_v3", // Choose the desired VM size for the master nodes
  },
  workerProfiles: [
    {
      name: "worker", // Define at least one worker profile
      vmSize: "Standard_D2s_v3", // Choose the desired VM size for worker nodes
      count: 3, // Specify the number of worker nodes
    },
  ],
  tags: {
    project: "kafka-services",
  },
});

// Pulumi Kubernetes provider that uses the kubeconfig to connect to the Openshift cluster
const k8sProvider = new k8s.Provider(clusterName, {
  kubeconfig: openshiftCluster.kubeconfig.apply((c) => c),
});

// Deploy the `services-kafka` helm chart
const kafkaHelmChart = new k8s.helm.v3.Chart("services-kafka", {
  chart: "services-kafka",
  // You might need to specify the repository if it's not a stable chart from 'https://charts.helm.sh/stable'
  repositoryOpts: {
    repo: "http://my-helm-chart-repo/",
  },
  version: "0.1.0", // Replace with the actual chart version
}, { provider: k8sProvider });

// Export the cluster's kubeconfig
export const kubeConfigOutput = openshiftCluster.kubeconfig;
```

### Important Notes:

- When selecting the Azure region for the cluster, ensure that OpenShift is supported in that region.
- To deploy a chart from a custom repository, you will need to specify the `repo` field within `repositoryOpts`.
- Define the appropriate VM sizes and node count that suit your workload and budget.
- You must replace placeholder values (like `version` in the `clusterProfile` or `repo` in the Helm chart definition) with actual values according to your requirements.
- Gather the necessary pull secret from Red Hat for the `clusterProfile` which is needed to pull OpenShift container images. This is not included in the code example for brevity and security reasons.
- The `kubeconfig` obtained from the `openshiftCluster` resource is used by the Pulumi Kubernetes provider to deploy the Helm chart.
- The above example assumes you have the necessary permissions to create resources within your Azure subscription.

This Pulumi program creates an Azure Managed OpenShift Service cluster and deploys the `services-kafka` Helm chart to it. Remember that running this program will result in actual resources being created and will incur costs in your Azure account. Ensure you have adequate permissions and your Pulumi access tokens and Azure credentials configured on your local machine.

Deploy the services-kafka helm chart on Azure Managed Openshift Service

Setting up an Azure Managed OpenShift Cluster

Deploying the services-kafka Helm Chart

Important Notes:

Deploying the `services-kafka` Helm Chart