Deploy the voting-app helm chart on Rancher

Question

Pulumi · Accepted Answer

To deploy a Helm chart on Rancher using Pulumi, you must interact with the Rancher Kubernetes engine directly. Given that Pulumi does not have a direct means to deploy a Helm chart specifically via Rancher, you would typically use the Pulumi Kubernetes resource to deploy a Helm chart to the cluster managed by Rancher.

The deployment of a Helm chart on a Kubernetes cluster typically involves the following steps:

1. If not already available, add the required Helm repository that contains the chart.
2. Install the Helm chart into the Kubernetes cluster with the specific configuration required for your application.
3. Verify the resources created by the Helm chart in the cluster.

Here's how you might set up a Pulumi TypeScript program to deploy a Helm chart to a Kubernetes cluster managed by Rancher:

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

// Define the configuration for deploying the Helm chart.
// You would adjust 'chartVersion', 'chartName', 'releaseName', and 'namespace' according to your needs.
const config = new pulumi.Config();
const chartVersion = config.require("chartVersion"); // The version of the chart e.g., "1.0.0"
const chartName = config.require("chartName"); // The name of the chart e.g., "voting-app"
const releaseName = config.require("releaseName"); // The name of the release e.g., "voting-app-release"
const namespace = config.require("namespace"); // Kubernetes namespace e.g., "default"

// Assuming your Pulumi program is already configured to talk to your Kubernetes cluster,
// we proceed to deploy the Helm chart using the pulumi/kubernetes provider.
const helmChart = new k8s.helm.v3.Chart(releaseName, {
    version: chartVersion,
    chart: chartName,
    namespace: namespace,
    // If the Helm chart requires specific values, they can be provided here.
    // For example:
    // values: { 
    //   service: { type: "LoadBalancer" }
    // }
});

// Export any relevant data, such as the Kubernetes service endpoint, if the chart creates it.
// For example:
export const appServiceName = helmChart.getResourceProperty("v1/Service", `${releaseName}-voting-service`, "metadata").apply(m => m?.name);

// To run this Pulumi program, you would need to set the configuration parameters for example,
// with these commands:
// pulumi config set chartVersion "1.0.0"
// pulumi config set chartName "voting-app"
// pulumi config set releaseName "voting-app-release"
// pulumi config set namespace "default"

// Then, you execute `pulumi up` to deploy the Helm chart.
```

For a Helm chart deployment, you use the `@pulumi/kubernetes` package, which provides direct integration with Kubernetes to manage resources through Pulumi.

In the code above, `chartVersion`, `chartName`, `releaseName`, and `namespace` would be set to match the voting app's characteristics. When calling `new k8s.helm.v3.Chart()`, you detail the version of the chart, its name, and the namespace in which you want to deploy the chart. You can provide additional custom values to configure your Helm chart according to your needs.

Note that this program assumes you have a Kubernetes cluster to which you've configured `kubectl` access, and that the cluster is being managed by Rancher. Typically, that would mean using `rancher2` as a context or similar in your `kubeconfig` file.

After deploying your infrastructure with this Pulumi program, that infrastructure's state will be stored in Pulumi's servers. This way, subsequent updates to the infrastructure will be incremental and managed through the Pulumi CLI. Remember to run `pulumi up` to deploy the configuration.

Lastly, please replace the placeholder chart repository and chart name with the actual details of the "voting-app" Helm chart. If "voting-app" is a chart in a custom repository, you'll need to add that Helm repository to your configuration before you can successfully deploy the chart.