Deploy the stackdriver-metrics-adapter helm chart on Rancher

Question

Pulumi · Accepted Answer

I'll guide you through deploying the Stackdriver Metrics Adapter as a Helm chart on a Rancher-managed Kubernetes cluster. To achieve this, you'll need to set up Rancher, create a Kubernetes cluster, then deploy the Helm chart to it.

We'll be using the `rancher2` Pulumi provider to interact with Rancher. This provider allows us to manage Rancher resources such as clusters, catalogs, apps, among others. In this scenario, we're interested specifically in deploying a Helm chart through Rancher, so we'll focus on that.

First, ensure that you have Pulumi installed, along with the necessary cloud provider CLI (in this case, it could be GCP for Stackdriver, which we'll not detail here) and have authenticated to both Pulumi and Rancher.

You would also need to input or configure your Rancher credentials securely, such as API tokens. Take note that the code given here assumes that you've already set up a project and a Kubernetes cluster in Rancher and that you've either obtained or managed the token outside of this code for security purposes.

Next, here's a Pulumi TypeScript program that simulates deploying the Stackdriver Metrics Adapter Helm chart on a Rancher-managed Kubernetes cluster:

```typescript
import * as rancher2 from "@pulumi/rancher2";

// Provide the cluster ID and project ID where the Helm chart will be deployed.
const clusterId = "CLUSTER_ID";
const projectId = "PROJECT_ID";

// Create a Catalog V2 resource to manage Helm chart repositories in Rancher. 
// You would add the Stackdriver Metrics Adapter Helm repository here.
const stackdriverMetricsAdapterCatalog = new rancher2.CatalogV2("stackdriver-metrics-adapter-catalog", {
    clusterId: clusterId,
    url: "HELM_CHART_REPOSITORY_URL", // Substitute with the actual Helm Chart repository URL.
    gitBranch: "BRANCH_NAME", // Substitute with the repository branch name if applicable.
});

// Deploy the Stackdriver Metrics Adapter Helm chart using the AppV2 resource.
const stackdriverMetricsAdapterApp = new rancher2.AppV2("stackdriver-metrics-adapter-app", {
    clusterId: clusterId,
    projectId: projectId,
    // Specify the chart name, version, and repository from the catalog we've just created
    chartName: "stackdriver-metrics-adapter-chart",
    chartVersion: "CHART_VERSION", // Substitute with the correct chart version.
    namespace: "METRICS_ADAPTER_NAMESPACE", // Substitute with the namespace where the app will be installed.
    releaseName: "stackdriver-metrics-adapter",
    repoName: stackdriverMetricsAdapterCatalog.name, // Use the catalog name as the repo name.
    // Values to configure the Stackdriver Metrics Adapter.
    values: `values in YAML format`, // Provide the necessary configuration for Stackdriver Metrics Adapter.
});

// Export the app's name which confirms the deployment or can be used to track the state.
export const stackdriverMetricsAdapterAppName = stackdriverMetricsAdapterApp.name;
```

Before running the above code, make a few necessary substitutions:

- Replace `CLUSTER_ID`, `PROJECT_ID`, `HELM_CHART_REPOSITORY_URL`, `BRANCH_NAME`, `CHART_VERSION`, and `METRICS_ADAPTER_NAMESPACE` with your specific details.
- The `HELM_CHART_REPOSITORY_URL` should point to the URL where your Helm charts are stored. This could be a GitHub repository URL if the charts are hosted there.
- The `values in YAML format` string should contain your specific configuration values for the Stackdriver Metrics Adapter, encoded in YAML format. This is equivalent to what you would pass to the `helm install` or `helm upgrade` commands via a `-f values.yaml` argument.

This code gives you a starting point to manage the Stackdriver Metrics Adapter Helm chart deployment with Rancher through Pulumi. To run it effectively, you'll first need to have Pulumi set up and configured on your local machine, as well as the required permissions and configurations in Rancher to allow for the deployment of applications via Helm charts on Kubernetes clusters.