Deploy the simple-mssql helm chart on Kubernetes

Question

Pulumi · Accepted Answer

Certainly! To deploy the `simple-mssql` Helm chart on Kubernetes using Pulumi, you'll need to follow these steps:

1. **Set up Pulumi**:
   Ensure you have Pulumi CLI installed and you've logged in to your Pulumi account. You should also have `kubectl` configured to communicate with your Kubernetes cluster.

2. **Create a New Pulumi Project**:
   Start by creating a new Pulumi project using `pulumi new` and choose the appropriate template for TypeScript.

3. **Install Required Packages**:
   You need to have the Pulumi Kubernetes SDK installed in your project to work with Kubernetes resources. This is done via the package manager for your language; in this case, we'll use npm for TypeScript.

```bash
   npm install @pulumi/kubernetes
   ```

4. **Write the Pulumi Program**:
   Write the deployment code that specifies the Kubernetes `Chart` resource for `simple-mssql`.

Below is a TypeScript program that performs the helm chart deployment:

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

const chartName = "simple-mssql";
const chartVersion = "1.0.0"; // Replace with the correct chart version
const releaseName = "simple-mssql-release";

// Deploy simple-mssql helm chart
const simpleMssqlChart = new k8s.helm.v3.Chart(releaseName, {
    chart: chartName,
    version: chartVersion,
    fetchOpts: {
        // Here you can specify options like the repo URL where your chart is located
        // For example:
        // repo: "https://charts.bitnami.com/bitnami"
    },
    // If the chart requires you to override values, specify those here:
    values: {
        // You might need to set required parameters for your chart here, as an example:
        // service: {
        //     type: "ClusterIP"
        // },
        // persistence: {
        //     enabled: true,
        //     size: "8Gi"
        // }
    }
}, { provider: k8sProvider }); // Ensure you pass the correct K8s provider if you're working with multiple clusters

// Export the chart's service `loadBalancer` IP if appropriate
export const sqlServiceIP = simpleMssqlChart
    .getResourceProperty("v1/Service", `${releaseName}-${chartName}`, "status")
    .apply(status => status.loadBalancer?.ingress[0].ip)
```

In this piece of code:

- We import the `@pulumi/kubernetes` package to interact with Kubernetes resources.
- `simple-mssql` is specified as the chart we wish to deploy. Make sure to specify the correct version of the Helm chart in the `chartVersion` variable. The Helm chart version should match the version you want to deploy.
- We create a new Helm chart instance using `new k8s.helm.v3.Chart`. The `releaseName` variable is the name you want to assign to your Helm release.
- We pass a configuration object that includes `chart`, `version`, and `values`. The `values` object is where you can override default chart values as needed. It's common for Helm charts to accept various configurations through this parameter. Refer to the [`simple-mssql` Helm chart documentation](https://artifacthub.io/packages/helm/some-repository/simple-mssql) to determine the required and available configurable values.
- In the `fetchOpts` object inside the config, you'll specify the storage options, such as where to find the chart if it's not in the default Helm repository.
- Optionally, we export the service IP address which would be relevant if the service that's created by the Helm chart is of type `LoadBalancer` and you want to make it accessible externally.
- The last parameter `{ provider: k8sProvider }` is where you pass your provider details. This is especially useful when you are working with multiple Kubernetes clusters; otherwise, it will automatically use the default provider configuration.

To run this program, you will typically execute `pulumi up` to preview and deploy the changes. Pulumi compiles the TypeScript code, figures out what needs to be deployed or updated, and shows you a preview. If you confirm, it then proceeds to make the necessary changes to your Kubernetes cluster.

Don't forget to replace the placeholder values in the `fetchOpts`, `values`, and `k8sProvider` with actual values that match your specific scenario.