Deploy the inferenceservices-controller helm chart on AWS EKS

Question

Pulumi · Accepted Answer

To deploy the `inferenceservices-controller` Helm chart on AWS EKS using Pulumi, you'll first need an EKS cluster. If you haven't already created an EKS cluster, the program will include a definition for an EKS cluster as well. After the cluster setup, we will use the Helm chart resource from the Pulumi Kubernetes provider to deploy the `inferenceservices-controller`.

The following TypeScript program includes:

1. Setting up an AWS EKS cluster using the `eks.Cluster` class from the Pulumi EKS package.
2. Creating a Kubernetes provider that connects to the EKS cluster.
3. Deploying the `inferenceservices-controller` Helm chart using the `Chart` resource from the Pulumi Kubernetes provider.

In this program, we will create an EKS cluster with some default configurations, which you may adjust according to your needs such as `desiredCapacity`, `minSize`, and `maxSize`.

Please note that for a specific `inferenceservices-controller` Helm chart, you might need to provide additional configuration parameters or specify the correct Helm repository. This program assumes that the chart is publicly available in a Helm repository. If it's not available, you would need to add the Helm repository using the `helm repo add` command before using this Pulumi program.

Here is the comprehensive Pulumi TypeScript program that achieves this:

```typescript
import * as pulumi from '@pulumi/pulumi';
import * as aws from '@pulumi/aws';
import * as eks from '@pulumi/eks';
import * as k8s from '@pulumi/kubernetes';

// Create an EKS cluster.
const cluster = new eks.Cluster('my-cluster', {
    // Specify the desired number of cluster nodes.
    desiredCapacity: 2,
    minSize: 1,
    maxSize: 3,
    // Pulumi EKS Cluster documentation: https://www.pulumi.com/registry/packages/eks/api-docs/cluster/
});

// Create a Kubernetes provider instance that uses our EKS cluster.
const k8sProvider = new k8s.Provider('k8s-provider', {
    kubeconfig: cluster.kubeconfig.apply(JSON.stringify),
});

// Deploy the 'inferenceservices-controller' Helm chart to the EKS cluster.
const inferenceservicesControllerChart = new k8s.helm.v3.Chart('inferenceservices-controller', {
    chart: 'inferenceservices-controller',
    // If your Helm chart is in a specific repository, specify the 'repo' property.
    // repo: "https://example.com/helm",
    // You can pass custom values to the Helm chart by using 'values' property.
    // values: { /* Custom values here */ },
    // Kubernetes Helm Chart documentation: https://www.pulumi.com/registry/packages/kubernetes/api-docs/helm.sh/v3/chart/
}, { provider: k8sProvider });

// Export the cluster's kubeconfig and the Helm chart's status.
export const kubeconfig = cluster.kubeconfig;
export const chartStatus = inferenceservicesControllerChart.status;
```

In this program:

- We first define our EKS cluster with minimal configuration required for a quick setup.
- We then create a Kubernetes provider that knows how to communicate with our EKS cluster using the `kubeconfig` output.
- We deploy the `inferenceservices-controller` Helm chart using the `Chart` resource from the `@pulumi/kubernetes` package.

Make sure you have Pulumi installed and AWS credentials configured. If you are new to Pulumi, you can follow their [Getting Started guide](https://www.pulumi.com/docs/get-started/).

Please remember that deploying cloud resources with Pulumi can incur costs. Always review the pricing details for the resources created and ensure that you clean up resources with `pulumi destroy` if they are no longer needed.