Set the default GKE cluster type for ray to GKE Autopilot.

Also add instructions to use a standard cluster if preferred.

applications/ray/README.md

@@ -3,22 +3,58 @@
 This repository contains a Terraform template for running [Ray](https://www.ray.io/) on Google Kubernetes Engine.
 See the [Ray on GKE](/ray-on-gke/) directory to see additional guides and references.
 
+## Prerequisites
+
+1. GCP Project with following APIs enabled
+    - container.googleapis.com
+    - iap.googleapis.com (required when using authentication with Identity Aware Proxy)
+
+2. A functional GKE cluster.
+    - To create a new standard or autopilot cluster, follow the instructions in [`infrastructure/README.md`](https://github.com/GoogleCloudPlatform/ai-on-gke/blob/main/infrastructure/README.md)
+    - Alternatively, you can set the `create_cluster` variable to true in `workloads.tfvars` to provision a new GKE cluster. This will default to creating a GKE Autopilot cluster; if you want to provision a standard cluster you must also set `autopilot_cluster` to false.
+
+3. This module is configured to optionally use Identity Aware Proxy (IAP) to protect access to the Ray dashboard. It expects the brand & the OAuth consent configured in your org. You can check the details here: [OAuth consent screen](https://console.cloud.google.com/apis/credentials/consent)
+
+4. Preinstall the following on your computer:
+    * Terraform
+    * Gcloud CLI
+
 ## Installation
 
-Preinstall the following on your computer:
-* Terraform
-* Gcloud
+### Configure Inputs
 
-> **_NOTE:_** Terraform keeps state metadata in a local file called `terraform.tfstate`. Deleting the file may cause some resources to not be cleaned up correctly even if you delete the cluster. We suggest using `terraform destory` before reapplying/reinstalling.
+1. If needed, clone the repo
+```
+git clone https://github.com/GoogleCloudPlatform/ai-on-gke
+cd ai-on-gke/applications/ray
+```
 
-1. If needed, git clone https://github.com/GoogleCloudPlatform/ai-on-gke
+2. Edit `workloads.tfvars` with your GCP settings.
+
+**Important Note:**
+If using this with the Jupyter module (`applications/jupyter/`), it is recommended to use the same k8s namespace
+for both i.e. set this to the same namespace as `applications/jupyter/workloads.tfvars`.
+
+| Variable                    | Description                                                                                                    | Required |
+|-----------------------------|----------------------------------------------------------------------------------------------------------------|:--------:|
+| project_id                  | GCP Project Id                                                                                                 | Yes      |
+| cluster_name                | GKE Cluster Name                                                                                               | Yes      |
+| cluster_location            | GCP Region                                                                                                     | Yes      |
+| kubernetes_namespace        | The namespace that Ray and rest of the other resources will be installed in.                                   | Yes      |
+| gcs_bucket                  | GCS bucket to be used for Ray storage                                                                          | Yes      |
+| create_service_account      | Create service accounts used for Workload Identity mapping                                                     | Yes      |
+
+
+### Install
+
+> **_NOTE:_** Terraform keeps state metadata in a local file called `terraform.tfstate`. Deleting the file may cause some resources to not be cleaned up correctly even if you delete the cluster. We suggest using `terraform destory` before reapplying/reinstalling.
 
-2. `cd applications/ray`
+3. Ensure your gcloud application default credentials are in place. 
+```
+gcloud auth application-default login
+```
 
-3. Find the name and location of the GKE cluster you want to use.
-   Run `gcloud container clusters list --project=<your GCP project>` to see all the available clusters.
-   _Note: If you created the GKE cluster via the infrastructure repo, you can get the cluster info from `platform.tfvars`_
+4. Run `terraform init`
 
-4. Edit `workloads.tfvars` with your environment specific variables and configurations.
+5. Run `terraform apply --var-file=./workloads.tfvars`. 
 
-5. Run `terraform init && terraform apply --var-file workloads.tfvars`

applications/ray/variables.tf

@@ -39,7 +39,7 @@ variable "ray_version" {
 variable "kubernetes_namespace" {
   type        = string
   description = "Kubernetes namespace where resources are deployed"
-  default     = "myray"
+  default     = "ml"
 }
 
 variable "enable_grafana_on_ray_dashboard" {
@@ -105,7 +105,7 @@ variable "private_cluster" {
 
 variable "autopilot_cluster" {
   type    = bool
-  default = false
+  default = true
 }
 
 variable "cpu_pools" {

applications/ray/workloads.tfvars

@@ -17,11 +17,16 @@
 ## Need to pull this variables from tf output from previous platform stage
 project_id = "<your project ID>"
 
-## this is required for terraform to connect to GKE master and deploy workloads
-create_cluster   = false # this flag will create a new standard public gke cluster in default network
+## This is required for terraform to connect to GKE cluster and deploy workloads.
 cluster_name     = "<cluster name>"
 cluster_location = "us-central1"
 
+## If terraform should create a new GKE cluster, fill in this section as well.
+##    By default, a public autopilot GKE cluster will be created in the default network.
+##    Set the autopilot_cluster variable to false to create a standard cluster instead.
+create_cluster    = false
+autopilot_cluster = true
+
 #######################################################
 ####    APPLICATIONS
 #######################################################